Table of Altitudes, deduced from Barometric Observations in Marocco.

Date, 1871HourPlace of ObservationObserved BarometerCorrected BarometerCorrected Mogador BarometerThermometer in Air, Fahr.Altitude in MetresAltitude in English feetObservations
April1110 P.M.Nahum’s house, Tetuan, second floor754·4752·6—6184·8278Assumed pressure at sea level 760 mm.
„126 A.M.Do.756·2*753·8—5570·0230⎧
⎪
⎨
⎪
⎩		* As the river near its mouthcan be only about 3 m. above the sea, a correction of -0·6 mm. isinferred for April 12.
„„7 A.M.Ford, Tetuan River762*759·6—604·7816
„„10 A.M.Upper limit of Chamærops730*727·6—61374·01.227
„„3 P.M.Ridge of Beni Hosmar685·2*682·8—60926·03.038
„„10 A.M.Douar Arifi748·5†754·5—7064·5212† See note as to corrections of Secrétan’sinstrument after arrival at Mogador. Assumed special correction forApril 29 +·05 mm.
„309.15 A.M.Souk el Tleta724729·6—76360·31.182Assumed pressure at sea level 760 mm.
May15 A.M.Camp, Aïn Oumast724729·6—54345·51.134Do.
„„1 P.M.Well under Hank el Gemmel720725·6—77410·11.345Do.
„„2 P.M.Summit of Hank el Gemmel712·6718·2—78502·41.648Do.
„26.30 A.M.Camp, Sheshaoua724729·6—57347·81.141Do.
„„1.45 P.M.Aïn Beida720725·6—80412·41.353Do.
„42 P.M.Marocco: Palace of Ben Dreïs, 40 ft. abovePiazza712·5718·1—78503·61.652Do.
„54 P.M.Do.713718·6—79499·31.638Do.
„67 P.M.Do.710·5716·1—73523·71.718Do.
„77 A.M.Do.710715·6—72529·01.736Assumed pressure at sea level 760 mm.
„85 A.M.Do.712717·6—72504·01.654Do.
Do.————511·91.679Mean of five observations.
„„11 P.M.Camp at Mesfioua694·0699·6—73733·22.406Assumed pressure at sea level 760 mm.
„96 A.M.Do.693·5699·1—587292.392Do.
„„NoonOlive Grove below Kaïd’s house, Tasseremout667·5673·2—691,077·13.534Do.
„„10 P.M.Camp by Ourika river, below village ofAchliz681·0686·7—59874·2—Do.
„11NoonDo.681·5687·2—71887·0—Do.
Do.————880·62.889Mean of two observations.
„104 P.M.Camp, Ourika valley (Assghin)669·5675·2—721,044·43.427Assumed pressure at sea level 760 mm.
„116 P.M.Summit of pass to Reraya664·0670·05759·762§1,094·33.59⎰
⎱		§ Comparison with Mogadorobservations reduced to sea level.
„126 A.M.Camp Tassilunt, Reraya674·0679·35760·456§963·13.16
„137 A.M.Camp Hasni, in Aït Mesan valley652·0656·5761·4557§1,263·3—Do.
„1410 P.M.Do651656·04760·959§1,274·0—Do.
„159 A.M.Do647653·77758·856§1,297·4—Do.
„1710 P.M.Do651654·83762·260§1,292·6—Do.
Do————1,281·84.205Mean of four observations.
„13NoonAdjersiman, village in Aït Mesan valley622626762·159§1,6875.535§ Comparison with Mogador.
„„2 P.M.Village of Arround602606·16762·152§1,950—Do.
„156 P.M.House, Arround597605·5756·649§1,947·5—Do.
„166 A.M.Do.598606·5756·646§1,926·4—Do.
Do.————1,941·36.37Mean of three observations.
„132 P.M.Arround, by comparison with Hasni————1.968⎱
⎰		6,463Differences by St. Robert’smethod 686·2 m. and 690·2 m. respectively.
„156 P.M.Do.————1.972
„168 P.M.Arround, by boiling-water observation, at 202·2Fahr.——757·140§1,976·16.483§ Comparison with Mogador.
Same, height adopted on comparison of allobservations————1.976.463
„151 P.M.Highest olives in Aït Mesan632·5640·3757·659§1,4884.882§ Comparison with Mogador.
„169 A.M.Saint’s tomb below Tagherot pass568——39‡2,393·27.852‡ By comparison with Arround.
„„2.30 P.M.200 ft. below summit of Tagherot pass498·5——25‡3,439·411.284Do.
Estimated altitude of Tagherot pass————‡3,500·411.484Do.
„188 P.M.Camp, Sektana646684·1764·358§1,378·2⎱
⎰		4,523§ Comparison with Mogador.
„1910 A.M.Do.647649·3764·165§1,379·1
„„10 P.M.Camp, Amsmiz672675·3762·758§1,018·8—Do.
„209 A.M.Do.672674·9763·172§1,049·0—Do.
„219 A.M.Do.672·5676·55761·8570§1,024·2—Do.
Do.————1,030·73.382Mean of three observations.
„„2 P.M.Iminteli, Amsmiz valley647651·4761·664§1,345·2—§ Comparison with Mogador.
„226 A.M.Do.646·5651·0761·560§1,344·0—Do.
„238 A.M.Do.646651·6760·1565§1,348·1—Do.
„„—Do.————1,345·84.415Mean of three observations.
„„—Do.————**1,348·84.425** Altitude by comparison with Amsmiz onMay 21.
„„—Iminteli————††1,346·54.418†† Altitude adopted. Mean of fourresults.
„229 A.M.Halt at base of Djebel Tezah629——64§§1,7085.604§§ Comparison with Iminteli.
„„2.30 P.M.Summit of Djebel Tezah512·5518·0760·8560§3,359·711.023§ Comparison with Mogador.
Do. do. do.————§§3,340·510.961§§ Comparison with Iminteli.
Do. do. do.————3,350·110.972Altitude adopted.
„„4 P.M.Upper limit of Quercus Ballota565——59§§2,4908.17§§ Comparison with Iminteli.
„245.30 A.M.Kasbah at Amsmiz666·5672·3759·961§1,063·53.489§ Comparison with Mogador.
„258 A.M.Kasbah at Mzouda694699·5760·165§721·32.367Do.
„„2 P.M.Halt near Kasbah, Keira688692·7760·170§814·32.671Do.
„268 A.M.Camp Seksaoua682·5686·9761·464§879·5—Do.
„276 A.M.Do. do.683687·2761·660§868·5—Do.
Do. do.————8742.867Mean of two observations.
„296 A.M.Camp below Milhaïn670674760·858§1,035·33.397§ Comparison with Mogador.
„„5.30 P.M.Watershed between Aïn Tursil and Mtouga656·0661·15760·752§1,190·23.905Do.
„307 A.M.Camp by Kasbah, Mtouga677·0681·4761·458§940·33.085Do.
„„8 P.M.Camp, Mskala716·0719·8761·960§466·3—Do.
„318 A.M.Do. do.715·5718·5762·765§486·3—Do.
Do. do.————476·31.562Mean of two observations.
„„5 P.M.Room 6 m. above Court of Kasbah of Shedma720722·8762·968§437·4—§ Comparison with Mogador.
June18 A.M.Do. do. do.720·5722·0764·3567§446·1—Do.
„„—Do. do. do.————441·81.449Mean of two observations.
„„—Court of the Kasbah————4361.43Do.
„26 A.M.Camp, Aïn el Hadjar744·5746·4763·758§154·8⎱
⎰		504§ Comparison with Mogador.
„35.30 A.M.Do. do.744·0746·6763·056§152·1

travelling in Marocco, and before the necessity for a considerable correction to the readings of Secrétan’s aneroid had become apparent. The difference arising from this and other corrections applicable to the highest points reached by us is considerable, and requires a deduction of about 500 feet from the estimated height of the Tagherot pass, and about the same from the calculated altitude of Djebel Tezah. The corresponding error in the calculated altitudes for the low country stations, e.g. those between Mogador and Marocco, averages about 200 feet.

J. B.

APPENDIX B.

Itineraries of Routes from the City of Marocco through the Great Atlas.

The information respecting the routes here given was supplied by a Jew named Salomon ben Daoud, an inhabitant of the city of Marocco engaged in trading operations with the natives of the portions of the Great Atlas wherein the authority of the Sultan is recognised. In the absence of more accurate reports, it appears desirable to publish this slight contribution to the topography of a country altogether unknown to Europeans, excepting so far as we were able to visit a few of the places enumerated. To assist those who may hereafter seek to follow any of these routes, the names of places inserted in the French map of Marocco by Captain Beaudouin, or in the map annexed to this volume, are distinguished by an asterisk. The distances are reckoned by hours, one of which may be counted as equivalent to four miles in the plain, and to a somewhat lesser distance in the mountain. A day’s journey usually varies from eight to ten hours. The spelling of the names is made to agree with that adopted throughout this work, the vowels having the same sounds as in most European languages, and not those peculiar to our country.

ROUTE 1.
Marocco to Demenet, and Excursion from Demenet to places in the neighbourhood.

Marocco to Ain el Berda3 hours.
„ *Sidi Rahal5 „
„ *Oued Tessout (ford the river)4 „
„ *Tidli (a mountain)2 „
„ Draha1 hour.
„ *Demenet3 hours.
18 „

This road is undulating, with hills and valleys, or hollows.

From Demenet cross over the river Emhasser, and proceeding for one hour on the mountain you will reach a place called Iminifri, on a high mountain, which contains an opening or pass only just large enough for one person to creep through on his hands and knees, the length of the pass being about 100 yards; and when through it you will find open ground on the top of the high mountain. There will be seen remains (ruins) of old Christian buildings, in which live many birds. From the upper part of this mountain overhanging parts (or cliffs) branch out downwards against the mountain, towards the River Tor, of 500 yards long; but these overhanging parts do not reach the water of the river.

[The places here spoken of apparently lie N.W. of Demenet—the El Acchabi of the French map. The river Tor is probably the Oued Lakdeur of the same map.]

ROUTE 2.
Demenet to the Sources of the Oued Tessout.

Demenet to Aït Cid Hassan (between mountains)4 hours.
„ *Aït Emdoual1 hour.

In Aït Emdoual is a river one day’s journey long. There are inhabitants along the river. From this to Aït Affan one day’s journey over barren desert ground uninhabited.

From *Aït Affan to Ansai (contains someinhabitants)? hours.
„ „Aït Kassi2 „
„ „Tel Khedit2 „

Tel Khedit is a mountain, and contains the source of the river Tessout; and on this mountain the snow remains both in summer and winter.

[This route agrees in many respects with the indications of the French map. The name Aït Chihatchen, there laid down south of Demenet, is probably the Aït Cid Hassan of the itinerary. It appears, however, that the importance of the mountain chain on the north side of the upper valley of the Tessout must be much exaggerated by the hill-shading on the map. The main chain of the Atlas is undoubtedly that on the southern side of that valley. The head of the valley is, on the French map, united to the province of N’tifa. A priori probability and the wording of the itinerary suggest that it all belongs to Demenet.]

ROUTE 3.
Marocco to N’tifa.

Marocco to Zourt ben Sessy2 hours.
„ Ras el Aïn2 „
[Here is a mountain called Bou Surkar,or stony.]
„ *Tamlelt3 hours.
„ *N’tifa1 day.

[None of these places seem to be laid down on the French map, unless Tendalet be the same place as Tamlelt. N’tifa is properly the name of the province. The particular place so named by our informant, is probably the residence of the Kaïd, or Governor. It seems likely that this is very near the place marked Bezzou on the French map.]

ROUTE 4.
Demenet to N’tifa.

Demenet to Aït Mazan (valleys and hills)3 hours.
„ *N’tifa3 „
6 „

On this road is found the Gum Euphorbium plant, or tree, and the trees producing the brown gum arabic. From N’tifa forward is the country of the tribe Aït Attab where there is little or no government among the people.

ROUTE 5.
Tour in the Mountains East and South-east of Marocco.

Marocco to Zourt ben Sessy2 hours.
„ Ras el Aïn8 „
„ *Tagana2 „
„ Aït Zehad (on the mountain in Mesfioua)4 „
„ Iminterrat2 „
[Here are found some ruins of Christian buildingsof old times.]
Marocco to Tasselt2 hours.
„ Tel Eizrat2 „
„ Tighidoun Idioum4 „
„ Aït Izzel (high mountain)4 „
„ Assefrag (Lasfaour)6 „
„ Imin Gagar6 „
„ Imin Zadin2 „
„ Tasghinout (Tasseremout?)3 „
„ Aït Absalem3 „
„ Tidiren2 „
„ Ohamma? „
„ Aïn Hehia2 „
„ Gries (Gers?)3 „
„ Ohida3 „
„ Tigardoun2 „
„ Tigola (Tougla?)2 „
„ Tabia (Aït Tieb?)2 „
„ Tamzart4 „

These being the mountains of the Mesfioua country.

ROUTE 6.
Tour in the Mountains South of Marocco.

Marocco to Amreen (plain, orchards)4 hours.
„ Resmat2 „
„ *Ourika? „

There is the river Ourika passing between mountains. Crossing, and going up to the left, the first village is Achliz. The chief, or sheik, who governs Ourika lives here, he being under the orders of his superior, Ibrahim el Graoui, who lives in Marocco. From Achliz you go to Azrou Miloul, and from this to Tourit. Here are salt wells or springs. From Tourit you go to Agadir which is on the top of the mountain, from this to Timluzen, and from this to the Zaouia. The before last stages from Ourika are all half an hour’s distance one from the other. From Tourit to Sissag on high mountains. [Apparently Sissag is the name of the Zaouia.] These are the villages on the left side of the river until Sissag. The villages on the right hand side of the river are as follows:—The first is Alzli; from this you go to Tafzhia, and from this to Anrar, and from this to Amsin; from this to Assgher, and from this to Arzballo; from this to Egremon, and from this to Ashni, and from this to Esurgraf; this mountain is covered with snow. These are the Ourika mountains near to Marocco town, besides the higher mountains which are above these we have mentioned.

[All the places here mentioned are in the Ourika valley. The left and right sides are those on the left and right of the stream to a person ascending the valley, contrary to the usage in European countries where those terms are supposed to refer to one following the course of the stream.]

ROUTE 7.
Description of the Roads of Ghighaya.

Marocco to Tahanout6 hours.
„ Tasslamat½ hour.
„ *Souk el Ad (of Moulai Ibrahim)2 hours.

From this to El Anraz; here is a village called Amareen, three hours distant—it leads to Immaregen; and from this one hour’s journey will bring the traveller to a place called Agadir Tagadurt el Bour, and from this last is the commencement of the road or highway to the province of Sous.

[There can be no doubt that the district here rendered Ghighaya from the Hebrew, is the same as we wrote down as Reraya, the r in the latter name having a guttural sound without an equivalent in any European language known to us. The Souk el Ad, or Sunday Market, is of course somewhere near to the Sanctuary of Moülai Ibrahim. Although we fail to identify any of the villages named above, it seems probable that the place spoken of as Agadir Tagadurt el Bour, is the same as Arround, where we passed two nights.]

ROUTE 8.
Description of the Road between Ghighaya and Ourika.

Marocco to Tahanout.6 hours.
„ Tedroura1 hour.
„ Ourika1½ „

This being a road to a mountain containing snow, in the country of Ghighaya. From Tranghert, six hours’ journey to a mountain called Ousertik, within the jurisdiction of the Governor, Kaïd Ibrahim el Graoui.

[The writer here gives an alternative route from Marocco to Ourika, slightly longer than the direct way given in Route 6, and then refers, obviously not from personal knowledge, to a mountain path connecting Ourika with one or other of the two valleys included in the district of Reraya or Ghighaya. Tranghert is probably a village in the western branch of the Ourika valley.]

Salomon ben Daoud.

APPENDIX C.

Notes on the Geography of South Marocco.

By John Ball.

Some remarks upon the geography of South Marocco seem to be called for from a writer who has ventured to put forth a new map, largely differing from those hitherto published; but the subject is encompassed with so much difficulty, and the amount of accurate information available is so limited, that a prudent writer must be content to regard most of his own conclusions as merely provisional, and liable to be modified or set aside by the results of further exploration, whenever this shall become practicable. In the mean time, some good may be effected by clearing the ground of some received errors that are absolutely disproved by facts now ascertained.

Little need here be said of the slight contribution to the knowledge of South Marocco that can be gleaned from the writers of antiquity. The earliest document bearing on the subject was doubtless the record of the voyage of Hanno, set up in the temple of Saturn at Carthage. This is known to us only by the version, rendered by an unknown hand into Greek, which, with all the accumulated errors of the translator and the subsequent transcribers, has reached us under the title of the Periplus of Hanno. From this record the particulars to be gleaned regarding this part of Africa are scanty and of an uncertain character. Commentators have, with much probability, identified the Solois promontory of Hanno (Λιβυκὸν ἀκρωτήριον λάσιον δένδρεσι) with Cape Cantin. But what are we to make of the next statement that, having passed the cape, they sailed for half a day east, or south-east (πρὸς ἥλιον ἀνισχοντα), before reaching the great marshy lake, ‘where elephants and other wild beasts abounded’? True it is that south of Cape Cantin there are two slight indentations, mere coves, where the land for a short distance trends to the south-east; but the general direction for a mariner along this part of the coast is SSW., as far as Mogador. Agreeing with the commentators that the ‘great marshy lake’ was probably near the mouth of the Oued Tensift, we are led to believe that Hanno disembarked settlers at no less than five stations on the coast of what is now the province of Haha. If we may rely on the correctness of the Greek text we must infer that these were settlements established by the Carthaginians before the date of Hanno’s expedition.[1] The next place reached by Hanno was ‘the great river Lixus, flowing from Libya, about which dwelt a nomadic people,’ who are called in the text Lixitæ (Λιξίται). It is further stated that the river is said to flow from great mountains in the neighbourhood, around which dwell the Troglodytes, referred to in our text, [p. 301]. The only assertion that can be confidently made about the Lixus of Hanno is, that it was quite a different stream from that afterwards known to the Romans by the same name, the latter being the modern Oued el Kous, falling into the sea at El Araisch, and which Pliny makes fifty-seven Roman miles from Tangier. The learned commentator, C. Müller, identifies the Lixus of the Periplus with the Draha; but, unless we assume that great physical changes have occurred during the interval, this supposition is scarcely compatible with the existence of a numerous population near the mouth of the river. It may possibly have been the river Akassa (the native name of the river of Oued Noun); but it appears far more probable that it was the Sous, the only one of these rivers which is believed constantly to discharge a large volume of water into the sea. It may be, indeed, that there is an etymological connection between the names Sous and Lixus, as there undoubtedly is between some names still current and those used by the Romans.

