GEOGRAPHICAL APPENDIX.

NOTE I.

The Accounts of Gallego and Figueroa compared.—On carefully comparing these two accounts, I have no doubt that Figueroa derived almost all his information from the journal of Gallego. He, to a great extent, employs his own phraseology; but in the descriptions of the islands and of the natives, the words and expressions employed are often identical, and the mode and order of description are evidently supplied by the journal of Gallego. An indirect proof of the source, whence Figueroa drew his materials, is to be found in the circumstance that, after the two vessels were separated during the voyage back to Peru, he confines his account to the experiences of the “Capitana,” which was Gallego’s vessel; and here his account is substantially a condensed form of Gallego’s journal which is occasionally quoted literally. Figueroa, however, does not inform us of the source of his information; and he has evidently, in some measure, endeavoured to infuse his own method of expression into the account. There are not wanting proofs, however, that he was assisted from other sources, but only in a small degree. For instance, he occasionally intercalates a circumstance to which Gallego does not allude; and he varies in the accounts of the conflicts with the natives: thus he refers to some of the Spaniards having died at Estrella Harbour, to there being a foot and a half of water in the hold of one of the ships during the return voyage, to the ships being heaved-down at St. Christoval, and to a few other similar occurrences unrecorded by Gallego. The account of Figueroa differs in the date of the year of the voyage. It contains only a bare reference to the cruise of the brigantine to St. Christoval and its adjacent islands, whilst the vessels lay at the Puerto de la Cruz on the coast of Guadalcanar. It is from this cause that the names of all the islands visited and named during this cruise of the brigantine are not given in Figueroa’s account. Herrera, however, in his short description of these islands, gives a full list of the names of the islands, and, in this respect, his description is superior to that of Figueroa.

NOTE II.

Discrepancies in the Dates of the Years.—There is a strange discrepancy in the dates of the years during which this expedition was away from Peru. The year 1566, is given on the title-page of the British Museum copy of Gallego’s Journal; and the author expressly states that the expedition left Callao on November 19th, 1566; he carries this year on, naming the following year, 1567; but in August he gives the year as 1568, and makes the return to Peru to be in 1569. It is evident from the narrative that the ships were absent from Peru about nineteen months, from November of one year to June of the second ensuing year; and it is highly probable that the year of their departure was 1566, and that of their return 1568. . . . Figueroa differs strangely in the dates he gives.[393] In the first line of his account he says that the ships were dispatched in 1567; and in the succeeding paragraph he gives January 10, 1568, as the date of their departure from Callao, thus being quite at variance with Gallego, both as regards the day, the month, and the year. The ships reached the coast of Mexico on their return voyage in January 1568, according to Figueroa. From this inconsistency it may be inferred, that 1567 was intended as the date of the departure from Peru. . . . Herrera,[394] in his description of these islands, states that they were discovered in 1567, which accords with the narrative of Gallego. . . . Arias[395] in a memorial addressed to Philip III. of Spain, says that Mendana discovered San Christoval in 1565; but his account is both short and confused, and was evidently not derived from original sources. . . . Notwithstanding the conflicting character of the dates, the probable dates would appear to be as follows.—The ships left Peru on November 19th, 1566, discovered the Isles of Salomon on February 7th, 1567, and arrived at Peru on June 19th, 1568.

[393] “Hechos de Don Garcia H. de Mendoza,” por el Doctor Christoval S. de Figueroa. Madrid, 1613.

[394] “Descripcion de las Indias Occidentales.” (Madrid, about 1601.)

[395] “Early Voyages to Terra Australis,” by R. H. Major (p. 1). Hakluyt Society, 1859.

NOTE III. ([Page 199].)

The Isle of Jesus.—Burney[396] estimated the longitude of this island to be 172° 30′ East of Greenwich; Krusenstern,[397] on surer grounds, fixed it at 171° 30′: but both estimates were based on an erroneous longitude of the Candelaria Shoals. . . . I have shown in [note iv.] that these shoals are probably identical, not with the Roncador Reef as is implied in the present [charts], but with the islands of Ontong Java, to the northward; however, this correction affects but little the question of longitude. Taking the longitude of the centre of Ontong Java at about 159° 30′ E. (in lat. 5° 25′ S.), the longitude of the Isle of Jesus, 167 Spanish leagues to the eastward (in lat. 6° 45′ S.), would be about 169° E. The only island shown on the present charts in the vicinity of this position is Kennedy Island, also called Motuiti, the existence of which is stated to be doubtful. Its position, as determined by the “Nautilus” in 1801, was 8° 36′ S. 167° 50′ E.[398] However, in 1883, the German war-vessel “Carola” failed to find it in this position in the chart, and the initials E. D. are there attached to the name. The difficulty may, I think, be explained by the existence in this region of some atoll of no great size, the position of which has been never correctly determined. It would appear that a similar view is held by Captain Wharton, the present Hydrographer, since in the Sailing Directions for these seas issued in 1885, the island is still given prominent mention.[399] Not improbably the missing island will be found between the 6th and 7th parallels, and near the position assigned to the Isle of Jesus.

[396] “Chronological History of Vovages and Discoveries in the South Sea.” Vol. I. p. 289. London, 1803.

[397] “Recueil de Mémoires Hydrographiques.” St. Petersburg, 1824.

[398] Findlay’s “Directory of the Pacific Ocean.” Part II 999. (London, 1851.)

[399] “Pacific Islands.” Vol. I. p. 50. (Western Groups.) 1885.

Herrera gives the name of another island, “El Nombre de Dios,” which is said by him to lie in 7° S. lat., and to be 50 leagues distant from Santa Anna; Gallego does not refer to any island with this name; and since Herrera makes no reference to the Isle of Jesus, it is possible that this isle may be here alluded to, as its latitude corresponds somewhat with that of “El Nombre de Dios.” M. Fleurieu[400] identifies this island, however, not with the Isle of Jesus, but with an island off the north end of Malaita which was named Gower I. by Captain Carteret in 1767, and Inattendue I. by M. Surville, in 1769.

[400] “Discoveries of the French, 1768-1769, to the S. E. of New Guinea,” p. 181. (London, 1791).

NOTE IV. ([Page 199].)

The Candelaria Shoals.—The shoals were identified by Fleurieu with the Roncador Reef discovered by Maurelle in 1781; and Krusenstern subsequently confirmed this opinion. Gallego, however, describes shoals trending N.E. and S.W. for more than fifteen leagues, which cannot possibly be the Roncador Reef of the present [chart], which is not more than six miles across. These Candelaria Shoals, on the other hand, correspond in their size with the large atoll of Ontong Java lying about 35 miles to the north of the Roncador Reef, and being about 50 miles in width. The apparent difference in latitude between Ontong Java, which lies in about 5° 25′ S., and the Candelaria Shoals of Gallego, which were placed by him in 6° 15′ S., may be explained by the circumstance that the majority of Gallego’s observations of latitude in the Solomon Group were about two-thirds of a degree in excess of the true latitude.[401] By making this correction, the latitude of Ontong Java and of the Candelaria Shoals will be found to closely approximate. The bearing and distance of the Candelaria Shoals from the west end of Malaita (as given by Gallego on p. 205) and from Estrella Harbour (as given on p. 202) go to support my view that the Candelaria Shoals of Gallego and the Ontong Java of Tasman are one and the same.

[401] Vide [Note V.] of Geographical Appendix.

NOTE V.

The Latitudes of Gallego in the Solomon Group.—On making fourteen comparisons of the latitudes obtained by Gallego with the latitudes of the same places in the most recent Admiralty [charts], places about which there can be no doubt as to their identity, I find that all but two are in excess of the true latitude. The excess varies between 11′ and 1° 7′ (about); and since seven of the twelve latitudes vary between 38′ and 46′ in excess, we may take 40′ plus as about the probable and average prevailing error of Gallego’s observations of latitude in this group. A constant error points to some constant defect of observation; whether it may be instrumental or otherwise, I must leave to the judgment of my nautical readers. . . . . It may be inferred from his journal that Gallego did not endeavour to make his latitudes by observation accord with his bearings, as they are so often at variance. This circumstance should be borne in mind in order to explain the discrepancies that occur.

NOTE VI. ([Page 206].)

The Isle of Ramos and the Island of Malaita.—On referring to the account of Figueroa in the original Spanish, I find that, like Gallego, he applies the name of Ramos to Malaita. Pingré, who published a translation of Figueroa’s account in 1767 at Paris,[402] associates the two names together. Dalrymple[403] in his translation, published in 1770, laid the ground for future misconception, by so pointing the sentence that the name of Ramos might be taken as intended for one of the “two islets” in the middle of the passage between Malaita and Isabel. Fleurieu,[404] in his translation of Figueroa published in Paris in 1790, applies the name of Ramos to Malaita. Burney,[405] in his version (1803), apparently applies this name to one of the islets above referred to. The authority of Dalrymple and Burney would appear to supply an explanation of the circumstance that in the present Admiralty charts this name of Ramos is applied to an islet between Malaita and Isabel; but Dalrymple’s version is susceptible of two meanings, and may be urged with equal justice on either side. Gallego and Figueroa both apply the two names to the same island; so that circumstance alone is sufficient to justify the restoration to Malaita of the Spanish name of “The Isle of Ramos.” The original cause of the mistake is to be attributed to the first discoverers, who gave their own name and were not content with the native name. Herrera[406] has fallen into the opposite error, since, in distinguishing between Malaita and Ramos, he gives the latter a circuit of 200 leagues.

[402] “Mémoire sur le choix et l’état des lieux où le passage de Vénus du 3 Juin, 1769.” (Paris, 1767.)

[403] “Hist. Coll. of Voy. and Discov.,” London, 1770.

[404] “Discoveries of the French in 1768 and 1769.”

[405] “Chronol. Hist. Voy and Discov.,” vol. i.

[406] “Descripcion de las Indias Occidentales.”

NOTE VII. ([Pages 207]-[209].)

The Islands between Cape Prieto and Guadalcanar.—These islands which occupied the attention of Fleurieu and Burney, and excited the curiosity of Dentrecasteaux, and which D’Urville had intended to have completely explored, have long baffled the efforts of geographical writers, who have endeavoured to identify them with the islands mentioned by Figueroa in his brief account of Mendana’s discoveries in this region. His description is evidently derived from that of Gallego, of which it is but an imperfect and erroneous extract: and I will therefore disregard it. The island of Galera is apparently a small island, not named in the present [chart], which lies close to the north-west coast of Buena Vista. The neighbouring large island, a league distant, to which Gallego only applies the native name of Pela,[407] is, as I apprehend, the Buena Vista of the present [chart]: the Buena Vista of the Spaniards is apparently an island, not named in the [chart], which lies west of the present Sandfly Passage. The remaining four of the five islands may be in the future identified with the incompletely surveyed intersected mass of land to which the general name of Florida is applied in the present [chart]. The island of Sesarga is without doubt the volcanic island of Savo: but I must refer the reader elsewhere for further information on this subject of Sesarga.[408]

[407] At the present day the whole of the Florida sub-group is known to the natives as Gela. (Codrington’s “Melanesian Languages,” p. 522.)

[408] The evidence is given in my volume of Geological Observations.

NOTE VIII. ([Page 220].)

The Excessive Dimensions of Guadalcanar.—How could such misconceptions have arisen? They are totally inconsistent with the rest of the journal; and to such statements must be attributed the exaggerated reports which long prevailed with reference to the size of this island. The lengths of the islands of Isabel, Malaita, and St. Christoval, as given by Gallego, are greatly overstated; in the case of the two former islands they are at least double the true dimensions, and they completely disagree with the latitudes and bearing, which are noted in the journal.

NOTE IX. ([Page 233].)

The Consultation as to the Future Course of the Expedition.—The ignorance in which Mendana seems to have kept his officers with regard to the character of his instructions considerably hampered the captains and pilots in their consultation. We learn subsequently ([page 237]) that it was originally intended to prosecute the voyage westward in order to explore the extensive lands that lay in that direction. However, the protest made by the crews seems to have caused a change of plans. They were to steer northward for the Isle of Jesus, where Gallego apparently expected to find more land, as they provided themselves with natives as interpreters ([page 233]) before quitting the group. This northerly course found favour, when Gallego pointed out that it was on the track of their return voyage.

NOTE X. ([Page 234].)

Islands in the Solomon Group which do not at Present bear the Names given to them by the Spaniards:—

Present name.Spanish name.
UgiSan Juan
Three SistersLas Tres Marias
Ulaua (Contrarieté)La Treguada
MalaitaRamos (Isle of)
SavoSesarga
Ontong JavaCandelaria Shoals
ChoiseulSan Marcos
New Georgia (?)- San Nicolas
Arracises (Reefs).

NOTE XI. ([Page 237].)

Inigo Ortez de Retes and Bernardo de la Torre.—We learn from Galvano’s “Discoveries of the World,”[409] that in 1545 Captain Inigo Ortez de Rotha was dispatched from Tidore to New Spain. He sailed to the coast of Papua, and not knowing that Saavedra had discovered it in 1528, he assumed the honour of the discovery. Mr. Coutts Trotter in a recent article[410] refers to him as Ortiz de Retez or Roda, and he informs us elsewhere[411] that Antonio de Abreu was probably the first discoverer of New Guinea in 1511. According to Galvano ([page 234]), a Spanish officer named Bernaldo de la Torre started from the Philippines in 1543, on a voyage to New Spain.

[409] Hakluyt Society’s Publication, 1862, p. 238.

[410] Encyclopædia Britannica (Article on “New Guinea.”)

[411] Proceedings, Royal Geographical Society, 1884, p. 196.

NOTE XII. ([Page 238].)

The islands of San Bartolomeo.—The Musquillo Islands of the Marshall Group, with which I have identified this discovery of the Spaniards, were thus named by Captain Bond in 1792.[412] They form a double atoll about 38 miles in length and trending N.W. and S.E. The N.W. end is in latitude 8° 10′ N., and the S.E. end is in latitude 7° 46′ N. Captain Bond ranged along the coasts of above 20 small islands. At the N.W. end and isolated from the rest are two small islands about three miles apart. On comparing this description with that given by Gallego, the reader will have little doubt as to the identity of the Musquillo Islands with the Spanish discovery. It is probable that Gallego considered this discovery to be near the position of an island discovered in 1536 in 14° N lat. by Toribio Alonzo de Salazar,[413] 328 Spanish leagues from the Mariana Islands, and named by him San Bartolomeo. This discovery of Salazar is marked in Krusenstern’s General Atlas of the Pacific.

[412] Purdy’s “Oriental Navigator,” p. 689.

[413] Krusenstern’s “Mémoires Hydrographiques,” St. Petersburgh, 1827: Part II, p. 49.

NOTE XIII. ([Page 239].)

The Isle of San Francisco.—Wake’s Island, with which I have identified the Isle of San Francisco, was discovered in 1796 by the “Prince William Henry.” Commodore Wilkes, who fixed its position in 1840 (lat. 19° 10′ 54″ N.; long 166° 31′ 30″ E. of G), thus describes it. “Wake’s Island is a low coral one, of triangular form and eight feet above the surface. It has a large lagoon in the centre, which was well filled with fish of a variety of species; amongst these were some fine mullet. There is no fresh water on the island, and neither pandanus nor cocoa-nut tree. It has upon it the shrubs, which are usually found on the low Islands of the Pacific, the most abundant of which was Tournefortia. The short-tailed albatross is found here; birds quite tame though not as numerous as in other uninhabited islands. The appearance of the coral blocks and vegetation leads to this conclusion that the island is at times submerged or that at times the sea makes a complete breach over it.”[414] Wake’s Island is about the size of the island described by Gallego. Its latitude, its isolated position, and the close agreement of Wilkes’ description with that of Gallego, leave no room to doubt that Wake’s Island and the Isle of San Francisco are one and the same . . . Burney refers to a small island named San Francisco which is placed in the chart of the Galleon in Anson’s voyage in lat. 191⁄2 north and 84° east of the Strait of San Bernardino; but he adds that it is too far to the east to be identified with the island discovered by Mendana.[415]

[414] “Narrative of the United States Exploring Expedition,” vol V. p. 267.

[415] “Chronol. History of Voy. and Disc.” vol I. p. 291.

NOTE XIV. ([Page 251].)

The List of Islands in the Vicinity of Taumaco which was obtained by Quiros in 1606 from one of the natives.—They are as follows, Chicayana, Guantopo, or Guaytopo, Taucalo, Pilen, Nupan, Pupam, Fonfono or Fonofono, Mecaraylay, Manicolo, Tucopia, Pouro. More than half of these islands can be identified with certainty, even after an interval of nearly three centuries.

Chicayana may be without a doubt identified with Sikyana or Sikai-ana, the present native name of the Stewart Isles which lie about 250 miles to the north-west of Taumaco, or as the Taumaco people reckoned, four days’ sail in their large canoes. In fact, the native from whom Quiros obtained his information was originally from Chicayana, having been carried by contrary winds to Taumaco whilst endeavouring with a number of his fellow-islanders to reach the island of Mecaraylay. The Chicayana natives were described to Quiros as being very fair with long loose red hair, some, however, being darker like mulattoes, but with hair neither curled nor quite straight. They possess much the same characters at the present day.[416]

[416] These islands, as far as is known, were not visited by Europeans until nearly two centuries after the visit of Quiros, when Captain Hunter came upon them in 1791.

Guaytopo or Guantopo was a larger island than those of Taumaco and Chicayana. Since it is placed three days’ sail (native reckoning) from Taumaco and two days from Chicayana, it may have been one of the eastern islands of the Solomon Group. The inhabitants were said to have skins as fair as Europeans and red or black hair. They punctured their bellies in a pattern of a circle around the navel; and painted their bodies red down to the waist. The women were very handsome and were clothed with some light material from head to foot. The natives of Guaytopo, Taumaco, and Chicayana, were on very friendly terms and spoke the same language.

The islands of Pilen and Nupan are evidently the Pileni and Nupani of the adjacent Matema or Swallow Islands, which lie to the northward of the large island of Santa Cruz. Fonofono or Fonfono, which is stated to lie near Pilen and Nupan, may perhaps be the Lomlom of the same small group. It was described to Quiros as being “many islands, small and flat,” with a good port. The inhabitants were said to be dun-coloured, and very tall.

Tucopia was subsequently visited by the Spanish navigator. In later times it has obtained a melancholy interest in connection with the fate of La Pérouse. Mécaraylay is apparently in the vicinity of Guaytopo, but possessing a different language, its inhabitants being noted for the use of tortoise-shell ornaments. Its name suggests that of Makira, on the south coast of St. Christoval, in the neighbouring Solomon Group. Taucalo may perhaps be the volcanic island of Tinakula lying off the north coast of Santa Cruz Island. It is stated to be near Taumaco.

The “large country” called Manicolo is to be identified with the adjacent large island, named Vanikoro in the present Admiralty charts, which lies about 100 miles to the southward of Taumaco. It is referred by Captain Cook[417] to the Mallicolo of the New Hebrides, lying 4° further south, which he visited in 1774; but this view cannot be sustained. In the first place, it is stated to lie two days’ sail from Tucopia. The following evidence, however, is sufficient of itself to settle the point. When Captain Dillon[418] was on his way to Vanikoro in 1827, to ascertain the fate of La Pérouse, he learned from the natives of the neighbouring island of Tucopia that the island he was going to was called Malicolo: but he subsequently ascertained on visiting the island in question, that it should be more correctly called Mannicolo or Vannicolo. In his chart of the island, Captain Dillon calls it Mannicolo. The resemblance in name between these two islands in the New Hebrides, and Santa Cruz Groups has been a frequent cause of misconception in references to the narratives of the early navigators.

[417] “Voyage towards the South Pole and round the World,” vol. II., p. 146.

[418] “Discovery of the fate of La Pérouse,” London, 1829: vol. I., p. 33.

NOTE XV. ([Pages 100], [251].)

The Pouro of Quiros.—A native of Chicayana, whom Quiros had captured at Taumaco, told the Spanish navigator that there dwelt in Taumaco “an Indian, a great pilot,” who had brought from “a large country, named Pouro,” certain arrows, with points, in the form of a knife, which, from the native’s description, Quiros concluded were of silver. Pouro, he learned, was very populous, and its inhabitants were dun-complexioned.

When I first came upon this reference to Pouro, I at once recognised it as an allusion to the Bauro (St. Christoval) of the Solomon Group, lying rather less than 300 miles to the westward of Taumaco. Mr. Hale,[419] the philologist of the United States Exploring Expedition, under Commodore Wilkes, endeavours to identify the Pouro of the Taumaco natives with the Bouro in the Malay Archipelago, an island lying more than 2,000 miles further westward: and he refers to the circumstance of the silver arrows that were brought to Taumaco as supporting his view. Regarding Bouro as the island referred to in the traditions of the Fijians, Tongans, and Samoans, relating to the origin of their race, Mr. Hale finds in the Pouro of the Taumaco natives an allusion to this sacred island, and in the circumstance of the silver arrows he finds evidence of communication between these two regions. There can, however, be little doubt that by this Pouro the Bauro of the Solomon Group was meant. The presence of the silver arrows may be easily explained, when we remember that about forty years before, the Spaniards were exploring this island of Bauro, or Paubro as Gallego gives it ([page 229]).

[419] “Ethnography and Philology of the U. S. Exploring Expedition,” p. 195.

NOTE XVI.

The Eddystone Rock and the Simboo of Lieutenant Shortland.—For a considerable time after the re-discovery of the Solomon Islands by the French and English navigators, few islands were better known in the group than Eddystone or Simbo Island. In thus naming this island, however, there has been a singular misconception; and since the name of Simbo has been omitted in the latest Admiralty [chart] (August, 1884) of the group, some explanatory remarks may be of interest.

In August, 1788, Lieutenant Shortland,[420] whilst sailing along the south coasts of the Solomon Group on his voyage from Port Jackson to England via Batavia, approached “a rock which had exactly the appearance of a ship under sail, with her top-gallant sails flying;” and so striking was the resemblance that a signal was made to the supposed vessel. The ships did not approach within three or four miles of this rock. It was named the Eddystone and was placed in lat. 8° 12′ S., bearing S.S.W. a league from two remarkable hills which were named the Two Brothers. A point running south from these two hills was named Cape Satisfaction. Whilst the English ships were off the Eddystone, some natives came to them in their canoes, from whom Shortland learned that they had come from “Simboo,” a place which lay, as they indicated by their gestures, near Cape Satisfaction. In the chart of his discoveries, this officer assigns this name to some land lying east of the Two Brothers near the position of the island at present called Gizo, but it is evident both from his chart and from his narrative that he considered Simboo as the general name for the land to the east of Cape Satisfaction; and Fleurieu, when remarking on his discoveries, made the suggestion that the Simboo of Shortland might prove to be the Choiseul of Bougainville.[421]

[420] The narrative of Lieut. Shortland’s voyage is given in “The Voyage of Governor Phillip to Botany Bay in 1787”: London, 1789.

[421] “Discoveries of the French, 1768-1769, to the S.E. of New Guinea:” London, 1791, p. 196.

In what manner, we may now inquire, have the discoveries of Shortland been identified with the islands that are laid down in the latest charts of this group? For half a century and more the name of Eddystone has been attached, not to a rock such as that to which it was originally given, but to the adjacent volcanic island about four miles in length and about 1100 feet in height; and the name of Cape Satisfaction has been given to the south end of Ronongo which lies ten miles N.N.E. of Eddystone Island. This cape is stated by Shortland to run south from the two remarkable hills which he named the Two Brothers. The island of Ronongo, however, has a long and level summit destitute of peaks; and it is evident that we must look elsewhere for the Cape Satisfaction of Shortland. In Eddystone Island, there are two singular conical hills which might very fitly have been named the Two Brothers, and it will be seen from the sequel that it must have been to the south extremity of this island that the name of Cape Satisfaction was in the first place given. I shall also point out that the original Eddystone rock is represented at the present day by a bare rock which rises out of the sea at a distance of about a third of a mile from the south-west coast of Eddystone Island, and that the Simboo, from which the natives came to visit Shortland, was a diminutive island on the opposite or south-east side of this same island.

When, in July 1792, the French expedition under Dentrecasteaux arrived in this locality, the Eddystone rock was at once recognised by the description of Shortland. . . . “nous aperçûmes”—thus wrote Labillardière[422] the naturalist of the expedition—“le rocher nommé Eddystone. De loin nous le prîmes, comme Shortland, pour un vaisseau à la voile. L’illusion étoit d’autant plus grande, qu’il a à peu près la couleur des voiles d’un vaisseau; quelques arbustes en couronnoient la sommite.” In the Atlas of this voyage (carte 24), this rock is placed off the south-west end of the island at present named Eddystone Island, and exactly in the position of the bare rock above alluded to, which will be found marked in the plan of this island made by the surveying officers of H.M.S. “Lark” in 1882. Lieutenant Malan tells me that this rock at the time of the survey was quite bare of vegetation. It rises in two conical masses from the water between which a boat can pass in calm weather. Although it has a height of 30 feet, it is frequently washed over by the heavier seas. The change in the appearance of this rock, since the visit of Dentrecasteaux in 1792 when its summit was crowned with shrubs, has been probably due to a movement of subsidence which has affected the adjacent coast of Eddystone Island in recent years (vide [below]). To such a change must be attributed the confusion which has arisen with reference to the Eddystone rock; and cartographers, failing to identify it, have applied its name to the adjacent volcanic island on which they have also bestowed the name of Simbo. During his survey of this island in 1882, Lieutenant Oldham ascertained that this name of Simbo actually belonged to a small island bordering its south-east coast with which it was connected by coral reefs. The true native name of Eddystone Island, he found to be Narovo, and in the latest Admiralty [charts] it is thus designated; the name of Simbo is there attached to the small adjacent island which is, I have no doubt, the Simboo from which the natives came, who visited Shortland’s ships in 1788 as they lay off the Eddystone rock. At the present day the larger island of Narovo is but thinly populated, and its inhabitants are under the sway of a powerful chief who resides on the small island of Simbo. There he rules over a warlike and adventurous people who by their head-hunting raids have established the fame of their diminutive island throughout a large portion of the Solomon Group.

