Earthworms and Their Allies

by Frank E. Beddard

Marine Species.

There are a few, but relatively speaking very few, worms of the order Oligochaeta which lead a marine life. And of these the majority are shore forms not extending into the waters of the sea. The most salient example, at any rate the best known perhaps, is the genus Pontodrilus, the name of which fixes its habitat, and was naturally given to it on that account. It was originally found on the sea shore of the South of France, and I have myself examined examples from Nice. The worm lives among bunches of sea-weed cast up by the sea, and which are thoroughly salt. Besides the two forms that have been met with in this Mediterranean region but which are united by Dr Michaelsen into but one species, other Pontodrilus have been described from so many and such diverse parts of the world as the following. The West Indies (Bermudas, Jamaica etc.), the coasts of South America, of both West and Eastern Africa, the Red Sea, Christmas Island near Java, Sharks Bay in West Australia, the Hawaii Archipelago, Celebes, South West Australia etc. In fact there is no great tract of the ocean excepting the antarctic region where this genus is not to be found. It is possible however that this latter statement is not correct and that New Zealand ought to be added. But the species described from those islands, viz. Pontodrilus lacustris, is not a marine form at all as its specific name denotes; nor is it quite certainly to be included in the genus. On the other hand a form from the Chatham Islands in the same quarter of the globe, described originally as Pontodrilus chathamensis, is to be referred to the antarctic region. Altogether some dozen species of Pontodrilus have been described by different naturalists; but quite recently Dr Michaelsen has reduced these to three only, which are P. bermudensis (F. E. B.), P. litoralis (Grube) and P. matsushimensis (Izuka), with the doubtful addition of P. lacustris already referred to. Whatever may be the ultimate verdict upon this question of species it is clear that the genus is widely spread upon the sea shores of the world and that forms from different regions show some fixed variations, which others may eventually agree with their original describers in regarding as definite species.

It cannot be said that any salient characters in the organisation of these worms mark them out from either terrestrial or fresh-water Oligochaeta. There are no such important variations of structure as can be seized upon to characterise them as inured to salt water. The genus agrees with many aquatic forms in the fact that the nephridia are not present in the earlier segments of the body, not indeed putting in an appearance until about the thirteenth segment or even later. They are thin delicate worms; but there is nothing distinguishing about this, while the feeble or absent gizzard is a character which is really difficult to correlate with habitat. Still we have here a whole genus which is marine in its habit. Among the Megadrili or earthworms there are not many other examples of these 'euryhaline' forms as they have been named. On the shores of Patagonia however and Kerguelen shore-living species of the mainly antarctic genus of earthworms Notiodrilus have been met with. And there are a few allied cases among the antarctic genera of Acanthodrilinae.

In addition to these terrestrial forms there are a few limicoline genera which are partly marine in their habit. Thus several species of the prevalently arctic and antarctic family Enchytraeidae are shore living. There are also marine Tubificids such as Clitellio arenarius and Tubifex ater (not uncommon on British shores), marine Lumbriculids and a marine Naid from the Italian coast. These forms show no great differences from their fresh-water allies.

Earthworms originally purely aquatic
animals.

The very name Earthworm, so distinctive as it is of the habitat of these animals, seems to have been expressly invented in order to crystallise into one word the remarkable distributions of these creatures. They are with very few exceptions the most purely terrestrial animals that we know. There are a few Mammals like the mole and several underground Amphibians and Snakes in the tropics which share this habitat with the worms, probably because they chiefly prey upon them. But there is no group of animals that is characterised by a subterranean existence in the way that earthworms are. For we cannot put burrowing animals, such as the prairie dog and many rodents, into the same category. These make and seek their burrows for protective purposes and in order to bring forth their young in security. They do not feed beneath the surface of the ground or pass their entire lives in that situation. We have already in a previous chapter dealt with such exceptional forms of earthworms as do not lead an entirely subterranean existence; but as was pointed out in chapter I these exceptions are but few and the immense bulk of earthworms fully justify their name.

Nevertheless there are many arguments which tend to show that these purely land-dwellers have grown out of exclusively water-dwellers and even that the change from the one mode of life to the other has been accomplished comparatively recently. For there are here and there vestiges of structures which seem only fitted for an aquatic life; and in other cases structural changes have commenced which would appear to be in definite relation to the underground mode of life prevalent to-day. Let us consider for a moment the differences which obtain between the conditions of life in water or in soft mud at the bottom of pool or river, and those which are undergone by a dweller in stiff soil or vegetable débris. In the first case the medium is fluid or at most very soft, while the soil is at least stiffer and harder to traverse.

Secondly the transition between the very bottom of a pool and the top layers of the water is more or less gradual, while the stiff soil ends abruptly in the tenuity of the atmosphere.

A third point of difference is doubtless the smaller supply of readily available oxygen in the still pools and even rapid rivers, which in certain stagnant pools and in the bottom waters of deep lakes must produce a very vast difference in physiological conditions.

We have already dealt with the characteristics of the aquatic genera of earthworms, not only in detailing the general characters of the families which are found in this situation but also in studying the features which earthworms show in those cases where they have reverted to an aquatic mode of life. It remains in the present section to attempt to descry in the purely terrestrial forms the remnants of adaptations to an aquatic life which are no longer of service to them.