After Hanno, the next voyager along this coast of whom we known anything was Polybius. The original record of his voyage has, unfortunately, not come down to posterity, but a few particulars have been preserved by Pliny.[2] We learn incidentally that the Romans called Cape Cantin promontorium Solis, a name evidently suggested by the earlier name Solois of the Carthaginians, afterwards rendered in Greek by Ptolemy ἡλίου ἄκρον. Whether Polybius succeeded in reaching the Senegal, or some other river within the tropics, may be uncertain; but he undoubtedly visited many places on the Atlantic coast of Marocco. We hear for the first time of the rivers Subur (modern Sebou), and Salat (the Bouregrag, which falls into the sea at Sallee). He touched at the port of Rutulis, said to have been eight Roman miles beyond the mouth of the river Anatis, which was 205 Roman miles from Lixus (El Araisch). The river is doubtless the modern Oum-er-bia, and the port was the same which the Portuguese named Mazagan. The next port touched by Polybius was named Risadir, which has been with much probability identified with Agadir.[3] As for the rivers named by Polybius on the coast south-west of the Atlas, their identification with any known to modern geographers is purely conjectural.

Of Roman writers Pliny is the only one from whom any positive information as to the geography of this part of Africa is to be gained; but even this is very limited.[4] He complains that the reports as to the region beyond the narrow limits within which Roman power was established in his day were most fallacious, and censures the Roman authorities for indolently giving circulation to mendacious stories, instead of investigating the truth for themselves. In his day Sala (modern Sallee) was the most southern of the Roman settlements in Marocco. He describes it as ‘a town standing on a river of the same name, on the confines of the desert (solitudinibus vicinum), which was infested by herds of elephants, and still more by the tribe of the Autololes, through whose territory lay the way to the great mountain of Africa, the many-fabled Atlas.’ It appears elsewhere that Pliny had access to the manuscripts left by Juba, which, unfortunately, have not come down to posterity. That accomplished prince appears to have held control over the whole territory of Marocco as far as the base of the Atlas. It is to these lost pages of Juba that we probably owe the only fragment of moderately correct information as to South Marocco which is to be found in Pliny’s work.[5] The river Asana, whose mouth is said to be 150 Roman miles beyond Sala, is doubtless the Anatis of Polybius, and the Oum-er-bia of the Moors. The next river, which he calls Fut, is the Tensift. The distance assigned for the interval between the mouth of the Fut and the Atlas is excessive; but not largely so if Agadir be intended, that being the first place on the coast from which the high summits of the Atlas are habitually visible. The statement as to the existence of remains of vineyards and palm-groves about the ruins of ancient dwellings seems to lend probability to the belief that the Carthaginian settlements on this coast may have had a prolonged existence. The fall of the parent State would have had but an indirect influence on their destiny. Verbal resemblances are so often misleading that little weight can be attached to them; but it is natural to compare the word Dyris, said by Pliny to be the native name for the Atlas, with that now used by the natives—Idrarn—this being the plural form of Adrar, which means generically a mountain, both in the Shelluh and in several other Bereber dialects.

Besides what Pliny may have learned from King Juba as to the geography of the coast of South Marocco, he had access to contemporary testimony as to some part of the interior of the country. Suetonius Paulinus (the same who at a later date played a conspicuous part in Britain) being appointed governor of the provinces of N.W. Africa, then recently incorporated in the Roman Empire, resolved to penetrate southward beyond the Great Atlas, whether with a view to intimidate the native tribes, or for the mere satisfaction of carrying the Roman eagles into a new region. He appears to have left a written account of his expedition, which, like so much else of ancient geographical literature, has been lost. The particulars preserved by Pliny are unfortunately so vague as to be almost valueless.

In ten days from his starting point, wherever that may have been, we are told that he reached the highest point of his march. He reported the mountain to be covered with dense forests of trees of an unknown kind, and declares the summit of the range to be deeply covered with snow, even in summer.[6] From the summit of the Atlas Suetonius descended, and marched on through deserts of black sand, out of which rose here and there rocks that had the aspect of being burnt, to a river called Ger. Although it was the winter season the heat of these regions was found intolerable. The neighbouring forests abounded in elephants and other wild beasts, and with serpents of every kind, and were inhabited by a people called Canarians.

The controversies to which this passage has given rise are not likely to be definitively decided. The balance of opinion leans to the belief that Suetonius ascended the valley of the Moulouya, and traversed the Atlas by the pass now called Tizin Tinrout, leading to Tafilelt. This was the pass traversed by Gerhard Rohlfs in 1864, and to his narrative alone we can refer for information respecting it and the country extending southward towards the Great Desert. The existence in that part of W. Africa, on the south side of the Great Atlas, so far from the influence of the Atlantic climate, of vast forests capable of maintaining elephants and sheltering a native population, would apparently be irreconcilable with existing physical conditions, and is not readily admissible in the Roman period. Whatever vigorous vegetation exists in the region traversed by Rohlfs adjoins the banks of the stream; and, though sand may encroach here and there, and sun-burnt rocks are seen there, as elsewhere on the south side of the Atlas, the description is not what would occur to any one following the course of the stream. It seems, further, highly improbable that a prudent general, such as Suetonius Paulinus, would have undertaken to lead an aggressive military force along the tortuous valley of the Moulouya, some 250 miles in length, enclosed for the most part between lofty mountains; and it is also to be noted that at the period of his expedition the Romans held no station in the valley of the Moulouya, if indeed they ever penetrated far into it.

The few particulars quoted above lead to the conclusion that the Roman general in his southward march beyond the Atlas did not follow the course of a stream, but was compelled to cross a tract of desert before reaching the river of which he speaks, which, therefore, probably flowed from E. to W. On the whole, it seems to me that the brief record is more easily reconciled with the supposition that Suetonius Paulinus made Sala (Sallee), the farthest Roman station in Western Africa, his base of operations; that he marched thence across the open country towards SSW., and gained the summit of the Atlas range at the pass between Imintanout and Tarudant.[7] Between the course of the Sous and that of the Akassa, or river of Oued Noun, there are extensive tracts of sandy desert, where, even in winter, his troops may easily have suffered from heat and thirst; and the river (called Ger) may have been the main branch, or one of the tributaries of the Akassa flowing from the range of Anti-Atlas. The former existence of great forests, frequented by elephants, on the flanks of that range, is far more probable than on the parched southern slopes of the interior, where, as Rohlfs tells us, the rocks and hills are now absolutely bare of tree and shrub vegetation. Finally, it is more natural to look for the ancient Canarians in the country near the Atlantic coast than in the interior.

The solitary argument of any weight in favour of the Moulouya and Tafilelt route seems to be derived from the fact that in descending southward from the pass at the head of the Moulouya valley the traveller follows the course of a stream which now bears the name Gers, or Ghir. But it must be remarked that this name exists elsewhere in Marocco, there being at least three streams so denominated, and further that it is nowadays borne by the river of Tafilelt only during a short part of its course. Rohlfs, who is here our only authority, tells us that the stream first met in descending from the pass of Tizin Tinrout is called Siss.[8] After following this for seven or eight hours, it is joined by another stream which he called Ued Gers. The united stream bears the latter name for a distance of some six hours’ ride, and then resumes the name of Siss, which it bears throughout its subsequent course till it is lost in the sands of the Sahara.

The long period that intervened between the decline of Roman power and the establishment of Mohammedan rule in Marocco, is a blank to the historian and the geographer. It can scarcely be doubted that Roman authority and Roman institutions spread themselves throughout a great part of the open country between the Atlas and the Atlantic, although there is but little direct evidence to that effect.

Little reliance can be placed on the statement of Leo Africanus that the people of Barbary were converted to Christianity 250 years before the birth of Mohammed, or about A.D. 320, for, in a country so split up into independent tribes, the new faith must have made way irregularly and at various periods; while it is most probable that it never struck root among the mountain tribes of the Great Atlas. But the positive assertion of the same writer, that when the Arabs arrived in Marocco they found the Christians masters of the country, probably holds good of all except the mountain tracts.

Whether any reliable information as to South Marocco is to be gleaned from the writings of the eminent Arabian geographers who lived between the tenth and fourteenth centuries, I am unable to say; but it seems sufficiently certain that the period of European exploration leading to practical results commenced in the fourteenth century. The Genoese, the Catalans, and the Venetians appear to have despatched several expeditions along the coast, most of them intended to reach the gold-producing regions of tropical Africa. The Portuguese, who were destined to outstrip all their rivals in maritime exploration, were the first to establish themselves on the western coast of Marocco; and, at one time or other, they held most, if not all, the Atlantic seaports. Much information doubtless lies concealed among the mediæval records of Italy, Spain, and, especially, of Portugal; but up to the present time nothing has been published to show that any European was able, from personal knowledge, to give an account of the interior of Marocco, before Marmol, who, having been taken prisoner by the Moors, passed several years at Fez and elsewhere in North Marocco, about the middle of the sixteenth century. The earliest known document showing a moderately correct knowledge of the coast is a map (number 5 in the series), contained in the celebrated Portulano of the Laurentian Library in Florence, bearing the date 1351.[9] In this map, which, from internal evidence, must be of Genoese origin, the general outline of the Marocco coast is correct, and the positions of the few places laid down unmistakable. The now abandoned town of Fedala (Fidalah), Mefegam (Mazagan), and Mogodor here appear for the first time. Of early Portuguese maps there must be many not now known to geographers, and it was certainly from Portuguese authorities that Gerard Mercator partly derived the materials used in both editions of his Atlas. In the Atlas Minor, published by Hondins in 1608, a map of South Marocco is given in page 567, wherein for the first time an attempt is made to represent the positions of cities and mountains, and the courses of rivers in the interior of the country. The outline of the coast is here less correct than that given in the much more ancient Medicean map; but there is far more of detail, especially as to places which were evidently well known to the Portuguese. Thus, as mentioned in the text, we here for the first time find the island of Mogador with the name ‘I. Domegador.’ The places laid down in the interior appear for the most part to be taken (but with numerous errors) from the work of Leo Africanus; but the chartographer has spoiled his map by making the river Sous flow from SE. to NW., instead of from NNE. to SSW. Mountains are scattered pretty uniformly over the map; but what is made to appear as the loftiest mass, and is marked ‘Atlas M.,’ with a town named Tagovast at its foot, stands S. of Tarudant about the western extremity of the range of Anti-Atlas. The accompanying letterpress, page 566, is to a great extent derived from Leo Africanus, but with additions from other sources. It is curious to read that Tarudant, now a place which no Christian stranger dare approach, was then resorted to by French and English merchants.

The name of the remarkable man, who stands almost alone as a geographical authority for the interior of Marocco, has already been mentioned; but it is impossible to dismiss him so lightly. Leo Africanus, to give him the name by which he is known to posterity, was a Moor of Grenada, born in the latter part of the fifteenth century, who, with his kinsfolk, fled to Fez at or about the time of the siege of Granada in 1492. In those days Fez was the head-quarters of Arabic culture; Leo was an earnest and successful student, and, as a man of learning and intelligence, was taken into favour by Mouley Ahmet, the founder of the dynasty still reigning in Marocco. Either in company with the new ruler, or with his protection and authority, he travelled through almost every part of the empire, as well as nearly all the rest of Northern Africa, and evidently made copious notes. He wrote, in Arabic, various works on history and grammar which have not been preserved, and, in the same language, the original version of his description of Africa. It would appear that he carried this with him, in manuscript, when, in 1517, he was made captive by Christian corsairs, who took him to Rome. Leo X., hearing that a learned Moor had been brought a captive to Rome, sent for him, and treated him with kindness and liberality. A suggestion that he should undergo the rite of baptism seems to have encountered no obstinate prejudices, for he soon complied, receiving at the font the Pope’s own names, Giovanni Leone, and perhaps becoming as earnest a Christian as the Pontiff himself. He afterwards lived many years in Rome, acquired the Italian tongue, and translated his work on Africa into that idiom. This remained for some time unpublished, until it fell into the hands of Ramusio, who included it in his famous work ‘Delle Navigationi et Viaggi,’ of which the first edition, in three folio volumes, was printed in Venice in 1550. It is not easy to account for the numerous variations between the original text and the versions which appeared in various languages during the century following the original publication; but in the absence of satisfactory explanation it seems safest to accept the text of Ramusio as alone authentic.

Like most modern readers, the members of our party, when they resolved to visit Marocco, knew nothing of the work of Leo Africanus beyond the fact that he is occasionally referred to by writers on North Africa. The time for preparation was far too short for extensive reading, and we took with us only the works of Jackson and Gerhard Rohlfs. It has, however, since that time been a matter of frequent regret that we had not the opportunity, while travelling in the country, of referring to the only writer who had actually seen the greater part of it with his own eyes, and as to whose general truthfulness there is no room for suspicion. It is impossible here to enter into the many interesting details that abound throughout the text; but it is worth while to point out the more important changes that are disclosed between the condition of South Marocco as it was more than three and a half centuries ago, and that of the present day.

So far as regards the manners, ideas, habits, and mode of living of the inhabitants, the changes are quite insignificant, save in so far as these are affected by a general decline in material prosperity. The central authority was at that period much weaker, and the separate tribes led a more independent existence. Amongst the Bereber people of the mountains, and even in many of the larger towns, such government as existed was ordinarily of the democratic type. Thus we read that in Tarudant four chiefs were elected to manage the affairs of the city, holding office for only six months at a time.

If it were possible to doubt the results of the establishment of a system of grinding despotism, administered by officials who enjoy practical impunity so long as they satisfy the pecuniary demands of their master, the pages of Leo Africanus bring ample evidence. It is, indeed, true that a slight improvement has ensued as regards internal tranquillity. There is now rather less of habitual turbulence; the mutual encounters between neighbouring tribes may be somewhat less frequent; and brigandage, which appears to have been not uncommon in the open country, is now comparatively rare. It may be doubted whether this advantage, such as it is, is not as much due to diminished population as to the successful administration of the Moorish Sultans.

On the other hand, there is overwhelming evidence of a general and progressive decline in prosperity. Throughout the southern provinces, and especially in Haha and Sous, Leo Africanus found numerous flourishing towns, most of them visited and described by him. In each one of these he found people living in comparative ease, inhabiting good houses with gardens, and possessing, according to the standard of the age, some literary education. From the towns, and even from the inner valleys of the Atlas, students flocked to Fez, then the head-quarters of Arabic knowledge and civilisation. All the principal places were then local centres of production, the artificers being principally Jews.

It is notable that excepting the city of Marocco, then full of a numerous and active population, none of the towns mentioned owed their foundation to the conquering race. Leo, not likely to detract from the achievements of his own people, expressly attributes the origin of most of them to the ‘antichi Africani,’ by which designation he commonly speaks of the primitive Bereber stock; and, as regards the smaller towns lying in the low country north of the Atlas, he frequently speaks of the population being harassed by the Arabs, then, as at this day, leading a semi-nomad existence in the plains.

If we confront his description with the present state of the country we find comparative ruin and desolation. In all the southern provinces we now find but two inland cities of any importance, Marocco and Tarudant, and these dwindled to a mere tithe of their ancient wealth and population. Where the traveller in the sixteenth century found thriving towns at intervals of ten or twelve miles, there are now miserable villages whose wretched inhabitants maintain a bare existence, and are often unable to pay the imposts which leave no surplus behind. It does not appear that in the great province of Haha there is now a single place that can be called a town except the ruined seaport of Agadir, destined by nature to be the chief port of South Marocco, but closed to trade by the caprice of a Sultan. Throughout the interior we saw or heard of but two places that could by courtesy be called towns, Amsmiz and Moulai Ibrahim. Although no statistics are available, it seems a moderate estimate if we reckon that the present population of South Marocco cannot exceed one-third of what it was when Leo wrote.

Along with the decay of wealth and population, we naturally find that of everything that could raise the people in the scale of existence. In Leo’s day iron and copper mines were worked in many places in the Atlas, and various handicrafts exercised, of which there is now no trace. Education, such as it was, was widely spread; and in some parts of the Atlas where it was absent, the traveller noted the fact as a proof of the low condition of the population. He notes as a curious incident that when he visited the mountain district of Semele, where the people were ignorant of reading and writing, they forced him to remain nine days, hearing and deciding all pending cases of litigation; in doing which, as he records, he had to act both as judge and notary, there being no one competent to write down the decisions of the court.

Several incidental statements in the work of Leo Africanus suggest an inquiry of considerable interest. There is nothing in the published annals of the Portuguese wars with the Moors to suggest a belief that the former at any time established their authority in the interior of South Marocco, or even undertook any inland expeditions. From Leo’s narrative it appears, however, that, at the beginning of the sixteenth century, they had, at least occasionally, penetrated much farther into the interior than has commonly been supposed, and that the authority of the Portuguese king was in some places paramount. At Tumeglast, a place in the plain of Marocco, probably not far from the present village of Frouga, Leo lodged in the house with a Moor, named Sidi Yehie, who had come in the name of the king of Portugal to levy tribute, the same Moor having been made by the king chief (capitano) of the district of Azasi. Elsewhere he relates that the king of Marocco sent an expeditionary force against an independent chief in the district of Hanimmei, forty miles east of the city of Marocco (apparently in the present province of Demnet), and which force was accompanied by 300 Portuguese cavalry. The expedition was unsuccessful, the Sultan’s troops were defeated, and, according to the narrative, not one of the Christian horsemen returned from the disaster. It seems highly improbable that the Portuguese should have taken part in such an affair if their troops had not at the time been stationed somewhere in the interior.

After Leo Africanus but little of a definite kind is to be learned from subsequent writers as to the geography of South Marocco. In 1791 the reigning Sultan applied to General O’Hara, then Governor of Gibraltar, for the assistance of an English physician to treat his favourite son, Mouley Absalom, who was at the time governing the province of Sous. Mr. Lempriere, an army-surgeon, undertook the office, and travelled by the west coast to Agadir, and thence to Tarudant. After successfully treating his patient, he was partly induced, and partly forced, to travel to the city of Marocco, whence, after considerable delay and difficulty, he succeeded in returning to Gibraltar. Mr. Lempriere probably travelled across the Atlas by the road from Tarudant to Imintanout, but his narrative supplies little information to the geographer. He speaks of the distance from Tarudant to the northern foot of the Atlas as an easy journey of three days, and describes the track as leading beneath and along tremendous precipices.

Frequent reference is made in the text to Jackson’s ‘Account of the Empire of Marocco,’ of which the first edition appeared in 1809, and the third in 1814. This is undoubtedly the fullest and most correct modern work on Southern Marocco. Jackson spent sixteen years in the country, chiefly at Mogador and Agadir; he acquired the familiar use of the Moorish Arabic, and seems to have obtained merited influence among the natives. Either because he had but little taste for exploration, or because he found the difficulties too serious, Jackson has added little to our knowledge of the geography of the country. His map, though it contains some corrections, is on the whole inferior to that of Chénier, published a century earlier.

A definite contribution to the slight existing amount of positive knowledge was made by the late Admiral Washington, then a lieutenant in the navy, who accompanied the late Sir J. Drummond Hay on his mission to the city of Marocco in the winter of 1829-1830. His paper, published in the first volume of the ‘Journal of the Royal Geographical Society,’ is frequently referred to in our text; and in the accompanying map the positions of several points in the interior of the country were accurately laid down from astronomical observation.

A most important step towards extending our knowledge of the entire empire of Marocco was made in 1848, when the French War Department published the map compiled by Captain Beaudouin. Whatever errors it may contain—and these were unavoidably numerous—this must be regarded as a monument of intelligence and industry. Recognising the fact that the greater part of the territory is likely long to remain inaccessible to Europeans, the author applied himself to obtaining information from natives who were personally acquainted with various portions of the country. Hundreds of such informants, as we were assured, were separately examined by Captain Beaudouin; the information supplied by each was laid down on a skeleton map; and by the careful comparison of the separate materials the general map was compiled.

