Earthworms and Their Allies

by Frank E. Beddard

The Range of Genera.

We have seen, and shall again refer to the fact, that individual species of earthworms have not as a rule a range over a great extent of country, save only in those cases such as Pheretima heterochaeta which belong to that physiological section of these worms called 'peregrine' forms; these appear to possess some means of extending their range by the assistance of man which is denied to other forms. Apart from these instances, which do not come under the present category, it is only Lumbricus and its immediate allies, Helodrilus, etc., of which certain species are found to exist over wide tracts of land. There are however many genera which have a wide range and which may be contrasted with others in which the range is very limited. The two extremes are moreover connected by forms with an intermediate range. There is no doubt whatever that the genus with the widest range is Notiodrilus of which species are found throughout the antarctic region, viz., in Patagonia, the islands of the Antarctic Ocean, the Cape of Good Hope, New Zealand, and also further to the north, sometimes even to and beyond the tropics in America, Australia, and Africa. There is no other genus of which the genuine extension (i.e. not in any way due to man) is so great as this genus Notiodrilus. And this fact gains much significance from the now generally accepted view that in its anatomical structure Notiodrilus comes near to the original type of earthworm.

Perhaps the next most widely distributed genus is Helodrilus of the family Lumbricidae which occupies Europe and Asia to the extreme east, and is thought also to be indigenous to certain parts of North America. But this range, though equally wide perhaps in mileage, is less impressive than that of Notiodrilus, since the land areas inhabited by the genus are continuous—almost so if we accept North America as its real habitat. Here we have a case precisely the opposite of that of Notiodrilus; for while there are reasons for regarding Notiodrilus as an ancient form of Lumbricid, there are equally good reasons for regarding the Lumbricidae as the most modern family of earthworms.

To find other instances of widely spread genera we must recur to the great family Megascolecidae. There are practically no Geoscolecidae which have a really extensive range. The only instances are Criodrilus and its ally Sparganophilus which occur in America, whether North or South, and in Europe; but as these forms are at least largely aquatic the facts are not quite comparable with those now under consideration.

The genus Dichogaster (which includes as synonyms Benhamia, Millsonia, Microdrilus) is unquestionably indigenous to tropical Africa and certain parts of America including the West Indies. It has been also met with in the East; but as the species there occurring, such as for example the species originally described by myself as Microdrilus saliens, are of small size, an accidental introduction is quite possible, and it is by no means certain that it has not occurred. In any case the genus is known to possess species which are undoubtedly to be reckoned among peregrine forms—such as D. bolavi, which has turned up in Europe. Gordiodrilus and also Ocnerodrilus with its sub-genera have very much the same range as has Dichogaster. It is to be noted however that these forms are circumtropical, and that their distribution is thus less continuous than that of Notiodrilus; they do not however show the markedly discontinuous range of certain other genera of Megascolecidae. For instance Octochaetus is well known from New Zealand, and, not occurring in the intermediate tracts, is again met with in India. Hoplochaetella is believed by Michaelsen to present us with another precisely similar instance. Then also the genera Woodwardia and Notoscolex are to be found in Australia and again (absent from the immense tract lying in between) in Ceylon. Megascolex has much the same range, showing also this marked and remarkable discontinuity. Stranger still, perhaps, is the range of Plutellus and Megascolides, of which the former, chiefly found in Australia and Tasmania, not only extends its habitat to Ceylon but also to North America; it is there represented by Eisen's species Argilophilus marmoratus, referred by him, and not unnaturally, to a distinct genus, but placed by Michaelsen in Plutellus. Megascolides is Australian and from the North Island of New Zealand, where its species were regarded by Benham as of a distinct genus, Tokea. There is also one form, Megascolides americanus, in the western region of North America.

The two genera Yagansia and Chilota, closely related to Notiodrilus, have a range which is short of that of Notiodrilus, and we shall see later that there are reasons for regarding these genera as derived from Notiodrilus. They are met with only in the south of South America, and in the Cape of Good Hope region.

