Skip to main content

Full text of "A monograph of the order of Oligochaeta"

See other formats


"' ■ ^' 

f.i.".«j:,.-:*"I--- ■—'-■» ' - . i--; 

L-.wKb*.''":-.- ..I... •■-.•:;:;: .- :■ .-.. ■. 

2»3&A>f2K.>.;':L'.:- ■.-..!>„:.-•. ,.; v.r^ 







Cornell University Library 
QL 391.04B39 

A monograph of the order of Oligochaeta, 

3 1924 003 403 924 


Cornell University 

The original of this bool< is in 
the Cornell University Library. 

There are no known copyright restrictions in 
the United States on the use of the text. 



Oxford University Press Warehouse 

Amen Corner, E.G. 








M.A. (OxoN.), F.R.S. 









rriHERE are already two works dealing with the Order Oligochaeta. The 
-*- first of these, in time of publication, is Professor Franz Vejdovsky's 
'System und Morphologie der Oligochaeten,' which appeared in 1884. Six 
years later Professor L^on Vaillant contributed to the volumes on the 
Annelids of the ■ ' Suites h, Buffon ' a volume and a half dealing with the 
same group. It might appear, therefore, at first sight, that the ground has 
already been taken up, and that there is no occasion for the present work. 
Tt must be recollected, however, that it is nearly ten years since the 
publication of Professor Vejdovsky's treatise, and that during this period 
our knowledge of the Oligochaeta, more particularly of the earthworms, 
has increased enormously. Professor Vejdovsky's object, moreover, was more 
especially to give an account of the structure of the group from his own 
observations, and to spend less time in dealing with the results of other 
observers — not that the previous work upon the group was by any means 
neglected by him, but no great amount of detail was given as to the 
genera and species of Oligochaeta not occurring in his native country. 

M. Vaillant's work is rather more comprehensive in scope, but there 
is no account of any researches made after the year 1 886 ; so that the large 
amount of literature which has appeared since that date has not yet been 
incorporated into any general work. M. Vaillant's contribution also is rather 
more devoted to the systematic side of the subject than to the description 
of structure. 

Under these circumstances it appeared to me that there was room for 
a treatise of rather wider scope than those of either Professor Vejdovsky or 


of M. Vaillant, and one which should deal with the entire subject up to the 
date of publication. As will be seen from the bibliography which I have 
given in the Appendix, the literature of this subject is large and scattered ; 
so that to bring together under one cover all of importance that is as yet 
known about the group of the Oligochaeta will possibly be considered not to 
have been a useless performance. In attempting this labour I have received 
much kindly encouragement from Professor Ray Lankester, which I gratefully 
acknowledge. I have been so fortunate, through the kind influence of 
Dr. Burden Sanderson, as to secure the assistance of the Clarendon Press, 
to the Delegates of which institution I wish to express my great indebted- 
ness. The cost of the necessary illustrations, which has been considerable, 
has been largely defrayed by the generosity of Mr. J. P. Gassiot, F. Z. S., 
who, at the , suggestion of Mr. Sclater, has placed a sum of £ioo at my 
disposal for this purpose. Without this very liberal act of assistance on 
Mr. Gassiot's part this work could hardly have been undertaken. 

I may now say a few words about the scope of this monograph. The 
study of the group with which it deals is one which has occupied my leisure- 
time for the past fifteen years ; I have had the opportunity of dissecting 
and examining most, if not all, of the more important types ; so that while 
a good part of the volume is necessarily derived from compilation, a con- 
siderable proportion of it is the result of first-hand knowledge. Whilst 
I have amalgamated my recent papers into the present work, I have also 
incorporated with it a certain amount of new matter which I have not 
published elsewhere, and I have given, in the systematic part, descriptions of 
a few new species. I have not, however, made any lengthy investigations 
for the express purpose of this monograph, but I have in several cases verified 
the statements of others, and have corrected, in a few minor particulars, 
errors of observation on the part of myself, as well as. of other zoologists. 
Those who have not followed closely the progress of knowledge in this group 
of animals may be surprised at the large number of pages which it has taken 
me to set down the facts : I have erred, however, rather on the side of 
compression and omission than on that of undue prolixity. The omissions 
(with a few exceptions to be noted immediately) will not be found, I trust, 
to be of facts of much importance. They chiefly concern the progress of our 


acquaintance with the group. It has not appeared to me to be necessary 
to go at great length into the history of erroneous views or of misstatements 
of fact — at any rate compression and omission here seemed to be more per- 
missible than in other departments of the subject. I have deliberately omitted 
to give any account of the early stages of the development of the Oligochaeta ; 
the development of organs will be found treated of to a certain extent. 
I determined to do this principally on account of Prof Vejdovsky's recently 
published ' Entwickelungsgeschichtliche Untersuchungen,' an elaborate and 
finely-illustrated work, which goes into the matter with all details, and 
treats of the rest of the literature in a most thorough fashion. Moreover, 
I have personally no first-hand acquaintance with the early development 
of the Oligochaeta — another circumstance which leads me simply to refer to 
the work already quoted those who are desirous of ascertaining what is known 
about the embryology of the Oligochaeta. Another branch of the subject 
which I have thought it well to abridge is the section dealing with 
unrecognizable species. No name, I hope, has been omitted, and no reference ; 
but I have not, except in a few instances, gone at any length into the endless 
possibilities as to the identity of species imperfectly described and now 

While this work was in preparation two excellent revisions of two 
families of the Oligochaeta have appeared : I refer to Dr. Rosa's detailed 
account of the Lumbricidae, and to Dr. Michaelsen's ' Synopsis ' of the 
Enchytraeidae. The appearance of these two valuable papers caused me to 
hesitate a little before dealing with the respective families in this work. 
I have, however, thought it best to incorporate both of them, not, I hope, 
without critical examination. 

In the preparation of the systematic part of this monograph I received 
the greatest assistance from M. Vaillant's work ; I gladly acknowledge that 
it formed the basis of my preliminary (manuscript) account of many of the 
species of Oligochaeta, and that it has saved me a vast amount of labour in 
compilation ; I believe, however, that I have nowhere followed M. Vaillant's 
descriptions and synonymies without careful verification and criticism. 

In the general sketch of the structure of the group I have only dealt 
with generalities ; the details of particular , genera, or it may be of species. 


will be found under their respective headings. My own experience is that 
in reading elaborate monographs the multiplicity of details tends to throw 
one off the main course of the argument. Details of minor importance are 
much better referred to their proper place instead of being included in one 
long dissertation on the structure of the group. I have therefore attempted 
to give in my introductory chapter such an account of the anatomy of the 
Oligochaeta as may be sufficient to satisfy any person not interested in the 
minute details, but desirous of having the main facts stated in as few words 
as possible. It may be thought that I have erred in the brevity of this 
chapter : I claim, however, to have put before the reader a more complete 
account of the structure of the group than can be found in any treatise yet 
published, and I have done my best to avoid details not of special significance 
except as generic or specific characters. 

In the systematic part I have not treated each family in a precisely 
similar fashion. In the Perichaetidae, for example, the internal structure is 
dealt with after the definition of the family ; in the Geoscolicidae, anatomical 
details are reserved as generic characters. In the highly peculiar group 
Eudrilidae the method adopted is somewhat intermediate. In each case the 
plan followed is not, as might perhaps be suspected, the result of the dis- 
continuous preparation of this monograph, but has been deliberately selected 
as being, in my opinion, most appropriate to the family in question. 

For material used in the preparation of this work I am greatly indebted 
to Dr. Benham, Prof Glaus, Mr. W. T. Thiselton-Dyer, C.M.G., F.KS., 
Mr. Chamberlain, Mr. Gustav Eisen, Mr. Everett, Mr. F. Finn, Dr. Gregory, 
Prof Lovfen, Dr. Michaelsen, Prof Mlntosh, F.R.S., Mr. Alvan MiUson, 
Prof T. J. Parker, F.KS., Prof Poulton, F.R.S., Dr. D. Sharp, F.E.S., 
Mr. W. W. Smith, Mr. Sowerby, Prof Baldwin Spencer, Prof Yejdovsky, 
the Rev. H. W. Woodward, the St. Petersburg Museum, and the Zoological 
Society of London. 


London, February, 1895. 






HisTOBiCAii Note . . . . . 

The Body-wall and Exteenal Chaeactees 
The Neetotjs System . 


Alimentaey Canal 
Vasculae System 
Eespieatoet Oegans . 
Eepeoductive System . 
Gbogeaphical Disteibution 









The Classification op the Oligochaeta 155 

Phylogenetic Aeeangement oe the Oligochaeta . : . . . . . 162 

Desceiptions op Geneea and Species 174 

Geoup Apeaneuea 176 

Family Aeolosomatidae 176 

Geoup Miceodeili 187 

Family Pheeoeyctidae 187 

Family Moniligasteidae 192 




I Family Lumbeiculidab 
Family Tubificidae 
Family Naidomoepha 
Family Enchtteaeidab 
Geoup Mbgadeili 


/ Family Peeichaetidab 
Family Ceyptodbilidab 
Family Aoanthodeilibab 
Family Eudeilidae 
Family Geoscolicidab . 
Family Lumbbicidae 












Fig. I. Perichaeta indica. 
Dissected, sp, spermathecae ; g, gizzard ; sp s, sperm-sac ; dv, dorsal vessel ; wph, nephridia ; 
sp gl, spermiducal gland ; caec, caeca. 

Fig. 2. Perichaeta everetti. 
Ventral view of anterior segments, sp, spermathecal pores ; ? oviducal pore ; (J male pores. 

Fig. 3. Polytoreutus magUensis. 
Ventral view of anterior segments. (J male pore ; sp p, pore of spermathecal sac. 

Fig. 4. Octochaettis muUiporus. 
Dissected slightly from the side, gl, peptonephridium ; g, gizzard ; nph, nephridia ; dv, dorsal 
vessel ; sp gl, spermiducal glands ; n, nerve-cord ; v v, ventral blood-vessel. 

F'g- 5- Perichaeta posihuma. 
Ventral view of worm ; ? oviducal pore ; (J male pores ; p, genital papillae. 

Fig. 6. Libyodrilus violaceus. 
sp p, spermathecal pore ; ^ male pore. 

Fig. 7. Eudrilus eugeniae. 
$ female pore : (J male pore. 


Generative organs of various types of Oligochaeta diagrammatically represented. The male 
organs are coloured pink, the female blue. To ail the figures the following lettering applies : 
T, testes ; F, sperm-duct funnels ; sp s, sperm-sacs ; o, ovary ; od, oviduct ; e s, e^-sacs ; os, sper- 
mathecal sac ; sp, spermathecae ; sp gl, spermiducal gland. 


Fig. I. Libyodrilus. 
Dissection, o, orifice of spermathecal sac ; sps, sperm-sacs ; sp gl, spermiducal glands ; 
vv, ventral blood-vessel ; dv, dorsal blood-vessel ; n, nephridia. 

Fig. 2. Siphonogaster miUsoni. 

Tt, penial (?) processes. 

The remaining figures bear a legend. In all of them the following is the significance of the 
letters : g, gizzard ; ea, calciferous glands ; sp, spermathecae ; sp s, sperm-sacs ; h, ' heart ' ; sp gl, 
spermiducal glands ; vd, vas deferens ; od, oviduct. 

The clitellar segments are numbered ; the dorsal vessel with its branches are coloured red. 



Fig. r. StuMmannia. 
Longitudinal section through anterior segments, hr, brain ; n, ventral nerve-cord ; g, gizzard ; 
ca, calciferous glands ; im, ventral blood-vessel ; dv, dorsal blood-vessel ; s, masses of perivisceral 

Fig. 2. Octoekaetus multiporus. 
Transverse section through oesophageal region of young worm, dv, dorsal vessel ; vo, ventral 
vessel ; s, septum ; ca, calciferous gland ; n, nerve-cord. 

Fig. 3. Eudriloides cotterilli. 
Transverse section, sps, spermathecal sac with appended glands {gl). dv, dorsal vessel ; 
CO, calciferous glands ; i, oesophagus. 

Fig. 4. StuMmannia. 
Section through penial process, gl, glandular epithelium of penis ; m, septum dividing its 
cavity from general coelom ; vd, vas deferens ; n, nerve-cord ; dv, dorsal vessel ; I, intestine ; 
el, clitellar epithelium ; sps, sperm-sacs ; p, muscular sac opening at end of penis. 


p. 183. To synonyms oi Aeolosoma he.mprichn add Aeolonais hemprichii and Aeolonais decorum of 

p. 184. To synonyms of Aelosoma quaternarium add Aeolonais quatemariwm of Geevais. 
p. 214. To synonyms of Lumbriculus variegatus add Tuhifex yentilinus, Duoiis (fide Vaillant). 
p. 229, 1. 14. For Telinatodrilus read Telmatodrilini. 
p. 251, 1. 24. For bogdunovii read bogdemovii. 
p. 266, 1. 13. For Sem,itubifex ater read Hemitubifex benedii. 
p. 275. To list of genera of Naidomorpha add Ripistes. 
p. 285, 1. 4. For O.josinae read N.josinae. 
p. 288, 1. 12. For P. elinguis read N. elinguis. 
p. 313, 1. 15. For vermiculus read vermicularis. 
p. 332, footnote. For M. fusca read M. semifusca. 
p. 348, 1. 15. For E. hegemon read F. he.gemon. 

p. 394. A few species (lo) have been accidentally omitted from the list. 
p. 481. To synonyms of Pletcherodrilus unicus add Cryptodrilus pelewensis, Michaelsen. 
p. 498, 1. 13. For C. gravidis read C. grandis. 
p. 530, 1. 18. For A. commwnis read D. communis. 
p. 552, last line. For A. muUiporus read 0. multiporus. 
p. 608, 1. 23. For P. elongatus read P.finni. 
p. 661, 1. 29. For V. papillata read U . papillifer. 

Note.— Formal descriptions of BrachydrUus, Megascolex hallii, and of Perichaeta dubia, have 
unfortunately slipped out. But the principal facts in the anatomy of these species will be found 
scattered through the volume, and can be referred to from the Index. 


;dda.rds Monoqi^apli on OlicfocTiffita, 

University 'Preas 0.x 
E.Wilsondel Cambi 














.. od. 

-.'Sp V 













Beddard's Monograph on Olig-ochasta. 




















(S5> <E^- 

J{yJ^er^odri Ztcs 


iji jar 

sp.s XM 









sp xn 



U) 0.1 


w 1 at 


S^\ M 

1 \ 



a 'Jft^ /\ ^^[JjK, 













University Press Oj 



xy , 


^/> JX 



T:-1 // ^ 






















-Sp M 





p xm 

~° Mr 

-od ^ 


■SS xvm 

Beddard's Monoqraph on Oligochaeta. 


..sps ^ 



'"""""^ )^ 

— ^.^.- 

5JJ? 1 ^ . 

'•^l '^'^-i 


^ / / \. r^ 


fii_ Lit 



•fiJiiy- If i^^ 

c^ ^^^-- 












3gL W J^^ 


Tut if ex 






-■'Sp X 





■ ,/ 

) \- - " 





[[ m 

w w -- 

-- ■ Tl 




Universiby Press Oxft 

Umversitv Press Oxford 


o-Tcl's Monograph on (JlKjochasta.. 

UuiVb7-';Uy pvfss Oxford 


Class OLIGOCHAETA, Gkube. 

Def. Segmented ' worms ' invariably hermaphrodite ; all the segments of the body — 
except the first and occasionally an additional and varying number at anterior end — 
setigerous, the setae usually /-shaped but showing variation in form, never borne upon 
parapodia. Excretory organs paired tubes metameric, or numerous in each segment 
and dysmetameric. Branchial organs, rarely present. Gonads limited in number 
(normally never more than two pairs of ovaries and testes) ; sexual products nearly 
always matured in special sacs developed from the septa. Special ducts carry off the 
genital products. Development direct. Terrestrial or fresh water, rarely marine, in 



§ I. Historical Note. The Bibliography appended to the present work shows how very recent is our 
knowledge of the anatomy of the Oligochaeta. Of the exotic earthworms, which form so much the 
greater portion of the known species, there was absolutely no anatomical knowledge until the 
publication of Vaillant's memoir upon Perichaeta in 1868 (3) ; all that was known before this 
date from the contributions of F. S. Leuckakt, Rapp, Templeton, and Schmaeda relates to little 
beyond external characters. Peekiee's researches, particularly his first memoir (6), gave the 
first indication of the very great structural variation exhibited by the terrestrial Oligochaeta. Since 
that date (1871) our knowledge has rapidly accumulated, particularly during the last ten years, by 
the investigations of Hoest, Michaelsen, Rosa and others upon the continent, Spencee and 
Fletcher in Australia, Eisen in America, and Benham and myself in England. 

Before 1868 we were only acquainted with the structure of European Oligochaeta. The common 
earthworm (represented no doubt by several species) furnished material for a number of investigators ; 



Home, Leo, Moreen published their researches, but our present accurate knowledge of that 
animal may be said to date from the memoir of d'Udekem (4), who first discovered the ovaries. 
Subsequently to that Lankester (9), CLAPAEiiDE (1), Hoest (1), v. Mosjisovics (1, 2), 
Cebpontaine, and others have dealt with the structure of Lumbricus. The aquatic Oligochaeta of 
Europe were first investigated by 0. F. MtIller ; but d'Udekem's memoir upon Tubifex, and those 
of Claparede upon that and other forms, are the memoirs from which our modern knowledge dates. 
Since the publication of Clapaeede's two memoirs, the aquatic Oligochaeta have been principally 
studied by Lankester, Benham, and myself in this country; on the continent by Leydig, Dieffenbach, 
and others ; in Ailierica by Eisen. 

I. The Body- Wall and External Chaeactees. 

§ I. External fcn^m, and segmentation. The Oligochaeta are segmented worms of very 
variable size ; at the one extreme we have the minute species of Aeolosoma and certain 
Naids, I mm. or so in length ; at the other the gigantic Microchaeta rappi and Megascolides 
australis, which measure from four to six feet. 

In all the Oligochaeta, with the exception of Aeolosoma, where it is at any rate less 
marked, the external segmentation corresponding to the internal metamerism is very 
obvious. The grooves which separate the segments from each other are clearly defined. 
It often happens however, particularly among the larger earthworms, that there is 
a ' secondary ' annulation of the segments. The term ' secondary ' is used, because this 
division of the segment by transverse furrows into two or more parts appears to have no 
relation to other organs, whether internal or external ; at the same time this annulation 
of the segments and the number of annuli appears to be fairly constant for the species. 
The number of segments in the body of the Oligochaeta is very small (6-20) in Aeolosoma 
and some others among the lower forms. As many as 500 or 600 segments have been 
counted in some of the larger earthworms. There are at present no exact data as to the 
constancy of the number of segments among earthworms. ' In all probability the number 
is not absolutely fixed, but there appears to be a mean for each species round which 
there is a certain amount of variation. There is but little specialization among the 
segments of a worm's body ; it is an invariable rule that the first segment of the body, 
and it occasionally happens that a few of the following segments also, are devoid of 
setae ; this ' cephalisation ' is dealt with more at length on a subsequent page. As a rule 
the anterior segments of the body in earthworms are wider and marked by more 
numerous secondary annulations than those which lie behind the clitellum ; the clitellum 
itself is formed by a specialized set of segments. 