It is a noteworthy fact, that the continuous circle of setae which is met with in certain earthworms is by no means a character of such classificatory importance as it was at one time, perhaps, thought to be. It is true that we meet with this character in the genera Megascolex and Pheretima which are not very far from each other in the system and are at any rate members of the same sub-family, the Megascolecinae. But we also find the continuous circle of setae well developed in Plagiochaeta which is not so near to Pheretima, and an approach towards it in Dinodrilus and Dinodriloides which are equally remote perhaps from both Pheretima and Megascolex on the one hand and Plagiochaeta on the other. Still it may be urged that all of these genera are at least members of the family Megascolecidae and that the question of a character which thus merely shows affinity is not yet eliminated. It is therefore of particularly great importance that Dr Cognetti de Martiis should have met with the South American genus Periscolex which, undoubtedly a member of a totally distinct family, the Geoscolecidae, yet shows the same complete circle of setae. The reason for dwelling upon this particular anatomical character in the present connection is because it would seem to be a character specially suited to an underground life where there is an equal pressure all round the body and where progression would seem therefore to be best attained by a continual leverage round the circular body. And this view is strengthened by the sporadic occurrence of this modification in different families. We thus come to the conclusion that the opposite state of affairs is a remnant of an aquatic life, a conclusion which it is the object of the present section to discuss. More than this, it would seem that an equal disposition of the two bundles of setae on each side of the body was a less modified state of affairs than the restriction of the two bundles or pairs of setae to the ventral surface, such as occurs for example in the genus Dichogaster and which is very obvious in some of the larger-sized members of this extensive genus. For the restriction of the setae to the ventral surface obviously favours progression upon a surface and not through a medium. And it is only among the terrestrial Oligochaeta that this mode of progression occurs. It might also be urged, and with some reason, that the retention of rather longer setae upon the clitellum in the Lumbricidae and Geoscolecidae, and the possession of equally long or in many cases much longer setae corresponding to one of the two pairs of setae of the generative segment in certain Megascolecidae, is a feature in which an aquatic condition—so to speak—is retained. The setae would represent a vestige of the general presence of long setae over the body generally such as is convenient or at least not inconvenient to an Annelid living in water or soft mud. But probably it will be thought the modified genital setae are a recent development and not a retention.

There is no more thoroughly terrestrial family of earthworms than that of the Moniligastridae and yet this family in its general anatomical characters shows many points of likeness to aquatic forms as has been now pointed out by many observers. It is true that these characters are not those which might be associated at first sight with an aquatic life. But none the less they are characteristic of most of the families which live in the waters of the earth. Thus Moniligaster and its allies (Eupolygaster, Drawida, etc.) have quite short sperm ducts which open on to the exterior at furthest in the segment next to that in which their internal funnel lies. Again the simple structure of the terminal gland into which they open and which in its turn opens on to the exterior is very like that of such a family as that of the Lumbriculidae. Another fact is the simple undivided cavity of the sperm sacs which is unlike that of typical earthworms but again like that of all of the Limicolous families. We may fairly see in these worms evidence of origin from aquatic ancestors. Evidence of the same nature, i.e. not as showing the retention more or less of anatomical characters commonly associated with a life in water, but as affording indirect evidence of an origin from actually aquatic forms, is to be seen in certain members of the families Geoscolecidae and Eudrilidae. In both of these it not infrequently happens that the sperm sacs are but a single pair and that that pair consists of sacs of extraordinary length. Thus in Trichochaeta (or Hesperoscolex) barbadensis Miss Fedarb and I have shown that the long thin sperm sacs extend through no less than 109 segments, which is vastly in excess of the length of those of the majority of earthworms in which they are most commonly limited to a single segment. In the same way the Eudrilid worm Polytoreutus magilensis has a pair of long and thin sperm sacs which extend through some fifty segments. This elongation of the sperm sacs in the ripe worms is a very common feature of the Limicolous genera.