The range of Microscolex seems to be much the same as that of Notiodrilus; but it is a little uncertain how far the genus is really autochthonous in the countries where it occurs; and in any case it differs from Notiodrilus in occurring in Europe, where the species has been named for a long time M. phosphoreus. We do not positively know whether this is 'peregrine' in Europe or not.

The range of the antarctic Acanthodrilinae is in a sense continuous; for they argue the former northward extension of the antarctic continent and in any case they occupy neighbouring land masses. In Octochaetus and Plutellus the case is different and one of real discontinuity. There are however cases of wide range which is also actually continuous and such is afforded by the genus Pheretima. This genus appears to be possibly indigenous to Australia; in any case it reaches from the Solomon Islands on the east to India towards the west, being found in all intermediate continents, while it reaches Japan on the north side of this large area.

There are other genera which extend their range over a considerable area, but which are not so widely distributed as these which we have just been considering. Thus Diporochaeta is chiefly Australian but also reaches even the South Island of New Zealand and the southward lying antarctic islands. Desmogaster and Eupolygaster among the Moniligastridae range from Burmah in the east to Sumatra and Borneo further east, though they are not recorded from intermediate islands. Perionyx is found in Burmah, India, Zanzibar, Sumatra, and Java. There are other examples of genera which have much the same range as those enumerated. Finally there are those which are confined to one land mass and very often to a restricted region of that. Thus Kynotus is confined to Madagascar, all the genera of Eudrilidae to tropical Africa, some of them, e.g. Beddardiella and Euscolex, to very limited tracts, others to wider or less wide areas in that continent. Maoridrilus is only found in New Zealand, to the South Island of which also is confined the genus Neodrilus. To the Cape region of Africa is limited Microchaetus; and to a belt running across the northern part of the tropical region and extending down the Nile, the remarkable, partly aquatic, Alma.

As a kind of appendix to these facts and conclusions we shall next deal with certain widely spread forms that have been already referred to, with the range of different genera over great land masses of the world, and with the earthworms of oceanic islands.

CHAPTER VI
PEREGRINE FORMS

Dr Michaelsen has used this term to describe those species which possess some powers of migration over the sea, denied to the majority of worms, and probably due to the direct interference of man. Thus we find in collections of earthworms from various parts of the world not only examples of forms which do not come from other parts of the world, but also a few which occur in many or even most of such collections. It is for example to be actually expected that a collection of earthworms made in South America, the Philippine Islands, or Australia will contain examples of the apparently ubiquitous Pontoscolex corethrurus. This is what has actually happened in cases of which I have personal knowledge, as well as in many others recorded in the literature of the subject. I have myself received this worm from the three parts of the world mentioned, and also from Hawaii. Others have increased its known range to other parts of the South American continent, to Central America, the West Indies, the islands of Sumatra, Java, Borneo, Celebes, Mauritius, and Madagascar, etc. It is in fact found everywhere in the tropics. With this range may be contrasted that of another genus of the same family (Geoscolecidae), viz. Kynotus, which, though consisting of many species, is not found outside of the Madagascar district. It should be added that Pontoscolex does not appear to contain more than two species, the one not mentioned in the above survey of its distribution being P. insignis of Kinberg, which is apparently the same as P. liljeborgi of Eisen, and is limited to certain parts of America.

Before attempting to grapple with the remarkable facts implied by the distribution of this genus, it will be well to survey the whole group of Oligochaeta and to reduce to as short a space as possible the total series of facts which are of the same nature.

A case, even more striking than that of Pontoscolex, is afforded by the Eudrilid genus Eudrilus. As with Pontoscolex there are two species of this genus, one, E. pallidus, being confined to West Africa, the remaining one, E. eugeniae, being world-wide in range. This latter species has received the following names, viz. E. decipiens, E. lacazii, E. peregrinus, E. sylvicola, E. boyeri, E. jullieni, E. erudiens, and E. roseus; they appear to be all synonyms of the name originally given by Kinberg who however did not recognise the distinctness of the form as a genus. It is now known as Eudrilus eugeniae. The variety of names given to supposed different forms (for two of which I am myself responsible) is due to the fact that in earlier days when nothing was known about the geographical distribution of this group of animals it was thought by no means unreasonable that a given genus represented by several species should range over the globe. This fact coupled with imperfect description of structural details led to the multiplication of supposed species, a position which is no longer tenable. This worm is quite as abundant in gatherings from all parts of the world as is Pontoscolex corethrurus; and in addition to the countries inhabited by the latter, Eudrilus eugeniae has been met with in New Caledonia: tropical Africa is probably its original home.