§ 2. Prostomium. In the majority of Oligochaeta a process of the first segment of the 
body overhangs the mouth on the dorsal side. In a few forms it is of considerable length, 


and no doubt plays an important part as a tactile organ; in the majority of Oligochaeta 
it is not large, and in a few it seems to be totally absent. This ' prostomium,' ' prae- 
stomium,' ' buccal lobe,' or ' upper lip,' as it has been variously termed, is often separated 
by a furrow from the first segment of the body -the buccal segment, of which it is 
a process. Sometimes the separation is not marked at all, or hardly marked by two 
lateral furrows which converge towards the middle line but do not meet. The genus 
Lumbricus (e. g.) and a few species of Acanthodrilus exhibit a very curious condition of the 
prostomium ; it is separated from the buccal segment by a cross furrow, but from the 
angles of this arise two longitudinal furrows which end at the posterior extremity of the 
buccal segment and as it were continue on the prostomium over this segment. The 
prostomium has in this case the appearance of being a process of the second segment of 
the body. Friend has recorded in Allolobophora cklorotica (9) a remarkable extension 
backwards of the prostomium which reaches as far as the end of the fourth segment. 
This, it should be stated, is not a characteristic of the species in question, but an 
occasional variation. 

The genus Phreorydes has a prostomium which is rather elongated and is divided 
into two halves by a cross furrow at about the middle of its length. In this pai-ticular 
it recalls the Capitellidae. 

An elongated prostomium — longer than in Phreorycten — chsiracterizes Nais lacustris. 
The names Nais proboscidea, ' die geziingelte Naide,' given to it by various writers, are 
expressive of that peculiarity. 

The long prostomium of Rhinodrilus gulielrwi and of Trichochaeta hesperidum has 
a peculiar structure which has not been described elsewhere. 

In the former species, I erroneously stated that the prostomium was altogether 
absent ; it is however present, but at times retracted, so that it is not at all conspicuous ; 
at such times it protrudes from the buccal orifice as a slight conical projection. A 
prostomium of apparently precisely the same character exists in Trichochaeta ; in that 
genus I have investigated its structure and relations by means of sections. It protrudes 
in a fashion similar to that of Rhinodrilus from the mouth, and in sections is seen 
to arise from a slight invagination of the dorsal wall of the buccal cavity just in front 
of the brain and at a point posterior to the orifices of the first pair of nephiidia. 
According to Vaillant the prostomium of Rhinodrilus paradoxus is similarly an 
'extroversion' of the buccal cavity. It is quite possible that in Trichochaeta, when the 
prostomium is completely everted (necessarily along with the most anterior section of 
the buccal cavity), it may have the appearance of being merely a process of the buccal 
segment, and may indeed prove to be morphologically such, and therefore comparable 
to the prostomium of other worms. In the meantime I am inclined to think that it 

B 2 


is an organ of a rather different nature, possibly comparable to the introvert of the 

In a few earthworms the prostomium seems to be completely absent. Perriee and 
I myself did not find any trace of the prostomium in Pontoscolex, but Hoest has 
subsequently affirmed its existence. There appears however to be no prostomium in 
Deodrilus. Is the absence of the prostomium to be regarded as an archaic character or 
as due to degeneration 1 The two alternative views Jiave been put forward. They 
depend of course upon the moi-phological nature of the prostomium, concerning which 
again there are two current views. 

Wilson in his lately published account of the development of the earthworm holds 
that the prostomium is a segment; the only difference which it shows froin other 
segments is that its cavity is unpaired. Wilson explains this by looking upon it 
as the terminal segment of a body which represents an elongated ring, — the view ih 
fact that the Annelid's body can be derived from a Coelenterate pulled out lengthwise. 

Vejdovsky (9), on the other hand, found that the prostomium was a comparatively 
late development — undoubtedly an outgrowth of the first segment. The mouth is 
originally completely terminal. 

§ 3. External Apertures. In many earthworms various pores upon certain of the 
segments are visible ; these are (i) the dorsal pores opening into the body cavity, (a) the 
external pores of the nephridia, (3) and lastly the apertures of the reproductive ducts. 
The varying position of these pores will be found mentioned under the various organs 
of which they are the outlets, either in the general part of this work or in the systematic 
section. A few general observations however may find a place here. The older 
naturalists used the varying position of these different pores for the purpose of specific 
and generic definition — usually with questionably useful results ; nevertheless it is 
possible for a person conversant with internal structure to make a pretty accurate 
guess by the external characters alone as to the genus, or even in a few cases the 
species, which he is examining. Most conspicuous of all the external orifices are as 
a rule the male pores ; it is only in the Geoscolicidae and in some species of Lum- 
bricus that they are not obvious, for the reason that the sperm ducts do not terminate in 
a spermiducal gland ; in Allolobophora they are conspicuous, although they do not 
terminate in such a gland, for the integument is swollen and glandular at their point 
of opening : in this case the position is always the fifteenth segment ; if the pores are 
obvious and upon the eighteenth segment the genus will be certain to be one of the Crypto- 
drilidae or Perichaetidae (which can be further differentiated of course by the setae). 
If the male pores are double and upon segments seventeen and nineteen the worm 
will be an Acanthodrilid. In all Eudrilidae the male pores are exceedingly conspicuous ; 



they are in this family nearly always unpaired and median. The pores of the 
spermathecae are not always so visible ; they are however for the most part in the 
Eudrilidae, where the pore is single and median. The accompanying plate (PL I) 
shows how great the external differences may be between earthworms, though in 
general shape they are so much alike. 

§ 4. Setae. In all Oligochaeta, with the single exception of Anachaeta (woodcut, 
fig. i), there are chitinous rods formed by epidermic cells and arranged in a certain 
definite plan, which are most usually 
termed ' setae ' ; these setae are partly 
buried in the thickness of the body-wall, 
and are prolonged into the body-cavity ; 
the free extremity projects for a varying 
distance beyond the epidermis ; the setae 
are the organs of locomotion of the animals ; 
they are furnished with special muscles 
which enable them to be retracted or pro- 
tracted and pulled forwards or backwards ; 
progression is effected by their movements. 

It used to be believed that the setae 
were structures of mesoblastic origin; but 
it is now, through the researches of 
Vejdovsky and others, so firmly established 
that they are epidermic that it seems to be 
of no pai*ticular use to go into the history 
of the older and erroneous view. The setae 
are implanted in sacs which are diverticula 
of the epidermis ; at the margin of these sacs the chitinous cuticle is invaginated and 
forms a lining for the seta sac down to a certain depth ; beyond this the seta is imbedded 
in a mass of cells, each one of which can produce a new seta to replace the original one. 
The invaginated part of the epidermis consists of a row of low cells lined by the cuticle 
already mentioned, and terminating in the mass of cells without cell boundaries in 
which the seta is firmly imbedded ; often, as Vejdovsky has figured for Rhynchelmis, 
the sac is reinforced by a second sac lying near to it, also multicellular and containing 
a seta in course of development. The setae first appear as small cones of chitinous 
substance ; the apex of the seta, its free extremity, is first developed ; it then gradually 
grows in length. A single sac (which it must be remembered is multicellular) often 
contains, as in the Tubificidae, a large number of setae. The invaginated epidermis does 

(After Michaelsen.) 

t. Grland cells = dorsal setae of other forms. 2. Clitellar 
epidermis. 3. Lateral line. 4. Openings of sperm-ducts. 
5. Nerve-cord. 6. Chloragogen-cells. 7. Epithelium of 
gut. 8. Ventral blood-vessel. 9. Blood-sinus of alimentary 
canaL 10. Longitudinal muscles. 


not always appear to exist ; thus in Tubifex (Nasse) and Limnodrilus (Vejdovsky) no 
Buch hollow sac is figured ; the setae and the solid mass of cells in which they are 
imbedded reaching right up to the epidermis : in these cases however there seems to be 
an iQvagination of the chitinous layer ; hence it is possible that a tube of epidermis is 
also invaginated. Ceefontaine (1) has figured two sections out of a series through a seta 
sac parallel to the surface of the body ; near to the external surface the seta is seen to 
be surrounded by a tube whose walls are composed of about fifteen cells ; this region is 
the tubular invagination of the epidermis ; further down the number is reduced to four 
and then to two which are flattened and concave, closely embracing the setae, the 
special lining of chitin present in the more superficial part of the tube being here 

In Anachaeta, as has been already mentioned, there are no setae at all; but the 
recent existence of these structures appears to be shown by the presence of large 
sacs depending from the epidermis on the dorsal side of the body, not the ventral, or 
(Anachaeta eiseni) on both: these sacs are apparently the equivalents of the seta 
sacs, and consist of a large cell of a glandular appearance with one or two nuclei 
near to its free extremity. 

The form of the setae in the Oligochaeta is varied ; but they may be grouped into 
two divisions, (i) long slender setae gradually diminishing in diameter towards the 
pointed extremity, and (2) shorter setae of a curved form, something like an elongated 
S with a thickening at about the middle; the shape of these setae has been aptly 
compared to that of the mathematical sign /. The capilliform setae only occur in the 
aquatic Oligochaeta, and not in all of them. 

The sigmoid setae vary very much in the details of their shape. The simplest form 
is that which characterizes the vast majority of earthworms, the Phreoryctidae and many 
Lumbriculidae. These setae end in a pointed extremity. In Onychochaeta the extremity 
(fig. 2 a. 1 ) may be markedly hooked. In Pontoscolex among earthworms, in certain 
Lumbriculidae, in all Tubificidae and Naidomorpha the sigmoid seta is cleft at the free 
extremity (woodcut, fig. z b. 4). A further complication is seen in Tubifex and other 
Tubificidae, where there are a few subsidiary prongs arising between the two main 
prongs into which the extremity of the seta is cleft (woodcut, fig. a &. 6) ; this later 
form of seta may be termed ' pectinate,' the simply cleft seta * uncinate.' A remark- 
able variation characterizes Heterochaeta, and to a less extent Psammoryctes and 
Spirosperma (woodcut, fig. 2 b. 3), where the widely divergent prongs are united by 
a ribbed membrane. This form of seta has been termed * palmate.' Another form of 
the uncinate seta characterizes certain Naidomorpha. The seta (woodcut, fig. a b. 7) is 
straight instead of being curved, and has a cleft extremity. This form is termed 


' hastate.' The Enchytraeidae show another form of sigmoid seta which is nearly 
perfectly straight but has not a cleft extremity. 

Another modification of the sigmoid seta is especially characteristic of the family 
Geoscolicidae among earthworms, though not unknown elsewhere. The end of the 
seta, which protrudes from the body, is ornamented (woodcut, fig. a a. 2, 3) with 
transverse ridges [Pontoscolex) or with minute spinelets (Trichockaeta). 

The capilliform setae, (fig. 2, h. 2) which are exceptionally {Lophochaeta) covered with 
fine processes (woodcut, fig. 3 &. 5), compar- 
able perhaps to the ornamentation of the 
sigmoid setae in Trichockaeta, to some 
extent graduate into the sigmoid setae. 
The intermediate condition is seen in 
Phreodrilus, where the capilliform setae 
are shorter, stouter, and more curved than 
in other Tubificids. As Phreodrilus is in 
other respects a link between the Tubifi- 
cidae and Lumbriculidae, the fact gets 
an additional intei'est. 

It sometimes happens that a worm has 
setae of one kind only from head to tail. 
This state of afiairs however is the less 
usual. Among the aquatic forms certain 
species of Aeolosoma, and of the Tubifi- 
cidae (Limnodrilus), Enchytraeidae, and 
Lumbriculidae, are the only instances. 
Among earthworms the Moniligastridae, 
many Megascolicidae, and some Lumbri- 
cidae and Eudrilidae are examples. 

The commencement of a diversity 
in the form of the setae is seen in 
Phreoryctes, where some of the setae are 

^g. 2 a. I. OnycTiochaetavnndlei: -posterior seta. z. Ornamented 
longer than the others — the dorsal longer seta of Geosooleoid. 3. IVichochaeta he^eridum. 3 b. End of 
_ , 11' J.T. same more highly magnified. 

than the ventral, or mce versa, or the pos- ^.^ ^ ^ ^ ^^^^^ ^^^^ „f AmnmodHius georgianus. =. Spiw- 
terior longer than the anterior: in many «'«''™»-''«''<'^- i. spiromrma fenx. 4. iiyodHiuscoccmeus. 5. 

® Lophochaeta ignota. 6. Tuoifex rivulorum. 7. Nazs ehnguis. 

Perichaetidae the ventraJmost setae are s. Nais eUnguU. 9. BohemUia comata. 

longer than the others ; in some those upon a few of the anterior segments are longer 

than the rest. 

Fig. 2, h. 

(After Vejdovsky, Stole, Miohaelsen.) 


Another stage is seen in many Geoscolicidae and Lumbricidae, where the setae upon 
the clitellum are longer than the others and of a slightly different shape ; it occurs 
for example sometimes in the Geoscolicidae that the clitellar setae are ornamented, 
while those formed elsewhere are not. In the Eudrilidae the setae are apparently 
similar throughout the body except in the neighbourhood of the male pore. These 
penial setae {see under description of reproductive system) occur in many Oligochaeta. 

It is however among the aquatic genera that the greatest specialization of the setae 
occurs. Among earthworms a single segment has never setae differing in form though 
they may differ in size, excepting only in the case of the segments which bear the 
specially modified penial setae, which may be a little different among themselves 
and are as a rule very different from the ordinary locomotive setae on the same 
segment. But the rule among aquatic Oligochaeta is that the dorsal bundles of setae 
should contain setae different in form from those of the ventral bundle. In most 
Tubificidae the dorsal bundles are made up of capilliform setae, with which may or may 
not be mixed uncinate setae ; the ventral bundles are entirely composed of the latter. 
The details in the distribution of the setae, which are most useful for classificatory 
purposes, will be found in the systematic part of this work. 

The arrangement of the setae among the Oligochaeta is varied. The most prevalent 
plan is the existence of four groups of setae, which of course recall the parapodia of the 
Polychaeta. The principal ways in which the setae are disposed are indicated in the 
following table : — 

1. Setae numerous in each segment and forming a complete ring. — Ex. Perichaeta, 
Pleionogaster, and other Perichaetidae. 

2. Setae numerous but arranged in two lateral masses leaving dorsal and ventral 
gaps. — Ex. Megascolex. 

3. Setae twelve in number in each segment. — Ex. Deinodrilus. 

4. Setae eight in each segment placed at equidistant intervals or grouped into four 
pairs, or intermediate, four setae being paired and four distant. — Ex. Acanthodrilus, 

5. Setae eight in number in anterior segments, and thence increasing to thirty or 
forty. — Ex. certain Perichaetidae. 

6. Setae in four bundles with more than two setae in each bundle. — Ex. Tubificidae. 

7. One pair of setae only on each side of segments. — Ex. Anachaeta. 

8. Setae, eight in number in each segment, but irregularly arranged. — Ex. Diachaeta. 
It will be seen from the above table that there is an almost unbroken series of stages 

in the arrangement of the setae ; we can pass from one extreme to the other without any 
difficulty ; thus assuming for the moment that the original condition is exemplified by 


Perichaeta, a gradual reduction in the number of the setae anteriorly leads to certain 
species of Megascolex ; a reduction upon all the segments of the body simultaneously 
leads through Deinodrilus to Acanthudrilus, &c. ; on the other hand, a crowding 
together of the setae into four bundles coupled with a reduction in numbers, brings 
about the arrangement characteristic of the Tubificidae. 

It is however by no means certain that the stages have been developed in the order 
suggested above. There appear to be three possible lines of development; either the 
' perichaetous ' condition is the most primitive, or the two pairs of setae in each segment, 
or the bundles with several setae (more than two) in each bundle. We shall consider 
these various possibilities seriatim. If we regard the Oligochaeta as to be derived from 
the Polychaeta it is clear that the last of the three alternatives is the one which would 
then appear to be the most probable ; the considerable specialisation in the individual 
setae of a bundle is a further point of resemblance to the conditions characteristic of 
many Polychaeta. But there is not after all much detailed resemblance between the 
setae of the Tubificidae and those of the Polychaeta ; the fact that in both groups the 
setae of the dorsal bundles are very often different in form from those of the ventral 
bundles is perhaps more striking as an analogy than valuable for the purposes of 
a strict comparison ; if the differences were of the same kind it would be different. 
Still it might be urged that in the Tubificidae there is a resemblance to the Polychaeta 
toned down by simplification. Furthermore, we have the extinct Pronaidites ; this 
worm has been referred to the Oligochaeta (Kusta), but it is not by any means 
convincingly an Oligochaet. As the creature is of Carboniferous age, this argument 
would be strongly supported if it were proved beyond doubt to be referable to the 

Unfortunately, this is the only piece of palaeontological evidence bearing upon the 
matter under discussion, and at best it is not conclusive in any direction. Besides, 
looking at the matter in another light, there are now undoubted Polychaeta known 
which inhabit fresh water, for example ManayunJcia. The only families of Oligochaeta 
which show this resemblance (such as it is) to the Polychaeta in the arrangement of 
their setae are the Tubificidae, Naididae, and Aeolosomatidae ; and they are all, it 
will be observed, aquatic in habit. Now it seems quite reasonable to suppose that 
long and delicate setae would be out of place in a worm having to force its way 
through dense soil ; we need not therefore be surprised at not meeting with such 
setae among the terricolous forms ; but there does not appear to be any valid a priori 
reason against finding bundles of short and strong setae in the land Oligochaeta ; and 
yet there are only the Enchytraeids among the terrestrial Annelids with such an 
arrangement of setae, and they are largely aquatic. One is accordingly inclined to 



suppose that after all the bundles of setae have some relation to the aquatic life; 
it is quite possible that a number of them radiating out in a fanlike way serve 
as an efficient swimming organ, and hence their development in Oligoehaeta which 
occasionally at least ' swim.' If there were any other reasons for associating together 
the families of Oligoehaeta mentioned with the Polychaeta, this question of the simi- 
larity in the disposition of the setae would have to be reconsidered; but it cannot 
be said that the Tubificidae and the Naids are nearer allies of the Polychaeta than 
any other families of Oligoehaeta; hence I am inclined for the present to put down 
the likeness in the bundles of setae in the marine and in the fresh water Annelids 
to a similar need. 

ExsiG has dealt with this subject in his Monograph of the Capitellidae (p. 574, etc.). 
In that group there are foiins in which the distichous arrangement, obvious elsewhere, 
is nearly lost; a state of affairs very like that of Perichaeta thus results, or more 
like Megascolex, since there are dorsal and ventral gaps. There can be no doubt, 
in Eisig's opinion, that the later arrangement is secondary, it being restricted to the 
beginning of the abdomen, and being preceded as well as succeeded by the ordinary 
paired bundles. 