Without noticing minor errors, which are, of course, inevitable in such a work, the most serious objection to be made to this map is that the orography is exhibited in a fashion primâ facie improbable, and which has been to a great extent negatived by subsequent evidence. The main range of the Southern Atlas is represented as a nearly straight wall, over 400 miles in length, with few and short diverging ridges, and, parallel to this on the south side, another equally straight and narrow ridge is made to stretch for nearly 300 miles. From near the eastern extremity of the main range two other straight ridges are shown, diverging abruptly at an acute angle, and enclosing a trench-like valley that extends north-west for fully 120 miles. If this were even approximately correct, we should be led to conclude that the structure of the Great Atlas is quite unlike that of any other known mountain region. The tendency of mountain ranges to follow a uniform general direction is always modified by the numerous secondary causes that have helped to fashion the earth’s surface.

The first recent traveller who succeeded in penetrating some considerable portions of the Marocco territory was M. Gerhard Rohlfs. Assuming the garb and professing the faith of a Mussulman, he traversed many districts where no Christian dare present himself; but the care necessary to prevent his real character from becoming known imposed severe restrictions on M. Rohlfs. Produced under conditions where it was impossible to be seen taking notes or using any scientific instrument, it is not surprising that the narrative of his adventurous journey is extremely meagre; but even for the little that he is told about a region so little known the reader is thankful. The chief geographical results of these journeys were embodied in the map annexed to G. Rohlfs’ first work[10] by the eminent geographer, M. Petermann. The scale of that map is small and admits of little detail; but, so far as regards the mountain country, I am disposed to think that the direct evidence, supplemented in some points by native report, requires us to depart more widely from the orographic features of Beaudouin’s map than M. Petermann has thought it fit to do.

In the map accompanying this volume I have ventured, in addition to the changes for which I had direct authority, to introduce a few others, avowedly conjectural, which must await further exploration before they can be either adopted or condemned. It is difficult to believe that in an age when the barriers that have closed the other least known regions of the earth are successively removed, Marocco, so close to Europe and so attractive, can alone resist the progress of modern exploration.[11]

FOOTNOTES:

[1]The phrase used is κατῳκίσαμεν πόλεις πρὸς τῇ θαλάττῃ καλουμένας Καριχόν τε Γύττην καὶ Ἄκραν καὶ Μέλιτταν καὶ Ἄραμβυν. When the author speaks of Thymiaterium, founded by Hanno in this expedition, he says, ἐκτίσαμεν πρώτην πόλιν.

[2]See Pliny, V. 1, § 8. His account is vague and confused, and the distances not to be reconciled with those given by him elsewhere.

[3]Not content with the indication afforded by the identity of the two terminal syllables in each name, C. Müller conjectures that the ancient name of the promontory near Agadir was Râs adir, Râs being the common Arabic designation for a headland. He apparently supposes that the natives spoke Arabic in the time of Polybius. Even now none of the headlands on this coast have the designation Râs.

[4]I am indebted for information as to several passages in Pliny’s writings to my friend, Mr. E. Bunbury, who will doubtless throw further light on the subject in an important work, ‘An Historical View of Ancient Geography’ which he is preparing for publication.

[5]‘Indigenæ tamen tradunt in ora ab Sala CL m. p. flumen Asanam, marino haustu sed portu spectabile: mox amnem quem vocant Fut: ab eo ad Dyrin (hoc enim Atlanti nomen esse eorum lingua convenit) CC m. p., interveniente flumine cui nomen est Vior. Ibi fama exstare circa vestigia habitati quondam soli vinearum palmetorumque reliquias.’

[6]This must have been from native report, as the expedition was made in winter. If he had said that the snow never quite disappears, and sometimes falls heavily, even in summer, his statement would have been accurate enough.

[7]Mentioned in the text at [p. 294.]

[8]This is evidently the river Ziz of Leo Africanus; and in his time, as at the present day, travellers going from Fez to Segelmese (modern Tafilelt) followed the course of the Ziz, or Siss. He also speaks of a river Ghir, which may possibly have been the affluent of the Siss mentioned by Rohlfs; but the particulars given are vague and scanty. It is interesting to remark that in Leo’s day the valley of the Siss was inhabited by a hardy and energetic Bereber tribe named Zanaga, probably the same as the Azanegues whom Cà da Mosto found about Oued Noun. They have since migrated across the Sahara, and still calling themselves Zanega, and speaking a Bereber dialect, are dangerous neighbours to the negro tribes of the Senegal.

[9]A portion of this map, containing the coast of Africa from the Straits of Gibraltar to the latitude of the Canary Islands, was published (in facsimile) by Count Baldelli Boni of Florence in his edition of Marco Polo, and is reproduced in Mr. Major’s valuable work, ‘The Life of Prince Henry the Navigator.’ London, 1868.

[10]Afrikanische Reisen, von Gerhard Rohlfs. Bremen, 1867.

[11]The scope of these remarks being limited to the geography of South Marocco, I have not noticed several recent publications, not devoid of interest and value, but in which no important contribution is made to our geographical knowledge. We have referred in the text to papers by MM. Beaumier, Balansa, and Lambert, in the Bulletin of the French Geographical Society. A more considerable work, entitled ‘Morocco and the Moors,’ by Arthur Leared, M.D., appeared in 1874. It contains much information carefully collected by the author, along with a lively account of his own experiences, but circumstances prevented him from entering on new ground.

APPENDIX D.

ON SOME OF THE ECONOMIC PLANTS OF MAROCCO.

By Joseph Dalton Hooker.

Gum Ammoniacum.

Our endeavours to obtain accurate information regarding the Marocco gum ammoniac plant were ceaseless and fruitless. Jackson, who gives a rude figure of a portion of a leaf and a scanty description (‘Account of the Empire of Marocco,’ 136, t. 7), says that it is the produce of a plant like Fennel, but larger, and called Fashook in Arabic, and that it grows in the plains of the interior provinces, abounding in the north of the city of Marocco, in a sandy light soil. Jackson further states that neither bird nor beast is seen where this plant grows, the vulture only excepted, and that it is attacked by a beetle having a long horn proceeding from its nose, with which it perforates the plant, and makes the incisions whence the gum oozes out. Under his description of the vulture, he states that, with the exception of the ostrich, this is the largest bird in Marocco; that it is common in all places where the gum ammoniac grows, as in the plains east of El Araiche,[1] where he has seen at least twenty of these birds in the air at once, darting down on the insects with astonishing rapidity (p. 118). Jackson’s figure (t. 8) of the so-called beetle apparently represents a dipterous insect resembling a Bombylius, with a very long straight proboscis.

Lindley (‘Flora Medica,’ 46) doubtfully refers Jackson’s Fashook to the eastern Ferula orientalis L.; and Flückiger and Hanbury (‘Pharmacographie,’ 289) say that, according to Lindley, the Ferula tingitana yields a milky gum resin, having some resemblance to Ammoniacum, which is an object of traffic with Egypt and Arabia, where it is employed like the ancient drug in fumigations. The authors go on to say that there can be but little doubt that the Maroccan Ammoniacum is identical with that of the ancients, and that it may well have been imported by way of Cyrene from regions lying farther westward.

Pliny and Dioscorides say that the Ammoniacum is the juice of a Narthex growing about Cyrene and Libya, and that it is produced in the neighbourhood of the temple of Ammon.

Dr. Leared (‘Morocco and the Moors,’ 356) was informed that the Fashook grows at a place two days’ journey from Mogador, on the road to the city of Marocco,[2] but states that the exudation from the roots of specimens which he obtained differed from the African Ammoniacum. We, on the other hand, were persistently assured that it grew nowhere along that route, nor nearer to it than El Araiche, north of Marocco city. And this is confirmed by information obtained by Mr. R. Drummond Hay to the effect that it is found near Marocco, and chiefly around Tedla. The Moors who gave us this information at once recognised the figure by Jackson, and called the plant Kilch (Kelth according to Leared). The roots presented to Kew by the kindness of Dr. Leared did not make any indications of growth.

The Maroccan Ammoniacum plant must not be confounded with the Persian Dorema Ammoniacum, or ‘Ushak,’ which is also bled by insects.

The Fashook gum is used by the Moors and by some Orientals as a depilatory, and in skin diseases; it is exported to the East from Mazagan, viâ Gibraltar and Alexandria.

Euphorbium, Furbiune or Dergmuse.

Euphorbia resinifera.—Berg. und Schmidt, Officinelle Gerwächse, v. iv. (1863) xxxiv. d.; Flückiger and Hanbury, Pharmacographia, 502; Cosson, in Bull. Soc. Bot. Fr. xxi. 163; Ball, in Journ. Linn. Soc. Bot. xvi. 661; Euphorbium, Jackson’s ‘Account of the Empire of Marocco,’ 134, t. 6 (left-hand figure only).

We have little to add to the description of the Euphorbium tree given by Jackson, and that in the ‘Pharmacographia’ cited above. As stated in the body of this work it is confined to the interior of the empire, and the only living specimens we met with were from a garden in Mesfiouia (see [p. 163]). Jackson confounded two plants under this name; one, the true species, growing in the Atlas, with 3-4-angled branches, the other a sea-coast plant, with 9-10-angled branches, which is carried to Marocco for tanning purposes, and of which he says, that during the three years of his residence at Agadir he never saw any gum upon it.

The true plant is figured and described by Jackson as an erect tree, with a stout short woody trunk, and very numerous upcurved long sparingly divided branches, the whole resembling a candelabrum. The angles of the branches are armed with short spines, and the flowers are produced from the tips of the young shoots. The thorns adhere to everything that touches them, and he supposes them to have been intended by nature ‘to prevent cattle from eating this caustic plant, which they always avoid on account of its prickles.’ The juice flows from incisions made with a knife, and hardens and drops off in September. The plants, he says, produce abundantly once only in four years, and the fourth year’s produce is more than all Europe can consume. The people who collect the gum are obliged to tie a cloth over their mouths and nostrils, to prevent the small dusty particles from annoying them, as they produce incessant sneezing.

The history of the Euphorbium as given in the ‘Pharmacographia’ is, that it was known to both Dioscorides and Pliny as a native of the Atlas, and was named in honour of Euphorbus, physician to the learned King Juba II. of Mauritania, himself the author of treatises on Opium and Euphorbium.

The prevalence of cactoid Euphorbiæ in Marocco, of which there are three species in the southern districts, is a similar instance to that of the Argan, of tropical forms advancing far north in the extreme west of the old world; and as the Argan has its nearest ally in Madeira, so have the Maroccan Euphorbiums close congeners in the Canary Islands. All these belong to the section Diacanthium of Boissier, of which the other species are Abyssinian, Arabian, Indian, and South African.

Gum Euphorbium was extensively used by early practitioners as an emetic and purgative, and was exported in large quantities; now, however, the trade in it is rapidly declining, and we were informed that it is chiefly used in veterinary practice, and as an ingredient in a paint for the preservation of ships’ bottoms.

Euphorbia resinifera is in cultivation at Kew, where specimens may be seen both in the Succulent-plant House and Economic-plant House.

The Arar, Thuja or Gum Sandrac Tree.

Callitris quadrivalvis.—Ventenat, Nov. Gen. Decad. 10; Richard, Conif. 46, t. 8, f. 1; Endlich, Synops. Conif. 41; Parlatore, in DC. Prod. xvi. pars 2, 452; Ball, in Journ. Linn. Soc. Bot. xvi., 670.

Thuja articulata.—Shaw’s ‘Travels in Barbary,’ 462, with a plate; Vahl, Symb. ii. 96, t. 48; Desf. Flor. Atlant. ii. 353, t. 252.

Frenela Fontanesii.—Mirbel, in Mem. Mus. xiii. 74.

This tree is a native of the mountains of North Africa, from the Atlantic to Eastern Algeria; but we are not aware whether its eastern limit has ever been accurately determined. It has no congener, its nearest ally being a South African genus of Cypresses (Widdringtonia), of which several species are recorded from the Cape Colony, Natal, and Madagascar, and which differ in having alternate leaves and many ovules to each scale.

The great interest attached to this plant arises from the beauty and durability of the wood, which, there is every reason to believe, was known to the ancients from the earliest times, under the name of Thuja. It is thus hypothetically, but probably correctly, identified with the θυῖον[3] of the Odyssey (ii. 6), with the θυῖον and θυία of Theophrastus (‘Hist. Pl.’ v. 5), and the thyine wood of the Revelations (xviii. 12). It is undoubtedly the Citrus wood of the Romans, and the Alerce of the Spaniards; the latter name being derived from the Moors of Marocco, for it is not a native of Spain.

The first botanical notice of the Callitris is in Shaw’s ‘Travels in Barbary,’ where it is figured and briefly described as Thuja articulata (462); and for its identification with the Alerce we are indebted to the late Mr. Drummond Hay when Consul of Tangier, who, further, sent a plank of the wood to the Royal Horticultural Society.[4] At about the same time, the attention of a most intelligent traveller, the late Capt. S. E. Cook (afterwards Widdrington), was attracted by the wood of the cathedral of Cordova (formerly a mosque built by the Moors in the ninth century) called Alerce, which differed from any Spanish wood, or any other wood now used in Spain. Coupling this name with the communication made by Mr. Drummond Hay to the Horticultural Society, Capt. Cook was enabled to identify the Cordova wood with the Callitris, which, as he assumes, was brought from Marocco, to roof a mosque intended to be second in sanctity only to that of Mecca.

Except in a garden at Tangier, we saw no specimen of the Callitris approaching a large size, or capable of yielding the beams which we were shown in the ceilings and roofs of buildings in that town and elsewhere, and which are considered to be indestructible. On the contrary, most of the native specimens we saw in Southern Marocco resembled small Cypresses, with very sparse foliage and branches, and were apparently shoots from the stumps of trees that had been cut or burnt down, though possibly their impoverished habit may have been due to the sterility of the soil. The largest were in the Ourika valley, and were about thirty feet high (see [p. 177]). In many cases the stem swelled out at the very base into a roundish mass half buried in soil, which is said to attain even four feet in diameter, though we saw none approaching that size.

It is the basal portion, whether the result of mutilation or natural growth, that affords the wood so prized by ancients and moderns, and which forms a most valuable article of export from Algiers to Paris, where small articles of furniture, &c., are made of it and sold at very high prices.

Under the name of Citrus wood, it is alluded to, according to Daubeny, by Martial and Lucan, and by Horace (‘Carm.’ lib. iv. Od. 1), who suggests its employment as the most precious commodity that could be selected for a temple in which a marble statue of Venus should be placed:—

Albanos, prope te, lacus

Ponet marmoream sub trabe citrea;

Also Petronius Arbiter, descanting upon the luxury of the Romans, seems to represent it as worth more than its weight in gold, when he says—

Ecce Afris eruta terris

Ponitur, ac maculis imitatur vilius aurum

Citrea mensa.

For a detailed description of what was known of this tree to the ancients, and of its value, we must refer to the description in Pliny (‘Nat. Hist.’ book xiii. chaps. 29, 30). This author describes it as the thyion and thyia of Homer and the Greeks, and adds that its wood was used with the unguents burnt for their pleasant odour by Circe; as also that Theophrastus awarded a high rank to it, the timber being used for roofing temples and being indestructible; as also that it is produced in the lower part of Cyrenaica, and that the finest kind grows in the vicinity of the temple of Jupiter Ammon.

Pliny himself gives Mount Atlas as the native country of the wood; in the vicinity of which, he says, is Mauritania, a country in which abounds a tree which has given rise to the mania for fine tables, an extravagance with which women reproach the men when they complain of their vast outlay upon pearls. He attributes the knots from which the tables are made to a disease or excrescence of the roots, of which the most esteemed are entirely concealed under ground, these being much more rare than those which are produced above ground, and that are to be found on the branches also.

The principal merits of the tables were to have veins arranged in waving lines, or forming spirals like whirlpools. The former they called ‘tiger’ and the latter ‘panther’ tables; whilst others, which are highly esteemed, have markings resembling the eyes on a peacock’s tail. In others, again, called ‘apiatæ,’ the wood appears as if covered with dense masses of grain. The most esteemed colour was that of wine mixed with honey.

In respect of their size, Pliny gives a little over 4 ft. as the average maximum, though one that belonged to Ptolemæus, King of Mauritania, was 4½ ft. in diameter and ¼ of a foot in thickness. It was formed of two semi-diameters so skilfully united that the joining was concealed. Another, made of a single piece, was named after Nonius, a freedman of Tiberius Cæsar, and was 4 ft. less ¾ in. in diameter, and 5¼ inches in thickness. And with regard to the price, Cicero paid a million sesterces (9,000l.) for one; two belonging to King Juba were sold by auction, one for one million two hundred thousand sesterces, and the other for somewhat less. Some of Pliny’s statements are probably fabulous; as that the barbarians bury the wood when green, first giving it a coating of wax, and that the workmen, when it comes into their hands, put it for seven days beneath a heap of corn, and then take it out for as many more, after which it is surprising to find how much it has lost in weight. More apocryphal still is his statement that it is dried by the action of sea-water, and thereby acquires a hardness and density that render it proof against corruption; also that, as if created for the behoof of wine, it receives no injury from it.[5]

In Marocco, where no ornament or article of luxury is known, it need hardly be said that the Alerce wood is employed only for building purposes and fire-wood; though the resin called Sandarach, which was once a reputed medicine, is collected by the Moors and exported from Mogador to Europe, where it is used as a varnish.

Gum Arabic.

Acacia gummifera.—Willd. Sp. Pl. iv. 1056; DC. Prod. ii. 455; Hayne, Arzneigew. x. t. 8; Benth. in Trans. Linn. Soc. xxx. 509; Ball, in Journ. Linn. Soc. Bot. xvi. 442.

Mimosa gummifera.—Brouss. in Poir. Dict. Suppl. i. 164.

Acacia coronillæfolia.—Desf. Cat. Hort. Par. ed. ii. 207.

Mimosa coronillæfolia.—Pers. Encheirid. n. 44.

Sassa gummifera.—Gmel Syst. ex DC. 1. c.