CHAPTER III
THE EXTERNAL FEATURES OF EARTHWORMS AND
THEIR RELATION TO HABIT AND ENVIRONMENT

To the very inexperienced eye all earthworms might appear to be quite similar in detail as they undoubtedly are in general form. But it needs not a great deal of examination to detect even salient characteristics whereby one kind may be distinguished from another; to the expert it is possible in very many cases to go no further than the outside before assigning its correct place in the system to a given example. The general external features of this group of worms have been already dealt with in another chapter. To some of these we again direct attention in a more elaborate fashion in order to emphasise the possible meanings of the variations met with apart from their use in systematic arrangement. It is difficult to say in comparing one worm with another what is the most salient external difference. There are however a few which may be regarded as equally conspicuous on a nearer examination of the specimens. The varying position and greater or less extent of the clitellum, the longer or shorter retractile or nonretractile prostomium, the position of the usually conspicuous male pores, and the existence of in the first place and—when present—the numbers and situation of the so-called genital papillae are among the most obvious. The setae and their position we treat of under the heading of the modification of the worms to lead a terrestrial life; and though these chitinous organs differ greatly they do not concern us in the present section. The girdle or clitellum ('eminentia quasi ulcerata') has been long observed as a character of these animals and it is one which distinguishes them from all other worms except the leeches and a very few marine Polychaeta. This modified region of the body is often of a different colour to the rest and has a glandular look which readily enables one to recognise its position and limits, though its obviousness is less in some cases. It either forms a complete ring round the body or is developed upon the dorsal surface and only to a slight extent upon the ventral surface. Its use, as is well known, is to secrete the cocoon in which the eggs are deposited; and the epidermis which forms it is thickened and more glandular than that in other regions of the body. Among earthworms it is doubtful whether the clitellum ever occupies less than three segments; it consists of three only in the great majority of species of the marked genus Pheretima. From this lowest level it extends in other forms, and in the partially aquatic African genus Alma it may occupy as many as forty segments. The position also varies from genus to genus and from species to species. It is sometimes further forward and sometimes further back. In the remarkable family Moniligastridae this organ is developed earlier in the body than in any other group of true earthworms, consisting of four segments or so commencing with the tenth. As a rule the clitellum begins further back than this—the thirteenth or fourteenth being a common place for the first commencement of the organ among the Megascolecidae, while among the Geoscolecidae and Lumbricidae it is generally much further back, commencing in Alma at the forty-fifth. These details might be increased to many pages; but enough has been said to emphasise the variability of the organ. What reason can be assigned to this variability, which might be supposed unnecessary in view of its functions? There are perhaps two suggestions that may be made, though many facts are lacking which might offer confirmation or refutation of either of these. It is to be noticed that on the whole the older types such as the Moniligastridae and the Megascolecidae (including for this purpose the Eudrilidae) have clitella which are short. There are a few but not many exceptions. These older types do not seem capable of extending their range with any rapidity. It is true that here again there are exceptions, notably many species of Pheretima which are considered under the section which deals with the migration of these animals. On the other hand the Lumbricidae have on the whole a more extensive clitellum and so have many Geoscolecidae. It is obvious that of all earthworms the Lumbricidae is the family which has the greatest capacity of migration and adaptation to new circumstances. The reason for this may be that in the latter case the more extensive clitellum produces a larger cocoon which in its turn can hold and cherish while they reach maturity a larger number of embryos. Much remains to be learnt under this heading. But the comparatively small clitellum of the large Ceylon Megascolex coeruleus only contains two embryos, while the also comparatively small cocoon of the large and restricted Octochaetus multiporus (limited to the South Island of New Zealand) only contains a single embryo. This latter fact may be regarded as fairly well established since I myself examined quite fifty cocoons.

On the other hand larger numbers seem to arrive at maturity in the cocoons of Allolobophora. The more extensive clitellum must produce a relatively larger cocoon, and it is interesting to note that the cocoon of the widely distributed genus Criodrilus (Europe and South America) is very long although not of great diameter. However the facts are not sufficiently great to dogmatise much upon this subject. Another conceivable reason for differences in the clitellum is—as I also think is the case with the genital papillae—to prevent hybridisation. That the sense of touch is delicate in these animals seems clear from the abundant development of epidermal sense-organs. It may be that the feel of the clitellum during union enables two individuals of a given species to come together and prevents those of different species from mating. In any case there is no positive evidence that hybridisation does occur in this group of animals. The astounding variability and yet constancy in a given species of the genital papillae is in favour of regarding these organs as tactile recognition marks; and it will be noted that they are not by any means characteristic of some of the older types of earthworms. Furthermore they are particularly conspicuous in such genera as Pheretima, Megascolex etc., which possess a large number of species. In these of course mutual recognition would otherwise be more difficult.

Fig. 12. Ventral view of Pheretima solomonis to show papillae which are to be compared with those of fig. 13. (× 2.)

Fig. 13. Ventral view of Pheretima sedgwickii. (x 2.)

CHAPTER IV
SENSE ORGANS AND SENSES OF EARTHWORMS

As this is not an anatomical treatise we shall not attempt any detailed anatomical and histological account of the sense organs in this group of worms. But a few facts must be given in illustration and explanation of the senses of touch and sight that the Oligochaeta undoubtedly possess. These Annelids, unlike their allies the marine Polychaeta, and even their allies on two other sides, the leeches and flatworms, have no complexly organised eyes or other sense organs. They have in fact no organs to which a definite sense can be attached on histological grounds. There is nothing comparable to an eye or to the auditory sacs of other low worms. There are only particular cells of the epidermis, often associated into small groups, and those again into larger associations of rows of such groups of cells. It is to be presumed that these modified groups of cells have a sense function; but no more can be said than that they are doubtless tactile and also to some extent receptive of the influence of light. True visual cells have been asserted to exist in earthworms, consisting of cells of which a part is clear and transparent and has been supposed to serve as a lens for the rest of the cell which represents a retina. But belief in the function of these cells is by no means unanimous. On the other hand many investigations have proved the existence of groups of epidermic cells of an oval form which are at present arranged in definite rows upon the segments of both terrestrial and aquatic forms, which are moreover connected with nerve terminations, or are at least—according to more modern views—in close contact with the terminations of nerve fibres. These are furnished often at their free tips with minute sensitive processes. There is nothing in the structure of these to associate them definitely with any sense in particular. They suggest of course tactile organs more than organs of any other sense. In addition to these are certain problematical organs which are found in the Eudrilidae and are present in the members of one section of that group, the section which is represented by the universally found Eudrilus.