One argument in favour of deriving the fewer setae per segment of most earth- 
worms from the perichaetous condition does not appear to me to have been considered. 
That argument is based upon the fact that in most Perichaetidae the number of setae 
is less in the anterior than in the posterior segments of the body. In a number of 
species, which I formerly proposed to place in a separate genus Anisochaeta. there 
are eight setae in a varying number of the anterior segments, while posteriorly the 
number becomes much greater ; this is an extreme case of the point at issue ; but 
in all Perichaetidae there seems to be a smaller number of setae on the pre-clitellian 
segments than upon those which follow, or there is a progressive increase for a few 
segments at any rate. In Megascolex this is, so far as we know, invariably the case ; 
but in Perichaeta the number, after increasing up to a certain segment (for example 
the seventh in P. taprohanae), diminishes. Now in such a foi-m as Onychochaeta the 
absence of setae upon the first few segments of the body (seen also in other 
Geoscolecids) must surely be regarded as an instance of what has been called 
'cephalisation'; it is at any rate a modification paralleled by other organs of the 
body. It would seem therefore likely that the smaller number of setae in the 
anterior segments of the body in the Perichaetidae is due to a reduction from 
a primitively greater number; as we know that this number is sometimes reduced 
to eight we have the origin of the eight setae per segment of the majority of the 
terrestrial Oligoehaeta suggested. Such general arguments as can be deduced from 


the primitive position of the Perichaetidae among earthworms can also be used ; 
this matter however is discussed later. Spenoek has referred to the existence 
in Megascolex of what would seem to he a regenerating tail ; and it is noteworthy 
that here the setae are more numerous than anteriorly ; it suggests a recurrence to 
an earlier condition. Why the number should have been reduced to eight exactly 
is a more difficult matter ; it is a problem comparable to that involved in the attempted 
explanation of why earthworms should usually have two pairs of testes and certain 
species four pairs of spermathecae. There is no obvious advantage to be discerned 
in either fact. 

Typically the setae are repeated from segment to segment, with or without 
modification in form in different parts of the body or in different regions of the same 
segment. It sometimes happens, however, that the setae are partially or entirely missing 
upon certain segments of the body. These segments are either the first few segments 
of the body or certain of the genital segments. I have already mentioned the entire 
absence of dorsal setae in Chaetoguster and of all the setae in Anachaeta. 

In the segments which bear the male pores the ventral setae are very commonly 
absent; this is the case for example with Bero and with a good many earthworms 
(e. g. Ocnerodrilus, Megascolides orthostichon) ; in Psavwioryctes harhata the dorsal 
setae also of the segment (eleventh) which bears the male pores are absent. 

A more remarkable instance of a specialisation of this kind occurs in a large 
number of species of Fer^haeta. In those species, for example in the common 
P. ind-ica, the three segments of the clitellum are quite without Various inter- 
mediate conditions between the total absence of setae and their presence to the full 
number are seen in other species. There seems also to be a tendency in other 
earthworms for the clitellar setae to disappear, though sometimes, as has been already 
mentioned, the converse occurs, and they become replaced by a series of a different 

§ 5. Cephalisation. Lankester has applied this expression to the specialisation of 
the anterior segments of the body so frequently seen among the Oligochaeta. As already 
mentioned, all Oligochaeta show cephalisation as regards the fii'st segment of the body, 
which never possesses setae. There are a few earthworms in which more than the first 
segments of the body are without setae ; these worms chiefly belong to the family 
Geoscolicidae, and the number of segments which are thus without setae differ in 
different species. There are as many as twenty in Kynotus. The Geoscolicidae are 
not the only family which show this character; in the genus Deodrilus, belonging 
to the Cryptodrilidae, the first segments are similarly devoid of setae. Among the 
Naidomorpha, the cephalisation affects the dorsal bundles of setae only ; in Chaeto- 



gatter, however, there are several segments following the first setlgerous segment, 
which have no setae ; in the other Naids a variable number of segments, varying with 
the genus, are without dorsal setae ; for the details the reader is referred to the 
special description of the family. The only other examples among the lower Oligo- 
chaeta are afforded by Enchytraeus monochaetus and ffesperodriius albus. 

The formation of a head or cephalisation is not however confined to the absence, 
or reduction in number, of the setae. Other organs show analogous modifications. 
It is common, for instance, to find the intersegmental septa not clearly definable in 
the first two or three segments of the body ; their place is taken or their existence 
is concealed by masses of muscular fibres which pass from the buccal cavity and 
phai-ynx to the parietes. In all Oligochaeta a certain number of anterior segments 
of the body are without nephridia, or the nephridia if present are modified. There 
are various other organs which show peculiarities at the anterior end of the body 
and contribute to the formation of a 'head.' 

§ 6. Eijidermiis. The epidermis of all Oligochaeta consists of cells which are 

separated from the subjacent muscles by ' a condensation of the connective tissue of 

the latter layer, which presents in parts the appearance of a veritable lamella' (Cee- 

fontaine). CLAPAEiDE, adopting Weismann's term — hypodermis, described this layer 

in Lumbricus as consisting of a nucleated meshwork in which no cell-outlines could 

be distinguished, enclosing spaces filled with a colourless substance; these bodies were 

regarded as of a glandular nature, but no nucleus was discovered. Leydig's earlier 

view (6), based upon a study of Phreoryctes as well as Lumh-icus, of the cellular 

nature of the entire epidermis was discarded, except for the prostomium, where 

CLAPAEtDE detected in osmic acid preparations its cellular character. 

This eiToneous view appears to have been first rectified for Lumbricus by Perkier 

(9) in a preliminary dissertation upon the structure of that worm, which precedes 

his account of the anatomy of Urochaeta. A few years later the cellular nature 

of the epidermis was slated by Lankester (12) ; he speaks of it as consisting of 

•varied forms of goblet cells and, excessively delicate, elongate, interstitial, or 

"packing" cells, instead of the altogether improbable syncytium of Claparede.' 

These results are mentioned as being confirmatory of those of HoRST and v. 

MosJisovics, which had been published previously. Since that date all observers 

have agreed in regardiDg the epidermis as distinctly cellular and built up of the 

two kinds of cells referred to in the quotation from Lankester's memoir. 

Cerfontaine's memoir, published in 1890 (1), contains the most detailed account 
that has yet appeared of the epidermis of Libtnbvicus. 

The cells are disposed in two rows, the cells of the innermost row beino- small. 


These latter are pointed at one extremity, the fine process running up between 
the other cells ; these cells are termed by Ceefoxtaine ' cellules de remplacement,' 
inasmuch as they appear to replace the other cells of the outer layer. The outer 
layer itself is composed of two sorts of cells, of large oval glandular cells and of the 
interstitial cells ; the former are large oval cells filled with numerous granules ; the 
basal portion of the cell, in which lies the nucleus, is more protoplasmic and is narrower 
than the swollen upper part ; the cell ends above in a fine prolongation which opens 
through a pore on the cuticle. The interstitial cells are of much less diameter; they 
often appear in cross sections to have excavated surfaces ; and in fact they are moulded 
to the form of the glandular cells which lie amongst them. The distal extremity 
of these cells is frequently prolonged or rather frayed out into a number of fine 
processes which seem to be in connexion with nerves. The cells of the epidermis are 
imbedded in a homogeneous connecting substance, which is more evident where the cells 
diverge from each other at the point of contact with the cuticle. This connecting sub- 
stance leaves a polygonal mesh work upon the cuticle, which is visible when the latter 
is viewed from the under surface. The cuticle seems undoubtedly to be a formation 
of the packing cells of the epidermis ; the pores upon its surface are the outlets of 
the gland-cells, and their existence appears to be simply due to the fact that the 
gland-cells do not secrete a cuticle like the other cells, their secretory activity being 
taken up in the formation of the granules with which they are laden ; hence at the 
points where they abut upon the cuticle there are gaps — the pores in question. The 
cuticle is sometimes distinctly to be seen as a double laj'er; viewed superficially it 
is seen to be traversed by two sets of striae, crossing each other at a right angle ; 
these striae correspond to two sets of fibrils, which are not upon the same plane, 
and therefore give rise to the double layer already referred to. 

The striation is thus not merely an optical efiect, but is due to the composition of 
the cuticle out of numerous fine strands of cuticular substance. 

Clitellar epidemiis. The epidermis on the clitellum is modified in structure. In 
Lumbricus it has been described, especially by Clapabede, Horst, Mosjisovics, 
Cerfontaine, and quite recently by Cole. 

The clitellum in the most fullj' developed portion, that is to say on the back 
of the animal, is made up of several layers of glandular cells. There are first of all 
elongated gland-cells of cylindrical form, which ai-e filled with granules, save at the 
inner extremity which is protoplasmic, and contains the nucleus. These gland-cells 
do not extend for a great distance into the thickness of the clitellar epidermis ; below 
them are the second sort of cells which are massed together into columns ; there 
are several layers of these cells, and the axes of the columns which they form are 


occupied by their processes which extend up to the cuticle. The columns are separated 
from each other by septa of a kind of connective tissue, which is fibriUated, and 
ends below upon the circular muscular layer, but not in actual connection with 
its fibres. In the fully developed part of the clitellum there are no non-glandular 
epidermic cells ; these exist, however, on the ventral surface of the body in the 
clitellar region. 

The epidermis, both clitellar and non-clitellar, of other earthworms seems to be 
like that of Lumbricus, which has just been described.' In Microchaeta, however, 
Benham has stated that the gland-cells seem to be more numerous than in Lumbricus ; 
such also appeared to me to be the case with Pontoscolex ; the clitellum of Microchaeta 
has been described and figured by Benham (2). In addition to the layers which have 
been described above for Lumbricus, and which occur in Microchaeta, there is an outer 
layer exactly like the epidermis of the general body-surface ; it has not only the 
interstitial cells which, as already mentioned, occur in Lumbncus in the less deeply 
modified clitellar regions, but also the oval gland-cells ; Benham does not state from 
what region of the clitellum his sections were taken. In the lower Oligochaeta there 
can generally be distinguished the two kinds of cells in the epidermis. In Enchytraeus 
mobii among the Enchytraeidae the gland-cells are not, comparatively speaking, very 
numerous ; in Phreodrilus and Pelodrilus I (21) recognised the same two varieties 
of cells ; in Aeolosoma the gland-cells contain a coloured substance which is frequently 
very characteristic of the species ; for example in Ae. quaternarium the oily substance 
is red ; in Ae. headleyi a bright green ; in Anachaeta bohemica there are three varieties 
of the glandular cells, colourless cells which are either elliptical or moi'e spherical 
in form, and larger cells with green contents, the substance being chlorophyll. 

The clitellum in all the lower Oligochaeta shows no great differences from the 
ordinary epidermis ; the same two kinds of cells are present, but the gland-cells are 
commonly more abundant and larger. The clitellum of the lower Oligochaeta has been 
described especially by Vejdovsky (24), but also by others — for instance Michablsen 
(Anachaeta), Stolc (Aeolosoma), myself (Moniligaster), etc. The morphological 
difference in the clitellum of the lower and the higher Oligochaeta is duly insisted 
upon later. In the former, as will have been observed, it is less modified and is only 
a single layer of cells thick. 

Sense organs of ejyiclermis. The cells of the epidermis are in parts modified to form 
sense organs. It is possible that the fine processes which arise from the epidermic 
cells in Aeolosovia and in many Naids have a sensitive function ; but there is apparently 
no particular modification of the epidermis to be traced in connexion with them, 
though in Bohemilla Vejdovsky figures ganglionic enlargements upon twigs which 


arise from the cerebral ganglia and end in these processes ; however, it is common 
among the Oligochaeta for the epidermis of the prostomium to be modified to the extent 
that it has no gland-ceUs and is formed of deeper cells than is the epidermis elsewhere ; 
in the genus AeolosoTtia and in the perhaps closely allied Ctenodrilus the under 
side of the prostomium is ciliated and there are a pair of lateral ciliated pits, 
possibly of a sensory nature. In the Naidomorpha we meet with the definite sense 
organs possibly of a tactile nature ; in the genus Slavina there are a series of these 
' Sinneshiigel,' as Vejdovsky has called them, on each segment, the actual arrangement 
differing with the species ; these elevations consist of specially elongated ceUs which 
tenninate in fine processes projecting beyond the cuticle. In Lumbriculus the same 
author has figured ' Becherfdrmige Organe/ which appear to be very similar ; they are 
hemispherical elevations of the skin due to the elongation of groups of epidermic cells, 
which terminate in the same way in fine processes. In BhyTtchelmis there are developed, 
at the breeding season, continuous zones of sense-cells arranged in groups ; the cells 
have the same elongated form that the sense-cells generally show, and appear also 
to possess the fine processes already mentioned ; Vejdovsky believed that he could 
trace nerves into connexion with these cells. 

Vejdovsky and Ceefontaine have described and figured groups of what appear 
to be sense- cells in Lwnihricus ; these consist of long cells which are so arranged 
as to project slightly from the general body-surface ; they are furnished at the extremity 
with fine processes and occur chiefly on the anterior part of the body, often in particular 
proximity to the setae. Among the exotic earthworms but little is known of sense- 
organs ; MiCHAELSEN has figured and described in Acanthodrilus georgianus a p%ir 
of papillae on the tenth segment which would seem to be sense-organs rather than 
glandular modifications of the epidermis ; they consist of the same kind of elongated 
cell that is usually associated with sense-organs ; and fine strands, apparently of 
a nervous nature, could be traced into them ; it is very possible that similar papillae, 
many of which, indeed most, have not been subjected to microscopic examination in 
other Acanthodrilides and other earthworms, are sensory in function. 

Eyes are present in a few Naids. They appear to consist of a lens-like body 
embedded in a cup of. pigment ; a strong nerve from the brain supplies each eye. 

In the genus Pontoscolex and also in the nearly allied Onychochaeta there are 
certain peculiar epidermal structures which may be of a sensory nature; these were 
first described by Peerier and have since been described and illustrated by Horst 
and by myself in the same genus. They consist of a large spherical deeply staining 
cell imbedded in a large sac which lies at the base of the epidermis ; the cell has 
a conspicuous nucleus, as is shown in the figures of both Perrier and HoRST. The 





^ (After Horst.) 

The cell is in the middle of the 

sac in which the cell in question lies seems to be connected with the surface 
by a canal ; such a canal is figured by Peebiee, but Hoest does not indicate 

it distinctly. Vejdoysky compared these cells with the large 
cells in Anachaeta, which represent the missing setae ; 
I supported that contention because it appeared to fit in 
with my view that the ' Perichaetous ' condition is the more 
primitive, and that these cells were the remains of -a Peri- 
chaetous condition in Pontoseolex ; 1 am not now convinced 
of the justice of this view — which is not shared by Hoest ; 
it seems probable, as he suggests, that the cells are of 
a sensory nature ; as to the special sense of which they are 
the organs that is a matter impossible at present to decide ; 
it is just conceivable that they may be rudimentary eyes ; the central cell appears 
to be highly refractive and a ray of light would pass through it to the tissue below ; 
but the entire absence of any pigment layer in connexion with the organs seems to 
negative the likehhood of that view. In the meantime, the accompanying woodcut 
expresses the general form of these organs which may be regarded as sense-organs 
of some kind. I did, but do not, compare them to the 'Pacinian bodies' of Eudrilus, etc. 
A peculiar form of sense-organ is to be met with in many Eudrilids. These 
organs were first described by myself in Eudinlus (62) and compared, erroneously 
as I now think, with the epidermal bodies of Pontoseolex that have just been described. 
Since then they have been studied by Horst in Eudrilus and by myself in several 

other genera, such as Hyperiodrilus 
and Heliodrilus. They are somewhat 
oval bodies, and lie just below the 
epidermis, though above the circular 
muscles ; the cells of the epidermis 
which cover them are shorter than 
the others and not glandular. These 
bodies consist (see fig. 4) of a deeply 
staining granular nucleated cylindri- 
cal core, round which are apparently 
wrapped a series of membranes, also 
nucleated, though with smaller nuclei than the core ; the whole structure reminds 
one most forcibly of a Pacinian corpuscle. In more than one case I fancied that 
I could detect a nerve fibre passing from the core, and I have represented one in the 
accompanying woodcut. There is nothing exactly comparable to these organs in any 


1. Nerre supplying it. 


other family of Oligochaeta ; but it does not seem to be too much to assume that 
they are really sense-organs, though their function must remain a matter of doubt. 

§ 7. Muscular Layers. In nearly all Oligochaeta the muscular layers of the 
parietes are arranged in two layers. There is an outer circular and an inner longi- 
tudinal. It is a little difficult in so delicate a form as Aeolosoma to recognize that 
both these layers are present. The only exception to this rule seems to be the 
Enchytraeid genus Fridericia, where there are two separate layers of longitudinally 
running fibres (as well as the circular layer). 

In Lumbricus (see especially Cerfontaink), the individual fibres of the circular 
coat form a layer of some thickness, variable, however, and particularly thin upon 
the intersegmental regions. The fibres are imbedded in a granular nucleated substance, 
and have a more or less strongly-marked columnar arrangement. The individual 
fibres are long and pointed at both extremities — a statement which applies to the 
muscular fibres of any part of the body — and longitudinally striate. This striation 
is apparently due to the presence of moniliform fibrils which makes up the substance 
of the fibre, and which sometimes give it a transversely striate appearance ; these 
fibrils are imbedded in a clear interfibrillar fluid, which is often particularly plain in 
the axis of the fibre, giving it a hollow appearance. 

The 'leech-like character' of the muscular fibres of the earthworm was first 
pointed out by Ratzel, who, however, believed that only certain parts of the body- 
wall had muscles of this kind, whereas Cekfontaine showed that they are general. 

The granular stroma in which the fibres are imbedded shows no cell outlines ; 
it contains clear spaces which are probably the 'lymphatic' spaces referred to on p. 30. 
The longitudinal muscles of the earthworm form a layer of much greater diameter 
than the circular layer. In many species, as was first pointed out by CLAPAEiiDE 
(1), a peculiar bipinnate arrangement of the fibres exists. The individual fibres 
.appear to be attached on either side of a central rhachis formed by a septum 
of connective tissue. This peculiar appearance does not exist in all species of 
Lumbricidae, but it does occur in a few earthworms not belonging to that family. 
Ude (3) explained the appearance not as due to a series of connective tissue 
septa with fibres regularly given ofi" on both sides, but as due to a series of 
compartments to the walls of which the fibres are attached all round. Cekfontaine 
proved, by histological methods more refined than those open to Claparede, that 
there is really no difference from the circular layer ; similar fibres are imbedded in 
an identical granular stroma ; the regularity of their arrangement, however, producing 
the appearance, wrongly interpreted by CLAPAEtDE, is more marked than in the 
circular muscles, where, nevertheless, it exists. Here and there, and more particularly 



at the insertion of the septa, radial fibres pass along the ground substance and 
between the longitudinal fibres ; these often give rise to the appearance of regular septa. 

The aquatic Oligochaeta are, with the exception of Phreoryctes and Pelodrilus, 
characterized by the possession of a longitudinal muscular layer consisting of flat 
flakes or lamellae imbedded in a granular substance. These fibres, or rather plates, 
show no axial core such as occurs in the Lumbricidae and earthworms generally. 
In Fridericia, as already mentioned, there is in addition to the lamellae and super- 
ficial to them, a single row of fibres with an axial core. Hesse has lately shown 
that in Fndericia aU these muscles (circular as well as longitudinal) are constructed 
upon the Nematoid, not Hirudinean, plan. Ratzel had previously stated this of 
the fibres of the inner longitudinal layer. 

The development of the muscles in Rhynchelmis has been worked out by Vejdovskt. 
The mesoblastic cells which are to form the longitudinal muscles become spindle-shaped, 
and the muscular fibre appears within them ; the original protoplasm becomes entirely 
used up or nearly so in the formation of the fibre, while the nucleus atrophies. 