[422] “Voyage à la recherche De la Pérouse,” par Labillardière: Paris, 1800: tom i, p. 215.

[In my volume of Geological Observations I have described the movement of subsidence, to which is due the confusion concerning the original Eddystone rock].

CHAPTER XIII.
BOTANICAL NOTES IN BOUGAINVILLE STRAITS.

My botanical collections were made during 1884 in the islands of Bougainville Straits; and in order to add to the completeness of this section of my work, I will briefly refer to the physical character of this locality. The principal islands of this sub-group are Treasury Island, the Shortland Islands, and Faro, or Fauro, Island; whilst around these lie numerous smaller islands and islets. The largest is not more than twelve miles in length, and none of them attain an elevation exceeding 2000 feet, Faro being about 1900 feet, Treasury about 1100, and Alu, the principal of the Shortland Islands, about 500 feet. In geological character they differ widely, Treasury being, for the most part, of recent calcareous formations, Faro of volcanic formations, whilst Alu is formed of rocks of both these classes. Of the numerous smaller islands and islets which dot these straits, some are of volcanic, and others of coral rocks.

In my botanical excursions in these islands, I received the greatest assistance from the natives; and I was particularly struck with the familiar knowledge of their trees and plants which these islanders possessed. They have names for not only nearly all the trees, but for several of the grasses; and, in the case of the former, when I was uncertain as to whether I had come upon any specimen before, they would obtain its flower, or fruit, or foliage, and point out to me its comparative characters. The superior knowledge, which these natives possess of each plant and its uses, has often led me to reflect on the meagre acquaintance with the commonest trees, shrubs, and herbs, which the ordinary white man can claim. Had my native companions asked me to instruct them in a similar manner on the vegetation of an English woodland—if such a rapid change of scene were possible—they would probably have regarded me as a very ignorant and unobservant fellow. They have names for and display a familiarity with many plants that can be of no service to them, a somewhat puzzling circumstance, which may be perhaps explained by their employing instinctively a method of exclusion in the selection of those plants that are of service to them. For the building of his house, the cultivation of his ground, the construction of his canoe, the manufacture of his spears, clubs, and other weapons, and for his many other wants, the native has to resort to the vegetable kingdom for the requisite materials. An extensive acquaintance with the vegetation of his island-home is unconsciously acquired by a native who has himself to provide for all his necessities: but his knowledge extends far beyond that limit which mere utility would appear to demand. In a paper published recently in an American serial,[423] Mr. Matthews combated the notion that savages are versed only in the knowledge of plants and animals that contribute to their wants. He found that the Indians are incomparably superior to the average white man, or to the white man who has not made zoology or botany a subject of study. In this respect, his experience accords with my own. The native of the Solomon Islands will point out by name, in some remote inland dell, an insignificant plant, which, he says, is of no service to him: he names all the weeds of his cultivated patches; and he is similarly acquainted with all the wild fruits, usually distinguishing them by their edible or injurious qualities. Yet, in arriving at such a conclusion, it behoves one to be wary, as I have sometimes found that the native applies the name of a useful plant to all other useless plants (usually of the same genus or family) that resemble it in their more conspicuous characters. Then, again, I have often been surprised at the singular holes and corners in the vegetable world which the native ransacks to supply his wants. A fern that clothes the higher slopes of Faro Island, and which is known to the natives as “sinimi,” and to the botanist as a species of Gleichenia, furnishes the material for their plaited armlets. For this purpose they employ narrow strips of the vascular tissue that forms the firm central portion of the stem. I had previously looked upon this fern as of little use to these islanders, and on learning of the ingenious purpose for which it was employed, I became very careful in the future when pronouncing on the utility or inutility of any familiar plant.

[423] “Bulletin of the Philosophical Society of Washington,” Vol. VII.

With these preliminary remarks, I will proceed to describe the general characters of the vegetation of these islands; and, in order to connect my observations together, I will treat of them in the form of a series of excursions made in different districts.

An ascent of one of the larger streams in the Shortland Islands.—In the lower part of its course, the stream follows a circuitous course amidst the gloom and dismal surroundings of a mangrove swamp. It is difficult to convey in words a true idea of such a scene. The features most imprinted on my memory are those of “a slow and silent stream” of dark turbid water, traversing a swamp of black, repulsive-looking mud, in which the crocodile finds a congenial home. The light of day is subdued into a depressing gloom by the foliage of the mangrove forest: the air, charged with the miasma of decaying vegetable life, is impregnated with a sour, unpleasant odour; and the silence that prevails is interrupted only by the fall of a branch, or by the startled cry of some wading-bird disturbed in its haunt. Nipa palms line the banks in places, and occasionally occupy the swamp for some distance on either side of the stream. Overhead, perched high upon the branches of the tall mangroves, occur the two singular epiphytes, Hydnophytum and Myrmecodia, both of which have been found to be species new to science (H. Guppyanum, Becc.; M. salomonensis, Becc.). From the following remarks, my readers will be able to observe the peculiar features of these interesting rubiaceous plants. The large swollen base of the stem, sometimes eighteen inches in length, is occupied by cavities which are usually infested by ants that actively resent any attempts to carry off their home. It has been considered that this swollen mass and its chambers are due to the irritation produced by the ants gnawing at the base of the young growing stem, and that the plant cannot thrive without the ants; but from observations made by Mr. H. O. Forbes,[424] in Java, on the origin of “this curious-galleried structure” in a species of Myrmecodia, it would seem that this swollen mass and its chambers are produced without the presence of ants, and that in their absence the plant may thrive vigorously. Not unfrequently, I found the ants in scanty numbers, and sometimes they were absent altogether. In the case of Myrmecodia salomonensis, and Hydnophytum inerme,[425] they are found in considerable numbers. The chambers of H. Guppyanum are usually nearly full of dirty rain-water, and contain scarcely any ants, a few cockroaches being generally found in the cavities. Those specimens which I examined of another species of this genus (H. longistylum, Becc.), that occurs on the coast trees, contained a few cockroaches, but no ants; and, on the outer surface of one of the swollen masses, I found a small crab. From my own cursory notes, it would therefore seem probable that these epiphytes may thrive without the presence of ants. . . . . With this digression, I return to my description of the ascent of the stream.

[424] “A Naturalist’s Wanderings in the Eastern Archipelago,” p. 81. (1885.)

[425] This species was obtained at Ugi.

Leaving behind the slime and gloom of the swamp, the rising ground is reached, at the base of which the vegetation is of the most luxuriant character, and often have I lingered here in my Rob Roy canoe to admire the luxuriance of plant-growth that surrounded me. For on account of the lowness of the district, it shares the dampness though not the infertility of the swamps below. The soft clayey rock, which is exposed in the banks of the stream, affords a rich and even too productive soil. Nature runs riot and becomes prodigal in her profusion; and thus growth is too often associated with decay to present on all occasions a pleasing picture to the eye. Here the tree-fern, the croton, the wild plantain, and numerous areca palms flourish; but the alpinias, heliconias, and other scitamineous plants form the chief feature of the vegetation on this gently rising ground.

Higher up the stream, tall forest trees rise on each side often enveloped partially by a drapery of runners and climbing plants, their leafy branches spreading over the water. Stout lianas hang in festoons across the stream. Partly hidden amongst the greater vegetation may be seen the fan-palm of the district (Licuala, the “firo” of the natives), and another pretty little palm known as the “sensisi,” Cyrtostachys, together with the handsome foliage of a Plerandra (“fo”) and numerous areca palms. An occasional Dolicholobium (“lowasi”) with white flowers distributes its fragrance around. Ferns abound along the banks, varying in size from the small Trichomanes to the tree-fern, twenty feet in height, and the Angiopteris with its magnificent spreading fronds fifteen feet and more in length. If one leaves the stream for a few minutes at the foot of the hills, a moist, low-lying district is traversed, the home of the scitamineæ and the areca-palms, which latter are distinguished amongst the natives as the “momo,” “niga-solu,” “niga-torulo,” and “au-au.”

Ascending the hill slopes towards the source of the stream, numerous palms rise up on either side. The Caryota (“eala”) with its branches resembling the fronds of a huge adiantum, the handsome “kisu” (probably a species of Drymophloeus) and a tall areca known as the “poamau,” are those which frequently meet the eye. Interspersed among them we notice the lesser areca-palms and the fan-palm before alluded to. On the crest of the hill, at a height of some 200 or 300 feet above the sea, are found tall forest trees, some of them of gigantic size and attaining a height of 150 feet and upwards. Amongst them occur the banyan (“chim”), other ficoid trees with the flange-like buttresses, and the “katari,” a species of Calophyllum which supplies the natives with a resin for their torches. In the following description of the interior of the forest in this region I have referred at some length to the larger trees.

The interior of the forest. . . . To obtain a true idea of the forest-growth in these islands, it is necessary to traverse one of the more level districts in the interior, which is removed from the vicinity of the cultivated patches of the natives. Entering the confines of the forest direct from the full glare of the tropical sun, one experiences a peculiar and often oppressive sensation, which may be attributed to the combined influences of the warmth, the humidity, and the effluvia arising from the decaying vegetation, to the impressive silence that reigns, and to the subdued light or dusky atmosphere that there prevails. Meeting overhead at a height of some 150 feet from the ground, the foliage and the smaller branches of the lofty trees form a dense leafy screen roofing over, as it were, a series of lofty corridors in which the palms and the lesser trees flourish. The gloom that there prevails is rarely lightened by the direct rays of the sun, except here and there through the gap left by the downfall of one of the huge trunks that now lies rotting on the ground. Nor is the silence that reigns often broken, except by the cooing of the fruit-pigeons overhead or by the rushing flight of the hornbill startled from its repose. Here the steady blast of the trade is no longer felt and is only perceptible in the movements of the foliage of the tallest trees. Yet there is little in such a scene that would strike the mind of the merely æsthetic lover of nature. Flowers he rarely sees: they are only to be found where the sunlight can reach them in the partially cleared spaces in the midst of the forest, or on the sides of ravines, or along the coasts. On the other hand, however, he cannot fail to be impressed by the luxuriance and magnificence of the vegetation in this conservatory of Nature.

Under such conditions the palms flourish. The Caryota, the “kisu” palm, numerous areca-palms, with the tree-fern, give the character to the lesser vegetation. Huge climbing stems, such as the “droau,” the “aligesi” (Aleurites?), the “nakia” (Uvaria), the “awi-sulu” (Lyonsia) lie in coils on the ground and rising vertically reach the lower branches of the trees some fifty to a hundred feet overhead. The large purple papilionaceous flowers of the “droau” sometimes strew the ground at the bases of the tallest trees. If the forest be situated on a hill-side, the slope is clothed by Selaginellæ which often display in the midst of their dark-green foliage pretty bleached fronds that form a striking contrast to the prevailing hue. Mosses, small ferns, and fungi, such as the massive expansions of Polyporus and the more delicate plates of Hexagona apiaria and others, conceal in some degree the unsightliness of the decaying log. A drapery of lycopods and of trailing and climbing ferns, such as Trichomanes and Lygonia more or less completely invests the lower portions of the trunks of the larger trees. Seventy or eighty feet overhead the wide-spreading fronds of the birds-nest fern (Asplenium nidus) appear half-suspended in mid-air, as they project from their point of attachment to the tree. Lower down the trunk, the handsome aroid Epipremnum may be observed. Epiphytic orchids form no marked feature in this forest-scene, preferring, as they do, those situations where the direct sunlight can reach them, as at the coast and on the sides of ravines. Terrestrial orchids, however, with inconspicuous and sombre-coloured flowers thrive in the gloom and moisture of the forest.

The larger trees, to which I have not yet referred, often attain a height of 150 feet and over. Here the banyan and more than one species of Canarium including the “ka-i” or Solomon Island Almond tree, together with a Ratonia (“nekale”), a Vitex (“fasala”), the “katari” (Calophyllum) before mentioned, and numerous ficoid trees known to the natives as the “uri,” the “ilimo,” and the “nie,” figure amongst the more conspicuous of the forest trees. Many of them possess at the base of the trunk large buttresses or flanges, which, as in the “tobu,” “ilimo,” “nie,” and “maranato” (Sapotacea?), may rise twelve to fifteen feet up the trunk and extend some twenty feet away along the ground. Some of the ficoid trees throw off at a height of from twenty to thirty feet, large flange-like buttresses, which, on reaching the ground, form natural arches. These lofty trees, as I have already remarked, meet together overhead to form a leafy screen, which, whilst it excludes the direct rays of the sun, admits and confines both the moisture and the heat. This conservatory of nature contains within its own precincts the conditions for its preservation. Here the young tree grows up, its safety ensured, until at length it becomes a pillar in the edifice in which it was itself reared. The open character of the wood and the absence of scrub and undergrowth, more especially on level ground, have often been a cause of surprise to me. I have often walked without impediment through the gloomy corridors of such a forest, brushing past the huge trunks of the tallest trees, and winding in and out amongst the palms that number as many years in age as their giant compeers count decades.

On first treading in such a forest, the visitor is much impressed by the imposing appearance and size of the banyans and the buttress-trees. With mingled feelings of awe and pity he will perceive that between these monarchs of the forest there is waged an unequal struggle, in which the huge buttress-tree always succumbs to the rough embraces of its foe. He will observe all the stages in the struggle. Here the buttress-tree may be seen in its prime, but in part embraced at its lower part by the tightly clasping offshoots of the young banyan. Further on, in the midst of the interlacing columns of the banyan, the buttress-tree may be seen partially strangled. Dry rot has attacked its trunk reaching almost to the core, so that a sheath-knife sinks readily up to the handle in its substance; yet, far overhead the wide-spreading branches of this forest potentate are covered with green foliage, and still wave defiantly in the trade. In the prolonged contest the buttress-tree is dying hard, and in fact it is the stout investing trunks of the banyan that alone hold its victim erect. Near by may be another banyan of larger size and presenting the appearance of a maze of columns which may cover an area thirty to forty feet across. Its victim has long since disappeared, and a hollow in the centre of the maze of stems alone marks the former situation of the huge buttress-tree.

What finer or more impressive simile could be employed to illustrate the gradual degeneration and final downfall of a nation under the choking influences of vice, corruption, luxury, and misgovernment? A mighty forest tree is slowly strangled by the caresses of an insidious creeper. With advancing decay its tottering stem is alone supported by the tightening grasp of its foe. Yet its higher branches retain their vitality to the last; and when the end comes, its ashes add fertility to the soil and vigour to the growth of its destroyer.

It is not to be surprised that this battle of the trees should be included in the mythical lore of some of the inhabitants of the Pacific islands. Dr. George Turner, in his recent work entitled “Samoa, a hundred years ago and long before,” gives the following legend of the banyan. . . . . . “A report reached Samoa that the trees of Fiji had fought with the Banian tree, and that it had beaten them all. On this the Tatangia (Acacia laurifolia) and another tree went off from Samoa in two canoes to right the Fijian champion. They reached Fiji, went on shore, and there stood the Banian tree. ‘Where is the tree,’ they inquired, ‘which has conquered all the trees?’ ‘I am the tree,’ said the Banian. Then said the Tatangia, ‘I have come to fight with you.’ ‘Very Good, let us fight,’ replied the Banian. They fought. A branch of the Banian tree fell, but the Tatangia sprung aside and escaped. Another fell—ditto, ditto—the Tatangia. Then the trunk fell. The Tatangia again darted aside and escaped unhurt. On this the Banian tree ‘buried its eyes in the earth’ and owned itself conquered.”

An ascent to the summit of the Faro Island. . . . In making an ascent to the higher districts of this island, which attains an elevation of about 1900 feet above the sea, a little may be learned perhaps of the vertical distribution of the coast flora in this portion of the Solomon Group. The cycad (Cycas circinalis) grows most frequently just within the trees that immediately line the beach and may be often observed at all heights up to 400 feet above the sea, but it is not usually found at greater elevations.[426] The following large trees commonly occur on the hill-slopes up to an elevation of a thousand feet, the “fasala” (Vitex), the “toa” (Elæocarpus), the “opi-opi,” the “ka-i” (Canarium), the “katari” (Calophyllum), and others; whilst the palms such as the fan-palm (Licuala), the Caryota (“eala”), the “kisu” (Pinanga) and the arecas, fill up the intermediate ground, the fan-palm growing in great numbers and often monopolising the slope.[427] The smaller trees, of a height usually of sixty or seventy feet, which are more frequently observed during the lower half of the ascent, are, a species of Cerbera (“anumi”), the “kunuka” (Gnetum), the “palinoromus” (Couthovia), the “poporoko,” and others; whilst on the hill-slopes below the elevation of 500 feet the small conifer Gnetum Gnemon (“meriwa”) may be commonly seen. In three different localities, at elevations of between 1,000 and 1,100 feet above the sea, I came upon brakes of fine bamboos (Schizostachyum?) 35 to 40 feet in length which are employed as fishing-poles by the natives. This bamboo, both in Treasury and Faro Islands, does not appear to occur below this height; whilst in the Shortland Islands, although found at a lower elevation, it selects the higher regions of the island.

[426] At Treasury Island I found a solitary cycad at a height of a thousand feet above the sea. As it was in the vicinity of a plantation of sago palms, it is probable that it had been planted by the natives who employ the fruits for medicinal purposes.

[427] This fan-palm, the “firo” of the natives, was in 1884 only represented in Treasury by a single individual which had been brought a few years before from Bougainville, where the leaves are employed in making a conical hat that is commonly worn.

Above a thousand feet, many of the trees and palms so frequent below become less common or disappear. The fan-palm (Licuala) which grows in such numbers in the lower levels did not come under my notice above this elevation. On account of the absence of large trees near the summit, the lesser vegetation receives more of the sun’s rays; and thus at 1,600 feet above the sea the alpinias, such as the “vitoko” and the “kokuru” re-appear, plants which usually abound in the lower levels in all open situations, as on the banks of streams. For the same reason, we find near the summit of the island at elevations of 1,600 to 1,700 feet the tall composite shrub, Wedelia biflora, which is one of the commonest of the plants that grow at the margin of the beach. On account of this absence of large trees, and the consequent increased exposure to the sun’s rays, the smaller trees with conspicuous flowers find a congenial situation at this elevation: here are found the species of Dolicholobium (“lowasi”), which is common on the banks of the streams in the lower levels, the Fagræa Berteriana (“bubulata”) which grows also at the coast, a wild nut-meg tree (Myristica), a species of Harpullia (“wawaupoko”), the “pakuri” (Eugenia), the “baimoloi” and others. In these higher regions tree-ferns grow to a height of thirty feet; and here the areca-palms, “momo” and “niga-torulo,” are also found. Here flourishes the Gleichenia, a fern which does not usually grow at elevations under 700 feet above the sea, and which is represented by two common species: it is the “sinimi” of the natives who, as I have already remarked, work the fine strips of its vascular tissue into armlets which they commonly wear. Near the summit and all down the slopes is found a species of Begonia, a genus, as I am informed by Baron von Mueller, not before recorded from islands east of New Guinea.[428] A dense growth of the trailing stems of a Freycinetia and of ferns clothes the rocky sides of the highest peak, which is almost bare of trees. Here however I found a new genus of the Pandanaceæ, which, like some other pandanus trees, is known to the natives as “sararang.” It grows to a height of fifty feet, and was only observed by me on the highest peak of the island and for two or three hundred feet below. It has a very conspicuous white “branching female spadix,” three to four feet in length; and I learn from Professor Oliver that the same or a near ally of it, though not in a condition to describe, was collected by Signor Beccari in Jobi Island off the north-west coast of New Guinea.

[428] A species of Ophiorrhiza is in Treasury Island usually associated with this Begonia and is found at all elevations.

The coast vegetation of the larger islands. . . . It is in the coasts of such an island as Treasury or Faro Island, where the strictly littoral and more inland plants become intermingled, that the Solomon Island vegetation in some degree redeems its character. Here the prevailing sombreness and inconspicuous inflorescences give place to bright hues and to a variety of flowers. Here are seen the handsome white flowers of a rubiaceous tree, a species of Bikkia; the yellowish flowers and bright red fruits of Harpullia capanioides (“koloa”); the crimson flowers of an Erythrina (perhaps indica); the yellow flowers of Cæsalpinia Nuga; the large pods of Pongamia glabra; and the fruits of a wild nutmeg (Myristica, sp.). Hernandia peltata and Clerodendron inerme may also be here found. The conspicuous flowers of Hibiscus tiliaceus, Thespesia populnea, and of other littoral trees such as Cerbera Odollam and Guettarda speciosa, add their brightness to the scene. Amongst the foliage of the trees twine a species of Ipomœa with handsome white flowers, and here are seen the wax-like flowers of more than one species of asclepiad (Hoya, sp). Orchids, some of striking beauty, hang from the trunks of the trees and form a conspicuous feature in the scene. Among them occur species of Dendrobium, Coelogyne, Cleisostoma, etc.

The littoral vegetation, as exhibited in a coral islet. . . . I will take the case of one of the many wooded islets that have been formed on the coral reefs by the action of the waves. On the weather side of such an islet, which may be termed its growing edge, the vegetation is scanty, and there are but few trees. A binding weed and more than one species of Ipomœa loosely cover a surface composed almost entirely of calcareous sand, broken shells, coral debris, and pumice pebbles; and it is on such an unproductive soil that two or more species of Pandanus and Casuarina angustifolia flourish. Here at the margin of the beach may be seen in profusion the tall composite shrub, Wedelia biflora, and another common shrub, Scævola Kœnigii. Two climbing peas prefer the sandy soil in this situation, one with yellow flowers, Vigna lutea, and the other with pink flowers, Canavalia turgida; whilst a dense growth of Flagellaria indica often conceals from view any rocky slope overlooking the beach. Just within the line of vegetation immediately bordering the beach, the following trees commonly occur, Ochrosia parviflora (“pokosola”), Heritiera littoralis (“pipilusu”), Terminalia catappa (“saori”), Cycas circinalis, and one or more species of Pandanus. Here also a species of Crinum (the “papau” of the natives) and the Tacca pinnatifida (“mamago”) may be usually found. (I hoped to have referred to the ferns of such a coral islet; but my endeavours to obtain any information of my collection have been unavailing).

On the lee side of such an islet, which is the oldest portion of its surface, the vegetation is much denser and of a different character. Here, the trees form a thick belt, their branches overhanging the rising tide. Those of most frequent occurrence are, Barringtonia speciosa, Calophyllum inophyllum, Hibiscus tiliaceus, Thespesia populnea, Guettarda speciosa, Morinda citrifolia, Cerbera Odollam, Pongamia glabra, Tournefortia argentea, and others. The trunks of the larger trees often lean over the beach or lie partly procumbent on the sand. Amongst the foliage of these coast trees, many of which have large conspicuous flowers, climbing asclepiads of the genus Hoya with their equally conspicuous flowers may be frequently observed. Orchids, often of considerable beauty, hang from the reclining trunks of the trees. Here, as in the case of the coasts of the large islands, we perceive how pleasant is the contrast which the littoral vegetation presents when compared with the gloomy and apparently flowerless forests, where the tallest trees possess but an inconspicuous greenish inflorescence.

In the interior of such a coral islet, huge banyans and other trees having wide-spreading buttresses are to be found. Many of them attain a height of 150 feet and upwards, and afford a home to numbers of fruit-eating pigeons which largely subsist on their fruits, and through whose agency the interiors of coral islets are stocked with these large trees. Conspicuous amongst the trees is a species of Canarium (the “ka-i” of the natives), the disgorged nuts of which frequently strew the ground beneath; a banyan (Ficus) with large oblong fruits and another species with small spherical fruits; other ficoid trees with large buttresses, such as the “uri”; a species of Eugenia, probably a variety of Eugenia jambos; together with several other trees.

This description of the vegetation of a coral islet brings me to refer to the manner in which such an islet, which is usually of very recent origin, has become stocked with its plants: and in so doing I shall be treating of a very important matter, the oceanic dispersal of plants. Fortunately for me, my notes and collections relating to this subject had an increased value at the time of my arrival in England, and in this respect I have been able to accomplish one of the principal aims of a young traveller, that of supplying trustworthy materials to those engaged in the particular line of research to which his notes and collections relate.[429]

[429] Mr. Botting Hemsley was on the point of completing his report on the oceanic dispersal of plants in connection with the Botany of the “Challenger” Expedition. Such of my collections, as referred to his work, were placed at his disposal by Sir Joseph Hooker; and my notes were incorporated in volume I. of the Botany of the “Challenger” (Part III. p. 309), to which I must refer my readers who are more specially interested in this subject.