The two families that have been hitherto considered offer no further instances, among their many species, of worms with so wide a range as those just dealt with. There are indeed one or two forms, e.g. Criodrilus and Glyphidrilus, which have a considerable range though not nearly equalling that of Eudrilus and Pontoscolex. These are, however, aquatic forms and the range of aquatic forms is determined as far as we can see by a different series of causes to that of terrestrial forms, which are referred to later.

Among the Moniligastridae we have apparently an instance of a peregrine form. The genus itself has its headquarters in Ceylon and extends a little way in other eastern regions; there is, however, one species, Moniligaster bahamenis, described some years since from the Bahamas which must surely be an example of a peregrine form, particularly since it is probably identical with M. japonicus whose name is indicative of its habitat.

Among the huge family of the Megascolecidae there are a considerable number of species which apparently possess the same facilities for making their way in different directions and across seas from the locality that is thought to be their real home.

Of the very many genera, however, of which this family is composed, a comparatively small number are thus peregrine in habit at times. All the species known which are distributed broadcast, more or less, over the tropics belong to the genera Pheretima, Microscolex, Dichogaster, Megascolex, Perionyx, Ocnerodrilus, Kerria. These several genera are placed in order of frequency of exotic occurrence. Indeed of the two latter genera their frequent life in fresh water may really remove them from the present category altogether. In addition to these are some perhaps more questionable instances, such as the genus Gordiodrilus which, prevalently West African, has also been found in the West Indies, in East Africa, and in India and Madagascar. These instances I propose to leave out of consideration in the present sketch. The most obviously peregrine genus of all those enumerated is Pheretima, which according to my experience turns up in almost all gatherings of earthworms from any part of the tropical or even sometimes temperate regions of the world. It seems to be fairly well settled that this extensive genus has its real home in the islands of the Eastern Archipelago, perhaps extending a little in various directions from that centre. But examples of the genus have been found in almost all other regions. And what is especially to the point in considering the facts, as will be pointed out with more emphasis later, the assumedly peregrine species do not differ from those found in the real district in which the genus is indigenous.

Dr Cognetti de Martiis enumerates in the Neotropical region, that is in South and Central America and the West Indies, the following species: Pheretima biserialis, P. californica, P. capensis, P. elongata, P. hawayana, P. hesperidum, P. heterochaeta, P. houlleti, P. posthuma, P. rodericensis, P. schmardae and P. violacea. Of these twelve species it is quite certain that the last six occur in the East, where they are doubtless indigenous. So too do the species P. biserialis, P. capensis, and P. hawayana. The synonymy of the different species of this large genus is not yet in a completely settled condition. But in the meantime it is in my opinion quite possible that both P. hesperidum and P. californica are identical with species also occurring in the East. There remains the somewhat doubtful P. elongata from Peru which has not been very fully described. There is thus no convincing evidence of species really indigenous to and confined to any part of America. Some of these species also occur in many other parts of the world. For instance, P. heterochaeta is very widely spread indeed, occurring as it does in Australia, New Caledonia, Madagascar, and even England (in hothouses). This species indeed is the most prevalent of all Pheretimas and seems to be abundant in gatherings of earthworms from various localities as are Eudrilus and Pontoscolex.