In the Lumbricidae (see Vejdovsky 9) there are two or three layers of cells which 
are converted into the longitudinal muscular layer. The muscular fibres appear first in 
the deepest cells, i. e. those nearest to the circular muscles ; a considerable number of 
fibres appear in one cell. In the second and third rows of cells the muscular fibres are 
only developed at the sides ; hence the pinnate arrangement in the adult. The cell 
boundaries are finally lost and the cell-substance remaining over after the formation 
of the fibres becomes the granular stroma lying between the fibres. The early stage 
in the development of the longitudinal muscles of Lunibricus therefore represents the 
condition which occurs in Rhynchelmis. 

The circular muscular layer has been generally put down as also a product of the 
mesoblast. It seems, however, fairly clear from the observation both of Bebgh (3) and 
Vejdovsky that epiblastic cells alone are concerned in its formation. On the other hand, . 
in the regenerating tail of Lumbriculus (see Randolph 4) it is stated that the circular 
muscles like the longitudinal are a product of the mesoblast. 

II. The Nervous System. 

The nervous system of the Oligochaeta is formed upon the same plan as that of the 
higher Chaetopoda ; the cerebral ganglia communicate with a ventral ganglionated chain 
by a circum oesophageal commissure— the entire cord lying in the body-cavity. 

In only one case is the primitive connexion of the cerebral nervous system with the 
epidermis retained. This occurs in Aeolosoma ; in all the species of that genus which 
have been microscopically examined, the cerebral ganglia, though projecting into 


the body-cavity, are in connexion with the epidermis. In this worm — and it is 
quite unique among the Oligochaeta — ^the entire central nervous system is confined to 
the cerebral ganglia ; the ventral nerve cord appears to be entirely absent, except in 
Aeolosoma tenebrarum, where it exists in a rudimentary form in the shape of a few 
scattered cells. 

In all other Oligochaeta there are not only the cerebral ganglia connected by the 
circumoesphageal commissures with the ventral chain, but there is in addition a system 
of small ganglia arising from the cerebral ganglia, and concerned with the nerve-supply 
of the anterior section of the alimentary canal ; in some forms also there is a ' lateral line 
system.' These various parts of the central nervous system may now be considered 
in detail. 

Cerebral Ganglia. These lie further forward in the lower than in the higher 
Oligochaeta ; in the Tubificidae and the lower forms generally they are situated in the 
fii-st segment ; in the earthworms, almost without exception, they have moved back 
to the third segment ; but the development shows that the former is the primitive 
position ; they are moved back in the earthworms by the invagination of the stomo- 
daeum. In most, if not all, earthworms the cerebral ganglia mark the junction of the 
buccal cavity and the pharynx. Why it should be so is mysterious, but it is a fact that 
in the more highly organised Oligochaeta the brain is smaller and simpler than in the 
lower forms. In the latter — in the Tubificidae and Naidomorpha for example — the 
brain is not only relatively large, but it is provided with accessory lateral lobes, and 
is often prolonged posteriorly into posterior lobes. The form of the brain in these worms 
is often highly characteristic of the genus or species. Some references to the particular 
form of the brain will be found in the systematic part of this work. In Phreoryctes 
the brain has the simple bilobed character that is characteristic of the higher Oligochaeta 
to which this worm is related. The Lumbriculidae also have a simple brain. Among 
the Tubificidae there is often an impaired anterior median prolongation of the brain, 
which sometimes comes to be detached and remains in connexion with the brain by 
a nerve-cord. The formation of this anterior ganglion is highly suggestive of the buccal 
ganglia in the Mollusca. Special muscles are often attached to the brain in the lower 
Oligochaeta, which are apt to be confounded with nerves derived from it ^. The cerebral 
ganglia give off a number of peripherally running nerves. 

The simplest arrangement of these again is found in the higher Oligochaeta — a fact 

which must, as it appears to me, be taken into consideration in fixing their position with 

respect to the so-called 'lower' forms. Vfjdovsky only finds one pair in the genera 

Lumbricus and Criodrilus ; Rosa figures an identical disposition for Hormogaster, and 

' They are attached of course to the connective tissue-sheath. 

D 2 


Pekriek for Pontodrilus; on the other hand, according to the last named author, 
Pontoscdex has two pairs of nerves arising separately from the braia, a pair of course 
on each side. Spencer figures only one pair of nerves in Megascoiides australis, which 
arise, as usual, near to the circumoesophageal commissures. Thus it appears general 
among the higher Oligochaeta for there to be only a single pair of cerebral nerves, 
which may however, and usually do, divide at once into two. I shall point out later, in 
describing the peripheral nerves given oflF from the ventral cord, the fact that there are 
there three pairs in each segment; one would suppose that there would be a corre- 
spondence between the cerebral ganglia and any one of the ventral ganglia, considering 
that they are developed as one continuous whole. As a matter of fact this correspondence 
exists, but it is masked by the origin of the third pair of nerves from the commissure, 
and not from the cerebral ganglia themselves. There is nothing extraordinary in this, 
for in the ventral ganglia one of the three pairs of nerves also arises from the commis- 
sural part of the cord. There is thus really a correspondence between the ' cerebral ' and 
the ' spinal ' nerves. How for does this hold good in other groups 1 The most careful 
figures known to me of the nervous systems of the lower Oligochaeta are those of 
Vejdovsky and of Stoi-c of the Tubificidae. In Ilyodrilus the latter figures three pairs 
of cerebral nerves, and also three pairs in each segment arising from the ventral cord. 
In Spirospevma there are four pairs of nerves springing from the cerebral ganglia, and 
also four pairs from the nerve-cord in each segment ('i), one being commissural. In 
Monopylephorus, however, there does not, it must be admitted, appear to be the least 

The cerebral ganglia are united with the ventral chains by the commissures which 
embrace the gullet ; from the commissures arise the visceral ganglia. 

Visceral Nervous System. This appears to occur in most, if not in all, Oligochaeta. 
In the earthworms it has been figured and described in Pontodnlus, Pontoscolex, 
Megascolides, Hormogaster, &c. I have never found it to be wanting in any earthworm 
where I have looked for it. It consists of either a solid mass given off' from the com- 
missure or of a plexus having a similar origin; the plexus, however, is not entirely 
formed of nerve-fibres ; there are also ganglionic cells ; these branches of the visceral 
system ramify in the coats of the buccal cavity and pharynx. Among the lower 
Oligochaeta the same visceral nerves are met with ; thus Vejdovsky figures in Chae- 
togaster a pair of ganglia on either side of pharynx, which are connected with the brain. 
Other groups have the homologous ganglia, and reference must be made to Vejdovsky's 
work for further details. 

Nerves of lateral line. In many Oligochaeta — and its occurrence is probably 
general — there is a nerve on either side of the body arising from the brain or from 


the oesophageal commissure, which has been compared to the nerve of the lateral 
line in fishes; it appears to consist especially of nerve-cells. This system of nerves 
has been described by Vejdovsky as existing in the Enchytraeidae, Phreoryctidae, 
Naidomorpha, Tubificidae, and Lumbriculidae^- Connected with it is a system of fine 
fibres which supply the walls of the alimentary canal ; in Chaetogaster the oesophagus 
has a ling of nerve-cells round it, which are apparently referable to this system of 
intestinal nerves, which may fairly be compared with the Sympathetic system of 
Vertebrates. This nervous supply of the alimentary canal will not of course be 
confounded with the visceral nerves ah-eady described as arising from the circumoe- 
sophageal commissure. The lateral nerve itself or rather ganglionic chain originates 
from the epidermis, and remains in connexion with the same, the longitudinal muscles 
being separated where it occurs. 

Ventral Nerve-Chain. The commissures which arise from the brain and embrace 
the gut unite below it to form a ganglionated chain. This runs from end to end of the 
body ; in the extreme posterior region, where a regeneration of segments is going on, 
the ventral nerve-cord may be often seen to lie in the thickness of the epidermis ; 
otherwise it always lies in the body-cavity ; but Vejdovsky states that it is for the most 
part naked ; that is, not covered by a continuous coating of peritoneum ; scattered 
cells of the peritoneum are attached to it here and there. The nerve cord is usually 
enclosed in a muscular sheath, which may be, or is sometimes not, continuous right 
round it; but this nerve-sheath is derived from the same embryonic cells as those 
which form the cord itself. The cord may be divided into the ganglionic and the non- 
ganglionic or ' connective ' part ; the degree to which these are diSerentiated varies ; in 
Chaetogaster for instance, they are sharply marked off from each other in the figures given 
by Vejdovsky ; on the other hand it is the rule among earthworms for there to be only 
a slightly marked distinction ; Spencer even went so far as to practically deny, in 
the case of Megascolides, any difference in the whole length of the cord; but Vej- 
dovsky found constrictions separating the pairs of ganglia. In any case it seems certain 
that the ganglionic part of the cord fades gradually into the connective region, and 
the latter when present is of so short an extent that it is hardly recognisable. It is 
remarkable that the higher Oligochaeta should in this respect also show more primitive 
characters than the lower forms. 

The primitively double character of the ventral nerve-cord is partly retained for 
life in the genus Chaetogaster. In the Plates illustrating Vejdovsky's work upon 
the Oligochaeta there are several figures of the nerve-cord of this genus ; it will be 

' Hesse however states that the supposed ganglion cells are only the non-modified protoplasmic portion 
of the ' nematoid ' muscular fibres in Enchytraeidae and Naidomorpha. 


seen that in Chaetogaster cristallinus the connectives uniting the first, second, third 
ventral ganglia are most distinctly double, the interspaces left being wide. The same 
holds good with the species Chaetogaster- diaphanus ; but here the ganglia themselves 
are distinctly separated, which is not the case with the other species ; each ganglion 
or each half gauglion is connected with its fellow by a short commissure which give 
to the anterior part of the nerve-cord a ladder-like appearance. 

For the remarkable specialisation of the nerve-cord in certain segments of Phreoi-yctes 
and various Enchytraeidae cf. under the descriptions of those families. 

The ventral nerve-cord gives off branches in each segment. These branches arise 
in two different ways ; in the earthworms and in many of the aquatic genera they 
arise on either side of the nerve-cord and lie in the body-cavity for a greater or 
shorter distance until they plunge into the thickness of the body-wall. In -the 
Enchytraeidae, on the other hand, in Phreodrilus, and in many if not all Lumbriculidae, 
the two nerves are so closely applied to each other that they appear to be only a 
single nerve given off from the ventral side of the cord ; this apparently single nerve 
plunges at once into the thickness of the body-wall, and then runs to right and left. 
This peculiar state of affairs has perhaps led some observers to abstain from figuring 
or to deny the existence of the branches of the nerve-cord; on a dissection of such 
worms as show this origin of the branches from the ventral surface of the cord no 
nerves would be apparent. 

In the Enchytraeidae judging from the figures of Michaelsen the unpaired character 
of the ventral nerves is the most marked ; in RhynchelTnis, on the other hand, the 
two nerves, although lying close together, are quite distinct as two nerves. They are 
not present in the middle segment. In the Tubificidae and among earthworms the 
nerves arising from the ventral nerve-cord do not at once enter the body-wall, but 
pass to a point at some distance from their origin before they enter the body-wall ; 
the distance varies in different species ; as a rule in those species with paired setae 
the nerves enter the body-wall near to the ventralmost seta ; when this is further 
away from the median ventral line the nerves have a longer course through the 
body-cavity than when it is nearer to the median ventral line. 

The number of nerves given off in a segment varies considerably in different genera 
of Oligochaeta. In Lumbriculus and Rhynchelmis Vejeovsky could only find a 
single pair; three pairs is a much more usual number. This occurs for example in 
apparently aU earthworms ; it is figured for instance by Peerier in Pontodrilus and 

In these and other earthworms the three pairs are not given off at equal distances 
from each other ; two pairs are quite close at their origin ; FriedlIndek has pointed 


out, and I confirm him (for Perichaeta), that in Lumbricus at any rate the two pairs of 
nerves which arise close together have a relation to the ventral nerve-cord, similar to 
that which the dorsal and ventral roots of the spinal nerves have to the spinal cord 
of most vetebrates ; the one in fact is situated more dorsally than the other. This state 
of aiFairs is remarkably distinct in certain species of Perichaeta which I have examined 
from this point of view. In Pontodrilus — in the segment which contains the spermi- 
ducal glands — there is a ganglion on one of these nerves, and just at the ganglion a 
branch arises which goes to the other nerve ; this recalls the ganglion on the dorsal 
root of the spinal nerves of Vertebrates and the branch which immediately after 
unites this branch with the ventral root. Although there is this resemblance between 
all earthworms in the number and position of the branches of the ventral nerve cord 
in all earthworms, there is by no means a close correspondence between the various 
genera of aquatic Oligochaeta. Moreover accounts are apt to differ in many cases. 
For example, in • Tubifex Vejdovsky figures no less than five branches of the cord 
in each segment ; d'Udekem gives three as the number, while Nasse only found two. 

Stolc figures five branches in Monopylephorus and Lophochaeta, two being dissepi- 
mental branches. I found three in Phreodrilus. 

There is not much information as to the course of the branches after they have 
left the nerve-cord; Pereiee carefully dissected out these branches in Pontoscolex ; 
he found that one only of the three branched considerably in the thickness of 
the body-wall. I found in the Perichaetid genus Biporoehaeta a considerable 
branching of these trunks within the thickness of the body of the wall, resulting in 
fact in the formation of a nerve-plexus ; on the other hand, I found in the same worm, 
and I have noticed similar appearances in other worms, that the branches arising 
from the cord were continuous • right round the body, apparently joining dorsally. 

The histology of the nervous system is a large subject and one which can only 
be treated very briefly in the present work. It has been investigated by a large 
number of obsei-vers, including Vejdovsky, Retzius, Feiedlandee, etc. In transverse 
sections of the nerve-cord of Lumbricus, three dorsal tubes are very obvious ; these 
'have received various names, and very various functions and homologies have been 
assigned to them. There is now no longer any question that these tubes, the 
' Neurochord,' are of nervous nature ; for they have been traced into connexion with 
nerve-cells ; there are generally three of them ; but occasionally four are present the 
tubes dividing and reuniting. These tubes in Rhynchelmis are developed out of 
a row of large cells which were formerly (and erroneously) regarded as being of 
mesodermal origin. The neurochord of the adult is single in the anterior segments 
and in the brain and the oesophageal commissures. In the middle and hinder part 


of the body there ai-e three tubes ; these consist of a central nerve-fibre enclosed in 
a double sheath. The outer sheath has a few scattered nuclei in it and has a fibrous 
texture ; the inner sheath has the same texture but fewer nuclei. The central fibre 
or rather bundle of fibres is the direct prolongation of certain large nerve-cells. The 
rest of the nerve-cord is made up of fibres and cells; the latter are ventral and 
lateral in position. In the brain of course the conditions are reversed. The fibrous 
part of the nerve-cord consists of a more or less transparent ' cytoplasm,' the remains 
of a poi-tion of the embryonic cell-mass ; this in places forms transverse and longi- 
tudinal canals dividing up the meshwork of nerve-fibres into different regions. The 
fibrous mass is surrounded and. the canals of the cytoplasm also, by a delicate sheath 
the ' Glia sheath ' ; in this are a few nuclei ; it appears to be of the nature of 
connective tissue and is not connected with the nerve-fibres. The ganglion cells are 
grouped into a medial and two lateral masses; the cells in Ekynchelmis are for the 
most part unipolar; only seldom are multipolar cells met with; in Lumbricus 
Cerfontaine figures multipolar and unipolar cells also. 

The development of the nervous system has been studied by Kleinenbekg, 
KovALEVSKY, WiLSON, Bergh, Ve.tdovskt, etc. Wilson discovered, and the subse- 
quent observers confirmed him, that the cells which form the nervous system, like 
those which will form the nephridia, originate from a single cell on each side, placed 
near the posterior end of the body and termed a ' teloblast ' ; continuous with this 
teloblast and forming a row of cells produced out of it is the layer which will ultimately 
become the nervous system ; the teloblast is an epiblastic cell and there is therefore 
no doubt as to the epiblastic character of the central nervous system in the Oligochaeta. 
It appears to be entirely formed by the proliferation of these cells, and Wilson declares 
that the cerebral ganglia are formed continuously with the ventral chain ; it follows 
from the mode of origin just referred to that the ventral cord and the cerebral ganglia 
are a double formation, that the nervous system is bilaterally symmetrical ; it has been 
held that it is a single formation laid down in one band. Bergh, while confirming 
Wilson, made the interesting addition to his facts that in the embryo there is a series 
of branched cells evidently of a nervous nature, which lie between the two nervous rows ; 
Bergh thought, and Vejdovsky confirmed him, that this plexus of cells and fibres is to 
be traced to the ventral epiblast and has no relation to the neuroblasts already referred 
to. Whether these cells have any relation to the definitive nervous system seems to be 
at present a matter of some doubt. Vejdovsky looks upon this primitive plexus as 
a remnant of the nerve-ring of the Medusa. The histological difierentiation is so 
special a matter that I do not enter into it here. For details the reader is referred to 
Vejdovsky's work (9). 



The coelom in the Oligochaeta is invariably spacious and nearly invariably divided 
into compartments which correspond -with the external metamerism. The division is 
effected by means of the intersegmental septa which are only wanting in Aeolosoma ; 
these septa are not as a rule applied to the parietes in such a way as to exactly 
correspond with the gi'ooves on the exterior of the body that mark the segments ; 
hence the internal metamerism is not precisely as the external metamerism. The 
coelom is lined throughout by the peritoneal epithelium, which is reflected over all 
the organs that lie within it ; from its wall are developed the gonads ; it communicates 
with the exterior directly by means of the dorsal pores, and indirectly by means of 
the nephridia and the genital ducts. 