The picturesquely wooded islets of these seas have been stocked through two principal agencies. Winds and currents drift to their shores the fruits and seeds of the littoral trees which ultimately form the margin of the vegetation; whilst the fruit pigeons disgorge the seeds or fruits of those often colossal trees which occupy the interior.

I will first refer to the former of these agencies. Lines of vegetable drift, intermingled with floating pumice, are frequently observable whilst cruising among the islands of the Solomon Group. The floating fruits commonly found belong to the most familiar littoral trees of this region, those of Barringtonia speciosa and Calophyllum inophyllum being especially frequent; and on more than one occasion the solitary fruits of the former were noticed at sea by Lieutenant Oldham and myself at distances of from 130 to 150 miles to the southward of the group, being probably derived from one of the islands of the New Hebrides to the eastward. Other fruits or seeds occurring frequently in the drift are those of Nipa fruticans and of two or more species of Pandanus; numerous beans (species of Mucuna, Canavalia, Dioclea), the long germinated seeds of the mangrove (Rhizophora), an occasional cocoa-nut, the cones of Casuarina equisetifolia, Terminalia catappa, Lumnitzera coccinea, Guettarda speciosa, Ochrosia parviflora, Heritiera littoralis and others.[430]

[430] Other fruits found floating were a second species of Calophyllum, a species of Gomphandra, Harpullia sp., and some Scitamineæ.

The foregoing seeds and seed-vessels with many others may be observed washed up by the waves on the surface of the bare sandy islets or sand-keys, which exhibit the first stage in the growth of those picturesquely wooded coral islets that are ultimately formed on the reefs. On such a sand-key, not more than some 25 or 30 yards across, I have counted as many as 30 different kinds of seeds and fruits, all collected together in the centre, which was only washed over at spring-tides. One of the first trees to establish itself is the mangrove (Rhizophora), which by its reclaiming agency adds to the area of the islet and enables other trees, such as Lumnitzera coccinea, to take up their abode. Pari passu with the seaward extension of the reef, the islet increases in size; and in time the winds and currents bring other fruits and seeds which germinate and form ultimately the belt of littoral trees bordering the beach. In this manner Barringtonia speciosa, Calophyllum inophyllum, Thespesia populnea, Hibiscus tiliaceus, Cerbera odollam, Ochrosia parviflora, Heritiera littoralis, Terminalia catappa, different species of Pandanus, Casuarina equisetifolia, and Cycas circinalis with many others referred to on a previous page, become established. It is worthy of note that the fruits of the great majority of trees which form the margin of the vegetation, whether on the lee or weather side of such an islet, float in salt water.[431] The small cones of the Casuarina however, require a certain amount of drying before they can be transported by the waves. The green fruits of the Cycas usually sink in salt water; but I found that one out of ten specimens floated, an exceptional circumstance which sufficiently accounts for the occurrence of Cycas circinalis on these coral islets.

[431] The results of some experiments I made are given on [page 305].

Whilst through the agency of the winds and currents the waves have stocked the islet with its marginal vegetation, the fruit pigeons have been unconsciously stocking its interior with huge trees, that have sprung from the fruits and seeds they have transported in their crops from the neighbouring coasts and islets. Perched up in the branches of the trees, these birds disgorge the seeds they have brought from other localities; and the rejected seeds and seed-vessels lie strewn on the ground beneath. The soft and often fleshy fruits, on which the fruit pigeons subsist, belong to numerous species of trees. Some of them are as large even as a hen’s egg, as in the case of those of the species of Canarium (“ka-i”) which have a pulpy exterior that is alone digested and retained by the pigeon. The fruits of the banyans and other ficoid trees, which with the Canarium are amongst the most conspicuous trees in the interior of the coral islets, are apparently preferred by the fruit pigeons, since they occur commonly in their crops. A species of Eugenia common in the interior of these islets possesses fruits found in the crops of these birds. Amongst other fruits and seeds on which these pigeons subsist, and which they must transport from one locality to another, are those of a species of Elæocarpus (“toa”), a species of laurel (Litsea), a nutmeg (Myristica), an Achras,[432] one or more species of Areca, and probably a species of Kentia. There is, however, another bird common on these coral islets, the ground pigeon Geophilus nicobaricus, known commonly as the Nicobar pigeon, which transports seeds in its gizzard cavity that on account of their hardness are not fed upon by the ordinary fruit pigeon (Carpophaga). The peculiar structure of the gizzard, which is described on [page 323], enables the Nicobar pigeon to crack seeds that can only be broken by a sharp blow with a stone. I have found inside this organ, the hard red seeds of a leguminous plant, very probably Adenanthera pavonina, of which one seed is occasionally found to be cracked. We may therefore consider that many small hard seeds and seed-vessels, which would be refused by the common fruit pigeon of these islands, are transported from one locality to another in the gizzard cavity of the Nicobar pigeon.

[432] I am indebted to Mr. Charles Moore of Sydney N.S.W., for the identification of this fruit. (Vide also “Journal and Proceedings of the Royal Society, N.S.W.,” XVII., p. 226.)

From the foregoing remarks it may be inferred that the pigeons in these islands play a very important part in the dispersal of plants, to which, as Mr. Botting Hemsley remarks in his report ([page 313]), they have perhaps contributed more than any other animal. In the Solomon Islands the fruit pigeons, as dusk approaches, frequent the islets on the coral reefs in great numbers, and from their unwillingness to leave their roosts in the trees they fall an easy prey to the sportsman. In one afternoon, on one of the islets of Choiseul Bay, 57 birds fell to the guns of Lieutenant Heming and Lieutenant Leeper; and it is to these two officers that I am indebted for my opportunities of collecting the fruits taken from the crops of these pigeons.

In drawing my botanical remarks to a close, it may be fitting to recall the more lasting impressions which I have received of the vegetation of these islands; and I may do so in a very few words. The characteristic features of the vegetation are to be found in the number and variety of the areca palms; in the abundance of the alpinias, heliconias, and other scitamineous plants; in the imposing size and form of the banyans and the buttress trees; and in the profusion of the ferns. I have not previously dwelt upon the important part which the ferns take in the vegetation of these islands, because I had hoped to have heard something of my collection which I presented to the British Museum eighteen months ago; but, to my great chagrin, I have been unable, after repeated application, to learn anything concerning it. I may here state that ferns abound everywhere; in moist and dry situations; in sheltered and exposed districts; now decking the tree-trunks with their draperies, or concealing the unsightliness of the decaying log; here covering the bare slopes of some lofty hill-top, or clothing the surface of some treeless tract. The tree-fern and the wide-spreading Angiopteris are to be found on the banks of streams or in some inland dell. The former avoids the coast, and occurs at all elevations up to 2000 feet and over: it flourishes at the heads of valleys.

LIST OF PLANTS COLLECTED IN THE ISLANDS OF BOUGAINVILLE STRAITS, SOLOMON GROUP, DURING 1884.[433]

[433] I am mainly indebted to the kindness of Professor Oliver for the list of the plants collected by me in the Solomon Islands, most of which were sent to Kew. The ferns are in the British Museum, but I can learn nothing of them. Fortunately, the fungi were not included; and for a list of them I am indebted to Mr. Baker. Most of the orchids, and some of the asclepiads, were given by me to Baron von Mueller, who intends to examine them in connection with subsequent collections. I take this opportunity of expressing my sense of the great kindness he showed me with reference to my plant collections. To Signor Beccari I am also indebted. Owing to my inexperience in botanical collecting, the specimens were often inadequate for descriptive and specific determination; but my deficiencies will appear more excusable when I state that I devoted my attention more particularly to the trees. Professor Oliver, however, informs me that, in spite of its defects, my collection gives an excellent conception of the flora of the islands visited.

ANONACEÆ.

Uvaria, sp. . . vulgo “Nakia.” A stout climber.

GUTTIFERÆ.

Ochrocarpus ovalifolius, T. And v. O (Calysaccion) tinctorium, Seem.? vulgo “Kokoilo.” A littoral tree about thirty feet high.

Calophyllum Inophyllum, L., vulgo “Bogoau.”

Calophyllum sp. . . vulgo “Katari.” Two tall trees apparently distinguished by the size of the fruits. (Flowers not obtained.) A dark resin oozes from the bark, which the natives burn in torches.

MALVACEÆ.

Hibiscus tiliaceus, L.: vulgo “Dakatako.”

Thespesia populnea, Corr.: vulgo “Kai-kaia.”

STERCULIACEÆ.

Kleinhovia Hospita, L.: vulgo “Lafai.”

Heritiera an H. littoralis, var. angustifolia? vulgo “Pipilusu.”

TILIACEÆ.

Triumfetta procumbens, Forst.

Elæocarpus sp. . . vulgo “Toa.” A tree about seventy feet high, with conspicuous blue fruits, eaten by fruit-pigeons.

OXALIDACEÆ

Oxalis corniculata, L.

SIMARUBEÆ

Soulamca amara, Lam.

RUTACEÆ.

Evodia hortensis, Forst.: vulgo “Luk-a-luk.”

Rutacea (§ Toddaliæ?). Detached leaves and flowers picked up from the ground at the foot of a tall forest tree. Flowers “4-meri; petala imbricata libera; stamina 4 libera, pet. alterna, ovarium liberum integrum, 4-loc?”

BURSERACEÆ.

Canarium sp. . . No flowers obtained. A tall forest tree, a hundred feet and upwards in height. Vulgo “Kai.” Known as the Solomon Island almond tree. The kernels afford a common article of food in August and September.

Canarium? vulgo “Nie.” A tree with buttresses, a hundred feet high.

Canarium? vulgo “Nie.” A tall forest tree, with buttresses, 100 to 150 feet high.

OLACINEÆ.

Gomphandra sp. . . vulgo “Ninilo,” or “Ningilo.” A tree thirty to forty feet high. Fruit eaten by wild pigs.

Lasianthera sp. . . nov? vulgo “Porutolo.” A tree sixty to seventy feet high.

Olacinea (dub): vulgo “Poporoko.” A tree sixty feet high, having a light reddish wood, and a dark red sap.

CELASTRINEÆ.

Salacia sp. . . nov.

RHAMNACEÆ.

Colubrina asiatica, Bngn.

AMPELIDEÆ.

Leea sambucina, L. (A Gr. U.S. Expl. Expn.)

SAPINDACEÆ.

Schmidelia aff. S. obovatæ, A Gr. A littoral tree, thirty feet high.

Harpullia cupanioides, Roxb.: vulgo “Koloa.” Littoral.

Sapindacea an aff. Harpulliæ? vulgo “Wawau-poko.” Growing 1400 feet above the sea.

Ratonia sp. . . vulgo “Nekale.” A forest tree, a hundred feet high and over, with inconspicuous buttresses.

Ratonia sp. . . vulgo “Nekale.” A forest tree, a hundred feet high and over, with buttresses.

ANACARDIACEÆ.

Mangifera indica. L.? vulgo “Faise.” Mango tree, growing in plantations. Fruit ripens in August. Height, thirty feet.

LEGUMINOSÆ.

Crotalaria quinquefolia, L.: vulgo “Kokila.”

Desmodium umbellatum, D.C., vulgo “Meki,” forma stenocarpa.

Desmodium ormocarpoides, D.C.?

Desmodium polycarpum D.C.

Erythrina: flowers only. E. monosperma perhaps, or E. indica.

Mucuna gigantea, D.C.? vulgo “Faso-gasuga.”

Mucuna sp. . . vulgo “Wassa-wassawa.”

Mucuna sp. . .

Papilionacea (dub); vulgo “Droau.” A stout climber on forest trees, with large purple flowers.

Canavalia turgida, Grah.

Vigna lutea, A. Gray.

Pongamia glabra, Vent.? vulgo “Ansapo.”

Sophora tomentosa, L.

Cæsalpinia Nuga, Ait.

Adenanthera Pavonina, L. (probably). Seeds only obtained.

Leucæna sp. . .?? vulgo “Gehala.” A tree thirty to forty feet high.

CHRYSOBALANEÆ.

Parinarium laurinum, A. Gr.: vulgo “Tita.” A tree about sixty feet high. From the fruit is obtained a resin used by the natives for caulking the seams of their canoes.

ROSACEÆ.

Rubus tilaceus, Sm.

COMPRETACEÆ.

Terminalia Catappa, L.: vulgo “Saori.” Seeds eaten by the natives.

Lumnitzera coccinea, W. and Arn.

MYRTACEÆ.

Eugenia sp. . . vulgo “Pakuri.” A tree thirty feet high, growing 1600 feet above the sea.

Eugenia clusiæfolia, A. Gray (allied to E. Jambolana).

Eugenia sp. . . vulgo “Tsugi.” A littoral tree.

Eugenia, aff. E. Richii, A. Gr.: vulgo “Malapo.” A tree eighty feet high, with buttresses, growing on coral islets.

Barringtonia speciosa, F.

Barringtonia cf. B. edulis, Seem. and B. excelsa, Huds. (New Hebrides): vulgo “Borolong.” A tree thirty to thirty-five feet in height, growing in plantations. Flowers gathered into very conspicuous pendent yellow spikes, two and a half feet long. Kernel of fruit edible.

Barringtonia aff. B. racemosæ, Bl.: vulgo “Misioko.” A tree forty feet high, growing near coast.

Barringtonia?? vulgo “Sioko.” A tree fifteen to twenty feet high, growing in plantations. Fruit edible.

MELASTOMACEÆ.

Medinilla sp. . . A climbing plant around the trunks of trees.

LYTHRACEÆ.

Pemphis acidula, Forst.

CUCURBITACEÆ.

Cucumis Melo, L., forma?

ARALIACEÆ.

Panax fruticosum, L.

Plerandra, near Pickeringii, A. Gray: vulgo “Fo.”

Araliacea (dub?): vulgo “Bubolio.” A littoral tree, fifteen feet high.

RUBIACEÆ.

Hedyotis Auricularia, L.

Ophiorrhiza aff. O. cantonensis, Hance.

Ophiorrhiza sp. . .

Dolicholobium aff. D. longissimo an D. longissimum, Seem. forma macranthus: vulgo “Lowasi.” A tree fifty feet high and under, common along the sides of streams.

Geophila reniformis, C. and S.

Morinda citrifolia, L.: vulgo “Urati.”

Guettarda speciosa, L.: vulgo “Orgoi.”

Myrmecodia salomonensis, Becc. A new species separated from M. samoensis, Becc. Noticed commonly on tall mangrove trees bordering the sides of streams in the lower part of their courses. The swollen tuberous stem measures as much as one and a half feet in length, and is usually occupied by many ants.

Hydnophytum longistylum, Becc. Found on coast trees. Those I examined contained no ants, but, instead, a few cockroaches. On the outside of one of them I found a crab.

Hydnophytum Guppyanum, Becc. A new species. Noticed commonly on the tall mangrove trees bordering the sides of streams in the lower part of their courses. The swollen tuberous portion of the stem has a characteristic scaphoid form; those I examined being nearly full of dirty rain-water, and almost free from ants; a few cockroaches occurred in all; some of them are one and a half feet in length.

(Hydnophytum inerme, a specimen I obtained from Ugi Island, at the east end of the group in 1882, and identified by Mr. C. Moore of Sydney.)

Psychotria sp. . . vulgo “Popotu.”

Psychotria, aff. P. Forsterinæ, A. Gr.

Bikkia sp. . . A littoral tree, twenty feet high, with large handsome white flowers.

COMPOSITÆ.

Vernonia cinerea, Less

Adenostemma viscosum, Forst.

Blumea aft. B. glandulosæ, D.C.

Eclipta alba, Hassk.

Bidens pilosa, L.

Wedelia biflora, D.C. A very common littoral plant, but in one instance I found it 1600 feet above the sea.

GOODENIACEÆ.

Scævola Kœnigii, Vahl. vulgo “Nano.” A very common littoral shrub.

SAPOTACEÆ.

Sapotacea? Seeds only obtained.

Sapotacea (dub): vulgo “Maranato.” A forest tree, a hundred feet high with large plank-like buttresses.

APOCYNEÆ.

Ochrosia parviflora, Hensl: vulgo “Pokosola.”

Ochrosia aff. 0. (Lactaria) calycarpæ (Miq.). Tree 30 feet high.

Ochrosia sp. . . vulgo “Bararai.” A tree 30 to 40 feet high.

Cerbera Odollam, Gærtn: vulgo “Lukapau.”

Cerbera sp. . . vulgo “Anoumi.” A tree about 50 feet high, growing away from the coast.

Lyonsia??: vulgo “Awi-sulu.” A stout climber: its bark supplies the fibres used for making fishing-lines.

ASCLEPIADEÆ.

Hoya australis, Br.? (H. bicarinata, A. Gr.) forma: vulgo “Alulu.”

Hoya sp. . . (narrow-leaved species).

Hoya Guppyi, Oliv. sp. nov. Ramulis ultimis parce hirtellis dense glabratis, foliis petiolatis coriaceis late ellipticis breviter acuminatis cuspidatisve base late rotundatis subcordatisve supra glabris, subtus præcipue versus basin plus minus hirtellis, umbellis pedunculatis, pedunculis pedicellisque glabris, calyce parvo tubo corollæ 2-4-plo breviore 5-partito lobis ovatis obtusis ciliolatis, corolla rotata lobis patentibus ovatis v. late ovato-lanceolatis acutatis medio depressis intus hirtellis extus glabris sinubus reflexis, coronæ foliolis cartilagineo-incrassatis disco ovato lanceolatis concavis obtusis basi angustatis dorso profunde excavatis marginibus lateralibus utrinque carinatis, folliculis subteretibus parce hirtellis.

Folia 31⁄2-41⁄2 poll. longis, 21⁄4-21⁄2 poll. latis; petiolo hirtello 1⁄2-3⁄4 poll. longo; umbella 10-14 flora; pedunculo 2 poll. longo, pedicellis 11⁄2 poll. longis. Corolla 11⁄2 poll. diam. rubro-purpurea. Follicula 8-9 poll, longa.

Faro Island: Bougainville Straits: “A climber on coast trees.”

LOGANIACEÆ.

Couthovia, nearly allied to C. Seemanni A.Gr., if not a variety with inflorescence throughout tawny-puberulous. Vulgo “Palinoromus.” A forest tree 70 feet high.

Fagræa Berteriana A.Gr.? vulgo “Bubulata.”

Fagræa morindæfolia, Bl. forma. Vulgo “Kirofe.”

Fagræa sp. . . vulgo “Mamuli.” A tree twenty-five feet high.

BORAGINEÆ.

Tournefortia argentea, L. f. vulgo “Diave.”

Cordia subcordata, Lam.

Cordia? (Corollas picked off ground.)

CONVOLVULACEÆ.

Ipomœa denticulata, Chy.

Ipomœa (Calonyction) grandiflora, Lam?

Ipomœa pes-capræ, Roth.

Ipomœa sp. . .

SOLANACEÆ.

Solanum repandum, F? vulgo “Kirkami.” -
Solanum repandum, F? vulgo “Kobureki.”

Natives distinguish these two plants, which grow in their plantations, as shrubs 4 to 6 feet high. Fruits edible when cooked.

Solanum vitiense, Seem. vulgo “Koriele.”

Physalis angulata, L.

SCROPHULARINEÆ.

Vandellia Crustacea, Bth.

CYRTANDREÆ.

Cyrtandra v. gen. nov. aff.

ACANTHACEÆ.

Adenosma cærulea, R.Br.?

Bæa Commersoni, R.Br. fide F. von. Mueller.

Hemigraphis reptans, T. And.

Hemigraphis reptans, forma.

Ruellia sp. R. arvensis. S. Moore var? v. sp. nov. aff. Growing beside a stream, 11⁄2 feet high, with light-yellow flowers.

Acanthus ebracteatus, V.

Eranthemum variabile, Br. var.? Very common in the waste ground of plantations and by the sides of paths: 11⁄2 to 2 feet in height.

VERBENACEÆ.

Premna obtusifolia, R.Br. an P. taitensis Schr? vulgo “Demoko.” A littoral tree 12 to 15 feet in height.

Vitex an V. acuminata, Br.? vulgo “Fasala.” A large forest tree, a hundred feet and over in height, with small buttresses, supplying wood for paddles and canoes.

Clerodendron inerme, Br.

Verbenacea dubia? vulgo “Au-au.” A tree fifty to sixty feet high.

LABIATÆ.

Moschosma polystachyum: Bth: vulgo “Pipituan.”

Ocymum sanctum, L: vulgo “Kiramma.”

Plectranthus v. Coleus? vulgo “Momauri.” Leaves, when crushed, give a reddish-brown stain, and used for staining the skin. Height 11⁄2 feet.

Plectranthus parviflorus, W.

AMARANTACEÆ.

Amarantus melancholicus, L.

Cyathula prostrata, Bl.

PIPERACEÆ.

Piper Betel var. (Chavica Siriboa, Miq.) vulgo “Kolu.”

MYRISTICACEÆ.

Myristica sp. . . vulgo “Ito-ito.” Coast tree fifteen feet in height.

Myristica sp. . . vulgo “Baimoloi.” A tree fifty feet high, growing 1600 feet above the sea.

LAURACEÆ.

Litsea sp. . . vulgo “Pitoponkano.” A tall forest tree.

HERNANDIACEÆ.

Hernandia peltata, Meiss: vulgo “Koli.”

EUPHORBIACEÆ.

Euphorbia pilulifera, L.

Euphorbia Atoto, Forst.

Phyllanthus (§ Emblica) sp., allied to P. bæobotryoides, Wall? vulgo “Mefuan.” A tree 15 to 20 feet high.

Mallotus tiliæfolius, M. Arg. M. acuminatus, Juss? Tree twenty feet high, growing at the coast on the border of swampy ground.

Macaranga sp. . . vulgo “Balako.” A tree forty to fifty feet high, with ringed trunk.

Aleurites? vulgo “Aligesi.” A stout climber on forest trees. Kernels of fruit edible.

Sapium indicum, Willd? vulgo “Aligesi.” A tree seventy feet high, growing on the verge of a mangrove swamp.

Excæcaria Agallocha, L.

Codiæum sp. . . (♂)

Codiæum variegatum. A. Juss: vulgo “Tiatakush.”

URTICEÆ.

Trema (Sponia) sp. . .: vulgo “Kio.” A tree seventy to eighty feet high.

Ficus nr F. theophrastoides. Seem? vulgo “Tutubolo.” Growing in plantations. Probably 10 to 12 feet high.

Ficus sp. . . vulgo “Uri.” A tree eighty to ninety feet high, with buttress roots. Growing on coral islets.

Ficus sp. . . vulgo “Sii.” A banyan growing at or near the coast and on coral islets. Multiple trunks, some cylindrical and erect, others plank-like and arching. Height eighty or ninety feet and over.

Ficus sp. . . vulgo “Chim.” A banyan often growing on the crest of inland ridges. The multiple trunks are all cylindrical and erect, and individually smaller than in the case of the Sii: they are also more closely arranged. Height a hundred and fifty feet and over.

Ficus? vulgo “Ilimo.” A tall forest tree over a hundred feet in height, with magnificent buttresses.

Artocarpus incisa, L. There appears to be but one variety of the Bread-fruit tree in the islands of Bougainville Straits. The fruit is stalked, seedless, and rough externally, the leaves pinnatisect, with smooth surfaces. Fruit ripens in August. Vulgo “Balia.”

Artocarpus sp. . . vulgo “Tafati.” Perhaps a variety of the Jack-fruit Tree (A. integrifolia). Sixty feet high. Fruit larger than the common bread-fruit, but more irregular in shape: seeded: edible.

Fleurya interrupta, Gaud. (F. spicata, var.)

Elatostemma integrifolium, Wedd.?

Elatostemma? vulgo “Obu-obu.”

Procris integrifolia, Don??

Pellionia sp. . .

Leucosyke an L. corymbulosa? Coast tree fifteen feet high.

Pipturus velutinus, Wedd? v. P. argenteus? vulgo “Dilipoa.” A tree thirty to fifty feet high; trunk partially ringed; aerial roots.

CONIFERÆ.

Gnetum Gnemon, L. vulgo “Mariwa.”

Gnetum sp. . .: vulgo “Kunuka.” A tree sixty feet high, prominently ringed. Kernels of fruits eaten by the natives.

CASUARINEÆ.

Casuarina angustifolia F.

ORCHIDEÆ.

Dendrobium hispidum, Rich. (fide F. v. Mueller).

Dendrobium sp. . . near D. dactylodes, R. fil?

Cœlogyne sp. . .

Cleisostoma sp. . .

SCITAMINEÆ.

Alpinia sp. . . vulgo “Karu.”

Alpinia sp. . . vulgo “Vitoko.”

Alpinia sp. . . vulgo “Konkoku.”

Costus or Alpinia sp. . . vulgo “Makisa.”

Alpinia Boia, Seem? v. sp. aff. vulgo “Pai-yang-pipiula.”

Riedelia curviflora, Oliv? vulgo “Kokuru.”

Canna indica, L.? vulgo “Sati.”

Marantacea aff. Phrynio? vulgo “Sinoili.” Flowers in two collateral pairs in each spathe with linear bracts between the pairs. Ovary shortly stipitate, ovule erect. Fruit 3-locular, cells 1-seeded, seeds with crustaceous muricate testa.