From the island of Madagascar and neighbouring islands the following species of Pheretima have been obtained and identified by Dr Michaelsen: viz. Pheretima pentacystis, P. peregrinus, P. heterochaeta, P. biserialis, P. rodericensis, P. houlleti, P. robusta, P. mauritiana, P. taprobanae, and P. voeltzkovi. It will be noticed that the majority of these are also included in the list from South America, and that many of them are also found in other parts of the world, and nearly all of them in the East. There remain a few which are doubtful. It is quite possible that P. mauritiana is the same as P. hawayana and P. bermudensis, in which case it has a world-wide range. P. taprobanae is well known as a Ceylon species. P. robusta also occurs in the East Indian islands. There remain P. pentacystis, P. peregrinus, and P. voeltzkovi. P. peregrinus is known from Australia and also from Sumatra, so that that species need not concern us in enumerating those which are possibly endemic. In fact it is only P. pentacystis and P. voeltzkovi which may be really Mascarene.

Another peregrine genus belonging to the sub-family Acanthodrilinae is Microscolex. But the limits of this genus may be regarded as at present rather uncertain. And this difficulty somewhat affects the bearing of the facts to be related, though it hardly affects the value of the facts themselves. Dr Michaelsen referred to the genus in his great work seven well-defined species, and four others not so plainly distinct. Of these eleven, two are confined to New Zealand, four to North and Central America, one to Hawaii, one to Madeira, one to Algeria, while the remaining two are found pretty well over the whole surface of the world. More recently the same authority has somewhat extended his view of the generic characters, so as to include a number of forms found in Patagonia, Cape of Good Hope, and the antarctic region generally, while he has lumped together into two species only, viz. M. phosphoreus and M. dubius, the eleven forms just mentioned, which species therefore are absolutely world-wide in range, and thus form an excellent example of a peregrine form. These species moreover differ from Pontoscolex and some others in that they have been able to establish themselves in Europe. Dr Michaelsen also relates that in the cultivated lands of South West Australia, Microscolex dubius and Helodrilus caliginosus are actually the commonest species; and he calculates that they form together quite 90% of the earthworms gathered in any locality belonging to this region.

Some of the other Megascolecid peregrine forms will be referred to later. There is no doubt that the family Lumbricidae offers by far the greatest number of peregrine forms and that these are most abundant in collections from extra-European countries, where the collector has searched in cultivated lands. There are at least eight or nine species which are common in many parts of the world though their original home is undoubtedly Europe.

This is a brief review of the facts, more detailed in some cases than in others. It remains to review and compare the results arrived at.

The first general statement that may be made is that this faculty of extending their range beyond the limits assigned by nature is not confined to any one family. For all the chief sub-divisions of the terrestrial Oligochaeta seem to possess it, though in unequal degrees. But the inequality may be more apparent than real. For if it be recollected that the species of the family Megascolecidae are very much more numerous than those of the Eudrilidae or even the Geoscolecidae, the fact that there are more peregrine Megascolecidae will lose some of its importance. With the Lumbricidae the case seems to me to be different. Here the preponderance, not only in species (relatively speaking) but in individuals, is much above that of other families. This preponderance I should be disposed to assign to the newness of the family coupled with the vigour seen in new races. That this is a possible explanation is borne out by the fact that the 'Perichaetidae' (i.e. the genus Pheretima) is the most salient race of peregrine Megascolecidae, and it is now generally held that this group is the most modern of that enormous family.

Another general statement may be made with even more confidence, viz. that it appears to be an undoubted fact that some species are more capable of extending themselves than others. Thus Eudrilus eugeniae occurs everywhere on the great land masses of the globe, except in Europe; it is in fact circummundane in the tropical zone, as is also Pontoscolex. Dichogaster bolavi is again a trifle more restricted in its range, having been recorded from tropical Africa, South America, West Indies, Madagascar, and India. Its occurrence near Hamburg in Europe is also to be noted. A little more restricted still is Nematogenia panamaensis whose range is in Central America, tropical West Africa, and Ceylon. Lastly there are cases such as Pheretima taprobanae which, a native of Ceylon, is also found in Madagascar.