The coelom of the Oligochaeta is developed out of the paired mesoblastic masses 
which are formed early in the embryo ; each pair joins its fellow in the mid-dorsal 
and mid- ventral line ; but there is a nearly complete fusion above and below. No 
longitudinal septum remains to mark the division of each compartment of the coelom 
into right and left halves, and on the ventral side there is the mesentery supporting 
the ventral vessel only, which represents the ventral part of the line dividing the 
right and left halves. It has been stated that in Criodrilus the dorsal septum is 
persistent, but this appears to be an error. As a general rule, the coelom shows 
no signs of division into ditferent cavities, except of course by the intersegmental 
septa; but this rule has a few exceptions. In the first place, the sperm- sacs an^ 
egg-sacs are undoubtedly portions of the coelom enclosed by special walls, which 
are set apart for the maturation of the sperm and ova respectively ; these structures 
are dealt with under the description of the reproductive organs. In addition to these 
there are the remarkable sacs which envelop the gonads and the spermathecae in 
many Eudrilids, and which form so marked a character of that family. These 
sacs for the most part play the part of spermathecae, otherwise for the most part 
wanting in the family Eudrilidae, and they are dealt with under the description of the 

The lining membrane of the coelom varies greatly in its characters in different 
regions of the body; but it is nowhere ciliated, as is the case with other worms 
(e. g. the Archiannelida). The parietal layer is composed of flattened cells, the nuclei 
of which can as a rule be alone recognized in transverse sections ; this epithelium 
can be demonstrated by the silver method with great success (see Powee). 
Claparede too has figured the flattened coelomic epithelium upbn the nerve cord, 
though, as already mentioned, Vfjdovsky has stated the non-continuity of the coelomic 



cells upon the nervous system. The layer of peritoneum covering the alimentary 
canal is greatly modified, particularly upon the intestinal region. To this layer the 
name of ' Chloragogen-cells ' was applied by Claparede. The layer was at one time 
thought to be a digestive gland; and the fact that the cells end in a finish thread 
which is closely applied to the wall of the intestine favoured the supposition. It 
was pointed out however by Claparede that these cells are not so much connected 
with the intestine as with the blood-vessels upon its surface. According to Kukenthal 
their function is that of extracting waste substances from the blood and setting them 
free into the body-cavity whence they are removed by the nephridia. These chloragogen- 
cells contain greenish to blackish granules, the pigmentation being more marked in 
some species than in others. Upon the nephridia the coelomic epithelium often shows 
a difierent modification ; the nephridia are of course always enveloped in a layer of 
this tissue as are all the organs lying in the coelom (except the gonads, which are 
peritoneum) ; but it is frequently a thin and barely discernable layer ; this is not 
the case with the aquatic Oligochaeta ; but those worms which appear to have 
the densest peritoneal layer round the nephridia are the Eudrilidae. The cells 
a]"e often loaded with round granules of various sizes ; the presence of these 
gives the nephridia, when viewed with the naked eye, a very conspicuous and 
white appearance. The coelomic cells which line the spermathecal pouches in the 
Eudrilidae are also very large ; they are pear-shaped and are often apparently in 
a condition of rapid proliferation. It is quite otherwise with the cells lining the sperm- 
sacs and the ovisacs ; in these sacs the peritoneal layer is but little conspicuous. 
The peritoneum, if it be really so, which lines the large spermathecal sac of 
Polytorewtus, has quite the appearance of a columnar epithelium. There is a further 
peculiar modification of the coelomic epithelium enveloping the remarkable calciferous 
glands of many Eudrilidae for which see the description of those glands. 

The coelomic cavity of the Oligochaeta also contains free corpuscles. In the 
higher Oligochaeta these are apparently of two kinds ; there are smaU. amoeboid 
corpuscles and large spherical corpuscles loaded with granules ; in addition to the 
corpuscles there is also a certain amount of fluid which is coagulated by alcohol. The 
two kinds of corpuscles referred to are probably merely stages in growth ; when 
the cell becomes loaded with excretory (?) products it naturally loses its activity 
of movement and assumes the spherical form referred to. Very remarkable are the 
elliptical corpuscles of the Enchytraeidae ; these have a fiixed outline ; sometimes 
there are round as well as oval corpuscles in the same species. 

In the lower Oligochaeta the corpuscles are often extremely abundant; this is 
particularly the case with the Enchytraeidae and the Naidomorpha ; in the latter 


worms one would be disposed to put down the immense numbers of free corpuscles 
to the process of asexual reproduction with which their presence seems to be con- 
nected. But this will obviously not do for the Enchytraeidae, where there is of 
course no asexual reproduction^. Among earthworms there is generally not such 
a great abundance of corpuscles ; but in many Eudrilids there is — a circumstance 
which gives to these species in many cases a milky white appearance (e. g. Megachaeta 
alba). In the embryos of Octochaetus I noticed a very large quantity of corpuscles, 
the presence of which may be related to rapid growth and excretion. There are in 
most, if not all, Oligochaeta apparently the equivalents of phagocytes ; these were 
first noticed by Hoffmeisteb. Generally in the posterior region of the body are to 
be seen masses of brownish cells enclosing old and broken setae ; the latter are 
evidently in the process of removal by disintegration; possibly this goes on until 
they can be thrown out of the body by the dorsal pores or the nephridia. 

§ I. Perihaemal spaces. Besides these chambers, formed by a sub-division of the 
cqelom, there exist others which for the most part involve various portions of the vas- 
cular system. The first structure of the kind to be described occurs in the Acanthodrilid 
Deinodrilus ; in this worm the dorsal vessel is seen on a dissection to present an 
obscured appearance, which is due to the fact that it is enclosed in a sac which 
completely surrounds it and separates it from the general body-cavity. In this worm 
the dorsal vessel is completely double, and corresponding to this is a separation between 
the two halves of the ' pericardium.' In transverse sections, through the dorsal vessel, 
the sac in which each of the two trunks lies, is seen to consist of a very delicate 
muscular layer, which is covered externally and lined internally by a cellular coat ; 
the external covering is formed of few and dehcate cells ; on the other hand, the 
internal lining consists of large cells, which are here and there heaped up into piles. 
At intervals delicate strands of muscular fibres pass from the walls of the sac to 
the contained blood-vessel, where they pass between the large chloragogen-cells which 
cover the blood-vessel and become lost in. its muscular layer. This perihaemal space 
seems to commence a little way behind the head of the worm, but I have not fixed 
the actual point at which it commences, nor its connexion, if any, with the general 
coelomic space. 

Spenceb (1) subsequently recorded the presence in Megascolides of a similar sac 
enveloping the dorsal vessel ; in this case, however, there is a further compKcation. 
It possesses in fact a series of diverticula, one more dorsal, the other more ventral 
in position ; these- diverticula — the dorsal ones especially as figured by Spencer, 
are crammed with free and slightly attached cells ; the main tube enveloping the 

' Cf. however Lemoine (2), by whom simple division is stated to occur. 

E 2 


dorsal vessel does not extend throughout the body ; 'it is connected with the general 
coelomic cavity by a ventrally placed slit situated anteriorly, just where it narrows 
to pass through the intersegmental septum. 

The dorsal vessel is not the only vessel which is enclosed in a special perihaemal 
space ; in the Eudrilid Libyodrilus, the two sub-intestinal vessels are enclosed in 
a space of the same kind ; two mesenteries arise from the ventral wall of the oesophagus, 
which meet below and shut oif a space with a crescentic outline ; in this space run 
the two vessels in question. For a part of their course the vessels are free in the 
interior of the space, further back they are attached to the walls of the space, and 
further back still they come to lie outside of them. The spaces in fact are each of 
them confined to a segment, and do not pass continuously from segment to segment. 

In the two closely allied genera Hdiodrilus and Hyperiodrilus, the supra-intestinal 
vessel is in the same way enclosed in a coelomic space, distinct from the general coelomic 
cavity. As in the case of the dorsal vessel of Beinodrilus, the walls of the perihaemal 
space are connected here and there with the walls of the contained blood-vessel by 
delicate strands of fibrous tissue. The interspaces of these are filled with corpuscles. 
It seems possible that the function of these perihaemal spaces is concerned with the 
formation of the coelomic corpuscles ; they were always found to be filled with 
corpuscles, and in more than one instance the corpuscles could be observed in the 
act of being budded off from the walls of the space. 

§ 2. Other subdivisions of Coelom. A subdivision of the coelom only paralleled 
in the Polychaeta occurs in the genus Libyodrilus and in the aquatic Branchiura. 
In the former worm the two pairs of seta of each side of the body arise from the 
floor of a chamber which is cut oif from the general coelomic cavity. There are thus 
a pair of chambers along the body like the parapodial chambers in certain Polj'chaeta. 
The membrane which forms the wall of these chambers is thin and presents no 
appearance of structure except externally, where it is covered by nuclei ; the nuclei 
are on both sides ; the membrane is continuous with the parietal peritoneum ; the 
band of muscles uniting the two pairs of seta lies well below the membrane, which 
in section is seen to be somewhat, though not greatly, arched. Something of a similar 
kind occurs in the Tubificid Branchiura. Here the body at least in the posterior 
region is hourglass-shaped in transverse section ; from the ' waist ' of the hourglass 
a septum runs across the body-cavity transversely, dividing it into an upper chamber 
which contains the gut, and a lower chamber in which lie the nervous system and 
both dorsal and ventral blood-trunks. 

In addition to the coelomic tubes which have been described as surrounding some 
of the blood-vessels in certain Oligochaeta there is in Allolohophora a ventral tube 


■which Vejdovsky, who discovered it, compared to a lymphatic trunk ; it runs on 
the ventral surface of the body between the intestine and the nerve-cord, and is of 
limited extent. 

All these tubular cavities are suggestive of lymphatic vessels. In the thickness of 
the body-wall there are often irregular spaces and clefts which are filled with corpuscles. 
Attention was first directed to these by Kukenthal (1) who saw in them a fore^ 
shadowing of the Vertebrate lymphatic system. They occur apparently in a good many 
difierent genera, but in none have they proper walls of their own ; they are merely clefts 
and crannies left between the muscles. 

The branchiae of the genus BrancMura are hollow structures containing what 
I presume to be an extension of the coelom. This cavity however is traversed by 
anastomising fibres with nuclei at the nodal points ; whether it is lined by a definite 
coelomic epithelium or not I am uncertain ; the cavity is however shut off from that 
of the coelom by a muscular diaphragm which during life is in constant movement. 
It seems to be quite imperforate — to completely separate the coelomic and intra- 
branchial cavities. Very frequently this diaphragm was convex towards the body- 
cavity. If it were pulled out so as to form an ampulla lying in the body- cavity 
there would be a state of affairs comparable to that which is met with in the cephalic 
tentacles of Saccocirrus where Mabion and Bobbetzsky have described a cavity 
communicating with an ampulla lying in the body and have compared to the 
ampullae of Holothurians. 

§ 3. Codoviic organs of problematic nature. Attached to the anterior septum of 
segments x. and xi. in Sutroa are two bodies suggestive at first sight of sperm-sacs. 
These are of a racemose form and are hollow ; the cavity is however not single, but 
divided up by trabeculae into numerous subsidiary cavities. The walls are thin and 
appai-ently muscular. Enclosed within the meshes are many loosely packed cells. 
EiSEN first called attention to these bodies, but compared them to the albumen glands of 
Rhynchdmis. I could myself find no duct ; and the fact that in one case a diverticulum 
of the spennatheca lay within the sac led me to regard the cavity of the sac as 

In certain Perichaetidae there are a series of minute paired whitish bodies lying one 
on either side of the dorsal vessel in the middle region of the body, and springing from 
the septa (in Perichaeta indica) or from the dorsal vessel itself {Perichaeta dyeri). 
These bodies are quite solid, consisting of a mass of cells surrounding a few muscular 

In Acanthodrilus faldandicus there are a series of similar bodies commencing at about 
segment xx. and continuing to the end of the body. They are attached to the septa near 


to the nephi'idia and are not solid but hollow outgrowths of the septa ; they are often 
rather racemose in form and are chiefly muscular with a lining and covering of cells. 
The opening of their lumen into the cavity of the segment in front of that which contain^ 
them is visible. Clapaeede has described and figured in the common earthworm solid 
masses of cells enclosing a few muscular fibres and depending from the septa. Vejdovsky 
states that similar bodies occur in RhyTichelmis and Tubifex. He suggests that they are 
concerned with the growth of the septa. I found that the septal sacs of Acanthodrilus 
were rich in Glycogen. The structure of the septal sacs in this worm and in Sutroa is so 
like that of the sperm-sacs and egg-sacs that it is possible to see in them the remains of 
a segmentally arranged series of sacs out of which the sperm-sacs and egg-sacs were 
originally evolved. 

§ 4. Dorsal Pores. The coelom is placed in communication with the external medium 
in a large number of Oligochaeta by a series of pores, one to each segment ; in addition 
to these structures which are called the dorsal pores there is in a certain number, most 
of the aquatic Oligochaeta, a single pore on the prostomium which is generally spoken 
of as the head pore. The two coincide in the same species in the genus Fridericia 
alone. The dorsal pores are never developed upon the first one or two segments of 
the body, and the point where they commence is characteristic for the species ; in some 
forms for example the first one will lie between segments iv. and v., while in others 
the first pore lies altogether behind the cliteUum. The dorsal pores were considered 
at one time to lead into sacs, the function of which was believed to be respiratory ; 
it is now known that the pores are simply perforations of the integumental layers 
just at the anterior boundary of the segment to which they belong; there is no 
lining of epithelium as has been erroneously stated to be the case ; there is simply 
a discontinuity of the muscular and epidermic layers where the pores exist. The 
structure of these pores has been more particularly studied by Ude (3). In the 
figure which this author gives of a section through a dorsal pore there are represented 
a heaped up mass of peritoneal cells in the immediate neighbourhood of the pore; 
the function of this is very doubtful. Dorsal pores are not present in by any means 
all earthworms; they are absent for example in most of the Geoscolicidae, in many 
if not in all Eudrilidae, and in a few species of Acanthodrilidae and of other families ; 
among the lower Oligochaeta they are only found in a few species of Fridericia 
(Enchytraeidae). Their structure in the latter has been studied by Vejdovsky and 
MiCHAELSEN ; in these worms the pore is bordered by large round glandular cells on 
each side; no such cells are visible in the case of the dorsal pores of earthworms. 
We are at present completely in the dark as to the morphological meaning of these 
pores. There seems to be no relation between them and any other organs; pores 


opening into the body-cavity from the exterior undoubtedly suggest nephridial organs ; 
but no relations are apparent between the dorsal pores and the nephridia. 

The head pore is present in the Enchytraeidae, Naidomorpha and Lumbriculidae ; 
its position differs: sometimes it is at the very tip of the prostomium, sometimes at 
the junction of the prostomium with the buccal segment. A head pore has not been 
described in any earthworm. As to the function of these pores Michaelsen thinks 
that the head pore acts as a kind of safety valve to prevent undue pressure upon 
the brain when the movements of the body force an unusual amount of coelomic 
fluid into the anterior end of the worm's body. The dorsal pores he thinks have 
the function of moistening the body and preventing its becoming unduly day ; it is 
certain that the coelomic fluid is pressed out through the pores ; and their occlusion 
is regulated by longitudinal muscles which pass from the margin of one pore to that 
of the pore lying behind. Spencee even thinks that it is used for the purpose of 
rendering the burrow of the worm sufficiently damp for it to move with comfort in; 
possibly some coelomic fluid is forced into the cocoon by the movements of the body 
when this is passed over the head. This latter function seems to be a likely one; 
the use of the coelomic fluid as a lubricant seems to be, so to speak, too expensive, 
especially when there are glands in the skin which appear to serve the same purpose. 
It is not impossible that there may be an analogy between the dorsal pores and the 
nephridia on the one hand, and the vertebrate kidney on the other. In the kidney 
there seems to be a purely filtering action at the extremity of the renal tubules and 
a secretory activity in the glandular section of the same tubes ; perhaps in the 
Oligochaeta the dorsal pores pass out the waste fluids while the remaining excretory 
products are elaborated and passed out by the nephridia. 

IV. Nephridia. 

The excretory organs of the Oligochaeta, to be treated of in the present section, wiU 
be termed ' nephridia ' after the convenient name introduced by Lankester ^ ; the older 
name of ' segmental ' organs — used by Williams, though it survives in many text 
books, is not so useful, since it tends to disguise the real nature of the organs in 
question ; moreover the term ' segmental ' suggests that they are always metameric in 
arrangement, which is not invariably the case ; and that they are present in every 
segment of the body, which is also as far from being the truth. 

§ I. Excretory organs in the embryo. Recent researches on the development of 
the Oligochaeta, particularly those of Vejdovsky (9), have shown that at various 

' Notes on Embi-yology and Classification, Q. J. Micr. Sci. 1875. 


periods in the life-history of these worms there are four sets of excretory organs ; the 
very young embryo is furnished with certain epiblastic cells probably of an excretory 
function ; the older embryo has a pair of larval pronephridia ; the older embryo a set 
of embryonal pronephridia, and finally there are the definitive nephridia of the adult ; 
the terms used are those introduced by Vejdovskt in his work already quoted (9). 
These various excretory organs will now be considered seriatim. 

(a) Excretory cdls. In the gastrula stage of various species of Lumbricus and 
Allolohophora there are a few large cells which are thus early set apart to perform an 
excretory function ; so at least it is believed from the fact that they contain canals in 
their interior which are often coiled in quite a complicated fashion. They are epiblast 
cells distinguished by their large siZQ and more granular appearance ; they always mark 
the anterior end of the embryo and only persist during the younger stages ; they do not 
exist in Allolohophora foetida, nor were they found in Rhynchelmis. There are three of 
these cells. But they get to be completely fused so that no cell-outlines, but only the 
three nuclei, are distinguishable. The canal appears to become lost in the primitive 
body-cavity, lying between the epiblast and hypoblast. When the cells are kept under 
observation the liquid contained in the canals is seen to be evacuated ; after this has 
taken place the canals are no longer visible, but the cell-boundaries come into view; 
for further details the reader must refer to Vejdovsky's work (p. 208 et seq.). 

(6) Larval pronephridia. These were first seen by Vejdovsky in Allolohophora at 
a stage when the blastopore was still large : it is a fine canal running in the primary 
body-cavity. There are ultimately a pair of these tubes ; these tubes do not exist in 
Rhynchelmis nor in Allolohophora foetida, but they are as described by Beegh (4) 
enormous in Criodrilus. The tubes are ciliated and open on to the exterior anteriorly ; 
in Lumbricus ruhellus alone did Vejdovskt find an iuternal ' flame-cell.' The organ 
opens on to the exterior through the lumen of the excretory cells ; under the description 
of these the extension of that lumen into the primary body-cavity was mentioned. 
These tubes persist during the first formation of the embryonal pronephridia. 

(c) The embryonal pronephridia have no relation at all to the last, but they give rise 
to the permanent nephridia. These occur in every segment of the body, and the first 
pair open on to the exterior by the headpore of the embryo Lumbricus ; this first pair 
commonly occupy two segments and their lumen is not always ciliated. They are 
developed before the others, and disappear early, in a number of forms such as 
Lumbricus and the aquatic Oligochaeta. The name 'headkidney' has often been 
applied to this first pair, and they have been supposed to be different from the pairs 
which follow them. The principal difficulty in comparing them is the fact that the 
external pore is differently placed, being dorsal instead of ventral, and that they occupy 


two or three segments. The latter fact is explained by Vejdovsky as due to the late 
appearance of the septum, which thus allows the nephridium to grow backwai-ds. As to 
the first point, the apertures in question, though dorsal in Lumbricus, are ventral in 
Rhynchdmis. There can be no doubt of the homology of the first pair of embryonal 
pronephridia in the two forms, and so the different position becomes a matter of sub- 
ordinate interest. Although the pronephridia of the first segment disappear in Luvi- 
bricus, this is not universally the case ; I found that in Octochaetus viultiporus the first 
pair of nephridia persisted and fusing with the next pair became the ' peptonephridia ' 
opening into the buccal cavity (see p. 46). The persistence of these nephridia in 
Octochaetus is of course an additional argument in favour of regarding the first pair 
of these organs in Lwmhricus and Rhynchdmis as equivalent to the pairs which follow 
and are converted into permanent nephridia. 