These bodies have been compared to a Pacinian body (a sense-organ found in Vertebrates) and they bear no little resemblance to it. For each consists of a darkly staining core surrounded by a layer of cells arranged like the coats of an onion. In any case it would appear that these bodies must be looked upon as of a sensory nature, though they do not reach the surface of the body but underlie the epidermis. Their function must remain purely a matter of guesswork at present, for nothing to the point is known of the habits of the Eudrilidae. It has been suggested by Dr Gustav Eisen that these cells are auditory and serve to warn the worm of the footsteps of birds and other enemies. That too is his view of certain peculiar but different cells found in the epidermis of Pontoscolex. In the latter something like an otosome has been found which is certainly lacking in the Eudrilidae, whatever may be the function of the cellular epidermic bodies here briefly referred to.

While there is thus nothing very positive to be gleaned from an examination of the structure of the Oligochaeta as to the senses which they may possess, experiment has done something towards an elucidation of their behaviour under stimuli and their reaction to light and to other forces which come into play during their lives. There is some evidence that earthworms can see, using that expression of course in a very broad way. At any rate they react to changes in the intensity of light. The gross experiment of flashing a lantern upon earthworms which are reaching out from their burrows with the tail end inserted in those burrows shows that they have an appreciation of light. More refined experiments have been conducted upon the sense of light. Dr Graber used a box with two compartments, the one of which was dark and the other illuminated with diffused daylight, i.e. not direct sunlight. The worms were allowed equally free access to both and were examined at the close of every hour, and their positions noted. The investigator found that on the average the dark half of the box contained 5·2 times as many worms as the light chamber, thus indicating a very marked preference for absence of light.

Not only is this the case, but the same observer proved that earthworms can distinguish between degrees of intensity of light. This obviously indicates a more complete 'visual' sense. He illuminated the light-box of the former experiment with light admitted through a ground glass screen, thus diminishing its intensity. The other chamber was left as before but the screen was removed, thus admitting full daylight. In this experiment the number of worms in each compartment proved to be different. The results were not so striking as before, since only rather more than one-half were found in the more dimly illuminated chamber. It is a well-known fact that if earthworms are abroad at all from their burrows, it is during the night that this movement takes place, the numbers decreasing towards morning though worms are often seen crawling about well after sunrise. Some experiments have been made in attempt of explanation of this apparent anomaly. It would appear from these experiments that while worms are negatively phototropic to strong and moderate light as has already been pointed out, they are positively phototropic to very dim light; hence the advent of evening calls them forth from their burrows. It will be noted that this perception is of very great advantage to the worm since its more active enemies above ground are diurnal. It was held originally that the head end of the worm only was thus sensitive to light; but more recent experiments have shown that this is not the case, and that all of the body is sensitive. This disposes of course of the existence of special light-receiving organs in the anterior part of the worm's body. Not only this, but an interesting extension of the view has been promulgated. It has been shown by Prof. G. Parker and a colleague that in the common Brandling worm, Allolobophora foetida, the response to light stimuli was related to the amount of the body exposed to its influence. This is very important as showing that the light perception in these creatures is probably not due to special organs having a limited position on the body, but is due to collective sense impressions of many cells scattered over the whole body, the impression being the greater when the whole body is exposed and less when only parts of it are exposed. Furthermore, and this has quite another importance, these observers noted that the reaction effects differed when only a part of the body was exposed; that they were greater in the front of the body, less in the middle, and less still at the tail end. Indeed they found that the reactions in the case of the front end of the body alone being exposed were rather more than one-third as compared with those which were shown when the whole body was subjected to the light stimulus. The fact that the least sensitive region of the body is the posterior end has, it is true, only been definitely proved in the case of the worm whose specific name has been mentioned. But it is possible that others are similarly affected. And it is highly important to note the prevalent habit among the Tubificidae of lying with the head end imbedded in the mud of the pool which they inhabit, while the tail end emerges and waves freely in the flood. The additional fact that this tail end occasionally bears gills (as in Branchiura sowerbii and Phreodrilus branchiatus) has a collateral importance not to be mistaken.

CHAPTER V
RELATIVE FREQUENCY OF EARTHWORMS IN
DIFFERENT REGIONS OF THE WORLD

It will be of use for various purposes to be considered later to arrive at a comprehensive view of the relative numbers of species and genera of earthworms in the four quarters of the globe. And in making this general census we shall not take into consideration the purely aquatic forms, but shall limit ourselves to the earthworms, sensu stricto, or Megadrili, of which, however, it is true that some members are actually lake and river dwellers. This latter fact will not, however, interfere with the usefulness of the comparative survey.

Two preliminary remarks are necessary. The opinions of naturalists vary as to the limits of genera; and a species may be a species to one and a mere variety to another. Thus it will be impossible to give a summary of the facts to be enumerated presently, which will be either absolutely accurate or which will satisfy everyone in every detail. But it is asserted that the following survey is substantially correct.

In the second place it is often possible to eliminate from the fauna of a given region those species and even genera which have been accidentally imported, a matter which will receive careful consideration on a later page. Such forms are therefore, in those cases at any rate where the evidence seems to be overwhelming, withdrawn from the list. In other cases, particularly in the Eastern region of the world, it has been found less easy to rectify the catalogues by removing what Dr Michaelsen has termed 'peregrine' forms.