Heliconia? vulgo “Kiari.” Clinogyne grandis Bth and Hook? (near C. dichotoma and affs) vulgo “Nini.”

Scitaminea (dub): vulgo “Temuli.” A plant 1 to 11⁄4 feet high growing in the waste ground of plantations. The roots have medicinal properties, according to the accounts of the natives, and they have a yellow juice which is used for staining.

Scitaminea (dubia): vulgo “Nakia:” a wild ginger.

AMARYLLIDEÆ.

Crinum sp. . . vulgo “Papau.” Grows near the beach. Height four feet.

Curculigo sp. . . vulgo “Bulami.” Growing 2 to 21⁄2 feet high on the banks of streams.

LILIACEÆ.

Cordyline sp. . . vulgo “Dendiki.” Tree twenty feet high; growing near the coast.

COMMELYNACEÆ.

Commelyna nudiflora, L.

DIOSCOREÆ.

Dioscorea sativa, L.? vulgo “Alapa.”

JUNCACEÆ.

Flagellaria indica, L. var.

TACCACEÆ.

Tacca pinnatifida, Forst.: vulgo “Mamago.” The natives do not appear to make use of the arrowroot-like starch obtainable from the tubers.

PANDANACEÆ.

Pandanacea: genus novum,[434](♀ flowers only and leaf collected). The only locality where I found it was the summit of Faro Island, where it grows to a height of fifty feet, and has a long white female branching spadix, three to four feet in length. The same, or a near ally, was obtained by Signor Beccari in Jobi Island, off New Guinea. (Vide [page 289].)

[434] I learn from Professor Oliver that Count Solins confirms the generic distinctness.

The natives distinguish several species of Pandanus trees, of which I was only able to obtain the fruit. The “darashi” “sararang,” and “pota,” grow at the coast, and have a height of from thirty to forty feet. The “darashi” has narrow leaves, and, if the ground is not rocky, aerial roots are often absent: the fruit is smaller than that of the two other littoral pandanus trees. The “sararang” has broad leaves, and always aerial roots: the fruit is often more than a foot in diameter. The “pota” has broad leaves, with contracted acuminate apices, two inches long: the fruit is about a foot in diameter: aerial roots are always present, and rise often fifteen feet from the ground. The segments of these pandanus trees all contain edible kernels. The broad leaves of the “pota” are employed in making mats. . . . There is another pandanus tree, the “samala” of the natives, which often grows away from the coast, as on the banks of streams: it has an erect, stout trunk, thirty-five to forty feet high, without aerial roots, and does not branch.

Freycinetia sp. . .

Freycinetia sp. . .

Nipa fruticans.

PALMACEÆ.

Cyrtostachys sp. . . vulgo “Sensisi.” Growing up to fifty feet high on the banks of streams.

Palmacea dub. (cf. Drymophloeus): vulgo “Kisu.” Growing seventy to eighty feet high. The tough sheathing at the bases of the branches is employed for making the native dishes.

Pinanga sp. . . vulgo “Kisu”: conf. the “Kisu” above. Growing up to seventy or eighty feet high.

Caryota sp. . . vulgo “Eala.” Growing up to fifty feet high.

Licuala sp. . . vulgo “Firo.” Grows up to thirty-five or forty feet in height. More common on volcanic soils. Absent from Treasury Island, with the exception of one imported tree. Said to be very numerous in the large adjacent island of Bougainville, the leaves being there used in making conical hats.

Palmacea dub.: vulgo “Poamau.” Grows up to seventy or eighty feet high. Its fruit, which is eaten by the women, is said to have a stimulant effect like the betel-nut. Its wood supplies the material for spears.

Areca sp. . . vulgo “Momo.” Grows up to thirty-five or forty feet high. Small fruits (1⁄2 inch) sessile on a branching stalk.

Areca sp. . . vulgo “Niga-torulo,” or “Torulo.” Grows up to thirty-five or forty feet high. Fruits larger (1-11⁄2 inch) sessile, and gathered thickly together on an undivided stalk.

Areca sp. . . vulgo “Niga-solu.” Grows up to fifty feet high. Fruits (1-11⁄2 inch) sessile, gathered thickly together on an undivided stalk.

[Note.—The three kinds of areca palms just mentioned are very common on the low ground at the foot of hills. They all have a general resemblance, and their fruits are often chewed in lieu of “betel-nuts.” They are distinguished from each other readily by the size and mode of attachment of the fruits, and by the number of ribs on the lateral pinnæ.]

Areca sp. . . vulgo “Poamau.” Growing up to eighty feet high.

Areca sp. . . vulgo “Au-Au.” Grows up to twelve feet high. Aerial roots rising from stem 11⁄2 feet above the ground.

Areca sp. . . vulgo “Olega.” The betel-nut palm of the Solomon Islands. Planted by the natives in the vicinity of their villages. Height up to thirty feet.

Sagus sp. . . vulgo “Bia,” “Nami.” Height up to sixty feet. Dry situations.

AROIDEÆ.

Schizmatoglottis sp. . . vulgo “Kuraka.” Growing wild along the banks of streams. The natives make a savoury vegetable broth of the leaves and unopened spathes.

Epipremnum cf. E. mirabile, Sch. Found on trees.

Scindapsus sp. . . vulgo “Kurricolo.” Grows on sandy soil near the coast. Pothos?

CYPERACEÆ.

Cyperus (Mariscus phleoides, Nees). Height two to two and a half feet.

Cyperus canescens, Vahl. Height two feet.

Cyperus (Mariscus umbellatus, V.). Height one foot.

Kyllinga monocephala, Rottb. Six to eight inches high.

Mapania sp. . . Three feet high.

GRAMINEÆ.

Eleusine indica, Gærtn.

Panicum (Digitaria) sanguinale, L.

Pan„cum radicans, Retz?

Pan„cum carinatum, Presl.

Pan„cum neurodes, Sch.

Pennisetum (Gymnothrix Thouarsii Beauv.?). Also

Pennisetum macrostachys, Trin. (fide F. v. Mueller): vulgo “Orsopa.”

Growing in waste ground of plantations to a height of eight or nine feet.

Coix Lachryma, L.: vulgo “Ken-ken.” The natives do not appear to make use of the seeds as beads. Growing in the waste ground of plantations.

Pollinia obtusa, Munro? Schizostachyum?? A bamboo usually found at elevations of 1000 or 1100 feet above the sea. The canes grow to a length of thirty-five to forty feet, and are used as fishing-rods.

MUSCI.

Octoblepharum (Leucophanes) squarrosum, Brid.

HEPATICÆ.

Marchantia linearis, L. and L.?

FUNGI.

Agaricus (perhaps mollic, Schff.).

Agar„cus (Inocybe) maritimus, Fr.

Hygrophorus metapodius, Fr. prox.

Lentinus submembranaceus, B.

Len„inus dactyliophorus, Lev.

Len„inus velutinus, Fr.

Polyporus (Mes.) xanthopus, Fr.

Poly„orus (Pleur.) affinis, Nees.

Poly„orus (Pleur.) luteus, Nees.

Poly„orus (Pleur.) lucidus, Fr.

Poly„orus (Placo.) australis Fr.

Hexagona apiaria, Fr.

Hex„gona similis, B.

Cladoderris dendritica, Fr.

Thelephora lamellata, B.

Hirneola auricula-judæ, Fr.

Lycoperdon gemmatum, Fr.

Bovista sp. . . (uncertain).

Wynnea macrotis, Berk.

The Flotation of Fruits in Sea-Water.—I made a few experiments on the fruits of this region, the results of which I here append. The fruits were all ripe and not dried.

(1.) Fruits that float in sea-water.[435]

[435] The following fruits and seeds, taken from my dried collection of plants, float in sea-water. I did not experiment on them in the green condition. . . . Pongamia glabra: Coix Lachryma: Scævola Kœnigii: Tournefortia argentea.

[436] Out of ten fruits experimented on, only one floated.

(2.) Fruits that sink in sea-water.

[437] This tree is widely distributed throughout the group, which may be due to the circumstance of its resin being generally employed in caulking canoes.

The weeds, rubbish-plants, and shrubs, commonly found in old clearings and in the waste-ground of the cultivated patches in Bougainville Straits.

One of the commonest plants in the islands of Bougainville Straits is the Eranthemum variabile, which is frequently found growing at the sides of the paths. The spurges, Euphorbia pilulifera and E. Atoto, are usually found in the waste-ground around villages. In the cultivated patches clumps of the handsome flowering reed, Pennisetum macrostachys (“orsopa”), which grows to a height of nine or ten feet, are often conspicuous. In one place may be seen the tall shrub, Kleinhovia Hospita (“lafai”), the inflated fruits of which are eaten by the cockatoos. In another place the botanist may recognise the Canna indica (Indian Shot: “sati”), and near by perhaps Coix Lachryma (Job’s tears: “ken-ken”), both of which plants have been probably introduced originally from the Malay Archipelago. Solomon Islanders occasionally wear the seeds of Coix Lachryma as a personal ornament. They are also used for this purpose by the Admiralty Islanders, and by the natives of some parts of New Guinea. Scented labiate plants are very frequent in the waste-ground of the plantations, and the natives are fond of wearing them in their armlets. Amongst them I may mention Moschosma polystachyum (“pipituan”) and Ocymum sanctum (“kiramma”). The “luk-a-luk” (Evodia hortensis), which is a favourite scented plant, is commonly found in the same situations. The tiny plant, Oxalis corniculata, may clothe a bare patch of ground; whilst in another part of the plantation, Commelyna nudiflora may similarly be observed. Numerous composite plants, such as Vernonia cinerea, Adenostemma viscosum, etc., form a conspicuous feature among the rubbish-plants in these cultivated patches. The Codiæum variegatum (“tiatakush”), with its very singularly-shaped leaves, is also to be seen: and, amongst other plants, I should refer to Solanum vitiense and Crotalaria quinquefolia. Tall sedges, such as Cyperus canescens and Mariscus phleoides, are to be commonly observed. Lastly, I should notice two small scitamineous plants, the “nakia,” a wild ginger, and the “temuli,” the root of which has medicinal properties, whilst its yellow juice is used for staining purposes.

A species of Pachyma ? ?

Whilst at the island of Santa Anna in October, 1882, my attention was directed by Mr. William Macdonald and Mr. Heughan to a curious vegetable substance, not unlike a yam in appearance, which is found lying loose on the soil. The specimens I obtained varied between one and five pounds in weight, but much larger examples have been obtained. The interior of the substance is white in colour, and sometimes has a waxy look. A large slab which had been whittled out by a native resembled a cake of compressed flour. There were many curious speculations as to the nature of these growths. In the estimation of the inhabitants of the island, they are poisonous, and they have received from them the name of “testes diaboli;” but I could gather but little information from the natives on the subject except to the effect that they are also commonly found on St. Christoval.[438] However, some time later I was informed by Mr. Stephens of Ugi that some mushroom-like growths were borne by a specimen that he kept, which after a few weeks fell away. I subsequently gave some of these singular masses to Mr. Charles Moore, the Director of the Sydney Botanic Gardens.

[438] In the event of it proving edible, Mr. Heughan cooked a specimen, but only a tasteless substance resulted.

Three years passed away and I had almost forgotten about the matter, when I accidentally came upon some substances, closely resembling these masses, which are exhibited in the Botanical Department of the British Museum. They are labelled Pachyma Cocos (Fries) from China. On my asking Mr. George Murray concerning their nature, I was pleased to learn that he had been taking a special interest in these growths; and he showed me a specimen obtained by the Rev. Mr. Whitmee in Samoa, from which a funnel-shaped fungus, about six inches high, was growing. This specimen was very similar to those of the Solomon Islands.

Very recently, Mr. G. Murray has embodied the results of his investigations of these growths in a short paper read before the Linnean Society, in which Mr. Whitmee’s specimen is figured (Trans. Linn. Soc., 2nd ser. Bot., vol. ii., part 11). From this source I learn that Rumphius was the first to describe these tuberous masses and their associated fungi from Amboina. The former, which he named Tuber regium, were stated to afford remedies useful in diarrhœa, fevers, &c. The fungi were said to shoot out from them during showers of warm rain on a fine day, or when there was thunder in the air. From the description and illustration given by Rumphius, Mr. Fries regarded the growth as a fungus belonging to the genus Lentinus, springing from a Pachyma (of which the Indian Bread of North America, Pachyma cocos, is an example). Strange to say, however, these tubers, which are found also in China and other parts of the world, have never been found with a fungus attached since the time of Rumphius. Mr. Whitmee’s specimen, therefore, had considerable interest. It is shown by Mr. Murray to correspond strikingly with Tuber regium and to have the structure of a true “sclerotium” (not of Pachyma), with a fungus of a species of Lentinus growing from it. All the facts go to prove that the fungus and the tuberous mass do not form part of the same growth, but are distinct organisms. A spore having germinated on the surface of the mass, its mycelium penetrates the interior, and becoming perennial, produces successive crops of fungi.

Residents in the Indian Archipelago and in the Pacific Islands might throw considerable light on the subject of these growths by careful notes and collections. It is important to discover the origin of the tuberous mass which becomes, so to speak, a convenient nidus for the fungus. How do such masses perpetuate themselves? A considerable number should be kept under observation, and the mode of appearance of the fungus carefully noticed. Experiments might be made with the spores of the fungus by dusting them over the surface of the masses. Such notes and collections should be forwarded to Mr. Murray, at the British Museum of Natural History.

CHAPTER XIV.
Reptiles and Batrachians.

In a memoir on the Reptiles and Batrachians of the Solomon Islands, which was read before the Zoological Society, on May 6th, 1884,[439] Mr. Boulenger remarked that very little was known about the herpetology of these islands until two important collections, which I sent to the British Museum in 1883 and 1884, brought to light several new and interesting forms, such as could hardly have been expected from this region. “The position of this group of islands on the limits of two great zoological districts,”—this author proceeded to observe—“renders the study of its fauna of special interest, as it is the point where many of the Papuasian and Polynesian forms intermingle. Curiously, all the Batrachians belong to species not hitherto found elsewhere, and one of them is even so strongly modified as to be the type of a distinct family.”

[439] Published in the Transactions of the Society; vol. xii., part i., 1886. The diagnoses of most of the new species in my collections were given in the Proceedings for 1884: p. 210. Vide also “Annals and Magazine of Natural History” (5) xii., 1883.

According to Mr. Boulenger, the Reptiles may be grouped under four headings, viz.:—

1. Species belonging to both the Papuasian and Polynesian districts.

2. Indo-Malayan or Papuasian species, not extending further east or south-east.

3. Polynesian species, not extending further north and west than New Ireland.

4. Species not hitherto found elsewhere than in the Solomons (and New Ireland.)

1

2

3

4

All of these 19 Reptiles were included in my collection, with the exception of Corucia zebrata, which, however, came under my observation. I will now proceed to refer more particularly to the Reptile-fauna of this region.

Crocodiles.—The species of Crocodile (Crocodilus porosus, Schneid), which is so common in the Solomon Group, ranges from India and South China through the Malay Archipelago and Papuan Islands to North Australia. In these islands crocodiles appear to frequent in greatest numbers the swamps and sandy shores of uninhabited coral islands, such as those of the Three Sisters, and the coasts of the larger islands in the vicinity of the mouths of the streams and rivers. I frequently surprised them basking on the sand under the shade of a tree. On one occasion I was standing on the spreading roots of a tree that were exposed on the beach, when one of these reptiles darted out from under my feet and dived into the sea. Of the marks that they make on the sand when lying at rest, an oblong shallow impression corresponding to the head, and a curved well-defined grove caused by the tail are alone specially recognisable. When they are not alarmed and move leisurely along, they leave a double row of footprints on the sand, with a narrow median furrow produced by the weight of the tail; but when they have been disturbed and make a rush to escape, they raise their tail and leave only the tracks of their feet on the sand. These crocodiles are equally at home in salt and fresh water. I have frequently passed them in my Rob Roy canoe when they have been floating as though asleep at the surface of the sea; and it was always in the sea that they found a refuge when my little craft intruded itself within their haunts. They came under my notice in the fresh-water lakes of Santa Anna and Stirling Island, and in the lower courses of the streams in several localities. They are apparently in no uncongenial conditions in the salt-water lagoon of Eddystone Island, although its waters receive the hot sulphureous vapours of submerged fumaroles.

These crocodiles do not apparently attain a greater length than 12 or 13 feet. Mr. Sproul shot one at Santa Anna which measured 91⁄2 feet. A female that I shot in the Shortland Islands, measured 11 feet. One of the seamen of the “Lark,” named Prior, obtained from the natives the skull of a rather larger specimen. Out of half-a-dozen individuals seen on the Three Sisters, not one measured more than 7 or 8 feet.[440] Mr. Bateman, a trader resident at Ugi, told me that at Wano on the St. Christoval coast he saw a very large crocodile which, from his description, appears to have been twice as long as any that I saw. It was, however, dusk at the time; and in connection with this circumstance I should add that I have found actual measurement to reduce the apparent length of a crocodile from 14 to 11 feet.

[440] A skull given to me by Mr. Nisbet, the government-agent of the “Redcoat,” at Ugi, was 12 inches long. It was obtained from the natives of Guadalcanar.

Natives are rarely attacked by these reptiles, and they show little or no fear of them. I have seen a full-grown crocodile dart under a line of swimmers without causing any dismay. Of the numbers I saw, all were but too anxious to get out of my way; and their cowardly nature is well shown in the account of my capture of a specimen which is given below. However, I came upon a man of Santa Anna who had had his leg broken by one of these reptiles. The natives of Rubiana hold the crocodile in veneration and work without fear in the places which it frequents. They believe that only faithless wives are seized and carried off by the monster. Pigs are occasionally the prey of the crocodile; but its usual diet appears to be opossums (Cusci), large lacertilians, and fish.

The following account of the capture of a crocodile may interest some of my readers. It was effected by no more formidable weapons than by a number of long staves and a small “bull-dog” revolver. Accompanied by six natives I was making the ascent of a large stream on the north-west side of Alu, when some of my companions espied a large crocodile at the bottom of a deep pool about 200 yards from the mouth of the stream. In setting to work to effect its capture my men proceeded very methodically to work, and evidently knew the tactics which the creature would employ. Standing in the water just below the pool, we stood awaiting the descent of the crocodile down the stream, whilst one of the natives was rousing it up with a long pole to make it leave its hiding-place. After a little time it began to get uneasy, and leaving the pool began to descend the stream. Where we were standing, the stream was only knee-deep, and as the reptile passed us in the shallow water some natives hit it on the head with their poles, whilst others hurled their poles sharpened at the ends, striking it in several places, and I planted a bullet behind its neck. The creature showed no fight and immediately hid itself in the pools near the mouth of the stream. During two hours, after we had been driving it from one pool to another by means of our pointed poles and staves, we seemed no nearer to its capture. At length there was a loud out-cry from the natives. The crocodile was making a final rush for life to cross the bar at the mouth of the stream and escape into the sea. We all followed, some in the canoe and some through the water; and for a short time I thought that the creature would escape. But being a little disabled by our previous attacks, its progress across the bar was somewhat checked; and the foremost of my men caught hold of its tail just as it was getting into deep water. Very quickly we all came up, and assisted in drawing it high and dry on the beach; and whilst two of our number kept hold of its tail, the remainder belaboured its neck with rocks and sticks until it died.[441] Its length proved to be 11 feet. Throughout the whole chase the reptile made no outcry, and even when we were belabouring it to death it only gave a kind of growl. In its stomach I found a large quantity of partially digested food with the remains of an opossum (Cuscus) and a large lizard 11⁄2 feet long (probably Corucia zebrata). It was a female, and, in the oviduct I came upon an egg, which my natives appropriated, saying that it was very good food; but they do not usually eat the flesh. I was unable from want of space to keep more than the head of the animal, which I cut off and carried back in my canoe to the ship. The skull is now in the British Museum.

[441] An illustration in Mr. Bates’ “Naturalist on the Amazons” represents a very similar scene.

Lizards. The Lacertilians are well represented in these islands. Those at present described are given in the subjoined list.

Geckonidæ

Agamidæ

Varanidæ

Scincidæ

The lizards, which most frequently meet the eyes of the visitors in the vicinity of the beaches, are the two skinks, Mabuia nigra and cyanura. As a rule those species that are common at the coast have a wide range, extending either into Polynesia or Papuasia or into both these regions (vide [page 307]). The species peculiar to these islands came less frequently under my observation. Thus, that of Lepidodactylus guppyi, is founded on a single (female) specimen I found in Faro or Fauro Island in Bougainville Straits. Corucia zebrata never came under my notice alive; it is said at Ugi to find its home in the foliage of the higher trees. Doubtless if I could have penetrated to the higher regions of the large islands, I should have obtained a large number of new species. My collections refer for the most part to the sea-border and its vicinity. In the elevated interior of such an island as Guadalcanar there is a region of great promise for the collector; but I shall have a further occasion to refer to this topic.

The Monitor, Varanus indicus, may be often seen at the coast, basking in the glare of the mid-day sun on the trunks of prostrate trees or on the bare rocks. It is considered edible by the natives of Bougainville Straits. Whilst we were anchored at Oima Atoll, Lieutenant Leeper captured a very large specimen (5 feet 73⁄4 inches long)[442] on the rocks close to the sea, and towed it off alive to the ship. After we had tried in vain to strangle it by a cord, a lead was fastened to it and it was sunk overboard, but an hour passed before we could say that the reptile was really dead. This Monitor is probably able to swim considerable distances. It very likely owes its wide range (from Celebes to the Solomon Group including Cape York) to the agency of floating trees. On examining the stomach and intestines, I found them empty. An enormous quantity of fat, developed in two large lobes in connection with the omentum or some other part of the peritoneum, almost filled the abdominal cavity. With this store of sustenance and heat, these reptiles must be able to live without food for a long time.[443]

[442] A specimen shot in the Florida Islands measured 3 feet 8 inches.

[443] As an instance of the tenacity of life that some reptiles possess, I may refer to the case of a young terrapin which I kept inadvertently for nearly five months on the coast of China without any sustenance except the dry rust of a tin can.

Snakes. Hitherto, the following six species of Ophidians have been found in the Solomon Group. All of them were included in my collection and one of them has been described by Mr. Boulenger as a new species.

Boidæ

Colubridæ

Elapidæ

Hydrophiidæ (Water-snakes)

[444] I was indebted to Lieutenant Symonds of H.M.S. “Diamond” for this snake.

One of the commonest of snakes throughout these islands is Enygrus carinatus, a harmless species of the Boa family. It often possesses considerable bulk in proportion to its length. One specimen which I obtained in Treasury Island measured 31⁄2 feet in length and 6 inches in girth. I handled a good many living snakes whilst in these islands, since the natives used to bring them in numbers to me both on board and on shore. The statements of the natives and of the white men resident in this region and the general appearance of the snakes had led me to believe that there were no poisonous species in the group. I was therefore somewhat surprised when, on my arrival in England, I learned from Dr. Günther that I had found a new species as poisonous as the Cobra. On being shown the specimen by Mr Boulenger, I at once recognised an old friend which had been brought on board in a bamboo by the natives at Faro Island and had got loose on the deck. Whilst the men standing round were preparing to kill it with more regard for their own safety than for my feelings, I caught it quickly around the neck and held it under water until it was dead. The natives certainly were not aware of its venomous character, nor was Mr. Isabell, who was my right-hand man in these matters, and used to manage the ticklish progress of removing the snakes from their bamboo-tubes in a manner only suitable in the case of harmless species. I only obtained one specimen of this snake, which was about 21⁄2 feet in length. It is named Hoplocephalus par and belongs to the Elapidæ, a family of poisonous colubrine snakes which possess the physiognomy of the harmless snakes of the same sub-order, and include the Indian and African Cobras with other well known venomous species. In the footnote I have quoted Mr. Boulenger’s description of its general appearance for the information of those who visit the group.[445]

[445] The upper surface of the head is uniform blackish brown. The body is crossed above by broad red-brown bands separated by narrow white interspaces. The lower surface of the head and body are uniform white, except on the posterior extremity of the body where the red and black extend as lines along the sutures of the ventral shields. On the tail the red forms complete rings. Nearly every one of the dorsal scales have a blackish brown border. The head is depressed, moderately large, and slightly widened posteriorly. The eye has a vertical pupil.

Batrachians.—The Spanish discoverers in 1567 remarked that the natives of Isabel worshipped the toad (vide [page 203]), and one of the officers of Surville’s expedition in 1769, described in his journal a remarkable toad from the same island;[446] yet it is only within recent years that any Batrachians have been collected in this region. Before I arrived in the group only two species were known to science, and to this number my collections, which were made in the islands of Bougainville Straits, have added seven new species, including a type of a new family. The following list represents the Batrachian fauna of the Solomon Islands, as far as it is at present known:

[446] “Discoveries of the French in 1768 and 1769,” &c., by M. Fleurieu: London, 1791; p. 134.

Ranidæ.

Ceratobatrachidæ.

(New family characterised by both jaws being toothed, and by the diapophyses of the sacral vertebra not being dilated.)

Hylidæ (Tree-frogs.)