The following pairs of pronephridia in both Rhynchdmis and Luvnbricus are 
short straight rows of cells without a lumen, but ending anteriorly in the case of 
Rhynchdmis, but not of Luvibricus, in a ' flame-cell ' provided with a long flagellum 
pointing backwards along the inside of the organ. These nephridia exist in the anterior 
segments of the body from which in the adult Rhynchelmis they subsequently disappear. 
In the Lumbricidae the pronephridia arise from a continuous string of cells (not proved 
in RhyTichelmis) ; this was first discovered by Wilson, it having been previously 
found by Whitman to hold good for the leech Clepsine. Wilson's results are so 
far confirmed by Vejdovsky (9). This string originates from a single large cell, the 
nephridioblast ; the row of cells is called by Vejdovsky the ' nephridiostich ' ; traced 
forward this row is seen to break up into oblique rows of cells each surmounted by 
a larger one which is on the ventral side of the body. The large terminal cell of each 
is the cell from which the funnel will ultimately be formed ; but it never shows the 
vacuole with the contained flagellum which is to be seen in the corresponding stage of 

From these pronephridia the permanent nephridia are developed ; but before de- 
scribing the way in which this development takes place it will be convenient to cast 
a glance over the general anatomy of the permanent nephridia in the Oligochaeta. 

§ 2. Nephridia of adult. 

Nephridia exist in all Oligochaeta. The only exception which there are good grounds 
to believe is really an exception in the Naid Uncinais littoralis ; Bouene, who 
carefully studied this species (5), was unable, after repeated observation, to discover any 
trace of nephridia. Another peculiar condition of the nephridia has also been described 
by Benham, and noticed by others, in certain Naids : in a few forms the nephridia are 




Fig. 5- 

limited to one side of the body ; with these exceptions all the aquatic Oligochaeta 
possess paired nephridia a pair to each segment ; ia all of these families, however, 
the nephridia are missing from a certain number of segments at the anterior end 
of the body. This state of affairs, however, is by no means confined to the Microdrili. 
It is met with in Pontodrilus, and the genera Glyphidrilus, Annadrilus, and 
Sparganophilus among the Geoscolicids ; the actual segment in which the nephridia 
commence is a matter which varies ; and as it is rather of classificatory interest, 
I refer to the description of species for more exact data. In the aquatic Oligochaeta — 
even if, as in the Lumbriculidae, the nephridia commence before the genital segments, — 
those segments never contain nephridia ; in earthworms, on the other hand, nephridia 
are present in the genital segments, except in certain of the genital segments in the 
few species mentioned above as resembling the aquatic families in the want of 
nephridia in the anterior segments of the body. The nephridia are always much 
coiled tubes ; and they always occupy two segments and two segments only ^. The 

internal aperture, the funnel, lies a segment in front of 
that which bears the external pore. In the lower Oligo- 
chaeta the nephridia have no blood supply^; they are always 
covered by a layer of peritoneum, the cells of which are 
often very large. 

The nephridium of Psanimoryctes barbatus will be 
selected as an example of a nephridium in one of the lower 
Oligochaeta ; it is fully described and figured by Vejdovsky 
(24, PI. ix, fig. i). The funnel which is composed of but few 
cells passes into a delicate spirally- wound tube decked with 
large clear vesicular peritoneal cells ; this passes into a thicker 
walled section of a yellowish colour ; this again passes into 
a clear walled tube which ends in a somewhat voluminous 
contractile bladder opening on to the exterior. In other 
NEPHEimuM OF MABioNiA aquatic forms the nephridia may be simpler ; but the same 

regions are generally recognisable ; not, however, in the 
Enchytraeidae whose nephridia (fig. 5) are very peculiar and 
resemble in many particulars the young developing nephridia of the higher Oligochaeta. 
A solid cellular mass, varying in shape according to the genus, is traversed by a coiled 
tube, the coiling of which again difiers in dififerent forms ; from this arises a duct 

(After Michaelsen.) 

* Doubtfully excepting Pluiellus. See also remarks on nephridia of Aeolosoma. 

^ Bhynchetmis is an exception (possibly other Lumbriculidae also) ; Vejdovsky has figured (9, PI. xxvi, fig. 20) 
a bloodvessel following the coils of the nephridium. 



which passes to the exterior and opens on to it through a small contractile bladder ; 
there appears to be no distinction here between the more or less glandular part of 
the organ in other Oligochaeta. Bolsius has lately discovered that the lumen is 
really a complex network. Among the Naidomorpha, the Lumbriculidae, and a few 
Tubificids, the funnel (totally absent in Aniphichaeta and Chaetogaster) is followed by 
an oval swelling coloured brown, and within which the nephridial tube appears to 
be branched and to form a small network; Vejdovsky describes a network along the 

Fig. 6. 

(From Benham.) 

a-A the narrow part of tlie tube partly ciliated, h-j ciliated wider tube expanding at C into ampulla. 
k-n. Wide tube. E. Muscular duct. I. Funnel. O. External orifice, t. Peritoneal layer, s. Muscular 
fibres, r. Nucleus of cell. F. G. 2nd and 3rd loops of nephridium. 

course of the nephridial tubes of Chaetogaster; this matter will be referred to again 
in considering the more complex nephridia of the earthworms. 

The nephridium, of Lumbricus. The most elaborate description of this organ 
is due to Benham (9), whom we shall here follow. The nephridium is divisible 
fii'st of all into two regions — one lying in the segment in front of that which bears 

F 3 



the external pore, which may be termed the preseptal portion ; the other the much 
more extensive postseptal portion. 

The preseptal portion consists of the funnel and of a short tube passing through 
the septum. The funnel is made up of a considerable number of tall columnar cells, 
which are ciliated over their entire inner surface; the connexion of the funnel with 
the tube is effected in the following way : the tube has the usual intracellular 

Fig. 7. 


(From Benliain.) 

E.F.G. ist, 2nd, and 3rd loops of nephridium. N. Nerve cord. S. Septum. SN. Subuerval bloodvessel. 
V. Ventral bloodvessel, a. b. c. Branches of the same. d. Commissure uniting dorsal and subnervian 
vessels, e. f. Its branches, g. Dilatations on capillaries. 

lumen ; the two sides of this (in optical section) diverge at about the centre of the 
funnel, 'each bending outwards, and then sharply backwards nearly parellel to its 
former course.' The lumen ceases at the point of divergence, the ceUs being only 
grooved; the cells are continuous with the marginal cells abeady referred to. The 
centre of the horseshoe-shaped funnel has now to be accounted for; this has been 


often figured as composed of a mosaic of numerous cells. It is, however, occupied by 
only a single large crescental cell. 

The 'narrow tube' which follows the funnel is the largest part of the whole 
nephridium ; the lumen is wide and of course intracellular ^ ; here and there the lumen 
shows slight indications of branching. Cilia are not universally present, but there 
is an alternation of ciliated with non-ciliated tract. This is followed by the ' middle 
tube,' which is of less extent and ciliated throughout ; its calibre is greater and the 
walls are more glandular. 

The 'wide tube' commences with a wider dilatation where it communicates with 
the middle tube. This part is also very glandular, but not ciliated. The last section 
of the nephridium is formed by the muscular duct. This has apparently a lin ing 
of large cells, so that the duct is intercellular ; it has numerous muscular fibres 
in the walls. This part of the tube opens directly on to the exterior. The actual 
course followed by the windings of the nephridium will be apparent from the 
illustrations and a special description is unnecessary. 

Nephrddia of other Genera. Lumhricus is really the only genus of earthworms 
whose nephridia have been carefully studied from the point of view of their minute 
structure ; there are, however, a few details to hand, which have been for the most 
part collected together in Benham's paper. In all the earthworms with paired 
nephridia the same regions of the tube can be distinguished ; but frequently there 
are differences in the relative development of the various parts. This is particularly 
the case with the terminal muscular section. In many genera of earthworms this 
section is very wide and large in proportion to the rest of the tube ; especially 
is this the case for example with Acanthodrilus dissimilis and a few other species 
of that genus, and with Microchaeta, etc. Moreover, in the species named, the 
muscular duct of the nephridium has a caecum given off near to its external pore ; 
the presence of such a caecum is very common. 

On the other hand, there are a good many species which appear to be without 
the terminal muscular duct, or in which at any rate it is but slightly developed ; 
this seems to be especially the case with small forms ; and it is perhaps a mark 
of degeneration ; such genera as Gordiodrilus and its allies show an apparently 
complete absence of the muscular sac. The above remarks, it will be understood, 
refer only to the genera with paired nephridia; in those with difiuse nephridia the 
terminal sac seems to be invariably wanting. The funnel too shows a certain amount 
of variation ; it is larger or smaller as the , case may be ; in Rhinodrilus gulidmi 

'' Vejdovskt (9, p. 349, etc.) does not admit intracellular nature of duct ; he believes it to be intercellular 


the funnels of the anterior nephridia are very large, and this condition is rather 
characteristic of the family Geoscolicidae ; it occurs also in Pontoscolex, where I have 
described it myself. In the last-mentioned genus the funnel is followed by a very 
wide section of the ' narrow tube.' The funnel varies in size in different worms, 
but in no earthworm with paired nephridia is it totally absent. 

The occasional branching, or rather the indications of branching, observable 
in the nepbridium of Luinbricus have already been referred to ; Bbnham has 
described in Microchaeta a complicated branching and anastomosis of the fine tube 
carried to such an extent that it formed a network round the other regions of the 
tube ; more recently KosA has met with the same thing in the genus Desmogaster ; 
in Eudrilus I have seen a certain amount of branching, but not so developed as in 
the genera mentioned. In all cases the nephridia of the higher Oligochaeta have 
an abundant blood supply ; this runs of course in the peritoneum which invests the 
nephridia externally; the only genera in which this vascular supply is absent (or, 
at the most, feebly developed) are Ocnerodrilus and Gordiodrilus. The actual course 
of the vessels supplying the nephridium will be described under the vascular system. 

Specialization of nephridia. Another matter which is worthy of note in 
connexion with the paired nephridia is the specialization which is occasionally 
shown in different regions of the body. Among the Geoscolicidae it is common for 
a variable number of pairs of nephridia, occupying the anterior segments of the 
body to differ in structure from those which follow and occupy the rest of the 
body. The very first pair of all, in Rhinodrilus ecuadoriensis for example, but 
,also in worms belonging to other families, often appear to have acquired a different 
function for they open into the buccal cavity ; but the consideration of these is 
deferred to a subsequent page. I am now concerned with those cases which 
may be exemplified by Microchaeta. In that worm the nephridia down to about 
the twenty-seventh segment are furnished with a long oval caecal appendage to the 
terminal sac. In the nephridia from the twenty-eighth segment onwards the terminal 
sac is larger and wider and is prolonged beyond the external orifice ; this prolongation 
corresponds of course to the caecum in the anterior nephridia, but it is hardly 
marked and is a continuation of the sac not being bent back upon it. 

In Acanthodrilus novae-zelandiae, one or two allied species, in Gryptodrilus fletcheri, 
and a few other earthworms, there is a very remarkable specialization of the nephridia, 
not connected as in Microchaeta with the cephalization of the anterior segment. 
There are in the worms now under • consideration two series of nephridia which 
open on to the exterior either in relation to the ventral or to the dorsal setae ; but 
although there are two series there is only a pair of nephridia in each segment. 


I speak of 'two series' because in the Acanthodrili at any rate the structure of the 
nephridium differs in relation to the varying position of its external orifice. When 
the nephridia open in front of the ventral seta they are provided with a very large 
caecum ; on the other hand, the dorsally opening nephridia have either no caecum 
at all or the large muscular terminal sac of the nephridium is prolonged a little 
way beyond the external pore. 

Alternation in position of external pores. In addition to the worms just referred 
to, a curious condition of the nephridia has been described by Hubrecht in the genera 
Lumbricus and Allolobophora ^ ; the position of the nephridio-pores shows an alternation 
similar to that of the worms already described, but there is no change in the position 
of the nephridia themselves within the body, such as occurs in Acanthodrilus for 

The bare fact of the alternation of these pores has been previously referred 
to by BoEELLi ; Borelli found that in a considerable number of different species 
this alternation from segment to segment was typical ; the following is a list of 
these species: — 

Lumbricus rubellus, HofFm. 
Lumbricus purpureus, Eis. 
Lumbricus herculeus, (Sav.). 
Allolobophora turgida, Eis. 
Allolobophora chlorotica, (Sav.). 
Allolobophora transpadana, Rosa. 
Allolobophora complanata, (Duges.). 
Allolobophora foetida, (Sav.). 
Allolobophora celtica, Rosa. 
Allurus tetraedrus, (Sav.). 

To this list others have since been added — for example Rosa's Allolobophora 
tellinii. In fact it may be taken apparently that the typical arrangement for the 
genera Lumbricus and Allobophora is the one that has just been referred to. The 
position of the pores does not show precisely the same kind of irregularity that has 
been mentioned in Acanthodrilus. Borelli found, and he has been confirmed by 
Hubrecht, that the pores may either occupy the position that is generally assigned 
to them in the text-books, i.e. just above the second seta, or they may lie in relation 
to the fourth seta or finally between the fourth seta and the dorsal pore. Moreover 
there is no regularity in the alternation from segment to segment and not always 

^ I can confirm from my own observations the accuracy of Hubbecht's statements and figures. 



Fig. 8. 


'1 ^B 

/I .MP 









/? .~3 

y_^ '1 "^ 


't ^^ 

a symmetry of aiTangement in the same segment ; in the latter facts these genera 
recall the condition so characteristic of certain species of AcantJvodriliis. 

Hubkecht's special contribution to this matter is his dis- 
covery that when the nephridium opens in an 'abnormal' position 
the duct enters the body-wall as if it were going to reach the 
exterior immediately by the ordinary course ; instead of which 
the tube bends to the left or to the right, as the case may be, 
and passing between the two muscular layers of the body-wall 
reaches its external orifice. The nephridium itself retains the 
same position in the body-cavity wherever the external pore 
may be situated ; on a mere dissection it would be impossible 
to say of any one particular nephridium where the external pore 
was placed. 

The tube as it passes along the body-wall lies so exactly 

between the circular and the longitudiaal layers of muscle that 

the appearances presented suggest at first sight merely a break 

in continuity of the body-wall along this line, due to a defect 

in the section ; this may possibly, Hubkecht thinks, have caused 

other cases similar to this to have escaped the attention of 

observers ; at any rate there is no doubt that so far as our 

present knowledge goes a similar state of affairs has not been 

described in any other genera than those mentioned. These facts have led Htjbeecht 

to suggest, though very tentatively, another theory of the variations of the nephridia 

in the Oligochaeta. 

It will be remembered that there are certain Cryptodrilids in which three pairs 
of nephridia are met with. Hubeecht indicates that the three positions which the 
nephridiopores occupy in many Lumbricidae may possibly be a reminiscence of the 
original presence of three pairs just as it has been suggested that the alternation 
in Acanthodrilus may be a trace of the former existence of two pairs of nephridia 
in the immediate ancestors of this genus. The same arguments that apply in the one 
case apply in the other. 

Diffuse nephridia. The nephridia are among those organs of the Oligochaeta which 
show most variation. For a long time it was thought that they agreed with other 
segmented worms in possessing a single pair to each segment of the body, a variable 
number of segments being without nephridia at all. It is now known that this state 
of affairs is by no means the only way in which the excretory system is developed. 
A very large number of genera are largely or entirely characterised by possessing 



fi. Setae, n. Nephridiopore. 


a nephridial system of a different kind. Pebrieu first called attention to the peculiar 
character of the excretory system of Perickaeta, which he described in the following 
words : ' Las organes segmentaires sont ici trfes rudimentaires, ce qui concorde avec 
I'absence d'orifice ext^rieur attribuable k ces organes.' This statement was made 
of Perichaeta posthuma; and later of Perichaeta robusta he wrote, 'Les organes 
segmentaires, sous forme de tubes extremement delicate, sont adherents aux cloisons, 
ou diss^min^es sur la membrane p^ritoneale que tapisse la cavity gdndrale'; further 
on, in the part of his paper devoted to a general resum^ of the anatomy of the group, 
he speaks of the nephridia forming a 'r&eau glandulaire,' which appeared to him to 
be an indication of an incomplete suppression of these organs. 

In a communication addressed to the Roj'al Society of London- (10), I pointed 
out that in Octochaetus multiporus there were more than a single pair of nephridio- 
pores to each segment of the body ; and that in the interior of the body the nephridia 
were divided into eight tufts in each segment, corresponding with as many external 
pores. In a later paper (47) I corrected the number, having found a much larger 
number of orifices. The next statement upon the subject was by Benham, who found 
in a species of Perichaeta a large number of small and separate nephridia. He 
referred in this paper to my own simultaneous discovery of a similar condition in 
another species of that genus. These results were published later ; I showed that in 
one species of Peridiaeta there, were a large number of external excretory pores, perhaps 
a hundred or so in a segment ; later still the funnels of these were discovered. My 
results and those of Benham were confirmed by Spencer for a large Cryptodrilid 
from Australia — Megascolides australis ; but Spencer, in addition to the network 
of small tubes with many external pores, found in the posterior segments of the body 
a series of larger tubes with funnels not possessed by the smaller tubes. Since these 
various papers were published a large number of species of earthworms have been 
described which possess an excretory system of this type, which has been called by 
myself 'diffuse' and by Benham ' plectonephric '. It characterises some or all of the 
species of the following genera (those in which aU the species have a plectonephric 
excretory system are marked by an asterisk) : Perichaeta*, Megascolex*, Octochaetus*, 
Beinodrilus*, Plagiochaeta*, Benhamia*, Trigaster*, Cryptodrilus, Megascolides, 
JDigaster*, Microdrilus*, Dichogaster*, Typhoeus*. All these genera, it will be noticed, 
are members of three families— Perichaetidae, Acanthodrilidae, Cryptodrilidae, which 
I unite here into one super-family Megascolicidae. In no other worms is this 
condition of the nephridial system met with, though I shall point out later, some 
Eudrilids are provided with an integumental nephridial network which is somewhat 





The naked eye characters of the excretory system of this kind are very obvious ; 
the absence of the large paired tubes cannot be passed over, and the delicate ramifying 
tubes especially attached to the septa can hardly be missed, at any rate in well- 
preserved specimens ; very often these tubes are massed into the semblance of paired 
nephridia, three or more pairs to a segment; this is often seen among the Acantho- 
drilidae; in the Perichaetidae, on the other hand, the network is more diffuse and 

not broken up into separate masses. In Mega- 
scolides and also in Megascolex armatus the diffuse 
network of minute tubules is reinforced by the 
existence of larger paired tubes, one pair to each 
segment. These large paired nephridia appear to 
be in connexion with the smaller tubes. 

In Octochaetus there seems to be no connexion 
from segment to segment of the nephridia, though 
the nephridial mass of, at any rate, each side of 
each segment forms a perfectly continuous net- 
work ; there are no funnels in the anterior region 
of the body, and the external pores are not so 
numerous as they are in Perichaeta. In the 
hinder region of the body the nephridia are also, 
as will be described more at length presently^ 
connected with the gut ; and the excretoiy system 
of this part of the body differs from that of the 
more anterior segments in two points : in the 
first place there are funnels, and in the second 
place there is no question at all about a network 
which is most clear ; I have figured this network 
(53), which often becomes so close that the inter- 
spaces are mere trabeculae of limited extent, 
traversing a kind of excretory sinus ; the net- 
work in the anterior part of the body has 
rather to be inferred ; I have never actually seen 
a branching of the tubules, but as the orifices 
to the exterior are numerous, and as there is 
nowhere any break of continuity, a network must exist; in any case there is no 
doubt whatever about its existence, in the hinder end of the body. The tubes which 
lead to the exterior often branch in the thickness of the body-wall, and often run 


(After Spencer.) 