We shall commence with a census of South America; the entire Continent will be divided for the present purpose into three divisions, viz. South America, Central and North America, and in the third place the West Indian Islands.

In South America we find that the bulk of the indigenous earthworms belong to the family Geoscolecidae and to a definite sub-family, viz. Geoscolecinae. These genera are Onychochaeta with one species, Hesperoscolex of which one species is known from the area, Periscolex with one species, Anteoides with two species, Pontoscolex one species. Opisthodrilus with two species, Rhinodrilus (including either as synonyms or as sub-genera, Thamnodrilus, Anteus, Tykonus, Urobenus and Aptodrilus) with no less than 49 species: Andiodrilus with five species, Holoscolex with one species, Glossoscolex ten species, Fimoscolex, Andiorrhinus and Enantiodrilus with one species apiece.

Thus of this sub-family of Geoscolecidae we have in South America a large number of genera and a much larger number of species. Of a second sub-family of Geoscolecidae there are three species of Criodrilus found in the South American Continent.

We now turn to the Megascolecidae of which a large number of species occur within the area now under consideration. The bulk of these belong to the sub-family Acanthodrilinae and they are as follows:

Of the genus Notiodrilus there are ten species, of Microscolex two species, of Chilota 19, of Yagansia 13.

A second sub-family Trigastrinae also occurs in this Continent and the following genera are permanent inhabitants, viz.:—

Dichogaster (= Benhamia) with only three species, of which two at least are found elsewhere, and of which therefore the autochthonism is doubtful.

Finally, there is the sub-family Ocnerodrilinae comprising the following genera: Kerria with ten species, Ocnerodrilus (with sub-genera Liodrilus, Ilyogenia and Haplodrilus) four species, which again are rather doubtful indigenes of the South American Continent.

Leaving aside certain species (of the genera Lumbricus, Pheretima, etc.) which are clearly not indigenous, the South American Continent harbours 150 kinds of earthworms which are distributed in some 19 genera. But of these a few species (e.g. Onychochaeta windlei, Kerria macdonaldi) stray into neighbouring regions, i.e. the West Indies and California. It is doubtful therefore whether they are to be referred to as limited to one of these regions and accidentally imported into the others, or whether they are genuine inhabitants of both.

The South American Continent shares with the West Indies the following genera, but the species (except in the case of Onychochaeta windlei, Glossoscolex peregrinus) are distinct; these genera are Hesperoscolex, Pontoscolex, Dichogaster, and Ocnerodrilus. Diachaeta is limited to the West Indies.

The following South American genera are also found in Central and warmer North America (Mexico, California), viz. Hesperoscolex, Periscolex, Rhinodrilus, Pontoscolex, Andiodrilus, Glossoscolex, Notiodrilus, Microscolex, Dichogaster, Kerria and Ocnerodrilus. But with the exception of Hesperoscolex, which seems to belong rather to Central America and the West Indies, Microscolex and Pontoscolex which are world-wide and whose original home is therefore difficult to fix, and Dichogaster and Ocnerodrilus which would seem rather to be rare immigrants (perhaps not truly indigenous) into South America, these genera are practically distinctively South American.

Thus we may fairly say that the genera Anteoides, Opisthodrilus, Andiodrilus, Holoscolex, Glossoscolex, Rhinodrilus, Andiorrhinus, Fimoscolex, Enantiodrilus, Notiodrilus, Chilota, Yagansia and Kerria are at least very distinctive of S. America and that they are represented by the large majority of species found in that continent, the total being 120 or slightly more.

Leaving the American Continent and adjacent islands and archipelagos, the next great land-mass to receive attention will be the Continent of Africa. In giving a census of the species and genera of earthworms which inhabit this quarter of the globe it must be premised that the facts relate only to Africa south of the Sahara. But little is known of the genera which occur in Algeria and Morocco, but from what little is known it is clear that they should come into consideration in connection with the fauna of Europe and not with that of tropical and south temperate Africa.

We have in the first place to consider an entirely peculiar family, of fair extent in genera and species, which is limited to this region of the world; that is to say with one apparent exception which is clearly only apparent. The genus Eudrilus is one of the few kinds of worms that turns up in collections from every tropical part of the world and even at times from more temperate countries. It is one of those 'peregrine' forms, as Dr Michaelsen has termed them, which possess unusual facilities for extending their range. Presumably its real home is Africa. This family is known as the Eudrilidae though by some it is only regarded as a sub-family of the Megascolecidae. In this family we have the following genera: Eudriloides with 11 species, Platydrilus with 11 species, Megachaetina with two species, Reithrodrilus with one, Stuhlmannia with five species, Notykus and Bogertia with one species each, Metadrilus also with but one species, Pareudrilus with perhaps five, Libyodrilus with one and Nemertodrilus with two species, Metschaina with two species, Eudrilus with two (or possibly more?) species, Parascolex with four species, Preussiella with two, Buttneriodrilus with two, Eminoscolex with 16 species, Hyperiodrilus, Heliodrilus, Alvania, Rosadrilus, Kaffania, Euscolex, Metascolex, Beddardiella, Gardullaria with only one species to each genus, Bettonia with three species, Teleudrilus with 15 species, Polytoreutus with 22 species, Iridodrilus with two, Malodrilus also with two, Neumanniella with eight, Eupolytoreutus with two and Teleutoreutus with only one species.