The natives of the islands of Bougainville Straits, where, as I have just remarked, my batrachian collection was chiefly made, have given frogs the general name of “appa-appa” in imitation of their noise, just as they have named the smaller lizards “Kurru-rupu” for the same reason. Amongst the particular species of frogs, I may refer to the large toad-like Rana buboniformis, which I found in Treasury Island, and on the highest peak of the island of Faro. Rana guppyi, according to Mr. Boulenger’s report, attains a larger size than any other species of the genus, with the exception of the Bull-Frog of North America. Rana opisthodon affords an instance of a Batrachian[447] which dispenses with the usual larval or tadpole stage, “the metamorphoses being hurried through within the egg.” On this subject I made the following notes. Whilst descending from one of the peaks of Faro Island, I stopped at a stream some 400 feet above the sea, where my native boys collected from the moist crevices of the rocks close to the water a number of transparent gelatinous balls rather smaller than a marble.[448] Each of these balls contained a young frog about 4 lines in length, apparently fully developed, with very long hind legs and short fore legs, no tail, and bearing on the sides of the body small tufts of what seemed to be branchiæ. On my rupturing the ball or egg in which the little animal was doubled up, the tiny frog took a marvellous leap into its existence and disappeared before I could catch it. When I reached the ship an hour after, I found that some of the eggs which had been carried in a tin had been ruptured on the way by the jolting, and the liberated frogs were leaping about with great activity. On placing some of them in an open bottle 8 inches high, I had to put the cover on as they kept leaping out. Mr. Boulenger remarking on this observation says, that there are no gills, but that on each side of the abdomen there are regular transverse folds (with an arrangement like that of the gill-openings of Plagiostomous Fishes), the function of which probably is that of breathing-organs. The tip of the snout is, he says, furnished with a small conical protuberance, projecting slightly through the delicate envelope of the egg, and evidently used to perforate that covering. In the instance also of Cornufer solomonis, another new species included in my collection, Mr. Boulenger remarks that there is every reason to believe that the young undergo the metamorphoses within the egg.

[447] Hylodes martinicensis affords another instance. Mon. Berl. Ac., 1876, p. 714.

[448] According to Mr. Boulenger, they measure from 6 to 10 mm. in diameter.

With regard to the interesting species, Ceratobatrachus guentheri, which forms the type of a new family, Ceratobatrachidæ, the same writer observes that it is remarkable for the numerous appendages and symmetrical folds which ornate its skin. It is, in fact, “all points and angles,” and may be truly termed a horned frog. There is great variation both in the coloration and in the integuments. “Out of the twenty specimens before me,” thus Mr. Boulenger writes, “no two are perfectly alike.” The development is presumed to be of the type in which the metamorphoses are hurried through within the egg. These horned frogs are very numerous in the islands of Bougainville Straits, and so closely do they imitate their surroundings, both in colour and pattern, that on one occasion I captured a specimen by accidentally placing my hand upon it when clasping a tree.

It is particularly important to notice not only that the Batrachians of the Solomon Islands, as far as we at present know, do not occur elsewhere, but also that in this region a distinct family has been produced. These facts support the conclusions deducible from the geological evidence that these islands are of considerable geological age (vide [page x.]). The insular and isolated conditions have been preserved during a period sufficiently extended for the development of a peculiar Batrachian fauna.

The modes of dispersal of frogs and toads, and, in truth, of the whole Batrachian class, are matters of which we are to a great extent ignorant. Frogs are usually stated to be absent from oceanic islands, a peculiarity of distribution which apparently accords with the circumstance that neither they nor their spawn can sustain submersion in sea-water. The occurrence, however, of three species of Cornufer in the Caroline and Fiji Islands, and of a species of Bato in the Sandwich Islands,[449] affects the general application of this conclusion. It may be suggested that these exceptions are due to human agency; but if so, it is difficult to understand why they have not been found in such a well explored island as New Caledonia.[450]

[449] Boulenger’s “Catalogue of the Batrachia Gradientia,” &c., 2nd edit., 1882.

[450] Perhaps the peculiar geographical distribution of the Batrachia may throw light on this subject. Ibid.

In concluding this chapter I will refer to the circumstance that my collections of the Reptiles and Batrachians of this large group have only in a manner broken ground in a region which promises the richest results to the collector. It cannot be doubted that in the elevated interiors of the large islands, such as those of Bougainville and Guadalcanar, there will be found a peculiar Reptilian and Batrachian fauna, the study of which will be of the highest importance for the furtherance of our knowledge of these geologically ancient classes of animals. I believe I am correct in stating that it was on account of the highly interesting Batrachian collections I sent to the British Museum, that I received a grant for further exploration from the Royal Society, which, however, I was unfortunately prevented from turning to account. The work has yet to be done, and there can be little doubt that the man who is first able to examine the lofty interior of such an island as Guadalcanar will bring back collections, the importance of which will amply recompense him for any hardship or personal risk he may have endured. My experience was confined to the sea-border and its vicinity. The future explorer will find his field in the mountainous interiors and on the highest peaks.

Note (April 19th, 1887).—Since I penned the above, further collections of reptiles and batrachians, made in these islands by Mr. C. M. Woodford, have been described by Mr. Boulenger at a recent meeting of the Zoological Society. I had the pleasure of meeting Mr. Woodford before he left England, and I hope that he has been able to accomplish his purpose of penetrating into the interior of one of the larger islands of the group.

CHAPTER XV.
GENERAL NATURAL HISTORY NOTES.

Amongst the numerous strange acquaintances which I made in the Solomon Islands, was that of the well known cocoa-nut crab, or Birgus latro; and I take this opportunity of giving my evidence towards the establishment of the fact of its cocoa-nut-eating propensity, for the following reason. When I read my notes on the subject before the Linnean Society of New South Wales on Dec. 27th, 1882,[451] I was surprised at the incredulity shown with reference to this extraordinary habit; and on inquiry, I learned that the evidence on the subject was deficient in one vital point, viz., the production of the writer who had witnessed this habit of the Robber Crab. Accordingly I referred to the various authors who have recorded this habit of the Birgus, and in no single account could I find that the writer had witnessed what he described. Neither Mr. Darwin, Dr. Seemann, Messrs. Tyerman and Bennet, Mr. T. H. Hood, the Rev. Wyatt Gill, nor the numerous authors whose accounts I also examined, seem to have actually witnessed the Birgus opening and eating a cocoa-nut. Herbst[452] was among the first to refer to this habit; whilst, long ago, M. M. Quoy and Gaimard[453] asserted, from their own observation, that the crab was fond of cocoa-nuts, and could be supported on them alone for many months, but they made no allusion to its capability of husking and opening them. The evidence on this point appears to have been always tendered by natives, excepting the account given to Mr. Darwin by Mr. Liesk, which is conclusive in itself.[454] Yet, credulous persons had fair grounds to retain their doubts, although in various works on natural history, popular and otherwise, this habit of the Birgus was described as an undoubted fact. I therefore submit my evidence; leaving to my reader to reply to the query—Can there be any reasonable doubt on the subject?

[451] Proc. Lin. Soc. N.S.W.

[452] Proc. Zool. Soc, 1832, p. 17.

[453] Freycinet’s “Voyage autour du Monde,” 1817-20: Zoologie, p. 536. (Paris, 1824.)

[454] “Journal of Researches,” p. 462.

The Birgus was to be found in most of the islands we visited. It is to be usually observed at or near the coast; but on one occasion, in St. Christoval, I found an individual at a height of 300 feet above the sea. Whilst traversing, in September, 1882, the belt of screw-pines, which borders the beach on the east coast of Malaupaina, the southern island of the Three Sisters, I came upon one of these large crabs, ensconced in the angle between the buttressed roots of a tree, with a full sized cocoa-nut within the reach of its pair of big claws. From the fresh-looking appearance of the shell, it had been evidently, but recently, husked, which operation had been performed more cleanly than if a native had done it. There was an opening at the eye-hole end of the shell of a somewhat regular oblong form, which measured 2 by 11⁄2 inches, and was large enough to admit the powerful claws of the crab.[455] The white kernel, which had the firm consistence of that of the mature nut, had been scooped out to the extent of from 1 to 11⁄2 inches around the aperture; small pieces of the kernel lay on the ground outside the nut, and others were floating about in the milk inside, of which the shell was about a fourth-part full.

[455] This shell was presented to the Australian Museum, Sydney.

I had, without a doubt, disturbed the Birgus in the middle of its meal; but, curiously enough, there were no cocoa-nut palms to be seen within fifty paces of the spot where the crab was found in its retreat. Not only had the shell been very recently husked, but it was evident, from the fresh condition of the milk and kernel, that an interval of less than a couple of hours had elapsed since the opening had been made. There was no possible explanation of the crab having got at the edible portion of the cocoa-nut, except through its own agency. The island is uninhabited, being only occasionally visited by fishing-parties of natives from St. Christoval, none of whom were on the island during the ship’s stay. There was, therefore, the strongest presumptive evidence that the Birgus had not only husked the cocoa-nut, but had also broken the hole at the end, in order to get at the kernel.

I kept the crab alive on board on a diet of cocoa-nuts for three weeks, when, one morning, to my great disappointment, I found it dead. Other foods, such as bananas, were offered to it but were left untouched, and its appetite for cocoa-nuts continued unimpaired to the last day of its life. Being desirous of observing the manner in which the husk was removed, I had a cocoa-nut with its husk placed in the coop in which the crab was kept. On one occasion the Birgus was surprised with the nut between its large claws; but, notwithstanding that no other food was offered to it for a day and a half, it did not attempt to strip off the husk. So the operation was done for it, and a small hole was knocked in the top of the shell. On the following day I found the shell—a young and somewhat thin one—broken irregularly across the middle, with the soft white kernel already removed and eaten. It was afterwards found necessary to break the nuts for its daily food.

In 1884, when the “Lark” was in Bougainville Straits, three of these crabs were kept on board with the intention of taking them down to Sydney. Mr. W. Isabell, leading-stoker of the ship, looked well after them, as he had also done in the case of the previous crab, but within three or four weeks they had all died. The cocoa-nuts had to be husked and broken for them, as they were in vain tempted to do it for themselves. One crab, however, was frequently observed clasping between its claws a full-grown unhusked nut, the upper end of which showed deep grooves and dents from the blows of its claws; and Mr. Isabell and I came to the conclusion that the coop, in which these crabs were placed, was too low to allow of the free play of the great claws.

My evidence alone would be sufficient to convict the Birgus of this offence: for an offender it would certainly be in the eyes of the owner of a plantation of cocoa-nut palms. I learned from Mr. Isabell that the first crab we had on board consumed, on the average, two cocoa-nuts in three days. A number of these crabs in a cocoa-nut plantation, might therefore prove a considerable pest: for, if this betokens the quantity of food which the Birgus consumes in a state of nature, a single crab in the course of twelve months would dispose of about 250 cocoa-nuts, which represent the annual production of three palms and between 20 and 30 quarts of oil.

As these crabs disliked observation, I was unable to gain much knowledge of their habits by watching. During the day-time they were sluggish, did not eat, and kept themselves in the further corner of the coop, as far from the light as possible. At night they moved about very actively and fed vigorously on the cocoa-nuts. The natives of the Shortlands, who were well-acquainted with the cocoa-nut eating habit of the Birgus, described to me the mode of husking and breaking the nut, just as Mr. Liesk described it to Mr. Darwin. They esteem as an especial luxury the fat which gives the chief bulk to the abdomen of the crab.

The habit of the Birgus, when surprised away from its burrow, is not to turn round and run away, but to retreat in an orderly manner with its front to the foe. Having reached some root or trunk of a tree which protects in the rear its less perfectly armoured abdomen, it makes a regular stand, waves one of the long second pair of claws in the air, and courageously awaits the attack. The attitude of defence is worthy of remark. The two large claws are held up close together to defend the mouth and eyes, but with the pincers pointing downward—the posture reminding me of the guard for the head and face in sword-exercise. One of the long second pair of claws is planted firmly on the ground to give the crab additional support; whilst the other claw is raised in the air and moved up and down in a sparring fashion. The whole attitude of the Birgus, when on the defensive, is one of dogged and determined resistance. The big pincers that point downward are ready to seize anything which touches the unprotected under surface of the abdomen; but on account of the position of these claws in front of the eyes, it can only foresee attacks from above, and it therefore cannot ward off a sudden thrust directed against the abdomen, although it may afterwards inflict severe injuries on the aggressor.

There seems to be some doubt whether the Birgus ascends the tree to get the cocoa-nuts or whether it contents itself with those that have fallen. Almost every author who refers to this crab alludes to its climbing the tree, and it is also said to climb the pandanus. The testimony in support of its climbing powers is almost conclusive, yet Mr. Darwin was informed by Mr. Liesk that in Keeling Atoll the Birgus lives only on the fallen cocoa-nuts, and Mr. H. O. Forbes,[456] who has recently visited this island, confirms this statement.

[456] “A Naturalist’s Wanderings,” etc.: London, 1885, p. 27.

My readers, after perusing the foregoing remarks, will agree with me that from the lack of actual observation on the part of the authors, who describe the cocoa-nut eating habit of this crab, there has been fair grounds for scepticism. Even now, we are but imperfectly acquainted with the mode of life of the Birgus, which is a subject I would commend to the attention of residents in the Indian and Pacific Oceans.

I may add that the Birgus is partial to other kinds of fruits beside cocoa-nuts. Different writers mention candle-nuts, nutmegs, figs, and other rich and oily nuts and fruits. In some islands it would seem that the Pandanus fruit is its only diet; and for breaking open these tough fruits, its heavy claws are well adapted, though from personal experience, I should remark that the crab would have its strength and ingenuity taxed almost as much as in the case of the cocoa-nut.

The handsome ground-pigeon, known as the Nicobar pigeon (Geophilus nicobaricus), is commonly observed in the wooded islets on the coral reefs of the Solomon Group. As I have remarked on [page 293], this bird is probably instrumental in transporting from one locality to another the small hard seeds and fruits which the common fruit-pigeon (Carpophaga) refuses. That it is able to crack such hard seeds as those of the leguminous plant Adenanthera pavonina,[457] is shown by the fact that I have found these seeds cracked in the cavity of the gizzard, which is in its structure and mechanism a veritable pair of nutcrackers. In this bird the muscular stomach or gizzard is of a surprising thickness, and is provided with a very singular mechanical contrivance to assist its crushing power. As shown in the accompanying [diagram], it is composed of two muscular halves, each having a maximum thickness of five-eighths of an inch and united with each other in front and behind by a stout distensible membrane, which is the proper wall of the organ. Developed in the horny epithelial lining membrane there are two cartilaginous bodies of hemispherical shape, one in each muscular segment of the gizzard, which measure about one-third of an inch in thickness and three-fourths of an inch in diameter. The outer or convex surface of each cartilaginous body fits into a cup-shaped cavity which is lined by a semi-cartilaginous membrane, the whole constituting a “ball-and-socket” joint with well lubricated surfaces. The two surfaces of this pseudo-articulation are capable of easy movement on each other, being retained in close apposition by the attachment to the subjacent tissues of the horny epithelial lining membrane in which the cartilaginous body is developed. The inner or free surface of each hemispherical body, that which looks into the gizzard cavity, is somewhat concave, and projects a little above the surface of the lining membrane; it is much harder than the opposite convex side of the cartilage and has almost the consistence of bone, the arrangement of the cells into densely packed rows with but little intervening matrix indicating an approach towards ossification.

[457] The Kuara tree of India, of whose hard seeds necklaces are made.

The firm consistence of these hemispherical cartilages combined with the mechanism of a moveable articulation must greatly assist the already powerful muscular walls of the gizzard; but there is an additional factor in the crushing power in the constant presence of a small quartz pebble, usually about half-an-inch across. With such a apparatus, I can well conceive that very hard seeds and nuts may be broken, as in the case of the seeds of Adenanthera pavonina already alluded to. The Nicobar pigeon is in fact possessed of a nut-cracking mechanism in its gizzard, by which nuts like those of our hazel tree would be cracked with comparative ease.

With reference to the small quartz pebbles found in the gizzards of these birds, I should remark that there is usually only one present, and that it varies in weight between 30 and 60 grains. I was sometimes able to say where the pigeon had obtained its pebble. Thus, in Faro Island the bird often selects one of the bipyramidal quartz crystals, which occur in quantities in the beds of the streams in the northern part of the island, where they have been washed out of the quartz-porphyry of the district. In other instances the pebble seems to have been originally a small fragment of chalcedonic quartz, such as composes some of the flakes and worked flints that are found in the soil which has been disturbed for cultivation. Sometimes the pebble is of greasy quartz; and now and then in the absence of quartz the bird has chosen a pebble of some hard volcanic rock. It is a singular circumstance that although these pigeons frequent coral islets where they can easily find hard pebbles of coral-rock, they prefer the quartz pebbles which are of comparatively rare occurrence. I never found any calcareous pebble in their gizzards, and was often at a loss to explain how the bird was able to ascertain for itself the different degree of hardness between the two pebbles, when the quartz was of the dull white variety. . . . I learn from a recent work on New Guinea by the missionaries, Messrs. Chalmers and Gill, that inside the gizzard of each Goura pigeon there is a good-sized pebble much prized by the natives as a charm against spear-thrusts and club blows.[458] The Goura pigeon resembles the Nicobar pigeon in habits; and I think it probable that its gizzard will be found to present a similar structure and mechanism for cracking nuts and hard seeds. The common fruit pigeons (Carpophaga) of the Solomon Islands, living as they do on soft fleshy fruits, and rejecting the hard seeds and kernels, have no peculiar structure of the gizzard, the walls of which are comparatively thin, and are thrown into permanent rugæ somewhat warty oh the surface.

[458] “Work and Adventure in New Guinea” (p. 317): London, 1885.

One of the most familiar birds in these islands is the “bush-hen,” which belongs to the family of the mound-builders (Megapodiidæ). They bury their eggs in the sand at a depth of between three and four feet. On one occasion in the island of Faro, Lieutenant Heming and his party found eight eggs, in different stages of hatching, thus buried: they were scattered about in the sand; and according to the account of the natives only one egg was laid by each bird. The eggs are sometimes found on the surface of the sand. The young birds are able to fly short distances soon after they are hatched. One that was brought on board astonished us all by flying some thirty or forty yards from the ship and then returning to the rigging.

The account recently published by Mr. H. Pryer of his visit to the birds’ nest caves of Borneo[459] has opened up the discussion as to the nature of the substance of which the edible bird’s nest is composed. Many and varied have been the surmises as to the source of this material; but nearly all of them have been based on mere speculation, and have been relegated to the limbo of sea-tales. Amongst the earlier explanations, I may allude to those which have been given by early writers. The swiftlets (Collocalia), which build their nests in this extraordinary fashion, were considered to gather a gelatinous material from the ocean-foam, or from the bodies of holothurians, or from the skin of the sun-fish. The Chinese fishermen assured Kæmpfer that their nests were composed of the flesh of the great poulpe. A more probable explanation, however, was found by Rumphius in the occurrence on the sea-coasts of a soft almost cartilaginous plant which he with confidence asserted was the material from which these swiftlets constructed their nests; but subsequently this naturalist inclined to the opinion that the substance of which the edible birds’ nests are composed is merely a secretionary product. In these two views of Rumphius we have the two sides of the controversy very much as it at present stands. On the one hand, there are those who hold that this substance is a secretionary product: on the other hand, the opinion is held that the nest is constructed of a vegetable matter, usually resulting from the growth of a microscopic alga, which is found in the caves and on the faces of the cliffs where the nests occur. All the weight of experiment and of actual observation tends to negative the view of the vegetable origin of this substance. Sir Everard Home in 1817 declared his opinion that certain peculiar gastric glands, which he found in one of these birds, secreted the mucus of which the nest was formed. In 1859, Dr. Bernstein[460], after having carefully studied the habits of the birds in question, came to the conclusion that their nests are formed from the secretion of certain salivary glands which are abnormally developed during the nest-building season. M. Trécul, who held the same opinion, showed that the bird constructs its nest by means of a mucus which flows abundantly from its beak at the pairing time.[461] This last view is strongly supported by Mr. Layard, who unhesitatingly pronounces his opinion that these swiftlets build their nests from the secretionary products of their own salivary glands.[462] However, when Mr. Pryer visited in March, 1884, the birds’ nest caves in British North Borneo, he considered that he had found the source of the material of which the nests were composed in the occurrence of a “fungoid growth,” which incrusted the rock in damp places, and which, when fresh, resembled half-melted gum tragacanth. Without at present expressing an opinion as to the validity of the inference Mr. Pryer drew from his observations in these caves, I may observe that the “fungoid growth” has been determined by Mr. George Murray,[463] of the Botanical Department of the British Museum, to be the result of the growth of a microscopic alga, a species, probably new, of Glœocapsa; whilst the edible nests from these caves, according to a chemical and microscopical examination made by Mr. J. R. Green,[464] have been shown to be formed in the great mass of mucin, which is the chief constituent of the mucous secretions of animals. After examining various specimens of edible nests from other localities, Mr. Green subsequently confirmed the results of his first experiments. The nest-substance, as he unhesitatingly states, is composed of mucin, or of a body closely related to it.[465] So far, therefore, there would appear to be but little evidence to support the view of Mr. Pryer that the species of alga, which he found incrusting the rock in the vicinity of the Borneo caves, supplied the material for the construction of the nests of the swiftlets. However, before proceeding to state my own opinion on the matter, I will refer briefly to my observations in the Solomon Islands relating to this question.

[459] Proceedings of the Zoological Society for 1884: p. 532.

[460] Journ. für Ornithologie, 1859, pp. 112-115; also Proceed. Zoolog. Soc., 1885, p. 610.

[461] “A General System of Botany,” by Le Maout and Decaisne: London, 1873, p. 983.

[462] “Nature,” Nov. 27th, 1884.

[463] Proc. Zoolog. Soc., 1884: p. 532.

[464] Proc. Zool. Soc., 1884, p. 532.

[465] “Nature,” Dec. 11th, 1884 and May 27th, 1886.

A species of Collocalia, which usually frequents inaccessible sea-caves and cliffs, is frequently to be observed on the coasts of the islands of this group. The natives of Treasury Island call this bird “kin-kin;” but they have no knowledge of the nutrient qualities of the substance of which it builds its nest, and they were much amused when I told them of its being a Chinese luxury. I only came upon the nests of this bird on one occasion, and that was in some caves on Oima Atoll in Bougainville Straits. A description of these caves will be here unnecessary. As in the instance of the birds of the Borneo caverns, these swiftlets shared their retreats with a number of large bats, the accumulation of whose droppings had produced a thick reddish-brown deposit on the floors of the caves. The nests, which were formed for the most part of fibres derived evidently from the vegetable drift[466] at the mouths of the caves, were thickly incrusted with the gelatinous incrustation which projected as winglets from the sides and fastened them to the rock.

[466] The husks of pandanus seeds more particularly.

A reddish soft gelatinous incrustation occurred on the faces of some of the cliffs in the vicinity of the caves. It was composed of an aggregation of the cells of a microscopic unicellular alga which measure 1⁄2500 of an inch in diameter. Unfortunately the specimens of this growth which I collected have been mislaid, but there can be little doubt that it is similar to the “fungoid growth” which Mr. Pryer describes in connection with the Borneo caves, and which, through the kindness of Mr. George Murray, I had the opportunity of seeing at the British Museum. On the faces of the coral limestone cliffs of some islands, such as on the east coast of Santa Anna, a like growth occurs in considerable quantity. In its freshest condition, it may be described as a reddish-yellow, gum-like substance forming a layer 1⁄4 to 1⁄8 of an inch in thickness. Where it incrusts the overhanging face of a cliff, it is more fluid in consistence and sometimes hangs in little pendulous masses, one to two inches in length, the extremities of which are often distended with water. This alga decomposes the hard coral limestone, making the surface of the rock soft and powdery. All stages in the growth of this substance may be observed. The older portions are very dark in colour and have a tough consistence; and in the final stage it occurs as a black powder covering the rock surface. On examining this alga with the microscope, I found it to be formed almost entirely of granular matter apparently resulting from the death of the cells; whilst the presence of a few cellular bodies alone gave me an indication of its true nature.

From my observations relating to the subject of the edible bird’s nest, it may be therefore inferred that in the Solomon Islands, as in Borneo, the occurrence of these nests is associated with the presence of a protophytic alga, which incrusts the rocks of the locality as a gelatinous or gum-like substance. Whether or not the birds employ this material in forming their nests, is a question which would appear to have been already answered in the negative; but it seems to me that those who hold that this material is used for this purpose might justly claim that the final judgment should be suspended, until a chemical examination of this vegetable substance has been made with the object of determining whether it might not yield a material closely resembling mucin. Amongst the nitrogenous constituents of plants occurs the so-called vegetable albumen, which in its chemical composition and in its behaviour with re-agents does not differ materially from the blood-albumen of the animal organism, of which in fact it is the source. In suggesting, therefore, that a vegetable mucin may be found in this low plant-growth, I do not pass beyond the bounds of probability.[467]

[467] Vide a letter by the writer in “Nature,” June 3rd, 1886.

Small scorpions came under my notice in Faro Island. They are not usually more than 11⁄2 in length and occur in narrow clefts of rocks and in the crevices of trees. I was stung by one on the thumb, but the pain was trifling and soon passed away.[468]

[468] Specimens of these scorpions were given by me to the Australian Museum, Sydney.