It may be asserted in the third place that there are no peculiarities of structure shared by all of these peregrine forms which might account for their physiological similarity, except indeed the somewhat negative feature which they have in common, that is of being of small or moderate size. Eudrilus and Pontoscolex are not isolated types in their respective families; nor do they seem to approach each other in any respect. Nor can it fairly be said that these peregrine species are marked by any great variability of structure as compared with other forms, which might allow for their suiting themselves to various climates and conditions. It is true that Eudrilus eugeniae has received many names which might at first argue some variability. But these names have been perhaps given by persons rather under the influence of the idea that remote habitat implied specific difference, and who were thus inclined to see minute differences, and who perhaps were furthermore led astray in the matter by imperfectly accurate descriptions on the part of others. Certainly some of the peregrine species of Pheretima are subject to some variation, particularly in the number and arrangement of their genital pupillae. But this feature is by no means confined to those species and cannot be utilised as in any way an adaptation to wide distribution.

But while we can lay down no general explanation of the phenomenon, it is possible to furnish some explanation of particular cases. Thus the genus Microscolex is the only exotic genus which appears to have established itself in Europe, from which country indeed it was early known as an apparently indigenous inhabitant. We must put this and some similar cases down to ability to do without great heat. It is probable in fact that the original home of Microscolex is the antarctic half of the globe; and this of itself would allow of its establishing a new home in the northern hemisphere, did other circumstances allow of it.

It might be urged that this genus has been able to establish itself in Europe because it has in fact had the chance denied to other species. There are a good many, however, which would in that case be in the same category. Some years ago I received from time to time a very large number of earthworms from the Royal Gardens at Kew which had been accidentally imported thither from many quarters of the globe, among which I described some eighteen or twenty new species including, for instance, the African genus Gordiodrilus. There are plenty of facts of a similar nature and Dr Michaelsen has pointed out that botanical gardens act as centres of dispersion for accidentally introduced Oligochaeta. We must therefore come to the conclusion that temperature is at least one of the causes of a difference in the capability of extending their range shown by the Oligochaeta, a cause which doubtless operates as a check upon extension of range in non-peregrine forms also, and prevents for instance the dispersion of the tropical African Eudrilidae into the region of the Cape.

We may, as it appears to me, confidently look upon indifference to varying temperature as a condition of ability to colonise new countries. But it is obvious that this is not of itself a sufficient cause to explain the facts. Otherwise this country and N. Europe would contain many antarctic earthworms; the only one that has been recorded to my knowledge is Microscolex.

Though an inability to endure a temperate climate may have rendered the movements of peregrine species more limited, the same or rather the exactly opposite cause does not seem to have played any important part in this direction. For it is above all the Lumbricidae, normally dwellers in temperate climates, that are so remarkable for their wide range over the world. Nor can it be convincingly asserted that the extra-Palaearctic Lumbricids are real indigenes of those—often tropical—countries. For if so we should expect them to be at least of different species. Lumbricids however from South America, Australia, etc., are specifically identical with European forms.

There is no doubt that wherever land has been at all long under cultivation in any part of the world that land will be found to produce species of the European genera Lumbricus, Helodrilus, Eisenia, etc. More than this the recently imported European forms will be found to have largely or almost entirely driven out the native species, which have retired more into the interior of the country. There is thus here no barrier placed by temperature. It should be remarked, however, that while these earthworms are most abundant in the less tropical regions, they occur in such tropical districts as Peru, though in less striking numbers. Whether those of North America are really indigenes or not remains perhaps a matter for discussion; but it is at least noteworthy that the vast preponderance of species occurring there are also European and even British. In this particular case, which is on the whole the most emphatic of all the cases of peregrine earthworms, some explanations are possible, or at least have been offered. In the first place it would appear that earthworms are more abundant as individuals in northern countries where the soil is rarely dry for prolonged periods. And as has been already pointed out there is a close relation between earthworms and agriculture. Dunghills are fertile gathering grounds for some species, and ploughed fields and gardens always swarm with several species. In the tropics these animals are not so evident; and the strong rays of the sun appear to drive them further underground and into marshes; this obviously lessens the chance of their accidental transference by man. Furthermore Dr Eisen has pointed out that the European species are apt to have clitella and to be fertile all the year round, which is not always the case with other genera. That naturalist has added to this observation the fact that in rich cultivated soils in California it is impossible to find anything but imported European species, since cultivation itself appears actually to drive away the native forms.