This completes the list of the Eudrilidae. We will take the huge family Megascolecidae next; and we find in tropical Africa that the genus Dichogaster alone contains at least 93 species confined to Africa, as well as a few which it shares in common with America, and the common and widely spread D. bolavi which has even made its way to Europe. Of the sub-family Ocnerodrilinae we have the genus Gordiodrilus with seven species, a genus which also occurs in Madagascar and the West Indies: Nannodrilus with three species, Diaphorodrilus with one species and a few examples of Ocnerodrilus and its sub-genera, some of which are also forms that occur elsewhere in the world, for example Nematogenia panamensis. Of Pygmæodrilus there are eight or nine species.

Of the sub-family Acanthodrilinae the Cape region of South Africa harbours some seven species of the genus Notiodrilus, of which one however is a West African form. The allied Chilota has 13 species, and there is a peculiar genus Holoscolex near to Yagansia with one species. We next have to deal with the Geoscolecidae, of which a sub-family, Microchaetinae, is mainly found in Africa, the rest being found in the neighbouring Madagascar and some few in the East. Microchaetus contains about 14 species, Tritogenia perhaps three, Callidrilus two, and the genus Glyphidrilus, mainly found in Asia, has one species in the region now under consideration. In addition to these Geoscolecids there is the peculiar and largely aquatic Alma with six or seven species in East, West, and Central, Africa, and in the Nile region.

Summing up the genera which are found in tropical and South Africa we find that there are 44 which are abundant in, or entirely confined to, the region. In addition to these four or so occur in Africa but are either more abundant elsewhere or (as in the case of Chilota and Notiodrilus) are equally distinctive of other parts of the world. The number of species may be estimated at 270, possibly rather more. Clearly therefore this part of the world is much richer than South America, both in numbers of genera actually found, and peculiar to the country, and numbers of species.

Passing from Africa the next definite quarter of the globe which will detain us here is the island of Madagascar, so remarkable for the Mammalian fauna which characterises it, for its lemurs, peculiar Insectivora and Carnivora, and above all by reason of the absence of the prevalent African types such as antelopes, zebras, rhinoceros etc. It is probable that a good deal still remains for discovery among the earthworms of this island; but a considerable number are already known and they are as follows:

The Eudrilidae are completely absent, a state of affairs which is paralleled by the absence of antelopes among mammals.

The sub-family Acanthodrilinae of the family Megascolecidae are represented by four species of the genus Notiodrilus, with which perhaps Maheina braueri from the Seychelles should be included as it presents small differences from Notiodrilus.

Among the Megascolecinae a good many species of Pheretima have been collected both on Madagascar and on certain adjacent islands; but these, with one possible exception, are forms which occur elsewhere and are often indeed very widely distributed 'peregrines,' so that it is hardly permissible to place them among the real inhabitants of Madagascar. The same arguments hold in the case of Lampito mauritii and the ubiquitous Eudrilus and Pontoscolex. But there is the peculiar Howascolex.

Among the Ocnerodrilinae a distinct species of Gordiodrilus occurs, and an obviously introduced Ocnerodrilus.

It is among the Geoscolecidae that the most characteristic forms are met with. These belong exclusively to a genus of the Microchaetinae, Kynotus, which is found nowhere else but in Madagascar, where it is represented by at least twelve different species.

We have therefore in Madagascar and the surrounding islands only two peculiar genera, only four genera which can be regarded as endemic, and only about seventeen peculiar species.

Passing eastward and crossing the Indian Ocean we come to the Continent of Asia, and we shall first direct attention to the peninsula of India and adjacent parts of Burmah and the island of Ceylon, of which there has been accumulated a great deal of knowledge concerning the Oligochaetous fauna.

This quarter of the globe differs quite as much from any that have been hitherto considered as they do among themselves. We have left behind us the Geoscolecidae with the exception of the ubiquitous and clearly peregrine Pontoscolex, and the genus Glyphidrilus, which, being aquatic at times, is perhaps hardly to be considered in the present survey.

We have also in this Indian region the equally ubiquitous Eudrilus eugeniae which need not of course detain us further. The Lumbricidae are for the most part of European forms with the exception of Helodrilus indicus conceivably an actual inhabitant of India. It is among the Megascolecidae that the vast majority of the forms endemic in India are to be found. We shall take this family according to its sub-families. In the first place we note that the sub-family Acanthodrilinae is totally unrepresented. The large sub-family Megascolecinae has very numerous representatives. Of the genus Megascolex itself there are some 30 species, of which, however, two or three are looked upon as varieties. Of the allied genus Notoscolex there are ten species and of Perionyx about 13. The genera Megascolides, Diporochaeta, Spenceriella and Woodwardia have only six species between them of which three belong to Megascolides. Plutellus has five species in this region. Lampito, which is a widely spread form with but one species L. mauritii, may or may not find here its original home. It also occurs in Africa and Madagascar. Pheretima is represented by no less than 12 species, of which at any rate the very great majority are peregrine forms, and not to be safely regarded as forming an integral part of the fauna of the Indian peninsula and adjoining countries.