Of this plant very little indeed is known, and we were unfortunately unable to find either flower or fruit of the only Acacia which we met with on our visit to Marocco, and which we were assured was the Gum Arabic plant (Alk Tlah) of that country. It is interesting as representing the northern limit of distribution of the immense genus Acacia in Africa. Our specimens, such as they are, coincide perfectly with the description of Acacia gummifera in Willdenow, and with the excellent figure in Hayne, which was taken from specimens collected by Broussonet near Mogador. We found the plant abundantly in the lower region of Southern and Western Marocco, occurring as a thorny bush, along with Rhus pentaphylla and other shrubs. That it was the plant producing the Marocco Gum Arabic the natives consistently testified, though this could not be inferred from the description in Jackson’s ‘Account of the Empire of Marocco’ p. 136, who says of the gum that it ‘is produced from a high thorny tree called Attalet, having leaves similar to the Arar, or gum Sandarac tree, and the Juniper.’ Jackson goes on to say:—

‘The best kind of Barbary gum is procured from the trees of Marocco, Ras-el-wed, in the province of Abda; the secondary qualities are the produce of Shedma, Duquella, and other provinces; the tree grows abundantly in the Atlas mountains, and is found also in Bled-el-jerrêde. The gum, when new, emits a faint smell, and, when stowed in the warehouse, it is heard to crack spontaneously for several weeks; and this cracking is the surest criterion of new gum, as it never does so when old; there is, however, scarcely any difference in the quality. The Attaleh is not so large a tree as the Arar, which produces the Sandarac gum, nor does it reach the size of the Auwar tree, which produces the gum Senegal. It has a low crooked stem, and its branches, from the narrowness of its leaves (long and scanty), have a harsh, withered, and unhealthy appearance at the time it yields the most gum—that is, during the hot and parching months of July and August; but although not an ornamental tree, it is a most useful plant, and will always be considered valuable. Its wood is hard, and takes a good polish; its seeds, which are enclosed in a pericarpium, resemble those of the Lupin, yield a reddish dye, and are used by the tanners in the preparation of leather. These seeds attract goats, who are very fond of eating them. The more sickly the tree appears, the more gum it yields; and the hotter the weather, the more prolific it is. A wet winter and a cool or mild summer are unfavourable to the production of gum.’

As observed in the body of this work, the gum does not seem to be collected in the western portion of its range in South Marocco, but in Demnet, whence it is brought to Mogador; and it may very well be that it is only in the hotter and drier regions of the interior that the gum is produced in sufficient quantities to be worth collecting.

It is remarkable that no notice whatever of Acacia gummifera occurs in Flückiger and Hanbury’s invaluable ‘Pharmacographia’ (1874), where the Marocco gum is supposed to be the produce of Acacia arabica Willd., a plant which extends from Nubia to Natal, and eastward to Central India, but which is not known as a native of Marocco. In another passage of the above work (p. 211), the ‘Marocco, Mogador, or brown Barbary gum,’ is described as consisting ‘of tears of moderate size, often vermiform, and of a rather uniform light dusky brown tint. The tears, which are internally glassy, become cracked on the surface and brittle if kept in a warm room; they are perfectly soluble in water.’

It is possible that the Acacia arabica, which is found in Senegal, may extend to the Sous Valley, and be the source of some of the Marocco gum; and that more than one species producing gum are confounded together by the Moors; this is the natural inference from Jackson’s account, itself anything but explicit. On the other hand, I am informed in a letter lately received from Mr. R. Drummond Hay, H.B.M. Consul at Mogador, who has kindly had inquiries made for me, that the Acacia arabica (Alk Awarwhal) is not found in Sous, no tree of the kind existing either north or south of the Atlas, but that its gum is brought from Soodan by Arabs, and is of inferior quality to that of the Acacia gummifera. Mr. Hay further informs me that the Acacia gummifera grows chiefly in the provinces of Blad Hamar, Rahamma, and Sous.

As stated above, the specimens which we collected of Acacia gummifera precisely accord with the published description and drawing; but we have others under this name from Mr. Cosson’s collector, Ibrahim, gathered near Mogador and at Ouanyna, which differ in having very short spines, ⅙ to ¼ in. long, whilst those of our plant are from ⅔ to ¾ in. long and much stouter.

Very small plants of Acacia gummifera are living at Kew, raised from seeds obligingly presented by Mr. Cosson. They grow exceedingly slowly, and several have been lost by damping off. They are not in a state fit for exhibition.

The Argan Tree.

Argania Sideroxylon.—Roem. and Sch. Syst. Veg. iv. 502; Alph. DC. Prod. viii. 187; Hook. in Kew Journ. Bot. vi. (1854) 97, t. iii. iv.; De Noé, in Rev. Hort. 1853, 125; Ball, in Journ. Linn. Soc. Bot. xvi. 563.

Sideroxylon spinosum.—Linn. Hort. Cliff. 69 (excl. syn. et loc.); Correa, in Ann. Mus. Hist. Nat. viii. 393.

Rhamnus siculus.—Linn. Syst. Nat. ed. 12, iii. 227, excl. syn., non Bocc.

R. pentaphyllus.—Linn. Syst. Nat. ed. Gmel. 398, fid. Dryandr. excl. syn. Bocc.

Elæodendron Argan.—Retz Obs. Bot. vi. 26; Willd. Sp. Pl. i. 1148, excl. syn. Jacq. and Bocc.; Schousboe, Iagttag. over væxtrig. in Marocc. 89.

Argan.—Dryandr. in Trans. Linn. Soc. ii. 225.

This tree is rightly regarded as the most interesting vegetable production of Marocco, being confined to that empire and to a very circumscribed area in it, belonging to an almost exclusively tropical natural family, yielding a most important article of diet to the inhabitants, and a wood that for hardness and durability rivals any hitherto described. The earliest account of the Argan tree known to us is a brief one by the celebrated African traveller Leo Africanus, who visited Marocco in 1510. Speaking of some of the customs of the Moors, Leo Africanus says: ‘Unto their Argans (for so they call a kind of olive which they have) they put nuts; out of which two simples they express a very bitter oil, using it for a sauce to some of their meats, and pouring it into their lamps’ (‘Purchas,’ ii. 772). And in another passage he describes the oil correctly, as ‘of a fulsome and strong savour.’ The further history of the Argan tree is given in a very full and careful account by the late Sir W. Hooker, in the ‘London Journal of Botany’ for 1854 (vol. vi. p. 97, Tab. iii. iv.), which, as the work is of limited circulation, we here introduce.

‘Through the kindness and by the exertions of the Earl of Clarendon, Chief Secretary for Foreign Affairs, the Royal Gardens of Kew have been put in possession of living plants and fresh seeds of a tree or shrub very little known in Europe, little known even to botanists, but highly esteemed by the Moors, in those parts of Marocco where it is a native, for its useful qualities, viz. the “Argan.” Its economical properties are best explained by the copy of a letter which his Lordship did me the favour to communicate along with the plants and seeds, from Henry Grace, Esq., British Acting Vice-Consul at Mogador, addressed to J. H. Drummond Hay, Esq., Her Britannic Majesty’s Agent and Consul-General at Tangier; both of which gentlemen spared no pains in procuring the information and seeds and living specimens; an example we should be glad to see followed by our consuls in other countries abounding in new and useful plants.

‘“Mogador, November 7, 1853.

‘“Sir,—The Argan tree grows more or less throughout the states of Western Barbary, but principally in the province of Haha, and south of this town. The soil in which it is found is light, sandy, and very strong; it is usually seen upon the hills, which are barren of all else, and where irrigation is impossible.

‘“I should imagine, from the appearance of some of the trees, that they are from one to two hundred years old; and a remarkably large one in this neighbourhood is probably at least three hundred. This individual measures 26 ft. round the trunk; at the height of three feet it branches off; the branches (one of which measures 11 ft. in circumference near the trunk) rest upon the ground, extending about 15 ft. from the trunk, and again ascend. The highest branch of this tree is not more than 16 ft. to 18 ft. from the ground, while the outer branches spread so as to give a circumference of 220 ft.: this is the largest I am aware of.

‘“The mode of propagation, in this vicinity, is mostly by seed. When sowing this, a little manure is placed with it, and it is well watered until it shoots; from which period it requires nothing further. In from three to five years after sowing it bears fruit, which ripens between May and August (according to the situation of the tree). The roots extend a great distance underground, and shoots make their appearance at intervals, which are allowed to remain, thus doing away with the necessity of transplanting or sowing. When the fruit ripens, herds of goats, sheep, and cows are driven thither; a man beats the tree with a long pole, and the fruits fall and are devoured voraciously by the cattle. In the evening they are led home, and, when comfortably settled in their yards, they commence chewing the cud and throw out the nuts, which are collected each morning as soon as the animals have departed upon their daily excursion. I have heard it remarked that the nut passes through the stomach; but this is only a casualty, and not a general rule. Large quantities of the fruit are likewise collected by women and children: they are well dried, and the hull is taken off, and stored for the camels and mules travelling in the winter, being considered very nutritious.

‘“The process of extracting the oil is very simple. The nuts are cracked by the women and children (and not a few fingers suffer at the same time, owing to the want of proper tools, for the nuts are very hard, and a stone is the only implement used); the kernels are then parched in a common earthen vessel, ground in handmills of this country, and put into a pan; a little cold water is sprinkled upon them, and they are well worked up by the hand (much the same as kneading dough) until the oil separates, when the refuse is well pressed in the hand, which completes the process. The oil is left to stand, and the sediment removed. The cake (in which a great deal of oil remains, owing to the want of a proper press) is generally given to the milch cows or goats.

‘“I never heard of any part being used as manure, but I have no doubt it would form an excellent one.

‘“Some of these Argans grow in clusters, others are in single trees.

‘“I have, &c.,

(Signed)‘“Henry Grace.

‘“To J. H. Drummond Hay, Esq., &c. &c.”’

‘Except a brief notice of the exportation into Europe of Argan oil by the Danish Councillor of State, Georges Höst, who travelled in the kingdoms of Marocco and Fez during the years 1766-1768, the only published account of the uses of the Argan is given in a very little known Danish work, published by P. K. A. Schousboe, entitled “Iagttagelser over Væxtriget i Marokko. Forste Stycke. Kiobnhavn, 1800, 4, 7 Tab.,” of which a German edition appeared in 1801, in 8vo, by J. A. Markussen. It gives an account of some Marocco plants; and, after an introductory sketch of the physical geography of Marocco, it contains descriptions of the plants of the country in Latin and German, with occasional observations in German. The account of the Argan under Retz’s name of Elæodendron Argan is long: first comes a technical description, followed by a history of its synonymy, and then the following notes (kindly translated for us by Mr. Bentham):—

‘“It is surprising that this tree should hitherto have been so little known; as it is found in a country near Europe, and visited by many travellers, who speak in their diaries and descriptions of oil of Argan and of Argan trees, these last as constituting a considerable proportion of the forests of the country. It is, however, not to be met with in the northern provinces, but only towards the south. All those persons, from whom I have sought more accurate information on the subject, are unanimous in stating that it only grows between the rivers Tansif and Sus—that is, between the 29° and 32° N. lat.—and there constitutes forests of considerable extent. It flowers in the middle of June, and the fruit remains on the tree the greater part of the year. The young fruit sets in the end of July or beginning of August, and grows slowly till the rainy season commences, towards the end of September. It now enlarges rapidly and attains its full size during that season, so as that by the middle or end of March it is ripe enough to be gathered for economical uses. Both the fruit and the wood are serviceable, but especially the former; for from the kernel an oil is extracted which is much employed for domestic purposes by the Moors, and is an important production of the country, as it saves much olive oil, which can thus be thrown into commerce, and made to bring money into the country. It is calculated that in the whole Argan region one thousand hundredweight of oil is annually consumed, thus setting free an equal quantity of olive oil for exportation to Europe. Our countryman, Höst, in his ‘Efterretninger om Marokos,’ p. 285, says that the Argan oil is exported to Europe, where it is used in manufactures. Such may have been the case in former times when it might be cheaper; but now there would be no advantage in doing so, as it costs almost as much as olive oil. At present, no Argan oil whatever is exported.

‘“As the practice in preparing this oil is somewhat different from that of common olive oil, it may be useful to enter into some details on the subject. I have myself been present during the whole operation, and consequently speak from experience.

‘“In the end of March the countryman goes into the wood, where the fruits are shaken down from the trees and stripped of their husks on the spot. The green fleshy pericarp, which is good for nothing else, is greedily eaten by ruminating animals, such as camels, goats, sheep, and cows, but especially by the first two. Therefore, when the Arab goes into the woods to collect Argan nuts, he gladly takes with him his herds of the above animals, that they may eat their fill of the green husks whilst he and his family are collecting and shelling the nuts. The horse, the ass, and the mule, on the contrary, do not like this food. When a sufficient quantity of nuts are collected they are brought home, the hard wooden shell is cracked between stones, and the inner white kernels are carefully extracted. These are roasted or burnt like coffee on earthen, stone, or iron plates; in order that they may not be too much done, they are constantly stirred with a stick. When properly roasted they should be all over of a brown colour, but not charred on the outside. The smoke, which is disengaged during the process, has a very agreeable odour. As soon as the kernels have cooled, they are ground in a handmill into a thick meal, not unlike that of pounded almonds, only that it is of a brown colour, and the meal is put into a vessel in which the oil is separated, which is done by sprinkling the mass now and then with hot water, and keeping it constantly stirred and kneaded with the hand. This process is carried on until the mass becomes so hard that it can no longer be kneaded: the harder and firmer are the residuary coarse parts, the more completely is the oil extracted. At the last, cold water is sprinkled upon it, in order, as they say, to expel the last particles of the oil. During the operation the oil runs out at the sides, and is from time to time poured into a clean vessel. The main point to be attended to in order to extract the greatest quantity and the best quality of oil, is that it should be well kneaded, and that the proper proportion of hot water for the extraction of the oil should be used; it is always safer to be sparing of it than to be too profuse. The residuary mass, often as hard as a stone, is of a black-brown colour, and has a disagreeable bitter flavour. The oil itself, when it has settled, is clear, of a light brown colour, and has a rancid smell and flavour. When it is used without other preparations in cooking, it has a stimulating and pungent taste which is long felt on the gums. The vapour which arises when anything is fried in it, affects the lungs and occasions coughing. The common people use it generally without preparation; but in better houses it is the custom, in order to take off that pungency, to mix it previously with water, or to put a bit of bread into it and let it simmer before the fire.

‘“The wood, which is hard, tough, fine-grained, and of a yellow colour, is used in house carpentry, and for other purposes.”’

‘We have been at some pains to distribute the seeds of this plant, with which we have been liberally supplied, to various parts of the East Indies, and to such of our Colonies as appeared suited to the growth of this tree, in respect of climate, &c. It is impossible for seeds to be in better condition; and though the surrounding hard portion of the nut is as thick and solid as that of hickory, those which we ourselves sowed sprouted in less than a month from the time they were put in the ground. The young trees bore the rough treatment of the voyage in midwinter remarkably well; and it is easy to see that this is a plant of ready culture in favourable climates.

‘The value of the husks of the fruit as food for cattle, and the uses of the wood, are mentioned in the above extracts. The nature of the oil seems only to have been considered in relation to olive oil. But vegetable oils are now so much in demand, especially by Messrs. Price & Co., for their great candle-works at Vauxhall, as well as at Birkenhead, near Liverpool, that I was anxious to know the opinion of Mr. G. F. Wilson, the scientific director of those vast establishments, on the nature of Argan oil. Some seeds were consequently communicated to that gentleman, and he lost no time in experimenting upon them, and assuring me that “they contain a large percentage of a very fine oil. We have tried it in several ways, in each case with a favourable result. Some is now being exposed to a severe test, to show how the air acts upon it: I have, however, little fear but that it will answer. Our city friends are inquiring for us the best means of getting a ton or two of the nuts for experiments on a large scale. The only unfavourable point I see is the small weight of kernel to that of hard shell:—

6 Nuts gave—kernel 30 grains
„ „ hard shell 350 grains
„ „ outer husk 193 grains.

The hard shell probably should be sent home with the seed when the kernels are required to yield a sweet oil; for unless prepared with great care, hardly to be expected in a wild country, the oil would not be nearly so sweet if sent home expressed, instead of in its kernel and shell. Perhaps if the kernel is pounded and rammed tightly into casks, we might obtain sweet oil without great waste in freight.”

‘In a botanical point of view this plant is scarcely of less interest than in an economical. It has had the hard fate, often the consequence of being with difficulty procured, to be much misunderstood, and, except by Schousboe, to be imperfectly described; and references are given in works to plants as being identical which have no relationship with it; or to descriptions which, if the same, exhibit little or no resemblance.

‘The first botanist who appears to have noticed this plant is Linnæus, who, in the Hortus Cliffortianus, in 1737, described it, from dried specimens, under the name of Sideroxylon spinosum. “From Clifford’s Herbarium,” observes Mr. Dryander, “now in the possession of Sir Joseph Banks, the Argan was taken up by Linné in his Hortus Cliffortianus; though most of the synonyms are wrong, and consequently the locus natalis (utraque India) which is deduced from them. The specimen in Linné’s Herbarium, under the name of Sideroxylon spinosum, is without flowers, and it is impossible to tell you with any certainty what it is. Clifford’s Herbarium is therefore the only authority by which this species can be ascertained.” Linnæus’s Rhamnus siculus, in the Appendix to the third volume of the twelfth edition of the Systema Naturæ, is, we are assured by Mr. Dryander, “the Argan, or Olive-tree of Marocco (see Höst’s ‘Efterretninger om Marokos,’ p. 284), as appears from the specimen in Linné’s Herbarium, which has a ticket affixed, with the name of Argan of Marocco, and which I have also compared with specimens in Sir Joseph Banks’s Herbarium from Marocco.” The description, too, of Linnæus is very correct. He errs only in considering the plant to be the same as the Rhamnus Siculus pentaphyllos of Boccone (Rhus pentaphyllum, Desf.), which has folia quinata, which latter he introduces into the specific character, but not into the description; and he erroneously followed Boccone in giving Sicily as the native country in addition to Africa, and in adopting the specific name Siculus.

‘In the Species Plantarum of Linnæus, Malabar alone is mentioned as the native country of the Sideroxylon spinosum. Nevertheless, with the exception of Willdenow, who rejects it altogether as “planta valde dubia, forte nullibi obvia,” most of the older authors adopt this name for the Argan of Marocco. Under it, it appears in the first edition of Hortus Kewensis, with the reference to Species Plantarum of Linnæus, and to Commelyn, Hortus Amstelod. tab. 83, where, however, nothing is said of its native country, further than may be surmised by the name adopted from Breynius’s “Lycio similis frutex Indicus spinosus, Buxi folio” (which, as already observed, Willdenow considered to be his Flacourtia sepiaria, from India), and of which the flowers and fruit were unknown to the author. If this were the Argan, it was in cultivation in Holland as early as 1697. At a period not much later, viz. in 1711, according to the Hortus Kewensis, it was introduced into England: “Cult. 1711, by the Duchess of Beaufort, Br. Mus. H.S. 141, fol. 39.” It is indicated as a stove-plant.

‘Sir James Smith, article Sideroxylon spinosum in Rees’s “Cyclopædia” (1819), throws no new light upon the subject; he omits the reference to Commelyn. Retz, in “Obs. Bot.” vol. vi. p. 26, refers the plant to Elæodendron, in which he is followed by Willdenow, and by Schousboe, which latter author has given by far the fullest and best account of the plant botanically and economically.

‘M. Corréa de Serra, “Annales du Museum d’Histoire Naturelle,” 1809, tom. viii. p. 393, tab. v. f. l., has published a very good analysis of the fruit, with very brief characters and no observations. At length Mr. Brown, “Botanicorum facile princeps,” in his invaluable Prodromus, under his Observations on Sapoteæ, says, “Sideroxylon spinosum, L., fructu valde diversum proprium hujus ordinis genus efficit;” and, acting upon his suggestion, Rœmer and Schultes, “Systema Vegetabilium,” vol. iv. pp. xlvi. and 502, have formed of this plant a new genus, Argania, in which they have been followed by Endlicher and Alphonse De Candolle. In this latter work a very full generic character is given, which need not here be repeated.’