A portion of execretory system of two segments. 
I. Nephridial tube in ■body-wall. 2. Funnel. 
3. Septum. 4-6. Nephridial tubes out in various 
directions. 7. External pores. 


for some distance in its thickness before opening on to the exterior ; in this case 
they generally run between the two muscular coats. 

In Perichaeta the nephridial tubules are of finer calibre than: in Octochaetus j 
and they are furnished with funnels, which are especially obvious in the anterior 
region of the body ; in this region, apparently in all true Perichaetae — it has at any 
rate been commented upon by many observers — the nephridial mass is much thicker 
than posteriorly ; the septa are here covered with a thick almost furry coating of 
tubes, which in sections are seen to leave but little of the available coelom free ; 
further back the nephridia are by no means so obvious, a fact which has led 
to their being described as absent by myself (in Megascolex coeruleus for- 
instance) ; in poorly preserved earthworms they would not been seen by anyone 
who was not already informed of their existence and therefore unprepared for their 
excessive minuteness. I have already mentioned that to Benham belongs the credit 
of having first discovered the existence of numerous minute nephridia in a species 
of Perichaeta; from his account it would appear that in this species the several 
nephridia are quite distinct from each other ; this does not seem to be the case 
in all the species of this genus ; in Perichaeta hermudensis, for example, it seemed 
to be undoubtedly the fact that there was a connexion between all the nephridial 
tubes of a segment, and that in addition to this there was a connexion between the 
nephridial plexuses of following segments ; the tubes were followed through the septa ; 
they appeared to pass into the plexus of the segments in front and behind. The 
exact arrangement of the nephridial tubes is very difficult to follow on account 
of their smallness and complicated course, but at any rate Spencer's investigations 
into the anatomy of the excretory system of Megascolides led him also to the 
conclusion that there was a connexion from segment to segment. That there are 
numerous external pores is a matter capable of being easily proved ; if a bit of the 
cuticle be stripped off, the pores can be easily seen, particularly in specimens which 
have been preserved in corrosive sublimate ; the pores, although small, are far from 
invisible; it is a matter of interest that there seems to be no regularity in their 
arrangement; there is no trace of any metameric disposition. The segmentation 
so clearly visible in most of the organs of the worm's body has here been lost. 

The minute structure of the nephridia in those worms which have an excretory 
system of the diffuse type has been chiefly studied by Spencer (1) and myself (45, 
47). Some details about the funnel are to be found in Benham's paper already 
referred to. As Benham has pointed out, the nephridia of Perichaeta evidently consist 
of the three regions described by himself in Lumhricus ; in Octochaetus only two kinds 
of tubes are to be found. 

G 2 


In all of these worms there is no muscular sac at the points of opening on to the 
exterior ; favourable sections are wanted in order to demonstrate the actual pores, 
since the tubes are of narrow dimensions up to the very .opening. Generally speaking, 
the cells of the epidermis show some peculiarities at the orifices ; in Perichaeta I have 
figured the conditions which are found ; and Spencek has figured them in Megascolides. 
The epidermic cells become slightly narrower and converge round the actual orifice, 
forming possibly a kind of constrictor for regulating the outflow of secreted matter. 
As already mentioned, the funnels may be present or absent in the diffuse nephridia; 
when present they do not seem to difler at all from those of the paired nephridia, 
except that they are generally smaller. Benham has carefully described the funnel 
in Perichaeta malamaniensis ; there are eight or nine marginal cells, but no central 
cell nor any centrifugal gutter-cells ; the intracellular duct of the tube appears to 
open at once into the midst of these cells. I have, however, described a horseshoe- 
shaped funnel in Perichaeta herrtiwlensis. 

Nephridia of this kind do not exist in any of the aquatic Annelids, nor have 
they been met with in any other groups of Annelids excepting only the leech 
PontoMella, where Bodene has stated that a network exists and is continuous from 
segment to segment ; this statement, however, has been denied. 

Integumental nephridial network. The network that exists in certain Eudrilidae is 
not, I believe, morphologically comparable to the diffuse nephridial system already 
treated of. This condition of the nephridial system appears to characterise a large 
number of Eudrilidae, and I have attempted a classification of that family partly based 
„ upon the presence or absence of the network. It has been studied most thoroughly in 
the genus Libyodrilus. That worm, like all the gther genera of the family to which it 
belongs, has paired nephridia — a pair to each segment. But the duct leading to the 
exterior, instead of passing at once to the exterior, branches and forms a complicated 
network in the integument ; in this worm the peritoneal layer which lines the body- 
wall is, in places at any rate, excessively thick; and in this layer tubes formed by 
the branches of the ramified external duct run ; these put into communication the 
nephridia of successive segments. The actual details of the way in which this complex 
network is formed are perhaps subject to some variation. It appeared to me, however, 
that there were four principal and longitudinally running trunks, symmetrically 
disposed two on each side of the nerve-cord (corresponding in position to the setae) ; 
from these branches arose which ramified in every direction through the longitudinal 
muscular layer and finally joined a circular vessel running right round the body 
between the two muscular coats. From this latter fine branches lead to the exterior. 
These tubes are nowhere ciliated and seem to be not comparable to the coelomic 




network of Ferichaeta, etc., but to be a branching of the epiblastic duct of the 
nephridium. The duct of the nephridium of Allolobophora is said by Vejdovsky (9) 
to branch in the integument, and I have already referred to the way in which 
that duct runs for a considerable distance between the two muscular coats before 
opening on to the exterior. In some few of the genital segments it seemed to me 
that the paired nephridia in Lihyo- 

drilus were absent, but that the net- ^" 

work was present. One cannot help 
being reminded by these facts of the 
excretory system in the Nematoda and 
the Acanthocephala ; in both these 
groups the excretory system seems to 
occur in the shape of tubes running 
in the integument ; the Lemnisci of 
the Acanthocephala (and an homo- 
logous structure has been described 
in certain Nematodes) are processes of 
the body-wall occupied by a quantity 
of tubes which are doubtless of an ex- 
cretory nature. Eliminate the paired 
nephridia of Lihyodrilus and the re- 
maining part of the excretory system 
would be exceedingly like that of the 
two groups of worms referred to. To 

a certain extent I have followed the development of the excretory organs in Lihyo- 
drilus s the integumental network is, as one would suppose, a secondary development ; 
the first part of it to appear is a contiauous longitudinal duct on each side connecting 
the nephridia of following segments. I imagine that Vejdovsky is correct in regarding 
the longitudinal duct connecting the nephridia of Lanice conchilega (described by 
Meyeb^ and Cunningham^) as formed out of the terminal epiblastic part of the 
nephridia — the ' bladder' of the nephridium of Lumbricus — which have become fused 
together ; it is therefore of some interest to note that a similar single connexion is 
first developed in Lihyodril-us. The fabulous (?) connecting duct of Polygordius is 
perhaps of the same nature. 

The genera Megascolex and MegascoUdes — probably others will be discovered when 



I. Dorsal blood-vessel. 2. Supra^oesophageal blood-vessel. 
3. Infra-oesophageal blood-vessel. 4. Ventral blood-vessel. 
5. Nephridium. 6. One of longitudinal trunks of integu- 
mental network. 7. Fine canals of network. 

' Quoted by Lang in his monograph on the Polyclada in Fauna u. Flora des Golfea von Neapel. 
'■' On some points in the Anatomy of Polychaeta. Q. J. Mior. Sci. vol. xxviii. 


the finer anatomy of earthworms is better known than it is at present — are in their 
excretory system to some extent intermediate between the Megascolicidae that have 
just been described and those in which the nephridia are paired; there is a network 
of fine tubes, but in addition to this a pair of larger nephridia which have a funnel 
opening into the segment in front of that in which the tube lies ; a more remarkable 
intermediate condition exists in those Cryptodrilidae which I place in a sepai-ate 
genus Trinephrus, the name being suggested from the fact that they have three 
pairs of nephiidia in each segment of the body ; at present, however, the finer anatomy 
of these worms is unknown, and we do not know the exact structure of these ne- 
phridia ; a second intermediate condition is ofiered by the Geoscolecid genus Brachy- 
drilus. In this worm there are two distinct pairs of nephridia in each segment of 
the body. 

Finally, we have those forms in which there are only a single paii- of these tubes 
and they form the majority ; they include all the aquatic families, the Lumbricidae, 
the Geoscolicidae (except Brachydrilus already referred to), all the Eudrilidae (where, 
however, there is an integumental plexus to be referred to), and many of the 
Megascolicidae, the following genera (those genera in which all the species have 
paired nephridia are marked with an asterisk): Diporockaeta*, Perionyx*, Acantho- 
dnlus*, Cryptodnlus, Megascolides, Ocnerodrilus*, QordiodrUus*, PygmaeodrUus*, 
Pontodriltis*, Microscolex*. 

Connexion of neph'idia with the alimentary canal. There are among the 
Oligochaeta various organs of a glandular nature, opening into the alimentary canal, 
which in some cases are cei-tainly, in others probably, more or less modified nephridia. 
These organs are invariably connected with either the front or the hind end of the 
canal ; that is to say, they never communicate with that section which is undoubtedly 
of hypoblastic origin. At the same time it cannot be said with absolute certainty 
that those sections of the gut into which they do open are either stomodaeum or 
proctodaeum ; unfortunately embryological data are at present too scanty to permit 
of a definite statement upon the point. In the present section I shall only deal 
with those glandular appendices of the alimentary tract which are certainly, or 
very probably, of nephridial nature; I deal later with such organs as the calci- 
ferous glands which cannot, with either probability or certainty, be referred to that 

I. Peptoncphridia. I accept Benham's term for nephridia opening into the 
anterior section of the alimentary canal; it seems probable that their function 
is in relation to that of digestion. Such organs occur in more than one family 
of the Oligochaeta. They were first made known in the Enchytraeidae, and 


have since been described in the Acanthodrilidae, Eudrilidae, Cryptodrilidae, and 

In the Enchytraeidae they are the so-called salivary glands, which have been 
described by Vejdovsky, Michaelsen, and others ; in certain species of this family 
there are a pair of tubes which are more or less branched, opening into the alimentary 
canal behind the pharynx ; sometimes the apertures are lateral in position, sometimes 
they are dorsal and ventral respectively. The principal reason for considering these 
tubes to be of nephridial nature is then- minute structure ; they have a lumen which 
is undoubtedly intracellular; it is, however, necessary to be careful not to be too 
much influenced by a consideration of this nature ; the calciferous glands, for 
example, are in some worms folded in so complicated a fashion that the lumen 
becomes intracellular ; the lumen of the blood-vessels is also intracellular ; it has 
been shown that minute vessels are formed by the canaliculisation of cells (by 
Lankestee in the Leech). The salivary glands of the Enchytraeidae, however, 
difier from nephridia in having no opening into the coelom ; there is not the least 
trace of a funnel. Perhaps this fact is not of first rate importance as an argument 
against their nephridial nature; but it must be considered. Moreover the tubes are 
not ciliated ; this again is against any homology with nephridia ; in no Oligochaetous 
worm are there nephridia which are entirely without ciHa ; it frequently happens 
that a greater or less section of the nephridia is devoid of cilia ; but the non-ciliated 
area is restricted. The segments which are occupied by these salivary glands are 
devoid of other nephridia ; this, at first sight, suggests the metamorphosis of the 
missing nephridia into the salivary glands ; but it must be remembered that in the 
Enchytraeidae, as in most other of the lower OUgochaeta, the nephridia are defective 
in the anterior segments of the adult worm. On the whole, it seems that the 
nephridial or non-nephridial nature of the salivary glands of the Enchytraeidae must 
be left an open question. It is quite otherwise with those earthworms in which 
similar salivary glands occur. The first description of nephridia opening into the 
anterior part of the alimentary canal in an earthworm was by myself in Octochaetus 
tnultiporus; in this worm there are lying along the pharynx a pair of tufted organs 
which end in a duct of some dimensions; this duct runs forward and opens into 
the buccal cavity ; here again one objection to the nephridial nature of these glands 
is the fact that they do not appear to be furnished with any coelomic pore. This 
statement, however, only applies to the glands in the adult worm. Fortunately in 
this particular instance the development has been traced and I have found that 
these compact glands are really formed by the fusion of at least two pairs of nephridia 
which are at first distinct and each provided with its own coelomic funnel. There 





is in fact no room for doubt that these ' salivary glands' in Odochaetus multiporua 
are nephridia, modified doubtless to perform some function other than that performed 
by the nephridia in the other segments of the body — though physiological observations 
are -wanting as to -what this function is. In Rhinodrilus ecuadoriensis Benham 
has described the first pair of nephridia as somewhat different in appearance from 
the remaining organs and as opening into the buccal cavity. Such glands also occur 
in Acanthodrilus annectens and in its near ally Acanthodrilus palvdosus. In certain 
earthworms there are a pair of large nephridia occupying the first two or three 
segments of the body, which appear to resemble those of Odochaetus already described ; 
such glands occur in Pontoscolex (see woodcut), where they were originally termed 

by Peekier 'glandes k mucosit^'; I showed that these 
organs are undoubtedly nephridia ; they do not, however, 
actually open into the gut, but so near to it that when 
the anterior segments are, as is occasionally the case in 
this worm, inverted, the orifice is actually sheltered by 
the inversion. These glands cannot be really distinguished 
from the ' peptonephridia ' of Odochaetus, etc. 

Spencer described in the Australian Megascolides 
australis a number of fine tubes opening into the buccal 
cavity, which are clearly nephridia ; in that worm the 
nephi'idial system is of the diffuse kind ; and the tubes 
lying in the anterior section of the body open into the 
buccal cavity. The difference between the condition 
which characterises this worm and that found in Odo- 
chaetus is simply that there are numerous openings 
instead of only a single one on each side. It seems 
reasonable to suppose that the two organs are referable 
to precisely the same category. Curiously enough the 
same presence of numerous nephridial tubes opening 
into the buccal cavity occurs in the Eudrilid Libyo- 
drilus; in that worm there is a fine network of tubes in the walls of the buccal 
cavity, opening here and there into the interior of that cavity. In all these cases 
it will be observed that the nephridia which are connected with the anterior part 
of the alimentary canal show precisely the same modifications as the nephridial 
tubes whicli are not so connected, but which open directly on to the exterior. 
There are in both cases paired and diffuse nephridial tubes. There can be no 
doubt, it would seem, that the buccal cavity of these tropical earthworms, like 


(After Perrier.) 

g. c. Cerebral ganglia. 6. Buccal cavity. 
p. k. Pliaiynx, os. Oesophagus, g. a. First 
nephridium Cglande & mucosit^'). o.g. 
External orifice of same. 0.8. Second 
nephridium. n. Nerve cord. /. Vacuities 
in muscular layer for implantation of 


that of Lumhricus, is stomodaeum ; hence there is nothing remarkable in the fact 
that nephridia open into the buccal cavity instead of the exterior, for the buccal 
cavity is morphologically external. In the case of Octochaetus the correspondence 
is even clearer; for in that worm the massive Peptonephridia, though they only 
possess a single aperture apiece into the buccal cavity, open also on to the exterior of 
the body and as in the case of the nephridia of the succeeding segment by numerous 
pores. The development, hovrever, happens to show that it is the openings into 
the buccal cavity which are the primitive openings, the numerous external pores 
being secondary. The relationship of the modified anterior nephridia to the 'head 
kidney' of the embryo has already been dealt with. 

2. Anal nephridia. Not only are there undoubted nephridia connected with 
the anterior end of the alimentary canal ; but in one species of Oligochaeta there 
are undoubted nephridia connected with the rectal region. In Octochaetus niultiporus 
I found that a few segments at the posterior end of the body are filled with 
a dense mass of nephridial tubes which open both directly on to the exterior and 
into the rectal part of the gut ; I am not, however, certain as to whether this part 
of the gut is or is not proctodaeum. On the whole, the facts which I was able to 
get together as to the development of this part of the nephridial system seemed to show 
that the section of the -gut into which they open is not proctodaeum; on the other 
hand, probability seems to urge that it is. A remarkable fact about these anal 
nephridia in Octochaetus is, that they are provided with numerous coelomic funnels. 
Elsewhere the nephridial system of this Annelid is not in the adult provided with 
funnels ; the tubes form a branching network, very easy to demonstrate as a network, 
which communicates from segment to segment and also communicates with the 
exterior through numerous pores upon the integument; there is thus a direct 
communication between the interior of the alimentary canal and the exterior through 
the nephridia ; the nephridial tubules, when they approach the lumen of the gut, 
open into wider passages, which have a lining of cells precisely like those of the 
gut ; it would appear that these latter are really diverticula of the gut, though here 
again the actual development has not been worked out. The occurrence of these 
nephridia opening into the gut is interesting when they are compared with the 
respiratory trees of Bondlia and its allies among the Gephyrea. I have also suggested 
a possible resemblance to the Malpighian tubes of the Arthropoda. The comparison 
here would be of course with the terminal wider tubes with an intercellular duct 
into which the actual secretory portion of the tubes open. So far as it is at 
present known Octochaetus multiporus is the only Oligochaet which possesses anal 




Fiar. 12. 

§ 3. Development of permanent nephridia. 

The development of the permanent nephridia out of the pronephridia has been 
followed by Vejdovsky (9) in Shynckelmis and in various species of Lumbricidae. 
In the former genus the postseptal part of the pronephridium gets thicker and a lobe 
(woodcut, fig. gl) is formed which lies to the side ; the distal part of the pronephridium 

remains unaltered ; in this condition the nephri- 
dium recalls that of the Enchytraeidae. The 
pronephridiostom, originally consisting of one 
cell with a flagellum, becomes multicellular, 
yJ ultimately consisting of eight cells ; these become 
ciliated and the long flagellum vanishes ; but it 
coexists for some time with the cilia. The 
vacuole of the pronephridiostom becomes the 
cavity of the funnel, which comes to open into 
the coelom and the tube following is gradually 
excavated. The last part of the permanent 
nephridium to appear is the epidermic, con- 
tractile, terminal bladder. Occasionally in the 
course of development secondary funnels appear, 
a course of development which recalls the 
formation of many funnels in some of the 
' plectonephric ' worms. In the Lumbricidae 
the distinction between the pronephridium and 
the nephridium is not so marked because of 
the absence of the vacuole and the flagellum in 
the pronephridiostom ; but in these worms, as 
in Rhynchelmis, the main part of the permanent 
organ appears as a lateral outgrowth of the 
the original straight tube lead- 
more or less 

unaltered, the alteration concerning the formation of a hollow tube out of it. It has 
been said by Bergh that there is no epidermic invagination to form the contractile 
bladder of the nephridium of Lumbricus ; but this is denied by Vejdovsky. There 
is evidently a greater break between the pronephridia and the permanent nephridia 
in Rhynchelmis than in Lumbricus. This is emphasised by the existence of the 
flagellum in the former. 

(After Vejdovsky.) 