Of the sub-family Octochaetinae the type genus Octochaetus has ten species in this region and Eutyphoeus, which is restricted to the region, has 15 species or perhaps rather more. Hoplochaetella has but one. Of the Trigastrinae there are four or five examples of the genus Dichogaster which occur within the region now under consideration; but it is doubtful whether they are truly indigenous. On the other hand Eudichogaster, closely allied both to Dichogaster and Trigaster, seems to be confined to this part of the world where it is represented by five species. Finally we come to the last sub-family—that of the Ocnerodrilinae, which is represented by a species apiece of the genera Ocnerodrilus, Nematogenia and Gordiodrilus. The latter species G. travancorensis is alone to be regarded as endemic and it is very near to the African G. zanzibaricus, described some years since by the present author.

The remaining family of terricolous Oligochaeta found in India is the family Moniligastridae which is practically limited to this part of the world and consists of at least twenty species distributed among the genera Moniligaster, Eupolygaster, Desmogaster and Drawida, the majority belonging to the last-named genus.

This quarter of the globe is therefore inhabited by 18 genera which are certainly truly endemic, and which comprise between them about 120 species. But only four or five genera are peculiar.

The remainder of the Asiatic Continent is not very well explored with regard to its earthworm inhabitants. It seems clear however that the southern and coastal region of China and Japan with the Malay peninsula are really continuous with the mass of islands which lie between India and Australia and form together a tract of land which is characterised by an Oligochaetous fauna differing from both that of India on the one hand and Australia on the other. We shall therefore consider this huge portion of the globe as one region comparable to the other divisions that have been hitherto considered. With reference to the Lumbricidae and Geoscolecidae the same remarks may be made as in the case of India. The indigenous forms of the latter family are to be looked upon as outside of the present survey since they are largely or entirely aquatic forms. Pontoscolex corethrurus, and Eudrilus eugeniae, need not detain us for reasons already amply stated. We now come to the great family Megascolecidae. Of this family the genus Pheretima stands at the head; and of the 200 or so species that have been or can be assigned to this genus all, with merely two or three exceptions, are natives of the Eastern Archipelago and adjoining mainlands of Asia. Of other Megascolecinae the region has yielded the following genera. In Java one species of Woodwardia (W. javanica) has lately been described. It is regarded by Michaelsen however as doubtfully indigenous. And the same remark may be made of Perionyx. Plionogaster, however, with four or five species, is limited, as far as our present knowledge goes, to the Philippines and to neighbouring islands. The Acanthodrilinae and Octochaetinae are totally absent from this part of the world, there being no record even of peregrine species of these sub-families. The sub-family Trigastrinae is not however unrepresented; for of Dichogaster several species occur, such as D. malayana and D. saliens. But these are by no means certainly to be looked upon as real natives of the situations within this area where they occur. This completes the scanty list of genera found in the region under consideration; for one Moniligastrid (M. barwelli) is hardly to be looked upon as indigenous. We have therefore to record here but five genera, of which only two are certainly indigenous and probably also confined to the region; they contain between them certainly two hundred species.

We next come to the Continent of Australia. The earthworm fauna is again quite without Geoscolecidae and of course Lumbricidae. As to the former there are not even doubtful cases like Glyphidrilus of the East; for we find only recorded Pontoscolex corethrurus, a species concerning whose extraordinary powers of migration there is no possible doubt. The same may be said of Eudrilus also recorded from Australia. Here the Megascolecidae are as conspicuous as in the old world generally. We find, however, a great many members of the sub-family Acanthodrilinae. Of the antarctic genus Notiodrilus there are some six species. Microscolex occurs; but the real habitat of this genus is very doubtful. Diplotrema, with one species, D. fragilis, is not only indigenous to, but confined to, Australia.

The most prevalent sub-family is that of the Megascolecinae. Of the genus Pheretima there are two species which may or may not be truly indigenous. Of Plutellus there are forty species; there are over thirty of Diporochaeta, while Notoscolex (with which Dr Michaelsen associates Digaster, Didymogaster, Perissogaster) includes more than forty species, Fletcherodrilus has but one species, Megascolex has seventy species, Woodwardia sixteen, Spenceriella five, and Megascolides seventeen. There are four species of Perionyx which were at one time regarded by Michaelsen as necessitating a new but allied genus Perionychella; the two are now merged. Thus there are not far short of 150 species and eleven genera represented, of which only two are limited to Australia.

Having completed the survey of the central and southern land masses of the globe we next direct attention to the northern land masses, viz. North and Central America on the one hand, and Europe and northern Asia on the other. With regard to America we find besides many species of Lumbricidae the genera Notiodrilus, Microscolex among the Acanthodrilinae, Megascolides and Plutellus of the Megascolecinae, no member at all of the Octochaetinae, a considerable number of species of Ocnerodrilus and sub-genera belonging to the Ocnerodrilinae, a good many species of Dichogaster and at least one of Trigaster among the Trigastrinae, while one sub-family, that of the Diplocardiinae, is only found here and contains two genera Diplocardia and Zapotecia with quite ten species between them of which only one belongs to the last-named genus. There are no Geoscolecids (except the chiefly aquatic Sparganophilus); this family stops short in the West Indies where their presence has been briefly referred to in considering the worms of South America. The vast majority of the genera enumerated here are only found in the warmer parts of the North American continent. We have therefore in this division of the world some nine genera of which at least one, viz. Diplocardia (and its close ally, hardly perhaps to be separated, Zapotecia), is confined to it, while Trigaster only extends as far south as to the West Indies.