‘It is singular that no further allusion to this tree should appear in Jackson’s “Account of the Empire of Marocco” than the following: “Oil Arganic is also in abundance in Suse; it is much used for frying fish and burning-lamps. When used for frying fish, a quart of it should be boiled with a large onion cut in quarters; and when it boils, a piece of the inside of a loaf, about the size of an orange, should be put in; after which it should be taken off the fire and let stand to cool, and when quite cold should be strained through a sieve; without this precaution it is supposed to possess qualities which promote leprosy.”—Dr. Barretta.’

The limited distribution of the Argan is one of its most noticeable features, for as a genus it is not far removed from Sideroxylon, a very widely spread tropical and subtropical genus of both hemispheres, and which reaches its northern limit in Madeira (in the same latitude as that attained by the Argan), where one species, S. Mermulana, Lowe, is found on the rocky heights of the interior. The order is not found in the Canary Islands, but reappears in the Cape de Verdes in a species of Sapota, and is well represented in the humid regions of Western Africa. It would thus appear that Argania and the Madeiran Sideroxylon are two outlying representatives of a very tropical order; and, considering the proximity of the areas they inhabit, and their position in the extreme west of the Old World, they are, in a Botanico-Geographical point of view, plants of a very high interest, as evidences of a relationship between the Floras of these areas, which must originally have been established under very different conditions from those which now prevail.

The Argan was, as stated above, introduced into England in 1811, and was long established on a south wall, but ultimately was killed in an unusually severe winter. Numerous plants were raised, from seed sent by Sir John Hay, by Mr. Grace, and from those brought by myself, and the plant may be seen in the Economic-plant House at Kew. It is of very slow growth, which has disappointed colonists and others, to whom the fruits have been largely distributed from Kew.

FOOTNOTES:

[1]Not El Araisch, SSW. of Tangier on the Atlantic coast, but some place in the interior, and N. of the city of Marocco.

[2]This is no doubt Elæoselinum humile (Ball), which we found near or at the above defined locality. Ball formed a very decided opinion that Jackson’s plant, whether the true Ammoniacum or not, was a species of Elæoselinum.

[3]It is mentioned under this name by Homer in his description of the Island of Calypso. See Daubeny On the Trees and Shrubs of the Ancients, p. 42.

[4]See Cook’s Sketches in Spain, vol. i. p. 5 (1831); and Loudon’s Gardener’s Magazine, Ser. ii. vol. iii. p. 522.

[5]See Bostock’s translation of Pliny, vol. iii. p. 194, &c.

APPENDIX E.

On the Canarian Flora as compared with the Maroccan.

By Joseph Dalton Hooker.

In respect of their botanical relationship to neighbouring Continents, Islands or Archipelagos may be roughly classed under two divisions: namely, those which are situated within a moderate distance of continents, and whose Floras are manifestly derived from them or have had a common origin with theirs; and those which are situated very far from any continents, and whose Floras differ so much either from that of the neighbouring continent or from that of those parts of the continent that are nearest to them, that their origin is a matter of speculation. Of the first division, the British Isles, and probably Vancouver’s Island, in North-West America, are conspicuous instances, their Floras being almost identical with those of the neighbouring continents. St. Helena, the Galapagos, Mauritius, and the Sandwich Islands are instances of the opposite extreme, for their Floras differ widely from those of any continents.

Between these extreme cases there are many intermediate ones; and there are others of an exceptional character, as Iceland, which, though far removed from any part of Europe, has but one flowering plant not found on that continent (Platanthera hyperborea); and Ceylon, which though it is almost united to the Peninsula of Hindostan, yet in many respects differs greatly from that peninsula in its Flora.

Amongst the exceptional cases to continental proximity being accompanied by close botanical relationship is the Flora of the Canarian Archipelago, which differs so greatly from that of the northern part of its neighbouring continent, namely, from that of Marocco,[1] that it demands notice in any work treating of the vegetation of the latter country.

This diversity between the Maroccan and Canarian Floras has been pointed out in Ball’s ‘Introductory Observations to the Spicilegium Floræ Maroccanæ,’[2] where it appears that whilst Marocco, out of 1,627 species of flowering plants, contains 165 endemic plants, it has only 15 which are confined to it and to the Canaries, or to it and Madeira. And Ball goes on to remark (p. 301), in respect of these few species common to both Floras: ‘I think it is safe to say that the facts rather tend to show the accidental diffusion of a few Macaronesian[3] species on the adjacent coast of Africa, than to indicate the direct connection between the continent and those islands within a geological period at all recent.’

Were this diversity due solely or chiefly to the Canaries wanting many Maroccan plants, the inquiry would not be a pressing one; but as to this deficiency is to be added the presence in the Canaries of many indigenous species, and even several genera[4] which are absent in Marocco, and in Marocco the great rarity of endemic genera, of which Argania only is arboreous, the inquiry becomes a very important one, inviting a much closer study than can here be given to it.

The Flora of the Canarian Archipelago, though consisting, like the Maroccan, for the most part of Mediterranean species, yet differs from that of Marocco, in containing many plants that may be classed under the following categories:—

I. It contains many non-Maroccan plants, obviously introduced by man, and not from Europe only, but from various parts of both the Old and New Worlds. This will not appear surprising when it is remembered that Teneriffe was for several centuries the Prime Meridian of Geographers and the resort of all the European ocean-navigators, who took their departure from it on their outward voyages, and made for it on their homeward ones. The Alternanthera achyrantha, a tropical American plant, was no doubt imported into the Canaries, and possibly from thence introduced into Spain (where it is now naturalised). Argemone mexicana is another, and there are still other as conspicuous examples of such foreign introductions. This maritime intercourse can, however, only partially account for the remarkable disproportion between the number of probably introduced plants in the Canaries and in Marocco; and we must take into account the isolation, barbarism, and exclusiveness of the latter country, and the absence of any commercial intercourse between it and the Canaries or the rest of the world.

In Webb and Berthelot’s ‘Phytographia Canariensis’ upwards of fifty plants are enumerated as to which we have little doubt that all have been introduced by man, and none of which have hitherto been found in Marocco. The list includes many weeds of the widest tropical and temperate distribution, as species of Sida, Waltheria, Siegesbeckia, Bidens, Lippia, Physalis, Nicandra, Euphorbia, Alternanthera, Commelyna, and various Cyperaceæ, and Grasses.

II. The Canaries contain many apparently indigenous plants, which, though not Maroccan, are widely distributed elsewhere; these form a large class, and the following are some of the most prominent of them:—

Delphinium StaphysagriaFragaria vesca
Hypecoum procumbensPyrus Aria
Biscutella auriculataPrunus lusitanica
Viola caninaEpilobium palustre
Silene BehenAnthemis fœtida
„ nutans „ coronopifolia
Rhus CoriariaCynara horrida
Spartium junceumLactuca sylvestris
Ulex europæusCressa cretica
Medicago arboreaCalamintha Nepeta
Trigonella hamosaAtriplex glauca
Trifolium striatumEuphorbia serrata
„ squarrosum „ obliquata
„ suffocatum „ Lagascæ
„ filiformeOrchis longibracteata
Lotus angustissimusOphrys tabanifera
Vicia hirsutaIris pallida
Lathyrus odoratusLilium candidum
Alchemilla arvensis

together with various Cyperaceæ, Grasses and water-plants, some of which, and of the above, will no doubt hereafter be found in Marocco.

III. They contain some quite peculiar plants which are more closely allied to endemic species of Marocco than to those of any other country, and may have been derived from species that originally were transported from that country. These are but few, and are almost confined to species of the genus Monanthes, which is limited to these countries and the Cape de Verde Islands, of Cactoid Euphorbiæ, of succulent Sonchi, and of the Kleinia division of Senecio.

IV. They contain plants not found hitherto in Marocco, and which are more closely allied to Mediterranean species than to any others; and these form a very large class. The data for a complete list would require a very careful comparison of the Maroccan species with the species described in the ‘Phytographia’ and discovered since, many of which are unquestionably founded on too slight or too variable characters.[5]

It will be sufficient for present purposes to contrast the results obtained from a selection of genera[6] taken for comparison from Ball’s ‘Spicilegium’ with the same from Webb’s ‘Phytographia’:—

GeneraCanary IslandsMarocco
Number of species in eachSpecies confined to CanariesNumber of species in eachSpecies confined to Marocco
Hypericum8770
Matthiola4330
Cistus2170
Helianthemum63140
Polycarpia6410
Sempervivum232311
Cytisus119114
Lotus106142
Dorycnium3310
Rhamnus3330
Ilex2200
Chrysanthemum1212114
Senecio95111
Doronicum5500
Tolpis5420
Sonchus171260
Convolvulus sect. Rhodorhiza5500
Echium121091
Micromeria[7]171710
Sideritis6572
Teucrium31114
Solanum6220
Scrophularia5391
Digitalis2220
Statice99133
Plantago103111
Beta3210
Euphorbia199226
Ephedra3220
Juniperus2140
Pinus1110
Ruscus2210
Asparagus5461
Scilla4490
Luzula3310
243187=¾20431=⅙

The disproportion between the two Floras in the case of these selected genera is thus well shown. It is most remarkable; the number of endemic species being in the Canaries three-fourths of the whole and in Marocco only one-sixth; and were the peculiar genera of the Canaries added, the disproportion would of course be increased.

The total number of Canarian species enumerated by Webb and Berthelot is about 1,000, of which 367,[8] or more than one-third, are regarded as peculiar to the Archipelago (a very few only of these being also Madeiran); whereas out of 1,627 Maroccan species only 165, or a little over one-tenth, are peculiar. Future discoveries will probably not materially increase the Maroccan proportion of peculiar species; whereas since the publication of Webb’s ‘Phytographia’ many peculiar species (especially of Statice and Crassulaceæ) have been discovered in the Canaries, and but few species common to other countries; and these additions will go far to neutralise any error introduced into the estimate, due to the great number of new species founded on insufficient data which the ‘Phytographia’ includes.

Under this head also should be included the peculiar Canarian genera that appear to be modifications of continental ones. They are Bencomia, closely allied to Poterium, of which there are two species, both confined to one Island (Teneriffe); one of these is also a native of Madeira, where only two individual trees, a male and a female, have ever been seen! Gesnouinia, allied to Parietaria; and Canarina, a monotypic genus allied to Campanula, but having a baccate fruit. Bosea, also a monotypic plant, is wholly unlike any known genus, and is, in some respects, intermediate between the two very distinct natural families—Chenopodiaceæ and Phytolacceæ.

V. Many Canarian plants are representatives of Floras more distant than those of Marocco or Western Europe, and are not found in those countries. These form an exceedingly interesting group, and may be classed according to countries thus:—

a. Oriental.—These are chiefly Arabo-Egyptian, but some of them extend even into Western India, and a few are representatives of tropical India. Some will no doubt yet be discovered in Marocco, especially south of the Atlas; and it is not unreasonable to suppose that such have crossed Africa in a subtropical latitude, and thus reached the Canaries under conditions now operating.

The most remarkable are the following. The genera in capitals have not hitherto been found in Marocco:—

Polycarpon succulentumCampylanthus salsoloides
Visnea MoccaneraTraganum nudatum
Gymnosporia cassinoidesApollonias barbusana
Trigonella hamosaEuphorbia Forskählii
Senecio flavusDracæna Draco
Ceropegia dichotoma

Of the above hardly any have been found west of the Levant, or anywhere between Egypt and the Canaries, except, possibly, in Southern Algeria. Traganum must be reckoned as an African and Oriental desert type, and will probably be found in South Marocco; but Ceropegia is mainly Indian, as is Gymnosporia (Catha cassinoides, Webb). Campylanthus consists of the Canarian species, of a variety or closely allied one in the Cape de Verde Islands, and of a third which extends from Southern Arabia to Scinde. The nearest ally of the Apollonias (Phœbe barbusana, Webb) is a Ceylon tree; and Visnea is nearly allied to the Malayan genus Anneslea. Dracæna Draco is the most interesting of all in the list; for, though the genus abounds in tropical Africa, the Canarian form, which is also a native of the mountains of the Cape de Verde Islands, has only one near ally, the D. Ombet, which is confined to Abyssinia, Southern Arabia, and the intervening Island of Socotra.

b. The peculiar species representing American types inhabiting the Canaries or Madeira, but not found in Marocco, are in some respects even more remarkable than the Oriental.

They belong to the following genera:—

Bowlesia[9] (Drusa oppositifolia, DC.), Clethra, five species of Bystropogon, and Cedronella. Of these Bowlesia is otherwise confined to the tropical Andes of America, one species only extending as far north as Mexico; the Canarian species, which according to Webb is found on rocky shaded places in Teneriffe, from the sea-level to the wooded region, is most closely allied to a Peruvian one. Clethra is a genus which extends from South Brazil to the Northern United States, and is also found in Japan and the Malayan Archipelago. The Macaronesian species most resembles a North American; it is found also in Madeira. Bystropogon is, like Bowlesia, an Andean genus, extending from Peru to Columbia. All the Canarian species belong to a different section from the Andean, and there is one species of the same section in Madeira. Cedronella is a North American and Mexican genus, and the Canarian species differs from all its congeners in its trisect leaves; it is also Madeiran.

Of the Canarian Laurineæ, Persea indica, also a native of Madeira and the Azores, belongs to an American section of that large genus.

c. Tropical and South African types in the Canaries. Of these the most noticeable are two forest trees, belonging to the large tropical genus Myrsine. One of these, M. excelsa (Heberdenia excelsa, Banks) is also found in Madeira; the other, M. canariensis, is confined to the island whose name it bears. The tropical order Sapotaceæ, to which Argania belongs, has no representative in the Canaries, but has one in the Sideroxylon Mermulana of Madeira.

The only almost exclusively South African genus[10] in the Canaries is a species of Lyperia, of which there are numerous Cape of Good Hope species, and one doubtful one in the Somali country (North-East Africa). The widely diffused Cape shrub, Myrsine africana, is found in the Azores and in Abyssinia, but not in the Canaries, Cape de Verdes, Madeira, or Marocco. The two singular shrubs Phyllis and Plocama, consisting each of a single species, of which the Phyllis is found also in Madeira, are representatives of the Anthospermeæ, a very large and conspicuously South African and Australian tribe of Rubiaceæ, and of which the only Maroccan representative is Putoria, a Mediterranean genus of a single species, and which is not Canarian.

The Oreodaphne fœtens of the Canaries and Maderia is now[11] referred to the American, Madagascar, and South African genus Ocotea, and is most nearly allied to a species found in the latter country.

The Maroccan flowering plants are thus grouped by Ball in his ‘Spicilegium Maroccanum’[12]:—

Total number of Maroccan species1.627
Species widely diffused, temperate or tropical467
Of which there are common to Marocco and the Islands300
Maroccan, but not Insular167
Mediterranean species in Marocco995
Of which there are widely spread species common to the Islands and Marocco254
Confined to Marocco and the Islands15
Mediterranean species in Marocco, but not in the Islands726
Maroccan species exclusively165

The proportion of Monocotyledons to Dicotyledons is in Marocco 1 to 4·6, in the Canaries 1 to 6—a very great difference.

The leading natural orders in Marocco and the Canaries respectively are:—

Marocco SpeciesCanaries Species
Compositæ208143
Leguminosæ189104
Gramineæ13477
Umbelliferæ8627
Labiatæ8159
Cruciferæ7329
Caryophylleæ6938

In each country these seven natural orders include nearly half the Dicotyledonous plants. But in the Canaries the Crassulaceæ with 31 species should replace the Cruciferæ, and the Umbelliferæ be excluded.

The natural orders which are indigenous to the three Archipelagos of the Canaries, Madeira, and the Azores, but which are absent in Marocco, and the reverse are:—

In the Archipelago, but not in Marocco.In Marocco, but not in the Archipelago.
Simarubeæ (Cneorum)Berberideæ
PittosporeæCapparideæ
TernstrœmiaceæPolygaleæ
IlicineæAmpelideæ
MyrsineæCoriarieæ
Phytolacceæ (Bosea)Saxifrageæ
MyriceæApocyneæ
Commelyneæ (introduced?)Lentibularieæ
Nyctagineæ
Ulmaceæ
Cupuliferæ
Ceratophylleæ
Alismaceæ
Juncagineæ
Melanthaceæ

In the above lists the Commelyneæ are most probably introduced by man into the Canaries, and the absence of Lentibularineæ, Ceratophylleæ, Alismaceæ, and Juncagineæ in the Archipelago may be due to the want of suitable localities. The total absence of Cupuliferæ in all the Macaronesian Archipelago is inexplicable; and of Quercus especially, a genus so prominently developed in number of species and individuals on both continents, and which further abounds in both the Pliocene and Miocene beds of Europe.

The apparently indigenous Macaronesian genera which are wanting in Marocco are the following. Those in capitals are confined to the Canaries, or to the Canaries and Madeira:—

Malvaceæ
Abutilon
Cruciferæ
Parolinia
Barbarea
Simarubeæ
Cneorum
Celastrineæ
Gymnosporia
Sapindaceæ
Melianthus?
Leguminosæ
Spartium
Ulex
Rosaceæ
Bencomia
Alchemilla
Fragaria
Aquifoliaceæ
Ilex
Pittosporeæ
Pittosporum
Ternstrœmiaceæ
Visnea
Umbelliferæ
Todaroa
Rubiaceæ
Phyllis
Plocama
Compositæ
Chrysocoma
Allagopappus
Vieræa
Doronicum
Serratula
Prenanthes
Campanulaceæ
Musschia
Canarina
Wahlenbergia
Ericeæ
Clethra
Asclepiadeæ
Ceropegia
Convolvulaceæ
Cressa
Boragineæ
Tournefortia
Labiatæ
Bystropogon
Cedronella
Verbenaceæ
Lippia
Solaneæ
Nicandra
Scrophularineæ
Campylanthus
Lyperia
Acanthaceæ
Justicia
Oleineæ
Notelæa
Myrsineæ
Myrsine
Sapotaceæ
Sideroxylon
Primulaceæ
Pelletiera
Chenopodieæ
Traganum
Laurineæ
Persea
Apollonias
Ocotea
Phytolacceæ
Bosea
Urticeæ
Gesnouinia
Myriceæ
Myrica
Aroideæ
Dracunculus
Liliaceæ
Dracæna
Cyperaceæ
Fimbristylis
Cladium
Gramineæ
Chloris
Tricholæna

There are in Marocco, out of a total of 517 genera, 202, included under 67 orders, that have no indigenous species in the Canaries or Madeira. Many of these, about a quarter, being North Maroccan, i.e. only found in parts of Marocco farthest from the Canaries, would not be expected to occur in those islands, were it not that the vegetation of islands near to large continents often most resembles that of a higher latitude on the continent than that in which the islands are situated.