A species of Iulus or Millipede, which attains a length of from 6 to 7 inches, is commonly found in the eastern islands of the Solomon Group on the trunks of fallen trees and amongst decaying vegetable débris. It is often to be seen amongst the rotting leaves that have gathered inside the bases of the fronds of the Bird’s-nest Fern (Asplenium nidus). These Myriapods seem to be less frequent in the islands of Bougainville Straits towards the opposite end of the group, as I do not remember seeing any large Iuli in that locality: their place appears to be taken by another Myriapod, apparently a Polydesmus, growing to a length of 21⁄2 inches, which I found amongst decaying vegetation at all elevations up to 1900 feet above the sea, as on the summit of Faro Island. But to return to the Iuli, I should remark that this genus of Myriapods evidently possesses some means of transportal across wide tracts of sea, since, amongst other islands similarly situated, it is found in Tristan da Cunha,[469] in the South Atlantic Ocean, and I have found it in the Seychelles, in the Indian Ocean. The habits of these Millipedes would render it highly probable that they have reached the oceanic islands on vegetable drift, such as floating logs. It is, however, a noteworthy circumstance that they do not seem to be able to withstand immersion in sea-water for any length of time. In experimenting on the Solomon Island species, I found that they were able to survive an hour-and-a-half’s complete immersion in sea-water, but that an immersion of three hours killed them. One individual, out of several experimented on, survived for twelve hours after it was taken out, but only in a half lifeless condition.[470] It may, therefore, have been that the Iulus has been transported to oceanic islands by such agencies as canoes and ships, rather than by means of floating trees.[471]

[469] Moseley’s “Naturalist on the Challenger,” p. 134.

[470] This species of Iulus was able to sustain a longer submersion in fresh-water, without apparently any injurious effects. Those experimented on recovered after being kept under water for four hours, but died after a submersion of six hours.

[471] As bearing on this point, it might be interesting to determine whether these large Iuli occur on islands far from land which are believed never to have been inhabited.

Like other species of the genus, the Solomon Island Iulus exhales a very pungent and disagreeable odour, which is caused by an acrid fluid secreted by small vesicles, of which each segment of the body contains a pair.[472] On holding my nose for a moment over the mouth of a bottle, containing two of these large Millipedes, I experienced a strong sensation in the nasal passages, reminding me much of the effects of an inhalation of chlorine gas. I had previously learned from resident traders that these Millipedes have a habit of ejecting an acrid fluid when disturbed, which, if it entered the eye, was liable to cause acute inflammation; and the instance was related to me of the captain of some ship, trading in these islands, who lost the sight of one of his eyes from this cause. Mr. C. F. Wood learned from the natives of St. Christoval, in 1873, that these Myriapods “could squirt out a poisonous juice, which was dangerous if it happened to touch one’s eye;” but he adds, “there seemed no great probability of their doing this.”[473] However, I usually found that native testimony, in such matters, was very reliable; and in the instance of this reputed habit of the Iulus, my personal experience has convinced me of its reality. Whilst handling one of these Millipedes as it lay on the trunk of a fallen tree in Ugi Island, I felt a sudden smarting sensation in the right eye, caused apparently by some fluid ejected into it. Remembering the injurious effect attributed to this habit of the Iulus, I at once plunged my head under the water of a stream, in which I happened to be standing up to my waist, and I kept my eye open to wash away the offending fluid. During the remainder of the day, there was an uncomfortable feeling in the eye and somewhat increased lachrymation; but on the following morning these effects had disappeared. At the time of this occurrence, my face was removed about a foot from the Millipede; and, although I was uncertain from what part of the body the fluid was ejected, I did not care, under the circumstances, to continue the inquiry.

[472] Hoeven’s Zoology. (Eng. edit.) Vol. I., p. 291.

[473] “A Yachting Cruise in the South Seas,” p. 131. (London, 1875.)

Amongst the first living creatures to greet the visitor as he lands on the beach of a coral island in the Pacific, is a small species of Hermit-Crab, belonging to the genus Coenobita, which frequents the beach in great numbers. The crab withdraws itself just within the mouth of the shell, where it forms a perfect operculum, by means mainly of the large flattened chelæ of the left great claw which is arched over by the left leg of the third pair, whilst the right claw and the right leg of the second pair serve to complete the shield The most plucky and pugnacious of these little crabs are those which occupy cast-off Nerita shells, a character which probably arises from their consciousness of the solid strength of the home they have chosen: and, strange to say, the tiny bosses on the surfaces of the large pincers, which are outermost in the improvised operculum, resemble similar markings on the outer side of the operculum of the Nerita (N. marmorata, Hombr and Jacq), whose shell they often inhabit. Mr. Darwin[474] observed that the different species of hermit-crabs, which he found on the Keeling Islands in the Indian Ocean, used always certain kind of shells; but I could not satisfy myself that such was the case in the instance of the Solomon Island hermit-crabs. In the case of the common beach species of Coenobita, I found, after carefully examining a number of individuals to satisfy myself of their being of the same species, that shells of the genera Turbo, Nerita, Strombus, Natica, Distorsio, Truncatella, Terebra, Melania, &c., &c., contained the same species of Coenobita, whether the individual was large enough to occupy a Turbo shell of the size of a walnut or sufficiently small to select the tiny shell of the Truncatella for its home. Another species of the same genus prefers usually the vicinity of the beach; but it may occur at heights up to 200 feet above the sea. It is rather larger than the beach species, and differs amongst other characters in the more globose form of the large claws and in the greater relative size of the left one. It occupies shells of different kinds, such as those of Nerita, Turbo, &c. A still larger species, which frequents the vicinity of the beach, usually selects Turbo shells, apparently because of their larger size. All the other species of Coenobita, which I met with, used, when I touched them, to withdraw themselves within their shells and close them up at once with their claws; but this kind, when I caught hold of the Turbo shell that it carried, left the shell behind in my fingers with apparent unconcern and crawled leisurely away, displaying, somewhat indecorously, the rudimentary plates on the back of its abdomen. These are the plates that attain their greatest development in the Cocoa-nut Crab (Birgus latro), which is thus able to dispense with a shell altogether. The greatest heights at which I found hermit-crabs were in the island of Faro on the two highest peaks, which are elevated respectively 1600 and 1900 feet above the sea. In both these localities, the crab had reached the very summit and could not have climbed higher. The species was apparently different from, though closely allied to, the common beach species, and frequented the shells of a land-snail (Helix). I was indebted to Lieutenant Heming for directing my attention to the hermit-crab, found 1900 feet above the sea. It appears to me likely that these hermit-crabs will be found at much greater heights in this group, since, in this island, their ambition to rise had carried them up as far as they could go.

[474] “Journal of the Beagle,” p. 457.

Other species of hermit-crabs, that are common in these islands, belong to the genus Pagurus. They are conspicuously distinguished from the species of Coenobita, above described, by their first pair of claws, which are small and weak and ill-adapted for defensive purposes. For this reason, these species are less able to look after themselves; and since they cannot form the operculum-like shield with their claws at the mouth of the shell, they always choose shells which will permit of their retiring well within it, so as to be out of the reach of their enemies. Some species are found in the stream-courses and in the brackish water near their mouths, when they often frequent cast-off Melania shells. Other species (?) prefer the sea-water on the reef-flats. I noticed one individual that displayed its eccentricity of disposition, in selecting, as its abode, the hollow tube of a small water-logged stick, about six inches long, which it dragged about after it during its peregrinations, and into which it retreated when alarmed. On one occasion, I observed a large Dolium shell, moving briskly about in a pool of salt-water, which, on picking up, I found to be tenanted by a Pagurus, so ridiculously small, in comparison with the size of the shell, that when frightened it retreated to the very uppermost whorl, and, notwithstanding the wide mouth of the shell, could not be seen. So light was the weight of the crab, that, on account of the buoyancy of its shell, it floated lightly on the surface of the water, on which I had placed it with the mouth of the shell uppermost, and was blown by a slight breeze across a pool of water, some twenty yards in width. While it was afloat, the shrewd little occupant retired to the innermost recess of its home; but as soon as the shell had grounded, it protruded its head and pincers and endeavoured to overturn the shell, which it finally succeeded in accomplishing.

In the case of these two genera of hermit-crabs, Coenobita and Pagurus, it was interesting to notice the relation existing between the defensive capabilities of the crab, and the relative size of the shell it selected as its home. The Pagurus, with its weak slender pincers, chooses large shells within which it can retire well out of reach when alarmed. The Coenobita, with its stout pincers, prefers shells much smaller, relatively speaking, and ensconces itself snugly in the body whorl, forming an operculum with its claws. As the hermit-crab, Coenobita, crawls along the dry sand of a beach, it leaves behind it characteristic pinnate tracks which may be often traced for several feet. The lateral markings are produced by the claws and legs working on each side of the shell; whilst a central groove is formed by the weight of the shell itself. As shown by the arrow in the [diagram], the lateral markings point in the direction of the course which the hermit-crab has taken. Sometimes only a single row of lateral tracks accompanies the grooves produced by the shell. Such markings were produced by a hermit-crab when frightened by my approach. It turned its front towards me, and crawled backwards, by working most of his claws and legs on one side of the shell. In the case of the larger hermit-crabs, which are much less frequent on the beach, each limb produces a distinct print on the sand; but with the small species of Coenobita which infests the beach, each lateral marking, as shown in the diagram, is produced by a single movement of the claws situated on the same side of the shell. The hermit-crabs only leave their tracks on the dry loose sand. One individual, that I placed on sand, still wet from the retreating tide, crawled along without leaving any impression. I have described these impressions with some care, as they bear on the origin of the surface-markings of rocks of shallow-water formation, a subject recently discussed in the geological world. It is highly probable that some of the larger and heavier forms of the Anomura (and, in fact, of the Decapoda generally) would produce prints such as I have here described, both on mud-flats left dry by the tide, and on the soft bottom in shallow depths. A cast of the impressions thus produced would have an unmistakeable plant-like form.

Whilst examining the island of Simbo, I noticed some singular Medusæ in a small mangrove-swamp, which is inclosed in the low point that forms the south shore of the anchorage. Numbers of these organisms of a large size (8 or 9 inches across the umbrella), and of a dirty-white colour, were lying on the mud with their tentacles, uppermost in depths of from one to three feet of water. I was struck by the handsome mass of arborescent tentacles which they displayed, and by the peculiarity of their lying upside-down. The dark mud which formed the bottom of the swamp was composed of decayed vegetable matter, confervoid growths, diatoms, and a few infusoria: but when I raised up these Medusæ, I found underneath each a patch of white sand corresponding with the outline of the organism, but completely concealed by the umbrella when the Medusa lay in its usual position. The sand was derived from corals, shells, and the volcanic rocks of the island; and the light patches formed a marked contrast with the dark mud around. I was unable to find any satisfactory explanation of these curious patches of sand; and I, therefore, proceeded to interrogate the Medusæ on the subject by watching them, but to no purpose. So I had my revenge by turning them all over on their tentacles, when each one immediately began to contract its umbrella in a most methodical fashion, and, after swimming a short distance, deliberately resumed its former position of tentacles upward. I had an extensive experience of mangrove-swamps after we left Simbo; but these self-willed Medusæ never came under my notice again.[475]

[475] I referred to the habits of these Medusæ in “Nature,” Nov. 9th, 1882.

With regard to these Medusæ, I should remark that they belong to a species of Polyclonia, and are classed amongst the Scypho-Medusæ.[476] Two species of Polyclonia seem to be known, P. frondosa (Agassiz) and P. Mertensii(Brandt), the first found in the Florida seas, and the latter in the Carolines. I am inclined to think that the Solomon Island species is more nearly allied to P. Mertensii. Both species, however, have similar habits, lying on the mud of mangrove-swamps, with their tentacles uppermost.

[476] I compared my notes with the description and figures given by Agassiz in his “Contrib. Nat. Hist. U.S.A.” (1862: vols. iii. and iv.). In the Solomon Island species, the dendriform mass resolves itself into 8 principal branches, each ramose, and all united at their bases by a common membrane. The umbrella, which was finely lobed or crenulated at its margin, displayed about 40 radiating canals, each communicating by an anastomosing network with the canal on either side of it.

The singular habits of these Scypho-Medusæ were noticed by Brandt in 1838. They have since been remarked by Mosely[477] in the Philippines, and by Archer[478] in the West Indies. L. Agassiz in his “Contributions to the Natural History of the United States,” describes and figures the Florida species (Polyclonia frondosa); and some additional notes on its habits have been made by A. Agassiz, to whose communication in “Nature” (Sept. 29th, 1881) I have been much indebted.

[477] Mosely’s “Notes by a Naturalist,” p. 404.

[478] “Nature” Aug. 4th, 1881.

Whilst we lay at anchor in Treasury Harbour, in April, 1884, a cetacean, unknown to the natives and to ourselves, got partly stranded in the shallow water, and was captured by the villagers. It was nine feet long, and possessed this remarkable character that, although no teeth showed through the gums, each lower jaw possessed a short, conical, hollow tooth an inch long, placed at the anterior extremity. I obtained the head from the natives, and placed it in a safe place, as I thought; but when we returned to Treasury a few weeks after, I found only portions of the skull with the lower jaw-bones, the wild pigs having held a feast over it. The remains, however, together with my notes and a sketch by Lieutenant Leeper were sent to the British Museum. I there learned that it is a species of Ziphius, probably unknown.

The Solomon Islanders believe in the existence of anthropoid apes in the interiors of the large islands, regarding them, however, like the Dyaks of Borneo in the case of the Orang-utan, as “wild men of the woods.” In Malaita they are said to be 41⁄2 to 5 feet high, and to come down in troops to make raids on the banana plantations. Captain Macdonald informed me that the natives allege that one of these apes was caught, and, after being kept for some time, escaped. Taki, the St. Christoval chief, told Mr. Stephens that he had seen one of these apes, and pointed out the locality. Tanowaio, the Ugi chief, also made a similar statement. In Guadalcanar, they are believed to live in the trees, and to attack men. Dr. Codrington refers to the prevalence of these beliefs throughout Melanesia (Journ. Anthrop. Inst., vol. x. p. 261). Such beliefs, as experience has shown in the case of the Gorilla and other anthropoids, have undoubtedly some foundation; but whether these mysterious animals are apes is quite another question.

CHAPTER XVI.
LAND AND FRESH-WATER SHELLS.

During my numerous excursions in these islands, I had in the majority of instances to follow up the stream-courses in order to examine their geological structure. I had therefore good opportunities in these regions of making a collection of the fresh-water shells, which, together with the land shells I collected, formed a total of between sixty and seventy species, amongst which there were 11 new species and at least 5 new varieties, whilst about 14 would appear to have been never previously recorded from the Solomon Islands, and there were in addition several from new localities in the group. The collection was sent to the British Museum and was examined and described by Mr. E. Smith, to whose paper on the subject[479] I am indebted for my acquaintance with the shells in question, and through whose kindness I have been thus enabled to supplement my other observations in these islands. A list of the shells with the descriptions of the new species is given on [page 344]. For its size, my collection presented a large amount of novelty, coming as it did from a region the land and fresh-water shells of which were previously considered to be fairly known. There can be no doubt, however, that in the Solomon Islands the conchologist has much work that remains to be done. Not only are the higher regions of the larger islands, entirely unexplored, but it would appear from the collections made up to the present date in this large group, that particular species may be not only confined to a special sub-group of islands but may be restricted to a single island, and that other species more widely distributed through the group may be represented in each island and in different districts of the larger islands by different varieties. Had I been aware of the extent of the influence of locality in this region, I might have made my collection of greater value. It would therefore seem necessary for future collectors in this group to make in every small island and in different districts of the larger islands special independent collections, disregarding the fact that they may have apparently met with the same shell very frequently before, because many of the varieties and some of the species can only be distinguished by the practised eye of the specialist, and a new locality for a previously well-known species may be often unwittingly found.

[479] “On a collection of shells from the Solomon Islands” by Edgar A. Smith. (Proceedings of the Zoological Society: June 2nd, 1885.) This paper is illustrated with two coloured plates of the new shells.

As an instance of the unexpected results, which may fall to the lot of others in this group, I may here add, that out of eleven land and fresh-water shells that I collected in the small island of Santa Anna, which is only 21⁄2 miles in length, four were new species, and besides there were some new varieties. The stations of these four species may be suggestive. Two of them—Helix (Videna) sanctæ annæ and Helix (nanina) solidiuscula—were generally found on the trunks of the cocoa-nut palms at the coast; whilst the other two occurred in situations far more likely to yield new species, Melania sanctæ annæ being obtained from a small stream in the interior of the island, and Melania guppyi being found dead in the stomach and intestines of a fish that frequents the fresh-water lake of Wailava. This last shell would appear to live in the deeper parts of the lake, as I only found one living specimen, all the others being obtained from the stomach and intestines of these fish. Mr. Smith describes it as “a very remarkable and distinct species.” Its length is about 11⁄5 inches; and its sharp-pointed spire was to be sometimes seen protruding through the vent of the fish, which evidently digests the animal and ejects the shell. These fish were usually 9 or 10 inches long; but the full-grown shells were found also in fish half this size, when the relation between the length of the shell and the size of the fish was truly alarming. Since the little fish actually swallow sharp pointed shells measuring a fifth of their own length and pass them out through the vent after they have digested the animal, we must credit them with a remarkable capacity for adapting their diet to circumstances.

To exemplify the variation which some species of shells display in this group, I will take the instance of Helix (Geotrochus) cleryi, Récluz. This species is probably distributed through the whole group; but considerable variation prevails in different islands. Amongst the several forms which I obtained, three were named as new varieties, var. meridionalis from Santa Anna, var. simboana from Simbo or Eddystone Island, and var. septentrionalis from the islands of Bougainville Straits, the localities of the two last varieties being only 80 miles apart. Mr. Smith remarks that this species is “subject to considerable variation in size, colour, and form, apparently resulting from difference of habitat. . . . . . . . . Whether these several varieties should take specific rank is questionable, for, although there is a considerable difference between the extreme forms, even in the series of nearly one hundred specimens under examination, the gradual transition from one form to another is observable.”

Amongst the more singular in appearance of the land shells, I may refer to the large Bulimus (B. cleryi) which I found on the north coast of St. Christoval. It attains a length of four inches. I was never able to get a living specimen, as they are said by the natives to live in the foliage of the high trees. The specimens which I obtained were empty shells which the natives of the Koofeh district on the north coast of St. Christoval are in the habit of throwing into heaps, each man when he picks up a shell throwing it into the next heap he passes. I was unable to learn the reason of this practice and the natives did not seem willing to tell me. . . . Two other Bulimi I commonly met with. One was the pretty Bulimus miltocheilus (Reeve), which, when the animal is young and the shell delicate, has a greenish-yellow hue resembling the colour of the leaves it feeds upon: as it grows older the shell becomes thicker and stronger, and in proportion as there is less need for protective resemblance, the greenish-yellow hue fades away, leaving a dull white colour behind. This species is found in St. Christoval and the adjacent islands. The other Bulimus (B. founaki, Homb. Jacq.) which I found in Faro Island, Bougainville Straits, and which had been only previously obtained at Isabel Island, attains a length of rather under three inches.

I come now to refer to the fresh-water shells of these regions. Stated in their order of frequency, the Neritinæ, Melaniæ, and Navicellæ are the common fresh-water shells of these islands. The Neritinæ were especially interesting to me. They abound in the streams: some of them preferring the moist rocks above the water, others finding their home in the waters of a quiet pool, whilst others, like the Navicellæ, prefer to buffet the full rush of the torrent. An important feature with reference to these fresh-water Nerites or Neritinæ is their wide dispersal. “Some of these species”—as Mr. Smith remarks in respect of those in my collection—“range not only through most of the islands of the Solomon Group, but have a considerably wider distribution.” Thus, Neritina subsulcata (Sowerby) and N. cornea (Linné), are not only found in the Solomon Islands, but also occur in the Philippines: N. macgillivrayi (Reeve) and N. petiti (Récluz) alike exist in the Fiji and in the Solomon Groups; while N. porcata (Gould) has been found in Samoa and in Fiji as well as in the opposite extremities of the group with which I am at present concerned. Being interested in the question of the mode of dispersal of these Nerites, I made the following experiment to test their powers of sustaining submersion in salt-water. One individual belonging to the species Neritina subsulcata[480]—a species which is also found as above stated in the Philippines, and at the same time is the most widely dispersed fresh-water Nerite in the Solomon Group—survived a submersion of twelve hours; but not one out of a dozen individuals was found alive after a submersion of five days, although the water was changed from time to time. The result was a surprise to me, as I inferred from the result of Baron Aucapitaine’s experiments as related by Mr. Darwin,[481] that their close-fitting stony opercula would have enabled them to resist the action of salt-water. Their death could have been scarcely due to want of food, since I have kept shells of this species for several months on a very scanty diet, and since the powers of endurance of other fresh-water shells are well known. The matter passed out of my mind until after my arrival in England, when Mr. Smith put the question to me, as to their mode of dispersal. I then remembered that their calcareous egg-capsules, which are so commonly seen on the rocky sides of the streams, are in all probability sufficiently thick to resist the action of salt-water. Here is therefore a probable mode of dispersal, and I see it is one which Mr. Smith refers to as such in his paper. These egg-capsules “if attached to floating timber, might be carried to considerable distances.” They are often to be observed on the outside of the shells of living Navicellæ, and I have seen them on the backs of the valves of a Unio which I discovered in the Shortland Islands.

[480] In his paper Mr. Smith refers to the species experimented on as N. cornea: but in my own list he named a shell belonging to one of the Neritinæ in question as N. subsulcata.

[481] Cyclostoma elegans was the species tested: vide “Origin of Species,” p. 353, 6th edit.

One common feature of these fresh-water shells, whether Neritinæ, Navicellæ, or Melaniæ, is the extensive erosion of the apices and surrounding parts of the shells. In some instances I have noticed that almost the entire exterior of the shell has been extensively eroded, particularly in the case of Neritina subsulcata, but I always found that the erosion was greatest in non-calcareous districts, where the free carbonic acid in the water is not all consumed in the solution of the limestone rocks. In volcanic islands the erosion of the fresh-water shells is greater than in islands of calcareous formation; and in streams, which, like those of the north coast of St. Christoval, flow in the upper portion of their course through a district of volcanic rocks and in the lower portion through a district of calcareous rocks, the same difference in the degree of erosion may be observed. I learn from a recent work by Professor Semper[482] that it is the boring of a minute fungus which first exposes the calcareous substance to the action of the carbonic acid, and that the mechanical action of the stream in forming tiny whirl-pools in the cavity probably assists in the erosion.

[482] “The Natural Conditions of Existence, etc:” London 1881: p. 212, circâ.

There are two common species of Neritina in these islands which I often confounded, viz., N. subsulcata and N. cornea; and I learn from Mr. Smith’s paper that these two species very closely approach each other. They, however, are usually to be found in different stations, N. cornea occurring on the trunks of palms and other trees away from the streams,[483] and N. subsulcata preferring the moist rocky sides of the streams a foot or so above the water.[484] Now and then they may be found encroaching on each other’s domain; for I have found them together on the trunks and branches of areca palms and tree-ferns in low lying moist districts, whilst, as at Choiseul Bay, I found them together in the streams.[485] Now it is a significant circumstance, that the specimen of N. cornea in my collection which was found by Mr. Smith to make the nearest approach to N. subsulcata was one which I obtained from a stream in Choiseul Bay. It had, in this case, not only intruded on the station of N. subsulcata, but had also assumed some of the distinctive characters of that species. It, therefore, seems to me probable that a graduated series of the shells of these two species might be formed, which would present the stages of transition from the one species to the other. If this be possible, then I would suggest that the fresh-water Nerite (Neritina subsulcata) may have been transformed into the tree Nerite (Neritina cornea) in the following manner.

[483] In St. Christoval I found this species on one occasion 150 feet above the nearest stream.

[484] This species often takes to the water. Some individuals that I kept alive on board used to spend a quarter of an hour at a time in the water eating voraciously all the while.

[485] According to Prof. Semper, these two species in the Philippines live a large portion of the year high up on the trees in mangrove swamps. (Ibid.)

I have already referred to the circumstance that in the higher portions of the St. Christoval streams, where the rocks are entirely volcanic, the fresh-water shells—and I may here add, especially those of Neritina subsulcata—suffer much more erosion than do shells of the same species in the lower parts of the streams where they flow through calcareous districts. Now, the geological structure of this island being mainly ancient volcanic rocks incrusted near the coast by recent calcareous formations, the time will come when these calcareous envelopes will have been entirely stripped off by denudation. How this will influence the Nerites of the streams may be thus explained. At present the normal characters of the species are preserved in the calcareous portions of the streams; but when all the calcareous rocks have been stripped off by denudation, the Nerite through its whole lifetime will be subjected to that extensive process of erosion, which now often denudes almost the entire surface of the shells of those individuals that live in the volcanic portion of the stream’s course. Here, Natural Selection may step in to favour the survival of any slight variation that makes the Nerite more suited to lead an entirely arboreal existence. Such a geological agency may in truth lead finally to the expulsion of the Nerite from the stream’s course. Varieties will survive only in proportion to their capability of adapting themselves to the new condition; and they alone will perpetuate their kind until a tree Nerite of distinct specific character is produced. . . . On this reasoning, tree Nerites ought to be more numerous in islands of volcanic formation; but this is a point on which I cannot pronounce from the lack of sufficient evidence.[486]

[486] Prof. Semper’s observations in the Philippines bear on this matter. (“Natural Conditions of Existence,” &c., p. 188.)