CHAPTER VII
THE EARTHWORMS OF OCEANIC ISLANDS

Oceanic islands are islands that have always been islands, a definition that seems tautological until we compare it with some other land masses that may be termed 'islands.' Geology teaches us in fact that from the point of view of their origin islands may be divided into two quite sharply contrasted classes. There are those detached land masses usually lying near to or comparatively near to some continent, which have been in the course of time detached by the action of the waves from that continent, such as for instance the British Isles, which undoubtedly represent a portion of the European continent which was once quite continuous with Europe. On the other hand we have the Hawaiian archipelago, St Helena, Fernando Noronha, and other similar islands, which are more remote in their position from continents and concerning which it seems clear that they have originated de novo by the action of submarine volcanos or of the growth of coral, combined with subsidence, following elevation, or from several of the causes combined. In any case the islands which are termed oceanic islands have never formed part of a continent. They are not relics of previously existing continents. It becomes a matter of great interest to compare the earthworms which are to be found upon oceanic islands with those which inhabit continental islands. Fortunately there are a good many facts at our disposal for this purpose; and we shall compare the earthworms of the Hawaiian archipelago with those which are found upon certain small islands lying to the south of New Zealand, viz. Campbell and Auckland islands and the more southern Macquarie islands.

The earthworms of the Hawaiian archipelago have been studied by a good many persons, and altogether a number of species have been described from that group of islands of which the following is a list: Pheretima hawayana, P. heterochaeta, P. peregrina, P. schmardae, P. hesperidum, P. morrisi, P. perkinsi, P. biserialis (= P. elongata), Allolobophora putris (= Kinberg's Hypogaeon havaicum), A. foetida, A. caliginosa, A. nordenskiöldi, A. limicola, A. rosea, and finally the well-known Pontoscolex corethrurus. Of these species there is only one which is even possibly a form limited to the Sandwich Islands, and that is Pheretima perkinsi, a species which I myself at first described as a new form, but which was afterwards regarded as identical with P. heterochaeta by Michaelsen, and later still resuscitated by Ude. All the others are found in many parts of the world and not only in the nearest mainland to the archipelago which we are now considering. I have had already occasion to speak of some of them as peregrine forms, especially of Pontoscolex corethrurus which occurs all over the world.

The conditions which have been recently revealed by an exploration of the antarctic islands mentioned above are totally different. Dr Benham has enumerated the following species from those islands, viz. Notiodrilus haplocystis, N. fallax, N. aucklandicus, N. campbellianus, N. macquariensis, Plagiochaeta plunketi, Rhododrilus cockayni, Leptodrilus leptomerus, L. magneticus, Plutellus aucklandicus, Diporochaeta heterochaeta, D. brachysoma, D. helophila, D. perionychopsis among the Megascolecidae, besides Phreodrilus campbellianus, Pelodrilus tuberculatus, P. aucklandicus and the Lumbricid Helodrilus constrictus. There were also four species of purely aquatic Oligochaeta which we shall leave aside from the present enumeration, as their range in space is a matter requiring a different explanation from that of the terrestrial forms. Here we have a series of worms, all of which, save the widely spread Lumbricid, are apparently absolutely indigenous to the islands mentioned since they are all different as species from those found elsewhere. Indeed there is a whole genus Leptodrilus, consisting, it is true, of but two species, which is a native of the Campbell and Auckland islands and of those only. The other genera are found in the antarctic region, while Pelodrilus is still more widely spread.

These facts as will be observed contrast about as strongly as they can with those supplied by the fauna of Honolulu and its adjacent islands. Not only are the worms of the antarctic islands different species from those found elsewhere, but the majority of them do not consist of widely ranging peregrine forms. It appears therefore most probable that these islands are not oceanic islands but a portion of the former existing northern portion of the antarctic continent. Were the species identical with those of New Zealand this conclusion would have of course to be reconsidered. The barriers to migration (see chap. VIII) explain the contrast recorded in the foregoing pages.