Europe and northern Asia, of which Europe alone and certain limited tracts of Asia are at all known, contain all the genera of the family Lumbricidae which are (according to Michaelsen) Eiseniella, Eisenia, Helodrilus, Octolasium, Lumbricus with various sub-genera of Helodrilus, such as Dendrobaena and Allolobophora. Besides these forms, which amount to at least 130 species, we find the genus Hormogaster, with two species, the only genus of the Geoscolecid sub-family Hormogastrinae; there is also a species apiece of the genera Criodrilus and Sparganophilus (sub-family Criodrilinae). But as these are at least largely aquatic they come under another set of rules and are not important in the present survey of the earthworms of the world. It will be thus seen that with these two last exceptions the area in question contains but six genera of which all are peculiar.

We have now completed the survey of the principal land-masses of the globe. We shall conclude with a reference to one of the largest islands of the world, viz. New Zealand, with which will be included a few outlying islands such as the Aucklands, Snares Island, etc. The reason for not ignoring these islands as we have ignored certain other islands of large size, such as Japan, will be apparent from the peculiarity of the earthworm fauna which they show. From New Zealand the Moniligastridae, Geoscolecidae, Lumbricidae, and, of course, the Eudrilidae, are absent, save the ubiquitous Eudrilus. The only family of earthworms which is here represented is that of the Megascolecidae. Of this family the Acanthodrilinae are very well represented. We have at any rate seven species of the genus Notiodrilus of which the bulk are from the small adjacent islands and not from the mainland. The genus Rhododrilus with nine species is actually limited to the New Zealand group and so is an allied genus consisting of two species only, viz. Leptodrilus. This latter genus is confined, so far as present information goes, to the Auckland and Campbell islands. Dinodriloides with two species is also limited to New Zealand and to the North Island. Maoridrilus with some ten species is another native and restricted genus. So too is the allied Plagiochaeta with numerous setae on each segment but with the alternating and single nephridial pores of Maoridrilus. There are several species of Plagiochaeta of which one has been lately regarded by Dr Michaelsen as really falling within the otherwise Indian genus Hoplochaetella, while for another he has formed the genus Pereiodrilus. Neodrilus with but one species N. monocystis is another peculiar New Zealand genus. The family Octochaetinae contains only four genera, of which one, viz. Dinodrilus (with four species), is limited to New Zealand, while Octochaetus has about five representatives. Hoplochaetella (if Dr Michaelsen's surmise referred to above be correct) has one species in New Zealand. The Megascolecinae are less numerous than the Acanthodrilinae, but there are ten species of the genus Megascolides (which includes Benham's genus Tokea), perhaps seven species of Diporochaeta, and two other species which Michaelsen has removed from the genus Diporochaeta and placed in Spenceriella. One Plutellus (which however may have been introduced) completes the New Zealand Megascolecines. We have therefore in this part of the world fifteen genera including between them some 58 species; eight of the genera are peculiar to the islands.

From this brief statement of facts some inferences of interest can be drawn. It is in the first place plain that every part of the world except the extreme north and south has a considerable fauna of earthworms. The one exception would appear to be the northern part of the North American continent. Here we meet with members of the family Lumbricidae which are however species that are met with in the Euro-Asiatic province and are thus to be regarded as possibly later immigrants introduced probably by man. Thus temperature short of a constantly frozen condition of the ground is not a bar to the existence of earthworms. Even a freezing of the ground for lengthy periods is not a complete obstacle to the existence of those Annelids; for I have myself received examples of Lumbricidae from the arctic island of Kolguev. Moreover the temperate regions would seem to be as fully populated in the way of individuals, and even of species, as are the tropical regions. Indeed as to individuals it seems that the temperate regions are more fully supplied than much of the tropics. This however is not quite the object of the present section to discuss. We are here concerned with the relative frequency of genera and species. There are according to a recent estimate of the Rev. H. Friend some forty species recognisable in Great Britain. And as already has been stated the earthworms of Europe amount to perhaps 130,—at any rate well over one hundred. In tropical America there are hardly more. But in the latter case the number of genera is very greatly in excess of that of Europe. We cannot however say that an abundance of generic types is quite characteristic of the tropics. For the Eastern Archipelago, though rich in species, is but poor in genera, not possessing more than half a dozen or so. And on the other hand the temperate climate of New Zealand has produced a very considerable series of genera, much more than those of the islands of the East and nearly as many as those of, for instance, Central America and the West Indies.

This conclusion is in its turn contradicted by the conditions observable in Chili and the temperate regions of South America, where the number of species is large but the number of genera small. In short no general laws, in the present state of our knowledge, can be laid down as to the connection between species and genera on the one hand and climatic conditions on the other. In this department of our subject we cannot do more than has already been done, i.e. to state the actual facts. One is tempted in comparing the rich fauna of tropical Africa with the very limited fauna of Madagascar to associate a richness of types with extent of land surface. In the two cases cited this conclusion is obvious. It may also be extended—if we confine ourselves to species and not to genera. For the two great islands of New Zealand have not between them more than fifty species of earthworms, while Australia has four or five times that number. It will be noticed however that we cannot associate poverty of generic differentiation with limited land masses; for New Zealand has a large number of generic types, very many more than the huge Euro-Asiatic tract of continent.