The following is a list of the Maroccan genera which are absent in Macaronesia;—those confined to North Marocco marked *; those which have been found in Macaronesia, but certainly introduced, marked ‖; those in italics have been discovered since our return from Marocco.[13]

Clematis
Thalictrum
*Anemone
Aconitum
Berberis
Rœmeria
Corydalis
Cardamine
Morettia
Anastatica
Draba
*Erophila
Malcolmia
Diplotaxis
Moricandia
‖Lepidium
Thlaspi
Iberis
Hutchinsia
Isatis
Ceratocnemum
*Cakile
*Hemicrambe
Cleome
Capparis
Caylusea
Fumana
Polygala
Velezia
Dianthus
Holosteum
Buffonia
Lœfflingia
Montia
Althæa
Malope
*Radiola
Peganum
Celastrus
Zizyphus
Acer
*Coriaria
Lotononis
Crotalaria
Argyrolobium
*Calycotome
Anthyllis
*Securigera
Coronilla
Colutea
*Glycyrhiza
Hedysarum
Ornithopus
Ebenus
*Pisum
Ceratonia
‖Acacia
Saxifraga
Parnassia
Ribes
*Drosophyllum
*Peplis
*Ecbalium
*Hydrocotyle
Eryngium
Deverra
Hippomarathrum
Kundmannia
*Magydaris
Sclerosciadium
Meum
Heracleum
*Peucedanum
‖*Coriandrum?
Thapsia
Elæoselinum
Gaillonia
Putoria
Callipeltis
Asperula
Crucianella
*Valeriana
‖Centranthus
*Fedia
Nidorella
Nolletia
Micropus
Leysera
Grantia
Anvillea
*‖Xanthium
Achillea
Cladanthus
Echinops
Xeranthemum
Onopordon
Stæhelina
Crupina
*Leuzea
Carduncellus
Catananche
Hyoseris
Phœcasium
Hieracium
Scorzonera
Jasione
Trachelium
*Calluna
Armeria
Limoniastrum
Plumbago
Coris
Argania
Fraxinus
Phyllyrea
*Vinca
Nerium
Dæmia
Glossonema
Boucerosia
*Microcala
Cicendia
Trichodesma
Echinospermum
Rochelia
Nonnea
Cerinthe
Calystegia
*Mandragora
Anarrhinum
*Pinguicula
‖*Acanthus
Vitex
Lycopus
*Satureja
Hyssopus
Cleonia
Zizyphora
*Betonica
Ballota
Boerhavia
Corrigiola
Scleranthus
Sclerocephalus
Polycnemum
Telephium
*Obione
Salicornia
Caroxylon
Passerina
Osyris
Andrachne
*Ulmus?
Celtis
Quercus
Populus
Ceratophyllum
Callitris
Cedrus
Aceras
*Serapias
*Cephalanthera
*Crocus
Leucojum
*Lapiedra
*Tapeinanthus
*Corbularia
Narcissus
Aurelia
*Alisma
Damasonium
*Triglochin
Chamærops
Gagea
*Hyacinthus
Anthericum
*Simethis
*Aphyllanthes
Colchicum
Erythrostictus
*Convallaria
*Schœnus
*Leersia
Lygeum
*Crypsis
*Alopecurus
Macrochloa
*Sporobolus
Ammophila
‖Arundo
*Ampelodesmos
Phragmites
Pappophorum
Echinaria
*Spartina
*Airopsis
Gaudinia
Glyceria
Secale
Elymus
*Lepturus
Anthistiria

These 202 genera, which are absent in the Canaries, comprise upwards of 300 Maroccan species, including Eryngium, with eleven species, Coronilla with eight, Diplotaxis with seven, Narcissus, Anthyllis, Polygala, Passerina, and Quercus five each, besides twenty other genera with three or four each. Not a few of them contain very common and wide-spread species, as do all the above-named, as well as Clematis, Malcolmia, Cardamine, Dianthus, Hedysarum, Heracleum, Asperula, Achillea, Onopordon, Hyoseris, Scorzonera, Phyllyrea, Fraxinus, Calystegia, Anarrhinum, Ballota, Populus, Chamærops. That no species of these or of many of the other genera should exist in the Canaries is inexplicable, considering the position and extent of the Archipelago, and the means of migration which must exist between it and the mainland.

The species common to Macaronesia and Marocco exclusively, are in so far as is at present known:—

Helianthemum canariense, Jacq.Sonchus acidus, Schousb. (In Lancerotte, only a single plant, possibly introduced)
Polycarpia nivea, Ait. (also occurs in C. de Verde)Lithospermum microspermum, Boiss.
Zygophyllum Fontanesii, WebbLinaria sagittata, Poir.
?Cytisus albidus, DC.Chenolea canariensis, Moq.
Ononis angustissima, Lam. (?A form of 0. Natrix)Salix canariensis, Chr. Sm. (rather uncertain)
Astragalus Solandri, Lowe (Madeira only)Romulea grandiscapa, Webb. (Perhaps only a var., but Baker keeps it)
Astydamia canariensis, DC.Asparagus scoparius, Lowe. (Not quite certain)
Bowlesia oppositifolia
Odontospermum odorum, Schousb.

Although it would be out of place here to discuss all the questions raised by this slight sketch of the peculiarities of the Canarian Flora, there are some of them which so intimately bear upon the Maroccan as to awaken attention.

The wonderful development in the Canaries of endemic species belonging for the most part to Mediterranean types, points to the very early introduction of the parent forms of these, and the long isolation both of the Archipelago and its separate islets. It is in accordance with generally accepted views, to assume that the endemic species of each genus have been derived from parent forms originally introduced into one or more of the islets; and that as the descendants of these species spread over the Archipelago they were exposed to different conditions in each islet, resulting in their varying, and in the segregation and conservation of different local varieties each in its own insular birth-place; a supposition which is in accordance with the fact that those endemic species are really very local, many being confined to a single islet. In Marocco the parent forms of its Flora would be exposed to no such diverse conditions, and the areas in which varieties occurred, not being isolated, would be exposed both to invasion on all sides by other plants, and to destruction by agencies that affected the whole surrounding country, as drought, floods, insects, and birds.

The tropical types in the Canaries, with the exception of the Egypto-Arabian and the trees mentioned under V. c., are chiefly weeds of wide distribution, which have not reached Marocco, because of its want of ports and its limited commerce.

Finally the Dracæna, together with the tropical trees of Myrsineæ, Sapotaceæ (in Madeira), and Laurineæ, and the Egypto-Arabian types, suggest the hypothesis that at a very remote period these and many other plants of warmer and damper regions flourished in the area included in North-West Africa and its adjacent islands, and that they have been expelled from the continent by altered conditions of climate, but have been preserved in the more equable climate and more protected area of the Atlantic Islands.

Ball, who has given me valuable aid on many points discussed in this article, directs my attention to the important differences that exist between the vegetation of the eastern group of the Canary Islands—Fuertaventura, Lanzarote, and the adjacent islets—the ‘Purpurariæ’ of authors, and the western group, including Teneriffe, Grand Canary, &c.

In the first place, nearly all the characteristic Canarian types are absent in the eastern group. Out of fifty-four genera above enumerated as present in the Canaries but wanting in Marocco, two are in the Canaries confined to the eastern islands: one of these, Traganum, is an African desert type, probably to be found in South Marocco; the other, Melianthus, a Southern African plant, and scarcely indigenous. Of the remainder Plocama alone is certainly present, and three other generic types probably exist in that group; while forty-eight genera, including eight out of nine peculiar to the Canaries, are apparently absent. In the next place several characteristic desert plants, such as Oligomeris subulata, Ononis vaginalis, Convolvulus Hystrix, and Traganum nudatum, are present in the ‘Purpurariæ,’ but absent from the western islands.

Although the Flora of the Purpurariæ is incompletely known, and our acquaintance with that of the neighbouring African coast between the rivers Sous and Draha is extremely imperfect, these facts tend to prove that there is a closer botanical relationship between the eastern islands and the adjoining continent than there is between them and the western portion of the Canarian Archipelago. Such relationship might be brought about in three different ways.

1. The greater dryness and heat of the eastern islands may have favoured the immigration of African forms, and at the same time led to the destruction, or weeding out, of the characteristic Canarian types. In this case the cause would be of a purely local and climatic character.

2. We may believe in the trans-oceanic migration of some African species to the nearer islands, along with the transport of some Canarian species (those enumerated in [p. 416], and others which may be hereafter found) to the neighbouring continent.

3. An ancient extension of the continent to the Purpurariæ, leaving the other islands separated by deep sea.

It is an objection to the latter hypothesis that a profoundly deep ocean bed lies between the lines of 100 fathom soundings that girdle the islands and the African coast respectively; and that while the 100 fathom line extends about thirty miles from the coast of the continent, it is never more than five miles, rarely more than one or two, from those of the islands.

In favour of the hypothesis of trans-oceanic transport it may be remarked that the distance between the African coast and Fuertaventura is not more than seventy miles, and that a moderate change of level of about 600 feet would reduce that distance by one-half, while it would but slightly affect the interval that separates the Purpurariæ from the other islands.

Among the possible causes leading to an interchange of species between the Purpurariæ and the African coast the agency of man must not be omitted. The fishermen of those islands were formerly in the habit of visiting some points on the opposite coast, although intercourse of this kind has almost ceased in recent times.

It must be observed that our knowledge of the vegetation of the Canary Islands is yet incomplete. Although several additions to the Flora have been published by C. Bolle and others, no supplement to Webb’s ‘Phytographia’ has been published. Several additional species exist in herbaria, besides those that may be hereafter found.

So little is known of the geology of Marocco, that there are no data for ascertaining whether during antecedent geological periods it contained a more tropical Flora than now; but evidence in support of such a hypothesis is forthcoming in Madeira, where fossiliferous beds which have been referred to ‘some part of the Pliocene period’[14] have been discovered, containing leaves referable in part to existing species of Madeiran plants, and in part to extinct ones of tropical aspect;[15] and it is well ascertained that during preceding geological periods Western Europe was clothed with a vegetation that suggests a very much warmer climate than now prevails, and of which vegetation the Laurus nobilis in the south-west of the continent has been supposed to be a surviving representative.

In Grand Canary, also, Upper Miocene beds exist, containing numerous species of fossil shells, of which one is an Oregon species, and another tropical African; and in more recent deposits of the same Archipelago many shells have been found which no longer inhabit the adjacent seas, including tropical West African, Mozambique, and Mediterranean species.

We can form no conception of means of transport from the American continent that would transfer the parent species of Bowlesia and of the Bystropogons from the Andes to the Atlantic islands; and we can but hazard the assumption that, at some very distant date, these genera existed in more eastern parts of America, from whence seeds were transported across the ocean. On the other hand, the transport of parent forms or existing species from the continents of Europe and Africa to the Atlantic islands may have been much facilitated by greater extensions of land in bygone ages. Madeira, the Canaries, and the Cape de Verde Islands, are all supposed to stand on a submarine platform which skirts the coasts of Western Europe and North-Western Africa, and whose submerged margin immediately to the westward of the position of the islands descends rapidly to a profound depth. The westward margin of this platform was possibly the coast-line in Miocene times. An elevation of its surface of a few hundred feet would approximate the islands to the mainland very materially, and greatly facilitate transport. That they were, however, ever united to the continent is opposed to the views of most competent geologists. Lyell, speaking of this, says: ‘The general abruptness of the cliffs of all the Atlantic islands, coupled with the rapid deepening of the sea outside the 100 fathom line, are characters which favour the opinion that each island was formed separately by igneous eruptions, and in a sea of great depth.’ Moreover, the Azores, whose botany in so many respects resembles that of the other Atlantic islands, as distinguished from that of the continent, are enormously more distant from the mainland; and these islands stand on a platform of their own, separated from the continental one by an ocean of profound depth; so that any theory of transport which applies to the Canarian and Madeiran Archipelagos, should apply also to the Azorean.

It remains a point of some nicety to decide whether the Macaronesian islands should be regarded as a Botanical province apart from the Mediterranean, or a sub-division of the latter. The assemblage of American and Oriental genera which their Flora contains, together with the arboreous representatives of tropical Laurineæ, all so entirely foreign to the European Flora, would give it a title to be called a Botanical province; and to this as a further title is the prevalence of a considerable proportion of North European plants, in the Northern Archipelago especially. On the other hand, fully two-thirds of the species are typical of the Mediterranean Flora, and by far the majority of the remainder are derivative species of the same origin; so that, on the whole, I am disposed to regard it as a very distinct sub-division of the Mediterranean province, which owes its peculiarities partly to the conservation of types once common to West Europe and North Africa, but which have been eliminated in those regions, and partly to the effect of isolation and climate on the progeny of species still existing in those regions.

FOOTNOTES:

[1]The Canary Islands are situated about 3° farther south, and 280 miles distant from Mogador. They are thus opposite a much more hot and arid part of the African coast than that north of the Atlas. The large island of Fuertaventura is only about 70 miles from the continent south of Oued Noun.

[2]Journ. Linn. Soc. vol. xvi. p. 297.

[3]A term first applied by Webb to the Flora of the Canarian Archipelago, but which should also include the Flora of Madeira (as Ball makes it do in the above mention of it), the Azores, and perhaps also of the Cape de Verde Islands, which together form either a distinct botanical province, or a marked subdivision of the Mediterranean province.

[4]No less than nine very distinct genera are confined to the Canaries or Madeira or both:—Parolinia, Bencomia, Visnea, Phyllis, Plocama, Canarina, Musschia, Bosea, and Gesnouinia. The only endemic genera in Marocco are Argania, Hemicrambe, Ceratocnemum, and Sclerosciadium.

[5]On the other hand, many peculiar species have been added to the Canarian Flora since the date of the publication of the Phytographia (1836-50).

[6]The genera, which are unduly multiplied in the Phytographia, are here reduced to the standard adopted in the Spicilegium. The species are, unfortunately, also inordinately multiplied in the former work, which seriously vitiates the table: this, however, it is impossible to set right. On the other hand, some of the Canarian genera have been largely added to by later explorers.

[7]In this, the most curious case of all, the species were elaborated by Bentham, and may, therefore, be depended upon. A second Maroccan Micromeria, allied to a Canarian one, has been found by M. Cosson’s collectors, as I am informed by Ball, whilst this sheet was passing through the press.

[8]This estimate is subject to the same deductions as I have referred to in note 2, [p. 407.] On the other hand, were the many obviously introduced species to be struck out of Webb and Berthelot’s enumeration, the proportion of peculiar species would be considerably augmented.

[9]Whilst this sheet was passing through the press, I am informed that M. Cosson’s collectors have found Bowlesia in South Marocco. No doubt this is another case of that accidental diffusion of Macaronesian species alluded to by Ball. (See [p. 405.])

[10]The Cape of Good Hope mountain plant, Melianthus comosus, found at the south end of Fuertaventura, must be assumed to have been introduced by man into that island.

[11]In Bentham and Hook, f., Gen. Plant. (vol. iii. ined.), the Laurineæ are described by Bentham, who has determined, for the first time, the proper position of the Canarian Laurels.

[12]Journ. Linn. Soc. Bot. vol. xvi.

[13]M. Cosson has published in the 22nd volume of the ‘Bulletin of the Botanical Society of France’ a list including the plants received from his collectors in South Marocco up to the year 1874.

[14]Lyell’s Principles of Geology, ed. 11, vol. ii. p. 410.

[15]Lyell’s Student’s Elements of Geology, ed. 2, pp. 538, 539.

APPENDIX F.

Comparison of the Maroccan Flora with that of the Mountains of Tropical Africa.

By Joseph Dalton Hooker.

As was to have been anticipated, the Maroccan Flora contains most of the European species which have been collected on the mountains of Abyssinia and of the Bight of Biafra, which alone of the tropical African Alps have been botanically explored. Of these the former have been visited by Schimper and various collectors; whilst the mountains of the pestilential West African coast, of Fernando Po, 9,500 feet, and the Cameroons Mountains, upwards of 13,000 feet, have been ascended for botanical purposes only by Gustav Mann, when employed for the Royal Gardens of Kew.

The results of the latter were published by myself in the ‘Journal of the Linnæan Society of London’ (vol. vii. p. 171), from whence the following observations are for the most part extracted. They included 26 European species, gathered at elevations between 5,000 and 10,000 feet. Most of them are also natives of the Abyssinian Alps, and two-thirds of them are also Maroccan, whilst others will probably yet be found in the latter country.

The following is a catalogue of all the European plants found in the upper regions of the Cameroons Mountains and Fernando Po:—

HeightWhere found
feet
Cardamine hirsuta7,000-10,000Marocco and Abyssinia
Cerastium vulgatum (viscosum Fr.)8,000„ „
Radiola Millegrana7,000Marocco
Oxalis corniculata7,000-8,500Marocco and Abyssinia
Umbilicus pendulinus7,000-10,000„ „
Sanicula europæa4,000-7,500Abyssinia
Galium rotundifolium7,000-12,000Marocco and Abyssinia
Galium Aparine7,000-10,000Marocco and Abyssinia
Scabiosa succisa10,500?
Myosotis stricta8,000-10,000Marocco and Abyssinia
Limosella aquatica9,000-10,000Abyssinia
Sibthorpia europæa7,000-7,500„
Solanum nigrum7,000-11,000Marocco and Abyssinia
Rumex obtusifolius7,000Abyssinia
Parietaria mauritanica7,000-8,000Marocco and Abyssinia
Trichonema Bulbocodium7,000-9,000„ „
Juncus capitatus7,000Marocco
Luzula campestris8,000-10,000Abyssinia
Deschampsia cæspitosa9,000-12,000„
Aira caryophyllea7,000-8,000Marocco and Abyssinia
Poa nemoralis7,000-10,000Abyssinia
Kœleria cristata8,000-12,000„
Vulpia bromoides7,000-10,000Marocco and Abyssinia
Festuca gigantea8,500Marocco
Brachypodium sylvaticum7,000Abyssinia
Andropogon distachyus7,000Marocco and Abyssinia

The most remarkable features of the Temperate vegetation of these West African tropical mountains are:—

1. Its poverty.

2. The preponderance of Abyssinian genera and species.

3. The considerable proportion of European plants.

4. The paucity of South African genera and species.

5. The great rarity of new genera.

6. The absence of St. Helena and Canarian types.

Upon each of these propositions I have a few general remarks to offer.

In the poverty of their Flora the Cameroons range and Peak of Fernando Po seem to partake of the characteristics of the Abyssinian Alps. We know far too little of the physical geography of either of these districts to hazard many conjectures upon this point, which must to a certain extent be dependent on the arid volcanic nature of the soil and the limited area of the Temperate region. Mr. Mann spent many weeks, and at various seasons, in his explorations, and yet 237 flowering plants were all that rewarded his toil. Geological causes have probably had, in the case of the Cameroons Mountains, much to do with the dearth of species, some parts of the range even now presenting evidence of subterranean heat.

The preponderance of Abyssinian forms is proved by almost all of the genera and half the species being natives of Abyssinia, and by many other species being very closely related to, or obvious representatives of, plants of that country. There are, further, several of the genera and many of the species peculiar to Abyssinia and the peaks of Biafra.

The number of European genera amounts to 43, and species to 26, the greater part of which are British. Very few of them extend into South Africa. The greater part are Abyssinian; the remarkable exceptions being Radiola, Scabiosa succisa, Luzula campestris, and Festuca gigantea, all of which, however, may have been hitherto overlooked in Abyssinia.