According to Professor Semper, we have in Navicella “a modified form of Neritina,” which genus it resembles in all essential anatomical characters, but “by long inurement to living in rushing mountain streams, it has had its shell modified in the way most suited to those conditions, while the operculum, in consequence of long disuse, has become a peculiar degenerate or rudimentary organ.”[487]

[487] Ibid, p. 212.

The growth of the fresh-water Nerites would appear to be slow. I kept a young individual of Neritina subsulcata for seven months in a bottle partly filled with rain-water, and supplied it with decaying leaves for food which it used to eat. Its weight was 37 grains both at the beginning and the end of the experiment, having only varied half a grain during the whole time; and its dimensions, as determined by measurement, were unaltered. This species, when it is first picked off the rock, ejects a watery fluid with a powerful musky odour, which effect accompanies the closure of the shell by the operculum. I kept some individuals of this species in rain-water, containing varying proportions of lime-water, for about three months. The lime-water was of the medicinal strength of the British Pharmacopeia. I began with water containing 64 parts of rain-water to one part of the lime-solution. By the end of the first month the proportion was increased to 32 to 1; by the end of the second month it was 16 to 1; towards the end of the third month the Nerites, having lived for over three weeks in the last solution, began to die; the survivors were placed in a solution containing the proportion of 8 to 1, but this amount of the lime-solution proved too much for them. It should be remarked that throughout the experiment, the Nerites used to descend to the water to get their food just as frequently as in the state of nature: they did not avoid the water; and after the experiment was over, there was no apparent alteration in the appearance of the shells. These observations were made in the north part of New Zealand during the latter part of the summer and the beginning of the autumn, a circumstance which may partially explain the death of the shells. The temperature there was about 20° below the temperature they are accustomed to in the Solomon Islands; this difference is of interest when it is remembered that Neritinæ are mostly found in the streams of tropical regions; and I may, therefore, infer that this species is capable of adapting itself to temperatures much lower than that to which it is accustomed, since some individuals survived the voyage to New Zealand from the Solomon Islands and lived in the climate of the former region for three months under very unfavourable conditions.

Professor Semper[488] remarks that some Neritinæ have the habit of detaching themselves from rocks on the slightest touch, by this means, as he considers, escaping the pursuit of their enemies. Some of them, however, as I observed, detach themselves spontaneously and independently of any alarm. The individuals of Neritina subsulcata that I kept in a large bottle in my cabin, used frequently in the course of a night to detach themselves from the sides and drop down into the water below. On one occasion when the noise woke me up, I found the culprit voraciously eating a portion of decayed leaf. In the daytime they sometimes dropped, and at other times crawled, down to the water. . . . . . It is probable that the musky water, which this Nerite ejects when it is picked off a rock, may cause a bird to drop it from its beak and thus save its life.

[488] Ibid, p. 210.

Amongst the new fresh-water shells that I found in this group was a species of Unio, to which Mr. Smith did me the honour of attaching my name, it being the first species of this genus of river-mussels that has been found in the Solomon Group. But its occurrence there means something more than a new locality, since, as I believe, I am correct in asserting, we have in it the first record of this widely distributed genus having reached the Pacific islands. I do not think that this species can be generally spread through the Solomon Group. I only found it in one locality, namely the Shortland Islands, near the western end of the group.

A very familiar shell, in low-lying moist and marshy situations throughout the Solomon Islands, is that of the auriculoid, Pythia scarabæus, Linné. Being usually accustomed to find it in the low-lying districts, I was surprised on one occasion to find it in the higher parts of Faro Island, which attains an elevation of 1,900 feet above sea. In the mangrove swamps and in the lower parts of the streams at Choiseul Bay, I found a species of Cyrena which has not yet been described, together with Cerithidea cornea (A. Adams: var.) and Pyrazus palustris, the last species occurring also in India. On the moist ground of the taro patches in the islands of Bougainville Straits thrives a species (S. simplex, var.) of that ubiquitous genus Succinea. The operculated land-snails, of which the Helicinæ are the most numerous, are found more frequently in calcareous districts.

LIST OF LAND AND FRESH-WATER SHELLS COLLECTED IN THE SOLOMON ISLANDS[489] DURING 1882 AND 1883. (EXTRACTED FROM MR. E. SMITH’S PAPER IN THE PROCEEDINGS OF THE ZOOLOGICAL SOCIETY OF LONDON, JUNE, 2ND, 1885.) THE DESCRIPTIONS OF THE NEW SPECIES AND VARIETIES FOLLOW:

[489] The habitats given are confined to the Solomon Group. I have added the new habitats of species in my collection to those previously ascertained from the collections of Brenchley, Macgillivray, Hombron and Jacquinot, etc.

(1) Helicarion planospira (Pfeiffer) Hab. Santa Anna, Ugi, St. Christoval, Guadalcanar.

(2) Helix (Nanina) nitidissima (nov. spec.) Hab. Treasury Island, a variety in Guadalcanar.

(3) Helix (Nanina) solidiuscula (nov. spec.) Hab. Santa Anna, found generally on the trunks of cocoa-nut palms.

(4) Helix (Corasia) tricolor (Pfeiffer) Hab. St. Christoval, Ugi, Santa Anna.

(5) Helix (Corasia) anadyomene, A (Adams & Angas) Hab. Guadalcanar, Ugi.

(6) Helix (Geotrochus) acmella (Pfeiffer) Hab. Faro Island, Bougainville Straits; Florida Islands, vide original paper.

(7) Helix (Geotrochus) gamelia (Angas) Hab. Isabel, Stephen Island, Shortland Islands, Treasury Island, Choiseul Bay.

(8) Helix (Geotrochus) hargreavesi (Angas) Hab. Faro Island in Bougainville Straits.

(9) Helix (Geotrochus) mendana (Angas) Hab. Shortland Islands, vide original paper.

(10) Helix (Geotrochus) motacilla (Pfeiffer) Hab. Simbo or Eddystone Island, also called Narovo.

Note.—In the original paper, Simbo and Eddystone are referred to as two different islands. This mistake arose from the omission of the name of Simbo in the latest charts; it is, however, the name usually employed.

(11) Helix (Geotrochus) guppyi (nov. spec.) Hab. Faro Island in Bougainville Straits.

(12) Helix (Geotrochus) dampieri (Angas) var. Hab. Choiseul Bay.

(13) Helix (Geotrochus) eros (Angas) Hab. Isabel, Stephen Island, Shortland Islands.

(14) Helix (Geotrochus) cleryi (Récluz) Hab. Santa Anna, Ugi, St. Christoval, Guadalcanar, Rua Sura Islets, New Georgia, Simbo or Eddystone, Treasury, Shortlands, Choiseul Bay. Three new varieties, var meridionalis (Santa Anna), var simboana (Simbo or Eddystone), var septentrionalis (Shortlands, Treasury, Choiseul Bay.)

(15) Helix (Videna) merziana (Pfeiffer) Hab. St. Christoval, New Georgia, Ugi; vide original paper.

(16) Helix (Videna) sanctæ annæ (nov. spec.) Hab. Santa Anna, living on the trunks of cocoa-nut palms.

Note.—This species, or a closely similar form, was observed by me in many other islands; but I neglected to collect it in any other locality than Santa Anna.

(17) Helix (Rhytida) villandrei (Gassies) Hab. St. Christoval, Ugi.

(18) Helix (Camæna) hombroni (Pfeiffer) Hab. Shortland and Faro Islands in Bougainville Straits, Isabel.

(19) Helix (Chloritis) eustoma (Pfeiffer) Hab. New Georgia, Ugi, Faro Island (Bougainville Straits.)

(20) Bulimus (Placostylus) cleryi (Petit) Hab. St. Christoval.

(21) Bulimus (Placostylus) founaki (Hombron and Jacquinot) Hab. Isabel, Faro Island (Bougainville Straits).

(22) Bulimus (Placostylus) miltocheilus (Reeve) Hab. St. Christoval (S.E. part), Ugi, Santa Anna: vide original paper.

(23) Partula, spec. Hab. Guadalcanar, Ugi, Treasury Island, Choiseul Bay: vide original paper.

(24) Succinea simplex (Pfeiffer) var. Hab. Treasury Island, Shortland Islands: living on the moist ground in the taro patches.

(25) Cyclostoma (Adelostoma) triste (Tapparone Canefri), var? Hab. Guadalcanar; Santa Anna; Faro and Shortland Islands and Choiseul Bay in Bougainville Straits.

(26) Leptopoma jacquinoti (Pfeiffer) Hab. Rua Sura Islets off the north coast of Guadalcanar: vide original paper.

(27) Leptopoma vitreum (Lesson) Hab. Santa Anna, Simbo or Eddystone; Shortland Islands.

(28) Omphalotropis nebulosa (Pease) Hab. St. Christoval, Guadalcanar, Ugi. I found this shell living on trees clothing a low tract of land skirting the beach.

(29) Pupina solomonensis (nov. spec.) Hab. Shortland and Treasury Islands in Bougainville Straits: living in the decayed trunks of fallen trees.

(30) Hargravesia polita, H. (Adams), var. Hab. Faro Island in Bougainville Straits: vide original paper.

(31) Helicina moquiniana (Récluz) Hab. St. Christoval, Guadalcanar, Ugi.

(32) Helicina egregia (Pfeiffer) Hab. Guadalcanar, Florida Islands.

(33) Helicina modesta (Pfeiffer) Hab. Guadalcanar, Shortland Islands, Treasury Islands, Choiseul Bay.

(34) Helicina solomonensis (nov. spec.) Hab. The Islands of Bougainville Straits (Faro, Shortlands, Treasury).

(35) Pythia scarabæus (Linné) Hab. Santa Anna, St. Christoval, which were the localities of my specimens; but I observed this species and its varieties in every island I visited. Stat: Moist ground usually near the sea.

(36) Melampus fasciatus (Deshayes) Hab. Isabel, Rua Sura Islets off the north coast of Guadalcanar. I found these shells in the crevices of a log on the beach.

(37) Melania amarula (Linné) Hab. Ugi, in the streams.

(38) Melania scabra (Müller) Hab. Ugi, in a stream.

(39) Melania salomonis (Brot.) Hab. Ugi, in a stream.

(40) Melania fulgurans (Hinds) Hab. Ugi, in a stream.

(41) Melania fastigiella (Reeve) Hab. Imbedded in a dark calcareous loam exposed in the banks of a large stream near its mouth at Sulagina, on the north coast of St. Christoval. I could not find any living specimens.

(42) Melania spec. Hab. The same as Mel. fastigiella.

(43) Melania verrucosa (Hinds) Hab. The same as Mel. fastigiella.

(44) Melania subgradata (nov. spec.) Hab. The same as Mel. fastigiella.

(45) Melania ugiensis (nov. spec.) Hab. From a stream in Ugi.

(46) Melania sanctæ annæ (nov. spec.) Hab. A stream in the interior of Santa Anna.

(47) Melania guppyi (nov. spec.) Hab. From the stomach and intestines of a fish living in the fresh-water lake of Wailava in the island of Santa Anna. This species probably frequents the deeper parts of the lake, as I only found one living individual.

(48) Cerithidea cornea A. (Adams) var. Hab. Mangrove swamps in Choiseul Bay.

(49) Pyrazus palustris Hab. Mangrove swamps in Choiseul Bay.

Note.—Not referred to in original paper.

(50) Nerita marmorata (Hombron and Jacquinot) Hab. Living just above high-water level on the surface of the coral-limestone coast, St. Christoval.

(51) Neritina cornea (Linné) Hab. Star Harbour, St. Christoval, on the trunks of trees 150 feet above the nearest stream; Choiseul Bay, from a stream; Shortland Islands, on the stems of tree-ferns and areca palms in a marshy district. I found this species in many other islands; but did not collect it except in the above three localities.

(52) Neritina subsulcata (Sowerby) Hab. Streams in St. Christoval and in the islands of Bougainville Straits. In the Shortland Islands I found this species on the stems of tree-ferns and areca palms in a marshy district. I only collected it in these localities; but I found it in many other islands. When first picked off the rock, it ejects a watery fluid possessing a powerful musky odour.

(53) Neritina dubia (Chemnitz) Hab. Shortland Islands, in a stream.

(54) Neritina adumbrata (Reeve) Hab. A stream in Choiseul Bay, and the rocky sides of a stream-course in Ugi.

(55) Neritina pulligera (Linné) Hab. Guadalcanar, St. Christoval, Ugi, Choiseul Bay, in the streams.

(56) Neritina petiti (Récluz) Hab. Treasury and Faro Islands, in the streams.

(57) Neritina olivacea (Le Guillou) Hab. Streams in Treasury Island and at Sulagina, St. Christoval.

(58) Neritina macgillivrayi (Reeve) Hab. Streams in Guadalcanar and in Faro Island, Bougainville Straits.

(59) Neritina asperulata (Récluz) Hab. Rocky sides of a stream-course in Ugi.

(60) Neritina porcata (Gould) Hab. St. Christoval, and Faro Island, Bougainville Straits, in streams.

(61) Neritina variegata (Lesson) Hab. St. Christoval, Ugi; Simbo or Eddystone; Faro Island, Bougainville Straits; Choiseul Bay; in streams.

(62) Neritina turtoni (Récluz) Hab. A stream in the Shortland Islands; streams in Guadalcanar; imbedded in a dark calcareous loam exposed in the banks of a large stream near to its mouth at Sulagina on the north coast of St. Christoval.

(63) Neritina brevispina (Lamarck) Hab. Streams in Shortland, Treasury, and Ugi Islands. The specimens from Treasury Island were destitute of spines.

(64) Neritina squarrosa (Récluz) Hab. Streams in Treasury Island.

(65) Navicella sanguisuga (Reeve) Hab. Streams in Faro Island, Bougainville Straits.

(66) Navicella suborbicularis (Sowerby) Hab. Guadalcanar, St. Christoval, Ugi, Treasury and Faro Islands in Bougainville Straits.

(67) Unio guppyi (nov. spec.) Hab. Streams in the Shortland Islands.

(68) Cyrena, spec. Hab. In the lower parts of streams and in mangrove swamps, Choiseul Bay. Note.—This species is not referred to in the original paper.

Descriptions of the new species and varieties,[490] by Mr. E. Smith.

[490] The numbers refer to the plates in Mr. Smith’s paper (Proc. Zool. Soc., June 1885).

(2) Helix (Nanina) nitidissima. (Plate XXXVI. figs. 1, 1 b.) Shell thin, transparent, very glossy, depressed, narrowly perforate, pale brownish horn-colour above, whitish towards the umbilicus, sculptured with very faint lines of growth. Whorls 4-5, slightly convex, impressed and marginate above at the suture; last whorl large, rounded at the periphery. Aperture obliquely lunate; peristome simple, thin, slightly thickened and reflexed partly over the perforation. Spire low, but very little raised above the last whorl, obtuse at the apex. Greatest diameter 14 millim., smallest 12; height 9.

(3) Helix (Nanina) solidiuscula. (Plate XXXVI., figs. 2, 2 b.) Shell very narrowly perforate, depressed, somewhat solid, dark chestnut-brown and a little glossy above, more shining and paler beneath, becoming almost white at the umbilical region; whorls 61⁄2, convex, separated by a deepish suture, and, with the exception of two or three at the apex which are smooth, sculptured with strong, close-set, arcuate, and oblique striæ on the upper surface, crossed with a few more or less distinct spiral lines. Body-whorl rounded at the periphery, or sometimes with the faintest indication of an angle, convex, and only exhibiting fine lines of growth below. Aperture obliquely semi-lunate; peristome simple, but, owing to the solidity of the shell, seeming slightly thickened, especially on the very oblique columellar margin, which is shortly reflexed above over the perforation. Spire depressed-conoid, having the least convex outlines and an obtuse apex. Greatest diameter 18 millim., smallest 161⁄2, height 12; aperture 8 long, 41⁄2 wide.

This species is well distinguished by its comparative solidity and strong sculpture on the upper surface.

(4) Helix (Corasia) tricolor (Pfeiffer). (Plate XXXVI., figs. 3, 3 b.) A specimen obtained on the north coast of the same island of St. Christoval, by Mr. Guppy, is worthy of special mention, and may be termed var. picta, on account of the undulating reddish-brown stripes which ornament both the upper and lower surfaces.

A similar example was also collected by Dr. A. Corrie and presented to the Museum. The markings on these two shells are very striking and distinctly visible within the aperture.

(11) Helix (Geotrochus) guppyi. (Plate XXXVI. fig. 4.) Shell elevately conical, thin, pale yellow, ornamented with conspicuous nearly black or black-brown spiral bands, one above and one below the sutures of the upper whorls, and three upon the last, one sutural, the second peripheral, and the third basal. Volutions 6, rather slowly enlarging, a little convex, sculptured with fine oblique striæ of growth, not glossy. Three first whorls livid purplish, the last rather sharply angled at the middle, not descending in front, having the basal band broad around the almost concealed perforation, and obsolete within the aperture. The latter is oblique, somewhat narrowed and pouting in front, banded within with three almost black and two white bands, the central one of the former being squarely truncate at the end, only the lower corner of it touching the margin of the lip, which is pale oblique, receding, a little expanded and reflexed in front and at the columellar margin, the upper end of which is spread over and nearly conceals the small umbilicus. Height 221⁄2 millim., greatest diameter 19, smallest 16.

This species is remarkable for the striking contrast of its colour-bands and the angular character of the last whorl.

(12) Helix (Geotrochus) dampieri, Angas, var. (Plate XXXVI. fig. 5.) Shell imperforate, subglobose, conoid, light brown or fawn-colour, here and there minutely dotted with dark-grey specks, with a broad white band around the middle of the penultimate whorl and two upon the last, one above and the other below the middle, also a narrow white line revolving up the spire beneath the suture, and a dark brown zone surrounding the pale or yellowish umbilical region. Whorls 5, a little convex above, somewhat glossy, obliquely and very finely striated by the lines of growth, the last more or less concentrically striated beneath, shortly descending at the aperture which is white within. Lip a little thickened, edged with reddish brown, only slightly expanded on the right side, more dilated below, produced into a thin transparent callosity over the umbilical region, united above to the upper extremity of the peristome. Columellar margin oblique, white or partly tinged with reddish brown, thickened and terminating below within the edge of the lip. Height 19 millim., greatest diam. 22, smallest 19.

The specimens from the Solomon Islands are smaller than the type with which, through the kindness of Mr G. F. Angas, I have compared them. They also have the peristome brown, and the basal band is darker.

(14) Helix (Geotrochus) cleryi, Récluz (Plate XXXVI. figs 6, 6 b). The specimens from Santa Anna (var. meridionalis fig. 6 b) are smaller than the type, pale brown above, with a white thread-like line at the suture, and the acutely keeled periphery, paler beneath, especially towards the centre, and have the aperture particularly acuminate at the termination of the keel.

The specimens from Simbo (var. simboana, fig. 6 a) are uniformly pale horn-colour, rather sharply carinate at the middle, and have the peristome white, considerably thickened and almost notched at the upper end of the columella, and the body-whorl is more contracted than in the typical form. The examples from Choiseul Bay, Shortland and Treasury Islands (var. septentrionalis, fig. 6) are all alike, of smaller dimensions than the normal form, thin pale brownish horn-colour, with rather more convex whorls than usual, the carina at the periphery being acute and thread-like as in the variety simboana.

Whether these several varieties should take specific rank is questionable, for, although there is considerable difference between the extreme forms even in the series of nearly one hundred specimens under examination, the gradual transition from one form to another is observable.

(16) Helix (Videna) sanctæ annæ. (Plate XXXVI. figs. 7. 7 b.) Shell depressed-conoid, deeply umbilicated, very acutely keeled at the periphery, light brown, sometimes with a few radiating pale streaks on the upper surface, sculptured with oblique lines of growth. Whorls 5, rather slowly increasing, slightly convex, depressed and margined above the suture, last not descending, compressed above and below the keel, a little convex towards the umbilicus, which is moderately large. Aperture transverse, flesh-tinted within. Peristome simple, a little thickened along the basal margin, with the extremities united by a thin callus. Height 7 millim.; greatest diameter 17, smallest 15.

(25) Cyclostoma (Adelostoma) triste, Tapparone Canefri, var.? Dr. Tapparone Canefri has kindly compared specimens from these islands with his C triste, and is of opinion that they may be considered a variety of it; and observes that the New-Guinean form is a little smaller, its spire a little more slender, its surface more glossy, the colour redder, and the apex of the spire darker.

The shells under examination are clothed with a very thin epidermis when in a fresh condition, exhibiting numerous very fine spiral thread-like lines, which entirely disappear in worn shells and can easily be rubbed off with a brush. For several species having a similar epidermis and an incomplete peristome, Dr. Tapparone Canefri has proposed the subgenus Adelostoma.

(29) Pupina solomonensis.(Plate XXXVI. fig. 9, 9 a.) Shell small and very like P. difficilis, Semper, and P. keraudreni, Vignard. It is of a reddish tint, especially the body-whorl; consists of 51⁄2 whorls, which are the least convex and exhibit a pellucid line, frequently brown, immediately beneath the suture. Last whorl very obliquely descending behind, narrowed below, and flattened somewhat above the aperture. Columella thickened with callus, white, parted off from the whorl above by an oblique circumscribing red line, truncated rather low down. Outer lip slightly thickened and effuse, and a little paler than the rest of the whorl, produced somewhat at its junction with the body-whorl, which in consequence has the appearance of rising suddenly after an oblique descent. Length 7 millims., diam. 32⁄3, aperture 2 long and wide.

(34) Helicina solomonensis. (Plate XXXVI. figs. 11, ll b.) Shell small, globose-conical, reddish or yellowish, pale at the apex. Whorls 4-41⁄2, the least convex above, sculptured with lines of growth and fine spiral striæ both on the upper and lower surfaces, very faintly margined above at the suture; last whorl rounded at the periphery, obsoletely angled near the junction of the outer lip and the least descending in front, so that the faint angulation is visible for a short distance above the sutural line. Aperture somewhat semicircular and oblique, small; peristome slightly expanded; umbilical callosity yellowish or pellucid whitish, defined towards the base of the columellar margin. Greatest width 42⁄3 millim., smallest 4; height 31⁄2.

(44) Melania subgradata.(Plate XXXVII. fig. 3, 3 a). Shell elongate, turreted, rather solid, covered with an (olive?) epidermis, and marked with fine longitudinal oblique red lines which extend from suture to suture. Whorls probably about 10, flat or even a little concave at the sides, shouldered above, usually with a spiral shallow groove and a few striæ near the shoulder, and marked with fine incremental striæ. Suture deep, slightly oblique. Last whorl long, finely transversely striated, most distinctly at the base. Aperture elongate-pyriform, acute above, effuse at the base. Outer lip thin, sharp, accurate, and prominent at the middle. Columellar margin rather thickly covered with callus, united above to the outer lip. Length of two specimens, consisting of five whorls 30 and 25 millim.; diameter 11 and 101⁄2; aperture 14 and 12 long, 6 and 5 wide.

(45) Melania ugiensis. (Plate XXXVII. fig. 4.) Shell subulate, acuminate, beneath the epidermis (which is wanting in the specimens at hand), of a dirty, pale, livid, or purplish tint. Whorls probably about 14 in number; the eleven remaining are a little convex, rather slowly enlarging, and sculptured with close-set, obliquish, fine riblets, which are crossed by crowded spiral striæ. Last whorl large, with the riblets rather obsolete below the middle, and very close together, much more numerous than those upon the upper whorls. Aperture obliquely pear-shaped. Length 25 millim., diameter 8; aperture 81⁄2 long, 41⁄2 wide.

(46) Melania sanctæ annæ. (Plate XXXVII. figs. 5, 5 a.) Shell small, acuminately pyramidal, somewhat eroded towards the apex, covered with a yellowish-olive epidermis, and sometimes marked with a few indistinct, reddish, irregular spots, and lines near the middle of the body-whorl. Whorls 5-6 remaining, flattish at the sides, divided by a slightly, oblique, distinct suture, all with the exception of the last one or two more or less distinctly, longitudinally, finely plicate; the plicæ are more conspicuous in some specimens than in others, being at times entirely eroded. The other sculpture consists of fine lines of growth, and a few rather distant spiral striæ, which cut across the incremental lines and produce a puckered appearance. Aperture elongate, pyriform, pale bluish within. Length of specimen consisting of six whorls 13 millim., diameter 5; aperture 5 long and 21⁄2 wide.

(47) Melania guppyi. (Plate XXXVII. figs. 6, 6 a.) Shell slenderly acuminate, covered with an olive-brown epidermis. Whorls about 14, divided by a very oblique, deepish suture, concave above the middle and somewhat convex below it, and then contracted; ornamented with a few spiral series of nodules (about five on the upper whorls) and rather indistinct, very oblique and flexuous, longitudinal ridges, upon which the nodules rests, also exhibiting very sloping and flexuous lines of growth; the most conspicuous rows of granules are near the middle of the whorls. Aperture pyriform. Outer lip thin, remarkably sinuated above towards the suture, and arcuately prominent below. Columellar margin oblique, straightish, covered with a callus, curving into the broad basal sinus. Length 31 millim., diameter 7; aperture 9 long, 4 wide.