CHAPTER VIII
MOVEMENT AND MIGRATION AMONG EARTHWORMS

That earthworms can move upon the surface of the ground at a rapid pace is probably well enough known to everyone, and that they can also burrow with considerable celerity. Multiplying the inches of progress in minutes of time by centuries with the resulting miles, it is quite clear that there is no reason to suppose that an individual earthworm might not enormously extend its range under favourable circumstances. Their powers of locomotion are such that they could in the course of comparatively few centuries people a continent. As a matter of fact these animals are frequently very widely spread upon a given land surface; but on the other hand they are sometimes equally limited. It behoves us therefore to enquire the reasons for the possibility of extended migration and the causes which have led to its restriction. We are now, it must be borne in mind, considering these animals as purely terrestrial animals moving over the surface of the land by their own unaided efforts. We leave out of consideration any possible assistance in crossing water, whether fresh or salt. We have to consider in fact in the present section the earthworm inhabitants of larger and smaller tracts of continuous land such as the African continent, which will serve as an excellent example wherewith to test the facts and inferences.

And as a 'control' we can compare this continent with the very different continent of Europe.

As an excellent instance, because of the certitude of specific and in most cases of generic distinctions, we may take the Eudrilidae as illustrative of the facts that are to be considered in the present section. That family consists, as will be remembered, of 33 genera at most, which have the following more exact range on the African continent. The genus Eudriloides occurs in British and German East Africa and has been met with as far south as Mosambique and even Durban, in which latter locality it has been thought that it is really an accidentally introduced stranger. Platydrilus is limited to eastern equatorial Africa, thus not having quite the range of Eudriloides.

The small genera (that is small in numbers of species) Reithrodrilus, Bogertia, Megachaetina, Metadrilus, Notykus have the same limitation of range as the last genus. Metschaina has a wider range from tropical North East to lake Tanganyika. Stuhlmannia has a wider range still being found as it is in the Tanganyika district, in tropical North East Africa, and in British and German East Africa near the coast. Pareudrilus reaches still further north while Nemertodrilus is limited to the Mosambique region and to the Orange River district further south. The only remaining genus of this sub-family of the Eudrilidae is Libyodrilus which is purely West African and equatorial.

Of the remaining genera which are usually grouped together into a second sub-family, five, viz. Malodrilus, Kaffania, Gardullaria, Teleudrilus and Teleutoreutus, are confined to tropical North East Africa. Eminoscolex occurs in the same district but also to the south in the great lake region. The most remarkable fact about this genus is that one species E. steindachneri comes from the Cameroons, and another E. congicus from the Congo, and thus the range of the genus is right across the continent. Neumanniella has much the same range. Polytoreutus is a purely equatorial East and Central genus, reaching from the coast to the lakes. Bettonia known by three species is from British East Africa.

The remaining genera, viz. Hyperiodrilus, Heliodrilus, Alvania, Iridodrilus, Rosadrilus, Euscolex, Parascolex, Preussiella, Buttneriodrilus, Beddardiella, Metascolex, are all West African and the vast majority equatorial. We thus see that with one exception the genera of East Africa are totally different from those of West Africa and that the family as a whole is restricted in its range to a comparatively small part of the vast African continent. It also obviously follows, and it is advisable to state this fact however obvious, that no species are common to the two sides of the continent except indeed the ubiquitous Eudrilus, whose range over the world has been more than once referred to in this book.