I find no other evidence of relationship between the Biafran mountain Flora and that of Marocco than what is afforded by the European species common to both. In most other respects the Floras differ totally, the other mountain plants of Biafra being Abyssinian or Cape types, or more nearly related to tropical African ones.

APPENDIX G

On the Mountain Flora of Two Valleys in the Great Atlas of Marocco.

By John Ball.

Although an attempt to discuss the character and relations of the Flora of a region so wide and so little known as the mountain region of the Great Atlas would as yet be quite premature, it appears that the materials at our disposal suffice for an examination of the vegetation of the valleys lying south and south-west of the city of Marocco, which may be an acceptable contribution to botanical geography. For this purpose it seems best to limit the discussion to the two valleys where our collections were sufficiently extensive to give a tolerably complete representation of the vegetation, as far as this was developed at the season of our visit, and to exclude altogether the plants found along the skirts of the great range below the level of about 1,200 metres above the sea. The Flora of the zone below that level is largely mixed up with extraneous elements, represented by plants of the low country that reach the base of the mountains, but do not penetrate the interior valleys, and if these had been admitted the special features of the mountain Flora would have become less apparent. The plants collected in the course of a somewhat hurried excursion from Seksaoua, when we reached a height of about 1,600 metres, have been designedly omitted. On such occasions attention is to a great extent monopolised by the new and rare species not hitherto seen in the same region, while comparatively familiar forms are less carefully noted. Collections made under such circumstances rarely give a moderately complete report of the vegetable population.

In ascending to the higher region of mountains that approach to the limit of vegetation the absolute number of species is so much smaller that this source of error is far less apparent; and it is not likely that in the two ascents which we made to the dividing ridge of the Atlas many species that came within our range of vision were overlooked. The following tabular arrangement shows that in the Aït Mesan valley, where we spent the greater part of six days, we collected 375 species of phænogamous plants, to which have been added three subsequently found there by MM. Rein and Fritsch; while in the Amsmiz valley only 223 species—or less than three-fifths of the above number—were collected. Of these 146 species are common to both valleys; so that our list does not in all exceed 455 species of flowering plants, to which I have added 10 vascular cryptogams, of which two only were found in the Amsmiz valley.

In the following list I distinguish a middle zone, extending from 1,200 to 2,000 metres above the sea, and a superior zone including all above that limit; the latter corresponding pretty nearly with the sub-alpine and alpine zones of the higher mountains of Europe. With reference to their distribution I have divided the species into four categories: 1, Mid-European, those extending to Central Europe, of which more than three-fourths belong to the British Flora: 2, wide-spread Mediterranean, extending beyond the bounds of the three adjoining regions, Algeria, the Spanish peninsula, and the Canary Islands: 3, confined to adjoining regions; that is, to one or more of those just enumerated: 4, endemic, known only in the Great Atlas, or the neighbouring provinces of Marocco.

Tabular View of the Mountain Flora of the Great Atlas, showing the distribution of the Species found in the Valleys of Aït Mesan and Amsmiz. S indicates the superior zone from 2,000 m. to 3,500 m. above the sea; M the middle (or mountain) zone, from 1,200 m. to 2,000 m.

Name of SpeciesAït MesanAmsmizMid-EuropeanWide-spread MediterraneanConfined to adjoining regionsEndemic
Ranunculus spicatus, L. var.—S——*—
„ atlanticus, BallMM—— *
„ bulbosus, L. var.MM*——
„ arvensis, L.M S—*——
„ muricatus, L.—M—*——
„ Reinii, nov. sp.S————*
Aquilegia vulgaris, L. var.—M*———
Delphinium Balansæ, B. et R. var.?SS——*—
Berberis cretica, L. var.—S—*——
Papaver tenue, BallM SM S———*
„ rupifragum, B. et R. var.SS——*—
Rœmeria hybrida, D.C.—M—*——
Hypecoum pendulum, L. M—* —
Corydalis heterocarpa (Dur.)M———*
Fumaria officinalis, L.S—*———
„ parviflora, Lam.MM*———
„ media, Lois var.M—*———
„ agraria, Lag. var.M——*——
„ tenuisecta, Ball.M————*
Nasturtium officinale, R. Br.S—*———
„ atlanticum, BallS————*
Arabis albida, Stev.S——*——
„ erubescens, BallS ——*
„ auriculata, Lam.SS*———
„ decumbens, BallSS ——*
„ conringioides, BallSS —*
Cardamine hirsuta, L. var.—S* ——
Alyssum alpestre, L. vars.MM* —
„ montanum, L. var. S*— —
„ campestre, L.M S—*—
„ calycinum, L.M S—*——
„ spinosum, L.S——*—
Draba hispanica, Boiss.—S——*
Sisymbrium Thalianum S*——
„ runcinatum, Lag. var.—M——*
Erysimum australe, Gay, var,M SM S*———
Brassica rerayensis, BallS——— *
Capsella bursa-pastoris, L.S—*———
Lepidium nebrodense, Raf. var.S —*—
Biscutella lyrata, L. var.S——*——
Thlaspi perfoliatum, L. et var.SS* ——
Hutchinsia petræa, R. Br.SS* ——
Isatis tinctoria, L. var.MM*— —
Crambe hispanica, L.—M— *—
Capparis spinosa, L.M——* —
Reseda attenuata, BallS—— *
„ phyteuma, L. M*— —
„ lanceolata, Lag.M—— *
Cistus polymorphus, Willd.M——*
Helianthemum niloticum, L. var.?S——*—
„ rubellum, Presl.—M *—
„ glaucum, Cav.—M *—
„ virgatum, Desf. et var.MM——*
Fumana glutinosa, L.MM—*——
„ calycina, Claus.M———*—
Viola tezensis, Ball—S———*
Polygala rupestris, Pourr.MM—* —
„ Balansæ, Coss.MM———*
Dianthus attenuatus, Sm.M S ——*—
„ virgineus, L.MM*———
Tunica compressa, Desf.—M——*—
„ prolifera, L.MM S*———
Silene inflata, Sm. var.M—*———
„ nocturna, L.M— *——
„ corrugata, BallM ———*
„ muscipula, L.M *——
„ italica, L.MM *——
Holosteum umbellatum, L.SS* ——
Cerastium glomeratum, Thuill.SS*———
„ brachypetalum, Desf.SS*— —
„ arvense, L.M SM S*——
Stellaria media, L.M—*——
„ uliginosa, Murr.S—*——
Arenaria pungens, Clem. et var.SS——*
„ serpyllifolia, L.SS*——
„ procumbens, Vahl.M —*—
„ fasciculata, GouanS—*———
„ setacea, Thuill. var.SS*———
„ verna, L. var.SS*———
Buffonia tenuifolia, L.M——*——
Sagina procumbens, L. var.SS* ——
„ Linnæi, Prese.S—*— —
Polycarpon tetraphyllum, L. M*——
„ Bivouæ, J. GaySS *—
Montia fontana, L.S *———
Hypericum perforatum, L.M—*———
„ coadunatum, Chr. Sm. var.M———*—
Malva sylvestris, L.M—* —
„ rotundifolia, L.M S—*— —
Linum corymbiferum, Desf.M———*—
Fagonia cretica, L.M——*——
Geranium malvæflorum, B. et R.SS——*—
„ pyrenaicum, L.MM*———
„ molle, L.S—*———
„ rotundifolium, L.MM*———
„ lucidum, L.SM S*———
„ robertianum, L. var.M—*———
Erodium Jacquinianum, F. et M.M S— *——
„ malacoides, L.M——* —
„ guttatum, W.M———*—
Oxalis corniculata, L.M—*— —
Ruta chalepensis, L.M S——*—
Rhamnus Alaternus, L.MM—*——
„ lycioides, L.—M——*
Acer monspessulanum, L.M—*———
Pistacia Lentiscus, L.MM—*——
Lotononis maroccana, BallMM—— *
Argyrolobium Linnæanum, Walp. var.M S——*—
„ stipulaceum, Ball—M———*
Adenocarpus anagyrifolius, Coss.MM———*
Genista dasycarpa (Coss.)M————*
„ myriantha, Ball M———*
„ florida, L. var.M———*—
Cytisus Balansæ, Boiss. var.SS——*—
„ albidus, D.C.—M———*
„ Fontanesii, SpachM———*—
Ononis atlantica, Ball—M———*
„ antiquorum, L.—M—*——
Trigonella monspeliaca, L.M —*——
„ polycerata, L. et var.SM——*—
Medicago lupulina, L.M—*———
„ suffruticosa, Ram.SS——*—
„ turbinata, W. vars.MM—*——
„ denticulata, W.—M*———
„ minima, Lam.M—*———
Melilotus indica, All.M——*——
Trifolium atlanticum, BallS ———*
„ glomeratum, L. var.S—*———
„ repens, L.M—* ——
„ humile, BallS *
„ tomentosum, L.M— * —
Anthyllis Vulneraria, L. et var.M S—*———
„ tetraphylla, L.M——*——
Lotus cytisoides, D. C.M——*—
Coronilla pentaphylla, Desf.M———*
„ ramosissima, BallM—— —*
„ minima, L.M—* ——
„ scorpioides, L.M——*——
Hippocrepis atlantica, BallS—— —*
„ multisiliquosa, L. var.M——*——
Psoralea bituminosa, L.M——*——
Colutea arborescens, L.M—*———
Astragalus sesameus, L.M——*——
„ Reinii, BallS————*
„ Glaux, L. var.M———*—
„ atlanticus, BallM————*
„ ochroleucus, Coss.S————*
„ incurvus, Desf.M———*
Vicia onobrychoides, L.S——*——
„ glauca, Presl. var.S——*—
„ sativa, L. vars.M S—*———
Lathyrus aphaca, L.M—*———
„ sphæricus, Retz.M —*—
Ceratonia siliqua, L.MM *——
Prunus prostrata (Labill.)SS *——
Poterium sanguisorba, L.M *———
„ verrucosum, Ehrnb. var.M —*——
„ anceps, BallSS———*
„ ancistroides, Desf. M——*
Rosa canina, L. var.MM*——
„ Seraphini, Viv.SS—*——
Saxifraga globulifera, Desf.M SM S —*—
„ tridactylites, L.S *———
„ granulata, L.SS*———
Ribes Grossularia, L.M SS*——
Cotyledon umbilicus, L.MM*———
Sedum modestum, BallMM———*
„ dasyphyllum, L. var.MM*———
„ acre, L.M SM S*———
Sempervivum atlanticum, BallM————*
Monanthes atlantica, Ball—S———*
Bryonia dioïca, Jacq.M—*— —
Eryngium Bourgati, Gouan, var.—S——*—
„ variifolium, Coss.M—— —*
Bupleurum spinosum, L.SS—* —
„ acutifolium, Coss.MM——*—
„ oblongifolium, BallM—— —*
„ lateriflorum, Coss.M—— —*
Deverra scoparia, Coss. et Dur.—M——*—
Carum mauritanicum, B. et R.MM——*—
Pimpinella Tragium, Vill.M——*——
Tinguarra sicula, L.—M—*——
Scandix pecten Veneris, L.M—*———
Kundmannia sicula, L.M —*——
Meum atlanticum, Coss. S ——*
Heracleum Sphondylium, L.S * ——
Bifora testiculata, L.M——*——
Caucalis latifolia, L.M—*———
„ daucoides, L.—M*———
„ leptophylla, L.M SM—*——
„ cœrulescens, Boiss.M———*—
Elæoselinum meoides, Desf.M——*——
Hedera Helix, L.M—*———
Sambucus nigra, L.MM*———
Viburnum Tinus, L.—M—*——
Lonicera etrusca, SantiM —*——
Putoria calabrica, L.M——*——
Callipeltis cucullaria, L.—M—*——
Rubia tinctorum, L.S— * —
„ peregrina, L. et var.M—*— —
Galium Poiretianum, BallM———*—
„ corrudæfolium, Vill.—S—*——
„ sylvestre, Poll. var.S—*———
„ acuminatum, BallM S ———*
„ noli-tangere, Ball—M———*
„ tunetanum, Lam.M———*—
„ parisiense, L. var.M SM S*———
„ tricorne, With.SM*———
„ spurium, L.M SS*———
„ murale, L.—M—*——
Asperula aristata, L. var. M *——
„ hirsuta, Desf.M———*—
Crucianella angustifolia, L.M *——
Sherardia arvensis, L.M—* ——
Centranthus angustifolius, D. C.M S— * —
„ calcitrapa, L.SS—*
Valerianella discoidea, W.M——*——
„ auricula, D. C.M SM*——
„ carinata, Loisel.—M*——
Scabiosa stellata, L.M——*——
Pterocephalus depressus, Coss.M SM———*
Bellis annua, L. var.—M—*——
„ cœrulescens, Coss.M SM S ——*
Evax Heldreichii, Parl.S *——
Micropus bombycinus, Lag.MM—* —
Filago germanica, L. var.—M* —
„ heterantha, Rafin.SS—*——
Filago gallica, L.MM*———
Phagnalon saxatile, L.MM—*——
„ atlanticum, BallM————*
Gnaphalium luteo-album, L.MM* ——
„ helichrysoides, BallS ———*
Inula montana, L.M —*
Pulicaria mauritanica, Coss.M——— *
Odontospermum aquaticum, L.M——*——
Anacyclus depressus, BallS— ——*
„ valentinus, L.—M—*——
Achillea ligustica, All. et var.MM—*—
Anthemis tuberculata, Boiss.—S— *
„ heterophylla (Coss.).MM— —*
Chrysanthemum Gayanum (Coss.) et var.M SM S——*—
„ atlanticum, BallSS———*
„ Catananche, BallSS———*
Senecio lividus, L. var.S——*——
„ giganteus, Desf.MM——*—
Calendula maroccana, BallMM———*
Echinops spinosus, L.M——*—
Xeranthemum modestum, BallMM———*
Atractylis cancellata, L.MM—*——
„ macrophylla, Desf.M———*—
Carduus macrocephalus, Desf.M——*——
„ Ballii, H. fil.M SM———*
Cnicus echinatus Desf.M——*——
„ ornatus, Ball—M———*
„ chrysacanthus, BallM— —*
„ Casabonæ, L.S——*——
Stæhelina dubia, L. var.—M *——
Centaurea incana, Desf. var.M— *——
„ Salmantica, L. var.MM *——
Carthamus cœruleus, L. var.M S——*——
Carduncellus lucens, BallM S————*
Catananche cærulea, L. et var.MM—*——
„ cæspitosa, Desf.—S— *—
Tolpis umbellata, Bert.M——*—
Rhagadiolus stellatus, L.MM—*
Crepis taraxacifolia, Thuil. var.M *—
„ Hookeriana, Ball—S———*
Phæcasium pulchrum, L.SS*———
Hieracium Pilosella, L.SS* —
Hypochæris glabra, L. var.M—*———
„ leontodontoides, BallS ———*
Leontodon autumnalis, L. var.S—*— —
„ Rothii, BallMM * —
„ helminthioides, Coss.—M ——*
Taraxacum officinale, Wigg. var.SS—*——
Lactuca viminea, L.M———*
„ tenerrima, Pourr.MM*—
„ saligna, L.M—*—
Sonchus oleraceus, L.—M—* —
„ asper, Vill.M—* —
Microrhynchus nudicaulis, L.M—*———
„ spinosus (Forsk.)M—*———
Scorzonera undulata, Vahl.M——*——
„ pygmæa, S. et S.S——* —
Jasione atlantica, BallS———*—
Campanula maroccana, BallM——*—
„ rapunculus, L.M————*
„ Lœfflingii, Brot.M——— *
Specularia falcata (Ten.)—M*— —
Trachelium angustifolium, Schousb.M—— *—
Arbutus Unedo, L.M * —
Armeria plantaginea (All.)? S———*
Asterolinum linum-stellatum, L.—M*———
Anagallis linifolia, L. et var.M S—*———
Jasminium fruticans, L.M——*——
Fraxinus oxyphylla, M. B.MM—*——
„ dimorpha, Coss. et Dur.M——* —
Phillyrea media, L.MM—* —
Olea europæa, L.MM— *—
Nerium Oleander, L.MM—*——
Convolvulus Cantabrica, L.MM *——
„ undulatus, Cav.M——*——
„ sabatius, Viv. var.M S——*——
„ siculus, L.—M—*——
„ althæoides, L.M— *——
Hyoscyamus albus, L.M——*——
Anchusa atlantica, BallM——*——
Lithospermum arvense, L.S— *——
„ incrassatum, Guss. var.SM S ——*
„ apulum, Vahl.M * ——
Myosotis sylvatica, Hoffm. var.SS—*——
„ hispida, Schlecht. var.S——*——
„ stricta, LinkSS*———
Cynoglossum Dioscoridis, Vill. et var.M S—*———
Rochelia stellulataMM*———
Verbascum calycinum, BallM— *——
Celsia maroccana, BallM ———*
Linaria ventricosa, Coss.MM ——*
Linaria heterophylla, Desf.MM—*——
„ galioides, Ball et var.S————*
„ arvensis, L. var.SM S—*——
„ marginata, Desf.S———*—
„ lurida, BallS —— *
„ Munbyana, Boiss. et Reut.M———*—
„ Tournefortii (Poir.)SS——*—
„ rubrifolia, Rob. et Cast.M——*——
Anarrhinum pedatum, Desf.M———*—
„ fruticosum, Desf.—M *——
Scrophularia canina, L. var.?—M*———
Digitalis lutea, L. var.M—*——
Veronica Beccabunga, L.S—*———
„ cuneifolia, Don. var.SS—*——
„ arvensis L. et var.SS*———
„ triphyllos, L.—M* ——
„ agrestis, L.S—*——
„ hederifolia, L. et var.M SM S*——
Phelipæa cærulea, Vill.M *———
Orobanche Hookeriana, BallM————*
„ barbata, Poir.M———*—
Lavandula dentata, L. et var.MM—*——
„ tenuisecta, Coss.M SM S———*
Mentha rotundifolia, L.S—*———
Thymus saturejoides, Coss. et var.MM———*
Thymus Serpyllum, L. var.S—*— —
„ lanceolatus, Desf. var.MM—*——
„ maroccanus, BallM————*
Micromeria microphylla, BenthM——*—
Calamintha graveolens, M. B.M——*——
„ alpina, L. var.SS*———
„ atlantica, BallM SM S— *—
Hyssopus officinalis, L.SS*———
Salvia Maurorum, Ball—M———*
„ clandestina, L. var.M S—*——
Nepeta multibracteata, Desf.M— —*—
„ atlantica, Ball—M— —*
Sideritis villosa, Coss.M SM S—— *
„ scordioides, L. var. S*———
Lamium amplexicaule, L.M SM S*———
„ album, L. var.M—* ——
Teucrium granatense, B. et R. var.—M——*—
„ polium, L. vars.MM—*——
Ajuga Iva, L.M SM S—*—
Globularia Alypum, L.—M—*——
Plantago albicans, L. var.—M—*——
Plantago coronopus, L. var.M SM*———
„ mauritanica, B. et R.—M——*
Paronychia argentea, Lam.M SM S—*——
„ capitata, Lam. var.M S * —
„ macrosepala, Boiss. et var.M —*——
Scleranthus annuus, L. var. S* —
Polycnemum Fontanesii, Dur. et Moq.M———*
Rumex scutatus, L. var.M S—*——
„ Papilio, Coss.M————*
Polygonum aviculare, L. *———
Daphne Gnidium, L.MM—*——
„ Laureola, L.M—*———
Thymelæa virgata, Endl. var.MM——*—
Osyris alba, L.—M—*——
Aristolochia Pistolochia, L.M——*——
Euphorbia rimarum, Coss.MM———*
Quercus Ilex, L. et var.MM S—*——
Salix purpurea, L. var.M—*———
Populus alba, L. var.MM*———
„ nigra, L.M—*———
Ephedra altissima, Desf.M——*——
„ procera, F. et M.M —*——
Pinus halepensis, Mill.—M—*——
Callitris quadrivalvis, Vent.MM——*—
Juniperus oxycedrus, L.MM * —
„ phœnicea, L. et var.M *——
„ thurifera, L.S *—
Orchis pyramidalis, L.M—*———
„ latifolia, L.S—*———
Ophrys apifera, Huds.M—*———
Iris germanica, L.M——*——
Chamærops humilis, L.MM—*——
Gagea foliosa, Schult.—S—*——
Muscari comosum, L.MM—*——
Scilla hispanica, Mill. S——*—
Ornithogalum comosum, L.S *———
„ tenuifolium Guss. S *——
„ orthophyllum, Ten. M *——
„ pyrenaicum, L. var.—M*———
Allium paniculatum, L. var.—M *——
Asphodelus microcarpus, Viv.M —*——
Anthericum Liliago, L. var.M *——
Colchicum Civonæ, Guss.—S—*—
„ arenarium W. K. var. ?S —*?——
Smilax mauritanica, Desf.M——* —
Asparagus acutifolius, L.M——*—
„ scoparius, Lowe?M— ——*
Juncus bufonius, L.MM*———
Scirpus Savii, S. et M.M—*———
Carex Halleriana, Asso.—M*———
„ ambigua, Link.M———*—
„ fissirostris, BallSM S———*
Phalaris nodosa, L.M——*——
Piptatherum cœrulescens, Desf.M SM—*——
Stipa parviflora, Desf.M——*——
„ gigantea, Lag.?M——*—
„ nitens, BallS————*
Agrostis verticillata, Vill.M——*—
Phragmites communis, Trin.M—*——
Echinaria capitata, Desf.M SM S * —
Aira caryophyllea, L. M S*———
Trisetum flavescens, L.M *———
Avena bromoides, Gouan. var.M —*——
Arrhenatherum elatius. L.—M*———
Poa annua, L.M SM*———
„ bulbosa, L.M SM S*———
„ pratensis, L.M SM S* ——
„ trivialis, L.M S—*———
Melica ciliata, L. var.M—*— —
„ Cupani, Guss. var.M SM—*——
Dactylis glomerata, L. vars.M SM S*———
Cynosurus elegans, Desf. M S—*—
Festuca rigida, L.M—*———
„ unilateralis, Schrad. var.M——*—
„ geniculata, L., et var.MM—*—
„ duriuscula, var.SS*———
„ arundinacea, Schreb.MM*———
Brachypodium pinnatum, L. var.M—*———
„ distachyum, L.MM—*——
Bromus tectorum, L.M S—*———
„ madritensis, L.MM*———
„ mollis, L. vars.M SM* ——
„ macrostachys, Desf. var.M——* —
Lolium perenne, L.M—*———
Triticum hordeaceum, Coss. et Dur.M ——*—
Secale montanum, Guss.M —*—
Elymus Caput-medusæ, L. var.M— *—
Hordeum murinum, L.MM*——
Ægilops ovata, L.MM—*——
„ ventricosa, Tausch.M——*——
Andropogon hirtus, L. var.M——* —
Cistopteris fragilis, Bernh.SS* —
Cheilanthes fragrans, L.M——*
Pteris aquilina, L.S * —
Asplenium trichomanes, L.M S—* —
„ viride, L.S *
„ Adiantum-nigrum, L.S *
Notochlæna vellea, Desv.M——*
Ceterach officinarum, Willd.M SM S*—
Equisetum ramosissimum, Desf.M—*——
Selaginella, rupestris, (Spreng.)M——*—
Total number of species 4653882251611686175