This is a very remarkable and distinct species, with a very drawn-out spire, peculiar granuled sculpture, and a deeply sinuated labrum. I have much pleasure in naming it after Mr. Guppy.

(66) Unio guppyi. (Plate XXXVII. figs. 8-8 b.) Shell elongate, very inequilateral, usually a little longer than twice the height, compressed, covered with a blackish-brown epidermis, exhibiting strong lines of growth, and very faint radiating substriation, and marked with fine wrinklings at the eroded beaks, which are small and placed quite near the anterior extremity. Dorsal margin behind the umbones almost straight or the least excurved for some distance, then at an obtuse angle becoming oblique before rounding into the extremity, which is a little more sharply curved than the anterior end. Ventral outline either faintly excurved, straight, or the least concave. Interior bluish-white, most iridescent at the hinder extremity, generally stained in parts with olive-brown. Cardinal tooth of the right valve moderately large, four or five-lobed at the top, situated just in front of umbo. Between it and the outer margin is a short ridge, the space between the tooth and the ridge receiving the single, smaller, roughened, and striated tooth of the left valve. Lateral tooth of the right valve long, obliquely truncate behind, fitting in between two teeth in the opposite valve. Anterior adductor scar deep, posterior superficial, squarish in front. Pedal scar in both valves under the cardinal tooth very deep. Ligament elongate, prominent.

Length80mm.;height38;diameter21.
„70mm.;„35;„18.

This species recalls to mind some of the forms from Australia and New Zealand. Its principal features are the elongate compressed form, dark brown colour, wrinkled apices, and coarse incremental lines. It is the only species as yet recorded from the Solomon Islands.

ADDITIONAL NOTE. (H. B. GUPPY.)

A species of the Litoritinidæ (Littorina scabra) is commonly found in this group on the leaves and trunks of mangroves, Barringtonias, and other littoral trees, the branches of which overspread the rising tide. These molluscs occur at heights varying from one or two feet to eight or nine feet above the high-water level; and they possess an unusually delicate operculum as compared with those of other species of the same family. They do not seem to be able to withstand immersion in salt water for any length of time, since out of six individuals kept submerged for twenty-four hours, three died. When first placed in the water, they were evidently very much out of their element, and tried in vain to creep out of the vessel. The delicate character of the operculum indicates a transitional stage between marine and terrestrial molluscs; and the experiment above referred to, throws a little light on this subject, since only the younger of the six individuals survived. One would have expected that the younger individuals would have been less able to withstand immersion in sea-water, but such was not the case, since they recovered from an immersion which killed the older individuals. In explanation of this unexpected result, I would infer that, on the theory of the inheritance of peculiarities at corresponding ages, the younger individuals would retain more of the marine habits of the original parent of the species, because in the first place only the adults of this parent species would have been modified to suit the new condition.

CHAPTER XVII.
THE CLIMATE OF THE SOLOMON ISLANDS.

Amongst the matters to which I devoted some attention in this group of islands, was the annual rainfall. As far as I know, there have been no continuous observations previously made there; and the only record of rain-measurement, which I have been able to find referring to this region, was an observation made on board the Austrian frigate “Novara” in the middle of October, 1858, whilst to the northward of St. Christoval, when three inches of rain were registered in five hours.[491] I therefore set myself to work to do what I could in this matter, making rain-gauge stations at Santa Anna and Ugi and keeping a register myself on board. Mr. Fred Howard undertook to make these observations at Ugi, and I supplied him with a rain-gauge for this purpose. His register, which extended over a period of fifteen months from October, 1882, to the end of the following year, was kept with great regularity; and as I was able to compare his observations with my own on board for a few days, I have every confidence in the accuracy of his observations. At Santa Anna, Mr. William Henghan, to whom I had supplied a gauge, undertook at first to keep the record, beginning in the last week of October, 1882; but he left the island two months after, when Mr. Charles Sproul voluntarily undertook to measure the rainfall, which he did with great regularity until the end of the following year. I regret to learn that Mr. Sproul has recently died at Sydney. He was one of those men who in a quiet inoffensive way have done much towards preparing the way for future settlers in this group. I have the greatest confidence in his observations, since for a few days at Santa Anna we were able to compare our daily measurements.

[491] Scherzer’s “Voyage of the ‘Novara,’” Eng. edit., 1861.

Before proceeding to consider the results of these rain-measurements, I will endeavour to convey to the mental eye of my readers a general idea of the most striking atmospheric phenomenon in connection with the rainfall of these regions. I refer to the oncoming of the black squall.

A clear and serene sky at first gives no token of the sudden change that is to quickly follow; but the stillness of the air and its increased dryness, together with the consequent greater scorching power of the sun’s rays and the apparent nearness of surrounding shores, give sufficient warning of the onset of the rain squall to those acquainted with these seas. In a short time a low black arch appears above the horizon, often in an unexpected quarter, and rising rapidly it sweeps majestically with great swiftness until it appears to span the heavens. Onward it rushes, quicker far than one imagines; and now must the navigator beware. Under yonder advancing arch a white line of foam marks its van. There, away towards one of its corners, a waterspout rises in fantastic shape; sea and cloud meet in mid-air and become intermingled in the whirling column. Lightning plays about beneath the arch and within its black mass, illuming for the moment its dark recesses and leaving it in the next far blacker than before. Peals of thunder herald on the advance of the black squall.

“Clear lower deck!” “Hands aloft!” “Shorten sail!” Such were the words of command which were almost daily issued during our cruises in these islands. In a few brief minutes, the ship is prepared to meet the squall. The temperature falls very perceptibly, and the officer of the watch gives a slight shiver as he dons his oilskins. The wind is freshening, a few large drops of rain fall, the men crouch under the bulwarks, and now the arch is overhead and we are in the thick of the squall. Down comes a deluge of rain which in less than a minute wets all who are unprotected through and through. The ship heels well over, even with her scanty canvas. There is nothing more to be done. We listen to the whistling of the wind in the rigging and patiently wait until the weather clears. In half-an-hour the arch has swept over us, and is pursuing its rapid course towards the neighbouring mountain-peaks, perhaps of Bougainville or it may be of Guadalcanar. The blue sky begins to show itself; and in less than an hour all is as before. With reefs shaken out and more sail made, the ship proceeds, plunging cheerily on under a fresh breeze as though glad to shake herself clear of the squall. The sea losing its murky colour reflects the bright hue of the sky now serene; and its white-topped waves sparkle in the sun. The wizard of the storm has shaken his wand, and the scene is changed, as though by magic.

All nature seems invigorated by this short battle of the elements and to be indebted to the bounty of the black squall. Whilst everything before was depressed and lowering, all is now bright and cheerful. Nature has in truth had its accustomed shower-bath, and the reaction that ensues does good to all; makes men the happier and the stronger, elicits a loud chorus from the lower creation in which bird, reptile, and insect, before hushed in the depressing gloom, now combine in strange medley; and the inanimate world shares in the bright change which has followed the storm.

If it be night, the increased luminosity of the sea may be the warning of the arched squall. The ship throws off a bright wave of phosphorescence on either side of the bow, and leaves a luminous track in her wake. Overhead the cloudless star-lit sky conveys its warning; for the stars shine with increased brilliancy, those of less magnitude usually invisible with the naked eye are now distinctly seen; and if the navigator, who has often tried in vain to count the six stars in the Pleiades, can do so now, let him look out for the black squall. Such are the warnings. Then sweeps along the lowering arched mass with its rain and its waterspouts, its wind and its thunder and lightning. On it comes, looking all the blacker as it spreads athwart the heavens and turns the star-lit night into a lightless gloom. Overtaken in the night by such a squall, unable to see more than half a cable’s length on either side, and perhaps in the vicinity of sunken reefs the position of which is uncertain, a sailor has need of all his wits. On one occasion, when in this situation, we came unexpectedly in soundings, whilst, as we thought, a hundred good fathoms and more lay beneath our keel. The time was anxious, but nothing could be done until the squall was over. When the arch has passed, the stars begin to show themselves, and in a short time they shine out with all their lustre.

With this description of the rain-squall, or black-squall, or arched-squall, as it may be also conveniently termed, I return to the consideration of the rainfall of this region; and first with regard to the observations at the east end of the Solomon Group. During 1883, 125·03 inches of rain were measured at Santa Anna, a small island lying at the extreme eastern limit of these islands. Two-thirds of the total amount fell in the live months between the beginning of April and the end of August. At Ugi, which lies nearly 60 miles north-east of Santa Anna, 146·24 inches of rain were registered during the same year. About one-third of the total rain for the year fell in the two months of April and July. On comparing the totals for each month at these two localities, there will be found to be but little agreement, which is due to the circumstance that the daily rainfalls of these two places have little relation one with the other, a heavy fall at one island being often only indicated by a slight fall of rain at the other. It is thus evident that locality has a great influence on the rainfall in this part of the group; and probably Ugi owes its greater rainfall to the proximity of the high land of St. Christoval. Here, as in other parts of this group, I often had opportunities of observing how the contiguity of land affected the rainfall in a single shower. I might have been in the interior of an island exposed to a deluge of rain for a couple of hours, and have found, as I did once in the Shortland Islands, that there had been very little rain on board. Another time, when in my Rob Roy canoe on the south side of Treasury harbour and not more than a mile from the ship, a rain-squall passed over me leaving scarcely a drop behind; but as it swept over the ship and was approaching the steep slopes of the island, a smart shower of 20⁄100 of an inch fell on the deck.

I cannot gather from the observations made in this eastern part of the group, that one season of the year has a heavier rainfall than another. On comparing the two records for 1883 of Ugi and Santa Anna, it might be thought that the closing months of the year would usually prove to be the driest; but on referring to the register kept on board the ship in this locality in the latter part of 1882 ([page 365]), which is one of the heaviest records we had in the Solomon Group, such an inference would be negatived. Nor do I find from these registers of rainfall that there appears to be any relation between the amount of rain and the prevalence or non-prevalence of the south-easterly trade, which usually becomes well established in May and lasts till the end of November or the beginning of December, when the north-westerly and westerly winds set in. These observations point towards the inference, therefore, that the distribution of rain through the seasons in this part of the group is capricious; and they do not warrant the conclusion that one season is wetter than another.

Perhaps a comparison of the number of rainy days, or days on which not less than 2⁄100 of an inch of rain were measured, may help us to form a more definite conclusion. It will be seen that at Santa Anna and Ugi there were much the same number of rainy days, 182 in the former island and 178 in the latter, or in round numbers about half the total number of days in the year were rainy.[492] At Santa Anna, during the prevalence of the trade wind, there were on the average 15 rainy days per month, and at Ugi 13 per month; whilst during the months from December to April inclusive, when westerly and variable winds prevailed, there were 18 rainy days per month at Santa Anna, and 19 per month at Ugi; so that we may infer that in this year of 1883 there were fewer rainy days per month during the prevalence of the south-east trade, i.e., from May to November, than during the period of westerly and variable winds, i.e., from December to April.

[492] From the record of the rainy days during the six months from June to November of the previous year (1882), it appears that at least 110 days were rainy. During the same months of the following year, only 84 days were rainy.

I come now to the subject of the greatest daily fall of rain in this eastern end of the group. On the 13th of June, 1883, 7·73 inches were registered at Santa Anna; whilst at Ugi on the same day only an inch and a half of rain fell, a circumstance showing how confined in their areas some heavy rainfalls may be. At Ugi the heaviest daily fall of 5·75 inches was recorded on the 28th of January of this same year; whilst at Santa Anna only a little more than two inches fell on this day; and here is another proof of the restricted locality of heavy rainfalls. On the 20th of November, 1882, when H.M.S. “Lark” was off the east end of St. Christoval, 5·74 inches of rain fell on the ship; whilst only a small amount of rain was measured at Santa Anna and Ugi. . . . . . With reference to the character of the rain in this part of the Solomon Group, I may remark that as in other tropical regions it is very heavy. A fall of an inch in an hour is very frequent during a rain-squall; but not uncommonly the rain falls far more heavily. Thus, on one occasion on board H.M.S. “Lark,” when in this part of the group, 2·90 inches fell in an hour; and at another time 1·03 inches fell in 25 minutes, and on another occasion an inch fell in half-an-hour.

But inasmuch as heavy falls of rain are not peculiar to the tropics, since far greater falls than those above named have occurred in temperate Europe, we can only judge of the character of the rainfall in this region by the total annual fall and by the frequency of heavy falls. Thus we find that at Ugi, in 1883, on 56 days the fall exceeded an inch; and that at Santa Anna, more than an inch of rain fell on 41 days. At Ugi, the daily records on eighteen occasions exceeded two inches; at Santa there was a lesser number of falls of over two inches, viz. 11.

If I were to estimate the probable annual rainfall at the coast in this part of the Solomon Group, I should place it at not far under 150 inches. Although only possessing the rain-register for a small portion of 1882, I am of the opinion, from having spent a large part of the year in this eastern end of the group, that the fall for 1882 was heavier than the rainfall actually registered for 1883;[493] although this is but a conjecture, it enables me to estimate the probable annual fall with some confidence at about 150 inches at the coast in this eastern end of the group.

[493] Vide footnote referring to number of rainy days in 1882 on p. 356.

The observations made on board the ship amongst the islands of Bougainville Straits (Treasury, Shortlands, Faro, etc.) during portions of the year 1883 and 1884 now claim our notice. As shown on [page 365], 60·43 inches of rain fell in the five months from June to October of 1883, this amount being a little under that which fell at Ugi (65·70 inches) and at Santa Anna (67·72 inches) in the same period, the two regions lying towards the opposite ends of the group. During the same period of the following year, we measured 67·66 inches of rain in Bougainville Straits, an amount a little in excess of that of the previous year. During the same periods, i.e., from June to October inclusive, in 1883 and 1884, there were the following number of rainy days, 120 in the one year and 118 in the other. At Santa Anna and Ugi, at the opposite end of the group, the total of rainy days for the same period in 1883, numbered only two-thirds of the amount in Bougainville Straits. During these five months in 1883 there were 16 daily records of over an inch of rain in Bougainville Straits; at Santa Anna and Ugi, in the same period, there were 23 and 26 daily records exceeding an inch. In the same period of 1884, in Bougainville Straits, there were 22 such daily records, but the total fall was about 7 inches greater than in the previous year.

I may now draw some inferences from the above observations. In the first place, it is probable that the annual coast rainfall of Bougainville Straits and that of the eastern end of the Solomon Group are much about the same, viz., about 150 inches: the chief difference between the two regions being, that in the former region, there are a greater number of rainy days and fewer heavy falls. The heavy falls, when they do occur, are not easily forgotten; thus, at Treasury we measured, in July, 1884, 11 inches of rain in 10 successive hours; but the daily record was only 8·09 inches, since the rain began in the evening of one day and lasted well into the following morning.

During the heavy rainfalls in these regions the streams swell in an astonishingly quick manner. Rivulets become turbid streams, the whole hill-slope discharges a continuous sheet of water, and the water rushes down the permanent stream-courses with the roar of a mountain-torrent. Large blocks of stone are swept some distance along the lower courses of the streams; and the trunks of trees are carried by each successive flood further and further towards the mouth of the stream.

It should be now remarked that the average rainfall for the year, which I have estimated from observations made in different parts of the Solomon Group at about 150 inches, only applies to the coast. It is probable that this estimate is generally applicable to the coasts of these islands, except on the lee sides of the loftier islands.[494]

[494] By the lee sides, I mean those sheltered from the prevailing S.E. trade.

This brings me to the question of the rainfall in the higher regions. The rainfall will increase with elevation until a certain height is reached, where the clouds attain their maximum density; at such a level the greatest rainfall will occur. I learn from an interesting paper by Mr. Bateman on this subject,[495] that it may be inferred that in the Lake District of England the greatest rainfall occurs at an elevation of 2,000 feet, which is the level of maximum cloud density. In India, an elevation of 4,500 feet represents the level at which the greatest rainfall occurs. In the Solomon Islands, a greater height will have to be attained before the level of maximum cloud density or that of the greatest rainfall will be attained. Probably I shall not greatly err if I assume it to be between 5,000 and 6,000 feet. I have already observed that the south-east trade, subject to its usual variations, is the prevailing wind in the eastern part of the group for nearly two-thirds of the year. Coming laden with its watery burden, it first strikes the eastern slopes of St. Christoval; but although the higher regions of this island must cause the rain-clouds to precipitate a large amount of their moisture, the higher peaks do not rise in sufficient mass to a height that would receive the greatest rainfall, the extreme height being 4,100 feet. The rain-clouds, with the bulk of their moisture, would therefore be driven over the higher regions of this island, and would deposit the greater part of their burden on the higher slopes of the mountainous eastern portion of Guadalcanar. Since this island, in its eastern portion, rises in mass to a height of some 5,000 feet and attains a maximum elevation of 8,000 feet, it does not seem probable that, during the prevalence of the trade for nearly two-thirds of the year, a considerable quantity of rain would be deposited on the western side of the island; and, that such is the case, is shown in the fact that the dense forest-growth that clothes the steep eastern and southern slopes of the island gives place, on the lee or west side of the mountains, to a vegetation which gives to the western portion of Guadalcanar, when viewed from seaward, the appearance of a savannah or a prairie.

[495] Journal of the Victoria Institute. Vol. XV. No. 59.

The lofty mountain-masses of the east end of Guadalcanar, which forms one of the finest specimens of coast-scenery in the world, are usually enveloped in rain-clouds at their summits. But occasionally one of the peaks is visible above the thick cloud-covering, marking by its elevation, as it were, the line of greatest rainfall lying below. In the same manner the high peaks at the east end of Bougainville, which have an elevation of between 7,000 and 8,000 feet, may be seen occasionally to project above the rain-clouds; but there is, probably, a smaller quantity of rain deposited on the higher slopes of this island than on those of Guadalcanar, because the mountains are more isolated, possess for the most part the tapering volcanic profile, and do not rise “en masse,” as in the case of the high lands of Guadalcanar. The greatest rainfall in the Solomon Group takes place on the steep southern and eastern slopes of this island of Guadalcanar. Huge mountain-masses appear to rise directly from the sea to a height of some 5,000 feet, ultimately attaining a height of 8,000 feet. The fall there must be tremendous, especially when, as is frequently the case, the land of St. Christoval does not interpose itself in the path of the moisture-laden trade-wind. Then, loaded with vapours after its passage across a wide expanse of ocean, and with but a thin tract of intervening lowland to rob it of its moisture, the trade strikes at once upon the precipitous mountain-slopes as against some Cyclopean rampart. There is no ravine or breach in the mountain-mass to ease the tension. There, on those mountain-slopes, a terrific precipitation must occur, which, if the annual rainfall of the coast is 150 inches, will here be three or four times that amount. This is no exaggerated language, but is the opinion I have formed, after having carefully considered the physical geography of these regions.

The subjoined rainfalls of a few localities in other parts of the world may be interesting to compare with that of the Solomon Group:[496]

[496] Somerville’s “Physical Geography,” 7th edit. pp. 331-334.

England32inches.
Singapore97„
Atlantic Doldrums225„
Western Ghats302„
Cherraponjee10„
Solomon Islands.
(a.) at the coast150„
(b.) on the higher slopes of Guadalcanar400to 500 inches probably.

Comparing the rainfall of the Solomon Islands with some results obtained in other parts of the Pacific, I would draw attention to the small rainfall of Port Moresby on the south-east coast of New Guinea, where 34·44 inches were registered at the Mission Station in 1875.[497] In Fiji the rainfall appears to vary between 60 and 250 inches per annum, according to the degree of elevation above the sea, and to the position of the station on the lee or weather sides of the islands, the greatest annual falls occurring in the interior of the large islands.[498] In Oahu, one of the Sandwich Islands, during 1873, the rainfall at the coast was 37·85 inches; whilst at a distance of 23⁄4 miles in the interior, it was 134·06 inches, the elevation being only 550 feet above the sea.[499]

[497] Stone’s “A Few Months in New Guinea,” p. 143.

[498] Rain-gaugers have been numerous in this colony, and their list would extend beyond the limits of a foot-note. (Vide Home’s “Year in Fiji,” &c., &c.)

[499] Mosely’s “Naturalist on the ‘Challenger,’” p. 497.

I will now make a few remarks on the barometric pressure, temperature, and other features of the meteorology of this group. They are based on the results of the observations made by Lieutenant Leeper on board the ship, and by Mr. F. Howard at Ugi. (Tables appended.)

As is usual in these regions of the Pacific, the fluctuation of the barometer, whether daily, yearly, or monthly, is very small. Thus, the range during the 22 months we passed in the group, was from 29·83 to 30·18 inches, or about a third of an inch; whilst the average monthly range was rather under a quarter of an inch, and the usual diurnal variation about ·04 of an inch.

Whilst endeavouring to compare the temperatures of the different seasons, I have mainly used the Ugi register, since it gives a continuous record for more than a year. At Ugi in 1883, the portion of the year from June to September inclusive was slightly the coolest, but the difference in the means was not 2°; and, in truth, taking all the thermometric observations into consideration, the seasons are scarcely distinguished by their temperatures. As Lieutenant Leeper[500] remarks in his report, the temperature varies but little all the year round, the monthly mean varying between 80° and 85°. The annual mean temperature may be placed at 82° to 83°, and the range from 75° to 95°. The daily variation is considerably affected by the exposed or protected position of any locality at the coast. Judging, however, from the data at my disposal, it is usually less than 10°, e.g., 79° at night, and 88° at mid-day.

[500] Vide Quart. Journ. Roy. Met. Soc. vol. XI., p. 309. The instruments used on board were previously verified at Kew. From want of leisure, Lieut. Leeper was unable to do much more than tabulate his observations. I have therefore extracted from them such general facts and inferences as they sustain.

From the hygrometrical observations, it may be inferred that the climate of these islands is generally very moist. The relative humidity, taking 100 as saturation, ranged at Ugi in 1883 from 54 to 100; but the monthly range was usually from 72 to 95, the mean for the year being 83.[501] This mean degree of relative humidity is much greater than that of Levuka in Fiji which would seem to average about 70;[502] but in truth there is little necessity for me to remark further on this well-known feature of the climate of these islands. Yet, I should add that this proportion of aqueous vapour would not necessarily be oppressive in a temperate latitude. In a tropical climate, however, any influence that retards the evaporation from the skin of the normally excessive perspiration, is a cause of personal discomfort, such as would not be experienced in a drier locality lying in the same latitude. The effects of this combination of heat and moisture are to be seen in the rankness of the vegetation, and in the rapid rusting of steel. Although the foregoing remarks may be taken as generally applicable to the group, it should be stated that on the lee side of a mountainous island, such as the western end of Guadalcanar, there is a comparatively dry atmosphere, and the difference is also shown in the character of the vegetation.

[501] There are no observations for January, but since the mean relative humidity varies with the rainfall, I have approximately estimated that for January to be 83.

[502] Lieut. Lake’s observations for 1876 and 1877. (Quart. Journ. Met. Soc.)

The moderate intensity of the sun’s rays in these islands is to be ascribed to the presence of aqueous vapour in quantity in the atmosphere. When, however, a thunderstorm and its accompanying rain-squall are portending, the air is unusually dry, and the sun’s rays are very fierce. At such times it often happens that the sky is overcast; and thus it comes about that the unwary traveller, by rashly baring his legs and arms, suffers severe sun-burns when he least expects it. Waterton and other travellers have, through ignorance of this fact, been laid up for several days, and even weeks. I was unable to walk any distance for about ten days, after experiencing a severe sun-burn of the legs as the result of baring them during an overcast day. The affection is peculiarly painful, though it often excites but little sympathy.

My remarks on the meteorology of this group will not be complete without a short reference to the prevailing winds. The South-east Trade Wind and the North-west Monsoon carry on a continual struggle for the mastery in these islands. However, for two-thirds of the year the Trade prevails, viz., from April to November. The appended record of winds, which extends over a considerable period, I have prepared from the observations made on board H.M.S. “Lark” in different parts of the group, and from the registers kept by Mr. Sproul and Mr. Howard at Santa Anna and Ugi. It will be there seen that at the eastern extremity of these islands, viz., in the vicinity of St. Christoval, the Trade announces its onset in April by unsettled weather, and frequent thunderstorms. In May, it becomes established, but, as Lieutenant Leeper remarks, it blows in fits and starts, is interrupted by calms, variable winds, and heavy rain-squalls, and does not blow home as in Fiji and the groups to the eastward. At the opposite end of the group, in Bougainville Straits, the Trade appears a month later, and does not become established until June. In this locality, however, it is more fitful than in the eastern islands, blows lighter, and is less to be depended on by the navigator.

It may be generally stated that the north-west and west winds set in about the end of November or the beginning of December, and prevail until the end of March. Although heavy gales accompany the frequent shiftings of the wind, especially when it is from the S.W., these islands are beyond the sweep of the hurricanes which in this season of the year occasionally devastate the groups to the eastward. The period of the westerly winds in the Solomon Islands is also characterised by calms and variable winds. The exhilarating freshness of the Trade then gives place to the enervating influence of the Monsoon; and, in consequence, the period of westerly winds is the sickly season.