On the other hand the genus Dichogaster offers quite different facts, which are in contradiction to those just enumerated. This genus as already said is very characteristic of tropical Africa, and a large preponderance of the known species are confined to that continent. Although there is some variation in structural characters among the many species which compose this genus, there is but little doubt that they are all rightly referred to one genus with perhaps some doubtful, though not very striking, exceptions. In any case the utmost divergence of structure between worms usually placed together in this genus is nowhere near to that which separates the genera of Eudrilidae from each other. Of the African members of the genus the species are pretty evenly divided between the eastern and western halves of the continent; they are, like the Eudrilidae, tropical in range, not occurring to the southward, where their place is taken by the Acanthodrilinae and Geoscolecidae. There are it is true a few species, such as D. gracilis and D. bolavi, which are common to the two sides of Africa; but in these cases we clearly have to do with those rather mysterious species which can apparently unduly extend their range and which are known as peregrine forms; for they also occur in other parts of the world besides Africa. We have therefore in Dichogaster the case of a genus which ranges all over the tropical parts of Africa, but whose species are not common to the Atlantic and Indian shores of that continent.

We will now contrast these conditions, which exemplify certain facts shown by the characteristic Oligochaeta of tropical Africa, with those which obtain in Europe. In this region of the world the prevalent and practically the only genera which need be taken into consideration in surveying the Oligochaetous fauna from the present point of view, are Lumbricus and the genus Allolobophora of Eisen which has been variously rearranged into genera and sub-genera known by the names of Helodrilus, Bimastos, Octolasium, etc. The structural differences which divide these genera and sub-genera are not great; in any case they do not exhibit such a wide range of variation from each other as do two such Eudrilid genera as Stuhlmannia and Hyperiodrilus. We find the genera mentioned not only in Europe but extending themselves over more or less of Asia, even occurring in Japan; while the North American continent contains also representatives of the same. Not only do we find this community of genera over vast extents of country greater in diameter than the African continent, but there are also many species which range as widely or nearly as widely as the case may be as the genus to which they belong. Thus the species of Allolobophora (we do not trouble about the newer sub-divisions as they hardly affect the facts to be emphasised), A. caliginosa, A. longa, A. rubida, A. chlorotica, A. octaedra, A. constricta, A. beddardi, Lumbricus terrestris, L. castaneus, have an enormously wide range over what is generally termed the Palaearctic region, extending also in some cases into the Nearctic. It is true no doubt that the majority, indeed perhaps all, of these are, like certain species of Dichogaster mentioned above, among those forms termed peregrine which have the capability of living in every quarter of the globe to which they have apparently been conveyed by man. But there remain many species which have a very extended habitat in the northern hemisphere, and in any case the genera and the species are there truly indigenous and widely spread.

It would thus appear that the capability for independent migration varies greatly among earthworms. Of the types selected for consideration the Eudrilidae are the slowest movers; the genus Dichogaster comes next, while the power of migration possessed by the genera Allolobophora and Lumbricus is very much greater. Assuming for the moment the correctness of this inference it is clear that it will influence many other propositions connected with the relative age of the families of these worms and with many problems of geographical distribution. It appears to us that this simple explanation is the correct one. But to show this it will be necessary to eliminate other possible explanations. It might be urged that the wider range of the genus Dichogaster and the still wider range of the genus Allolobophora (shown by community of species in widely distant localities) was evidence merely of relative age, that the older groups have had more time to travel and that the newer groups have not had so long a time to spread themselves over their habitat. On this hypothesis the genera of Eudrilidae would be geologically much newer than the genus Dichogaster and similar statements might be made for the other forms here under consideration. As already explained we cannot attempt to answer this question in the only way in which it can be really satisfactorily answered, by a reference to fossil forms; for there are no fossils to refer to. So far as comparative anatomy enables us to arrive towards a solution of the question, it would appear that the genus Dichogaster belongs to a more ancient race than either of the other two groups considered, and that of these latter the Lumbricidae are the most modern. Moreover we associate not only a wide, but also a discontinuous, distribution with an archaic race; and for this reason also we should place the genus Dichogaster in the position of being the most ancient of these Oligochaeta. For the genus occurs in Central America and in certain parts of the East as well as in Africa. So that we can fairly dismiss the view that the Lumbricids by virtue of their greater range over a given area are the most ancient type and that their range is associated merely with their antiquity. Nor does it appear that geographical or meteorological consideration can have had effect in the present instances. For conditions favourable to earthworms prevail in tropical Africa, as in Europe and much of North Asia.