Before discussing the inferences to be derived from this list, it may be well to notice some sources of error that, to a slight extent, affect the results. Although the season of our visit—the second half of May—was probably the best as regards the middle zone, it was too early to find the vegetation fully developed in the superior zone, especially on the highest ridges. It is probable that on this account the proportion of Umbelliferæ and Gramineæ found in the higher region is smaller than it would have been at a later season. At first sight it would appear that the shorter time that we were able to devote to an examination of the upper region, and the snow-storm which we encountered in the ascent to the Tagherot Pass, make the proportion of species found there, as shown by our lists, unduly small. There can be no doubt that we must have lost several species owing to these causes, but not enough to vitiate the results to a serious extent. In confirmation of this opinion it may be mentioned that although a native employed by M. Cosson has since made a large collection in the same part of the Great Atlas, and two German naturalists—MM. Rein and Fritsch—have visited the head of the Aït Mesan valley, very few species have been added to the Flora of the higher mountain region.

The first conclusion that strikes a botanist on examining the foregoing list is that the general type of the vegetation clearly marks this as belonging to the great Mediterranean Flora, which extends, with local peculiarities, from Persia and Belutschistan to the Atlantic Islands. Out of 248 genera represented in the Flora of these valleys there is not one which is not common to other portions of the Mediterranean region, and one only (Monanthes) is confined to the Great Atlas and the Canary and Cape de Verde Islands, all the others being types more or less widely spread. Further than this, the proportion borne by each of the prevailing natural orders to the whole vegetable population is pretty nearly the same that we are accustomed to find in the mountain regions of the Mediterranean region.

The materials for a comparison are unfortunately yet incomplete as regards many of the mountain districts which are best fitted for the purpose. The Flora of the Lesser Atlas of Algeria, as well as that of the rest of the French possessions in Africa, will be fully known only on the appearance of the important work promised by M. Cosson. The Flora of Spain by MM. Willkomm and Lange is yet unfinished, and there is the further difficulty that those authors have admitted a large number of plants to the rank of species which many botanists reckon only as varieties. M. Boissier’s great work, the ‘Flora Orientalis,’ is also unfinished, and no adequate materials exist for compiling lists of the plants of the Greek mountains, of those of Asia Minor, or of the Lebanon chain, all of which would afford interesting materials for comparison. In the following table I have taken for comparison the Flora of the Sierra Nevada, with the neighbouring mountains of the ancient kingdom of Granada above the level of about 800 metres, compiled from Boissier’s ‘Voyage botanique dans le Midi de l’Espagne;’ that of the Bulgardagh (the principal group of the Cilician Taurus), from a list published by M. Pierre de Tchihatcheff in the ‘Bulletin of the French Botanical Society;’ that of Dalmatia, from Visiani’s excellent ‘Flora Dalmatica;’ and that of the southern slopes of the chain of the Alps from Nice to the Karst, formed by myself from all available sources.

In the same table I have introduced, for the purpose of further comparison, separate columns for the middle and superior regions of the Great Atlas valleys, and in connection with the last I have added in a separate column the results for the higher zone of the Sierra Nevada. Under each heading I have stated

Table. I—Showing the number of species of each of the principal groups and natural orders of plants in two valleys of the Great Atlas compared with other mountain districts in the Mediterranean region.

Gt. Atlas valleys. 455 sp.Middle Zone, Gt. Atlas. 341sp.Superior Zone, Gt. Atlas. 176sp.Sierra Nevada, &c., above 800m. 890 sp.Superior[1]Zone, Sierra Nevada. 486 sp.Bulgardagh. 882 sp.Dalmatia. 2,002 sp.Southern side of the Alps. 2,545sp.
Dicotyledones39186·028683·915487·576285·641986·280891·6159479·6203580·0
Monocotyledones6414·05516·12212·512814·46713·8748·440820·451020·0
Compositæ6313·84613·52212·511913·46313·09711·023511·734313·5
Leguminosæ4810·53811·1148·0677·5326·69310·522211·11726·8
Gramineæ398·63710·8148·0596·6377·6384·31738·61766·9
Caryophylleæ265·7144·1158·5404·5296·0819·2743·71214·8
Cruciferæ255·572·12111·9495·5377·6849·5984·91395·5
Labiatæ235·0185·3116·3546·1285·8677·01005·0893·5
Scrophularineæ214·6133·8105·7374·2285·8394·4663·31094·3
Umbelliferæ204·4164·752·8505·6234·7333·71135·61134·4
Rubiaceaæ184·0154·474·0202·2122·5192·2261·3341·3
Papaveraceæ102·282·331·7101·161·2121·4140·7140·6
Geraniaceaæ102·282·342·3101·161·260·7170·8230·9
Liliaceæ102·261·842·3151·791·9232·6613·0522·0
Boragineæ92·051·563·4182·0112·3232·6402·0391·5
Ranunculaceæ81·851·542·3283·1153·1111·2532·6873·4
Cistineæ71·561·810·6232·6102·110·1110·5100·4
Rosaceæ71·541·231·7262·6204·1212·4572·8933·7
Campanulaceæ61·351·510·691·061·2212·4261·3461·8
Convolvulaceæ51·151·510·630·320·140·5100·590·4
Coniferæ51·141·210·6101·181·6151·7150·7110·4
Saxifrageæ (inclusive of Grossulariæ)40·941·242·3121·391·920·290·4522·0
Cyperaceæ40·941·210·6171·9122·570·8432·21194·7
Gentianeæ00000070·851·040·5140·07311·2
Primulaceæ20·420·610·670·861·280·9120·6602·4
Junceæ10·210·300111·291·90080·4311·2

the whole number of phanerogamous species included in the Flora of the region, and opposite the name of each natural order I have entered the number of species found in each region, and the percentage proportion which this number bears to the entire flora. Besides the orders which bear the largest proportion in the Great Atlas Flora I have enumerated those that usually characterise the vegetation of high mountains in this part of the world, though several of these are little, or not at all, represented in the Flora of the Great Atlas.

Confining the comparison in the first instance to the figures given for the Atlas Flora as a whole in the first column, and those given in the fourth, sixth, seventh, and eighth columns respectively, for the Sierra Nevada, the Bulgardagh, Dalmatia, and the southern side of the Alps, we remark in the first place that Monocotyledons bear about the same proportion to Dicotyledons in the Great Atlas that they do in the Sierra Nevada, the percentage here being much larger than it is in the Bulgardagh, and considerably less than in Dalmatia or the Southern Alps. In this part of the world this percentage in the Flora of a given region mainly depends upon the number of Gramineæ and Cyperaceæ. The abundance of the latter group in the Alps doubtless arises from the fact that at a former period physical conditions favoured the migration of a large number of northern species that have been unable to extend to the more southern mountain regions of the Mediterranean area.

In all the regions under consideration we find, with a single exception, that the same eight natural orders take precedence of all others as regards the number of species that they exhibit, the aggregate in every case exceeding one-half of the whole phanerogamous Flora. These natural orders are Compositæ, Leguminosæ, Gramineæ, Caryophylleæ, Cruciferæ, Labiatæ, Scrophularineæ, and Umbelliferæ. The exception arises from the prevalence, already noticed, of Cyperaceceæ in the Flora of the Southern Alps. In comparing the figures in the Great Atlas column with those for the other areas above enumerated, it is well to recollect that our materials are taken from a district much more limited in extent than the others, and are necessarily imperfect, because obtained from a single short visit to each valley at a season when many species are yet undeveloped. It is probable, for instance, that the proportion of Umbelliferæ would be increased if the whole Flora were better known. Subject to this remark, it will be seen, as might be expected, that the constituents of the Great Atlas Flora show more analogy with those of the Sierra Nevada and Bulgardagh Floras than with those of Dalmatia and the Southern Alps; but the proportion of Compositæ is larger than in any of them (nearly 14 per cent). In comparing the vegetation of a small district with that of a large one it must be recollected that a small natural group containing a few widely spread species, such as Geraniaceæ, is likely to show a larger percentage proportion to the whole Flora in the small district than in the larger one. It may happen that the same species are spread through both regions; but in one case the number is to be compared with a small total, in the other with a much larger one. This remark has a bearing on the fact that in the Great Atlas Flora the natural orders that bear an unusually large proportion to the total number of the Flora are Leguminosæ, Caryophylleæ, Rubiaceceæ, Papaveraceæ, Geraniaceæ, and Convolvulaceæ. On the other hand, there is a remarkable deficiency in the natural orders that especially characterise the Flora of the Alps, and in a less degree, the high mountains of Southern Europe. These are Ranunculaceceæ, Rosaceæ, Saxifrageæ, Primulaceæ, Junceæ, and Cyperaceæ; not to speak of Gentianeæ, which are here altogether absent.

If, instead of regarding the Atlas Flora as a whole, we examine separately the figures given in the several columns for the middle and superior zones respectively, we find very different proportions for the chief natural orders, except for Compositæ and Leguminosæ which are in both very numerous. In the middle region of the Atlas these two orders represent very nearly one-fourth of the phænogamous Flora. After these Gramineæ, Rubiaceæ, Papaveraceæ, Geraniaceæ, Cistineæ, and Convolvulaceæ are, in the middle region, unusually frequent, while Cruciferæ, Rosaceæ, Boragineæ, and Liliaceæ are remarkably deficient. In the superior zone, on the other hand, the proportion of Compositæ and Leguminosæ is less excessive, making jointly a little over one-fifth of the whole Flora of the upper region. The most marked characteristic here is the very large proportion of Cruciferæ, being less by one species only than the number of Compositæ. Taking into account the number of individuals as well as that of species, this must be regarded as the dominant element in the Flora of the higher region of the Great Atlas, affording as it does 12 per cent of the whole Flora. The only region in which this characteristic is approached is the Bulgardagh in Cilicia, where Cruciferæ supply near one-tenth of the whole list. Caryophylleæ also form an unusually large element in the Flora of the upper zone of the Atlas; but, unlike Cruciferæ, this order exhibits no endemic species, and four-fifths of the whole number are common plants of Central and Northern Europe. Rubiaceæ and Boragineæ have more representative species than is usual in mountain Floras; while there are but three species of Rosaceæ in our list; and Campanulaceæ, Primulaceæ, Coniferæ, and Cyperaceæ are each represented by a single species, and Gentianeæ and Junceæ are altogether absent from the higher zone.

Although statistical results, such as those given above, are not without interest, as throwing light upon the general characteristics of the Flora of a given region, any rational grounds for speculation as to the real affinities and past history of the vegetation must be derived from a closer examination of the individual species of which it is constituted. It is at least conceivable that two Floras should exhibit similar proportions of species belonging to the several natural groups, with no identical species, and with little or no indication of community of origin. The particulars given in our general list will have already led the reader to infer that the results of an examination into the distribution of the individual species that go to make up the Great Atlas exhibit some very peculiar features. Taking the totals at the foot of our list, and excluding cryptogams, it is seen that more than one-third of the species are plants of Middle and Northern Europe, while about one-sixth is made up of endemic species peculiar to Marocco, and, with few exceptions, not known out of the Great Atlas, more than half of the whole list belonging to one or other of these categories. The results, as shown in the following table, are still more remarkable when we separately examine the zones into which mountain vegetation is naturally divided. As in the former table the figures first entered in each column represent the number of species belonging to each category, those next given showing the percentage proportion borne by that number to the total proportion of each region.

Table II.—Showing the distribution of the species of flowering plants included in the Flora of the Great Atlas, and of the Sierra Nevada of Granada,[2] and the Bulgardagh in Cilicia.[3]

Mid-EuropeanWide-spread MediterraneanConfined to adjoining regionsEndemic
Great Atlas, including all species found above1,200 m. 455 sp.15433·816536·26113·47516·6
Middle Zone of Atlas, from 1,200 m. to 2,000 m.341 sp.10631·114141·34613·54814·1
Superior Zone of Atlas, from 2,000 m. to 3,500 m.176 sp.7844·34324·42011·43519·9
Superior Zone of Sierra Nevada, above 1,600 m.486 sp.20943·07415·210421·49920·4
Bulgardagh in Cilicia. 882 sp.15918·035940·715717·820723·5

From this table we see that while over one-third of the whole Atlas Flora consists of plants of Central and Northern Europe, the proportion reaches nearly to one-half in the higher region (above 2,000 metres); and also that the proportion of endemic species, which in the aggregate is one-sixth of the whole, rises to one-fifth in the upper zone. On the other hand, the proportion of purely Mediterranean species, which amounts to 55 per cent. in the Flora of the middle zone, falls below 36 per cent. in the upper region. Of these Mediterranean species the large majority (more than two-thirds) are widely distributed plants, several of them extending to the mountains of Asia Minor, and twenty species only are exclusively confined to the Great Atlas and to the mountains of Southern Spain, the Lesser Atlas, or the Pyrenees. There is nothing in the distribution of these latter plants to indicate any special connection between the Atlas and any one of the mountain regions above mentioned. Six Atlas species are common to Southern Spain and the Algerian Atlas, six more are known only on the mountains of Southern Spain, five have been hitherto supposed to be peculiar to the Lesser Atlas, and three are elsewhere confined to the Pyrenees.

Some further light may be thrown on the origin of the Great Atlas Flora by considering the affinities of the plants which are reckoned in our list as endemic in Marocco, nearly all being confined, so far as we know, to the chain of the Great Atlas. Although all of these, along with some that we have classed as mere varieties, would be counted as distinct species by many botanists, a considerable number, amounting to more than a quarter of the whole, are, according to the views expressed elsewhere by the writer,[4] to be ranked as sub-species. But here again we fail to discover indications of special relations between the Great Atlas Flora and that of neighbouring mountain regions. Ranking as sub-species twenty-one out of the seventy-five endemic forms enumerated in our list, we find that ten of these are allied to widely spread Mediterranean species, three are related to plants of Central Europe, three to species common to Algeria and Southern Spain, three more to species confined to the Spanish peninsula, and two to endemic Algerian forms.

If we scrutinise in the same manner the endemic forms of the higher region of the Great Atlas, we find that out of the thirty-five enumerated eight, or less than one-fourth, are to be ranked as sub-species. Of these, three are nearly allied to wide-spread Mediterranean species, one to a plant common to Spain and Algeria, two to endemic Spanish species, one to an Algerian endemic form, and one is related to a species indigenous in the Alps and other high mountains of Central Europe.

While recognising the fact that the relations between the vegetable population of the Great Atlas and that of the south of Spain are less close than might have been expected on theoretical grounds, we must yet admit that, on the whole, the Great Atlas is more nearly connected in a botanical sense with this than with any other mountain region that is known to us; and it becomes a matter of some interest to compare closely the list of species obtained by us in the Atlas, with the comparatively well known Flora of Southern Spain. The results of this comparison are given for the Great Atlas generally, and for the superior zone separately, in the following table, in which the Atlas species are distinguished under five heads: 1, those found in the higher region of the Sierra Nevada; 2, in the mountain region of Andalusia; 3, in the lower warm region below the level of about 2,000 feet; 4, absent from Southern Spain, but found in the central or northern provinces; and 5, those not included in the Spanish Flora.