The stages are numbered oonseoutively. gl Lote pronephridium 
forming chief part of Nephridium. 

ing to the exterior remaining 



The development of nephridia of the plectonephric type has been investigated 
by myself in Odochaetus muUiporus. There is no doubt that the numerous nephridial 
tufts of the mature worm with their numerous openings on to the exterior are 
produced by the growth of a paired series of organs, which, as Vejdovsky thinks, 
are probably the equivalents of the pronephridia of Luonbricus. These paired 
nephridia, however, are provided with well-developed and ciliated nephrostoms ; 
and in their course they are much coiled upon themselves before reaching the 
external pore. The first pair are anomalous from the very earliest stage in which 
I studied them. They occupy at least two segments as does the ' head- kidney' of 
other species, and they open, not on to the lateral body surface but into the 
commencement of the stomodaeum. A little later in the course of development 
these first pair become fused with the second pair and the pore is more definitely 
within the stomodaeal cavity, from which a tubular outgrowth appears to have 

Fig. 13- 


(After Vejdovsky.) 

J. Funnel. 2. Septum. 3. Nephridial tufts in various stages. 4. Connecting tubes finally 

(4 a) degenerating. 

arisen to meet the duct of the nephridium. Thus the ' mucous gland ' of the adult 
worm is a compound structure representing, two nephridia of successive segments. 
The funnels of all the pairs of pronephridia, at first functional, degenerate later ; 
the cilia disappear and the row of cells which forms the funnel becomes more than 
a single cell thick. The lumen of that part of the nephridium also which immediately 
follows the funnel vanishes and a solid string of cells is left. 

The development of such nephridia has also been studied by Vejdovsky (4) in the 
Australian earthworm Megascolides australis. There appears to be no doubt that in this 
species there is (see accompanying woodcut) to begin with a pair of nephridia to each 
segment ; these have a funnel and from the funnel leads a straight duct not perforate ; 
here and there the cells become larger and finally form loops ; these loops ultimately 
increase in size and become complicated coils, the connecting part of the original tube 

H 2 



degeneratiDg into a mere strand of connective tissue. The last step is the absolute 
severance of the connexion. Thus it appears firstly that the nephridial system of 
this worm originates from a pair of pronephridia to each segment; and that this 
becomes broken up into a large number of nephridia of which one only, the large 
nephridia described by Spencee, retains the funnel. If there is, as has been described 
by Spencee, a plexus formed by the interconnexions of the small nephridia it must 
be secondary ; but at any rate it is noteworthy that at first when the several 
nephridia are in course of formation out of the pronephridium there is a connexion 
at least between the numerous nephridia of the same segment. As to the continuous 
longitudinal ducts described by Spencee the most careful search failed to show them 
in the embryo ; they also must therefore be secondary structures. 

§ 4. Phylogeny of excretory system. 

The facts just detailed concerning the development of the plectonephric system 
might appear at first sight to argue for the theory that this condition is secondary 
and that the paired nephridia of Luvibricus, &c. are to be looked upon as the primitive 
condition. The facts seem to negative my view, supported by Spencee and Benham, 
that the paired nephridia of the majority of Oligochaeta are formed by reduction from 
a network such as now exists in FericJiaeta and many other genera. Before any 
developmental facts were known this course of evolution seemed to be supported 
by many considerations. In the first place a progress from a more generalized to 
a more specialized condition is seen in the evolution of other organs. Then there are 
certain resemblances between the network nepkridia and the excretory organs in the 
Platyhelminths. Traces of the supposed primitive condition also existed in those 
worms which are now provided with the paired form of excretory organ; thus in 
Anachaeta the nephridium has occasionally more than one funnel; in many forms 
there is a branching and anastomosis of the fine tubes of the nephridium, for example 
in Microchaeta, to which reference has already been made, and in other forms also. 
This view must evidently be now given up ; but, on the other hand, it is not by 
any means permissible to adopt the converse view already suggested. It does not 
follow that the diffuse nephridia are the outcome of a branching and specialization 
of the paired nephridia ; on the contrary the developmental facts absolutely disprove 
this. "What they do prove is that both paired and diffuse nephridia are formed 
out of similar pronephridia ; that in fact both kinds of excretory organs are equally 
ancient. This opinion, practically arrived at by myself after the study of the 
development of Odochaetus, was more definitely formulated by Vejdovsky (9). 


It is clear too from the fact that the diffuse and paired form of the excretory 
system occur in forms which are so nearly related, for instance in Acanthodrilua 
and Octochaetus, that there can be no profound gap between the two kinds of 
organs. The ' Plectonephrica' of Benham I agree with Vejdovsky in considering 
an artificial group. 

V. Alimentary Canal. 

The alimentary canal in all Oligochaeta consists of a straight tube running from 
the mouth which opens on to the first segment and is overhung by the buccal lobe' 
(when present) to the posteriorly situated anus ; with two exceptions the anus 
is surrounded by the last segment of the body. These exceptions are Criodrilus and 
Sparganophilus ; in the former worm Vejdovsky figures (24, PI. x, fig. ai), seven post- 
anal segments, and the anus itself as dorsal in position. There is, however, no fiexure 
of the intestine in CriodrUus ; it passes, as in other Oligochaeta, perfectly straight 
from the mouth to the anus. Developmentally the alimentary canal of the Oligochaeta, 
as of other animals, consists of three portions: (i) Mesenteron, hypoblastic in origin, 
(3) StoTnodaeum, (3) Proctodaeum, both formed by later invaginations of the epiblast ; 
of the two the proctodaeum is the later formation. The greater part of the alimentary 
canal is of hypoblastic origin. The actual extent of the stomodaeum seems rather 
doubtful. Vejdovsky, who at one time thought that the end of the pharynx m&rked 
its posterior limit, was subsequently inclined to think that the buccal cavity only 
was of epiblastic origin in Lumbricus and Rhynchelmis. 

In the adult worms the alimentary canal may be divided into the following 
regions: mouth and buccal cavity, oesophagus, pharynx, and intestine. The tnouth 
is nearly invariably ventral in position^ and it leads into the buccal cavity which 
is of limited extent; the buccal cavity of the Enchytraeidae is often provided with 
one or a pair of small tongue-like organs which spring from its floor. These are 
probably sense organs. Michaelsbn at one time put them down as playing the 
part of a sucker. These organs are furnished with minute hair-like processes, and 
appear to be entirely cellular; they are in fact a product of the lining epithelium 
of the buccal cavity; in some species they can be everted; in a few species the 
buccal cavity has a dorsal diverticulum in which the cells are more glandular ; 
this state of affairs occurs in Benhamia, in Microdrilus and a few more species 
In the higher Oligochaeta the buccal cavity is separated from the ensuing pharynx 
by a constriction on which lies the cerebral ganglia; in Aeolosoma there appears 

' A few exceptions where the intestine is spiral are noted under the description of that organ. 
^ Terminal where the prostomium is absent. 


to be no buccal cavity at all but the mouth leads at once into the pharynx. 
The latter has usually exceedingly muscular walls, the musculature being dorsal ; the 
lumen is folded and ciliated, in the lower Oligochaeta, and dorsally, as Benham has 
lately pointed out in the earthworms — at least in many earthworms. In Pei-ichaeta 
the buccal cavity is markedly capable of extrusion; to a less extent this is also 
the case with other Oligochaeta. The pharynx, on the other hand, is not to be 
exti-uded. After the pharynx comes the oesophagus ; this is a tube of moderate 
calibre which extends through a varying number of segments. The oesophayus 
is always ciliated in the lower Oligochaeta but not throughout its whole extent 
in the earthworms ; the cilia usually begin after the calciferous glands, if these are 
not present in the hinder region of the tube ; thus in Liiyodrilus the cilia begin in 
segment xiv. The oesophagus is the most specialized part of the alimentary tract 
in correspondence with the cephalization exhibited by the other organs which lie 
in the anterior region of the body. The pharynx is often supplied with glands 
which have been variously termed ' salivary glands,' ' septal glands ' ; they are not 
to be confounded with 'salivary glands' of apparently nephridial origin which are 
treated of under the excretory system. 

The septal glands (at one time mistaken for ganglia of the visceral nervous system) 
are especially conspicuous in the aquatic Oligochaeta, where they are attached to the 
septa, whence of course the name ; they are masses of pear-shaped cells, each cell being 
prolonged to form its own duct ; the ducts appear to enter the pharynx. Vejdovsky 
has figured very obvious glands of this kind in the embryo Allolohopfwra. 

These septal glands seem to me to be simply epidermic glands which have been 
invaginated along with the stomodaeum ; they are appended to a part of the alimentary 
tract which must be, though the actual proof is in most cases wanting, of epidermic 
origin ; in this case the glands will be entirely comparable to those which open on 
to the genital papillae of the Perichaetidae, or in fact to the glandular cells of the 
clitellum. On the other hand their position upon the septa and lying freely within 
the coelom seems to be against this interpretation, and in favour perhaps of regarding 
them as homologous with the septal sacs of certain species of Perichaeta and Acantho- 
drilus referred to above. 

Gizzard. In the majority of the Oligochaeta a part of the oesophagus is modified 
into an organ which is usually called the gizzard ; this part of the alimentary canal 
is distinguished by the immensely thickened muscular walls and by the thick chitinous 
layer secreted by the lining epithelium. As to the muscular layer, it is the circular 
fibres which are especially increased. The gizzard is absent in a large number of 
Oligochaeta, particularly among the mud-living forms; in no family of Oligochaeta 


Which habitually live in or at the bottom of streams, lakes, etc. is the gizzard present ; 
this fact might, and indeed has, led to the inference that its absence is to be accounted 
for by the soft nature of the food. Probable though this hypothesis is, it seems to be 
contradicted by the fact that, the gizzard is also absent in a number of terrestrial 
Oligochaeta whose food is presumably quite the same as that of other terrestrial species 
which possess a gizzard. Moreover a gizzard is wanting or rather represented only 
by a rudiment in a species whose habitual food is harder than that of any other 
Oligochaeta ; in Pontodrilus bermiidensis, which lives on the sea shore in coral debris 
(with which its alimentary canal is always full) there is only the trace of a gizzard. 
Microscolex too is a purely terrestrial form but it either has no gizzard or a degenerate 
one. It is evidently therefore not safe to lay down any such general statement about 
the cause of the presence or absence of the gizzard. In the greater number of earth- 
worms the gizzard only occupies a single segment ; but the segment in which it is 
found is not always the same ; in Luvibricus, for example, the gizzard is usually 
in the xviiith segment; in Megascolex, on the other hand, the vith segment is 
occupied by this organ. Very often there is more than a single gizzard ; when this 
is the case the gizzards are in consecutive segments ; the genera Bigaster and 
Dichoffa&ter have been so named on account of the presence of two gizzards ; there 
are two gizzards also in Benhamia and three in the genus called by Fletcher 
Perissogaster — a genus which is in the present work included in Bigaster. In the 
genera Moniliga&ter, Pleionogaster, Hyperiodrilus, Heliodrilus, and one or two others, 
there are a considerable number of gizzards— three to ten in number. 

The genus Perichaeta is remarkable for the fact that it is provided with only 
a single gizzard, which nevertheless occupies two segments. It seems to be quite 
possible that in this case there are really a pair of gizzards which have beconxe 
intimately fused so as to form a single one. 

An important point to be noticed about the gizzard is that it may occur in any 
segment or segments of the oesophagus ; it has no fixed position except of course 
for the species or genus as the case may be. 

As to the histology of the organ, comparative researches are as yet wanting. It 
is perhaps remarkable that the muscular tissue which enters into its formation is 
precisely similar to that of other parts of the body and not striated; it so often 
happens that the muscular tissue of organs of great muscular power is made up of 
striated fibres that the negative fact — that this is not the case with the Oligochaeta — 
is worth calling attention to. 

Calciferous glands. — Appended to the oesophageal region of the digestive tract 
of many Oligochaeta are a series of glands which have been variously termed 


' Glands of Morren,' ' Oesophageal glands,' ' Calciferous glands,' etc. The oesophagus 
in the higher Oligochaeta is usually divisible into two tracts, of which one is more 
richly supplied with blood-vessels than the other and has a more folded lining 
epithelium. Frequently this part of the oesophagus is constricted by the septa, and 
the sections of the tube which lie between the septa are broader than those which 
are nipped by the septa ; hence a moniliform appearance often exists. This is the 
case for example in the genus Perichaeta (s. s.). The specialization of a tract of 
oesophagus having these characters in a more pronounced fashion is characteristic 
of the genus Onychochaeta and of other forms ; in the genus mentioned that portion 
of the oesophagus which occupies segments xii. to xv. is wider than elsewhere 
and is provided with markedly regular and deep folds of epithelium ; this is 
a step in advance towards the existence of distinct diverticula of the oesophagus 
such as exist in a great many genera and species ; it is at the same time 
merely an exaggeration of the vascular tract of oesophagus commonly found in 
the higher Oligochaeta whether they possess or do not possess distinct calciferous 

In Octochaetus multiporus the oesophagus is swollen in segment xvii. ; in transverse 


sections of these swellings they are seen to be really diverticula of the oesophagus 
opening into it by wide apertures on either side. In Pontoscolex the three pairs of 
calciferous glands first investigated by Pereier are only attached to the oesophagus 
by their duct on either side ; they are otherwise quite separate from it ; every stage 
in fact appears to exist between mere dilatations of the oesophagus and diverticula 
of it ; in the most differentiated form of the calciferous glands, such as are found 
for example in Pontoscolex, the pouch communicates with the gut by a distinct duct 
whose walls are lined by cells different in character from those which line the gland ; 
it is commonly the case that this duct has a lining of ciliated cells while the cells of 
the gland itself are not ciliated. 

We shall return to a more detailed description of the minute structure of these 
glands later ; at present we are concerned with their distribution in the group. 
Calciferous glands in some form or other have been found in the following families ; 
those families in which they are nearly or quite universal are printed in capitals ; 
those in which a good many forms are without such glands are printed in clarendon ; 
finally, italics denote that the glands are rarely met with in the family. 

LUMBEICIDAE. Cryptodrilidae. 

GEOSCOLICIDAE. Perichaetidae. 




In all the remaining families of Oligochaeta there are no recognizable traces of 
these glands or of anything that can be compared to them ; they are absent in fact 
in the following: — 

Aeolosomatidae. Phreoryctidae. 

Naidomorpha. Lumbriculidae. 

Tubificidae. Moniligastridae. 

It will be seen that the presence or absence of glands appended to the oesophagus 
is broadly indicative of an aquatic or a terrestrial life ; it is by no means absolutely 
so ; for Moniligaster is, so far at any rate as is at present known, a purely 
terrestrial type, much more so than is the family Eudrilidae of which many members 
are largely aquatic in habit. 

There appears, however, to be a much closer relation between the presence of 
calciferous glands and comparatively large size and complex organization. This 
generalization is trammelled by fewer exceptions. Certain exceptions will at once 
occur, notably perhaps the smaller species of Benhamia which are among the smallest, 
if not the smallest, of earthworms ; it must, however, be borne in mind that this genus 
Benhamia is not typically composed of small-sized species ; on the contrary the 
average size is large, and some of the biggest earthworms are referable to the genus 
Benhamia. The smaller genera of a given family have frequently a reduced number 
of calciferous glands. This is best exemplified by the genera Ocnerodrilus and 
Pygm,aeodrilus, in which there are but a single pair of these glands ; so too in 
Kerria among the Acanthodrilidae and Gordiodrilus a genus of somewhat doubtful 
affinities ; the Geoscolecid Ilyogenia has but one pair of glands, and is withal 
a small member of its family. In fact it seems that everything points to a decided 
relation between size and absence or presence of calciferous glands. The relation 
may be more accurately stated thus: — 

Calciferous glands are absent or reduced in number in genera which are entirely 
composed of small-sized species. When the entire family contains only small-sized 
genera calciferous glands are completely absent. 

It has been asserted that the presence or absence of these glands is in accordance 
with the nature of the food of the worm ; the Limicolae of Clapaeede are, as their 
name denotes, mainly dwellers in soft mud or among weeds, and thus contrast with 
the boring earthworms. This view, however, does not seem to me to be consistent 
with the facts. A greater knowledge of the life habits of the Oligochaeta may perhaps 
reveal some relation between the two series of facts, but at present this knowledge 
does not exist. 



A further indication that there is a relation between the glands and the size of 
the worm where they occur is afforded by the simplification in structure which the 
glands show in the smaller Oligochaeta. In Ocnerodrilus the lumen of the single 
pair of glands is but slightly divided up by folds projecting into the lumen ; glands 
of this worm have been figured by myself and by ErSEN. In Gordiodrilus there 
is a similar simplicity in the minute anatomy of the single unpaired oesophageal 
pouch found in the worm. 

The calciferous glands are limited to the oesophagus ; nothing at all resembling 
them occurs in the intestinal region — except possibly the ' kidney-shaped glands ' of 
Megascolex coeruleus and of Typhoeus (see below). These latter however are not known 
to produce calcareous particles such as are secreted by the calciferous glands ; a 
diflTerence of function which this implies is not of course an objection to a serial 
homology ; and it is also true that both oesophagus and intestine are derivatives of 
the mesenteron. It seems however to be clear that the folded structure of the intestinal 
glands is not comparable to the folded structure of the calciferous glands, for the 
latter is simply an expression of the fact that the glands in question are diverticula 
of the already folded lining membrane of the oesophagus, while the epithelium of the 
intestine is not folded ; hence the complication of the intestinal glands is an independent 
formation. The calciferous glands vary in number from one to eight pairs ; they are 
nearly always in consecutive segments, but these segments are not fixed. Thus in 
Pontoscolex the glands are in segments vi.-viii., while in Benhainiia they occupy 
segments xiv. xv. xvi. as a rule. It seems that after the first segments of the oesophagus 
any segment is capable of developing calciferous glands. These organs are not always 
paired. In the Eudrilidae and in the genus Gordiodrilus there are median unpaired 
glands which have the structure of calciferous glands ; in Buchholzia there is a median 
dorsal gland. 

A remarkable fact about the unpaired glands of the Eudrilidae (as seen for example 
in the genera Eudrilus, Polytoreutus, and Hdiodrilus) is that they coexist with paired 
glands ; this does not however mean that the two glands are in the same segment, 
but they are so far independent that there is sometimes a break of a segment 
between the last unpaired pouch and the single pair of calciferous glands. These 
unpaired glands of the Eudrilidae were first discovered and described by myself in 
the genus Eudrilus (62). To these structures Michaelsen has given the name of 
' Chylustaschen.' He is of opinion that their function is difi'erent from that of the 
paired calciferous pouches. The ' Chylustaschen,' according to Michaelsen, are organs 
not of secretion but of absorption ; by their epithelium nutritive matters are supposed 
to be taken up from the blood. Here again a difference of function does not by any 


means necessarily imply a morphological difference ; and Michaels en is far from 
suggesting directly any such difference ; the fact however that he gives them a different 
name would perhaps lead to the inference that there was some difference of structure ; 
MiCHAELSEN does not in his descriptions of the minute structure of these glands (12) 
show any differences of importance from the paired calciferous glands of the Eudrilidae 
and of other Earthwoi-ms ; there is however a difference which I have noticed and 
described in Heliodrilus and Hyperiodrilus ; in these two genera, and in all probability 
in others, the periphery of the glands is occupied by a network of tubes whose lumina 
are intra-cellular ; the rest of the gland shows only inter-cellular lumina. This is 
however in my opinion not a matter of great importance ; the excessive folding of 
the epithelium of the pouch becomes at length so complex that the lumen inevitably 
becomes intra-cellular, the folds get to be smaller than the length of a single cell; 
this at least is my explanation of the matter. 

As to the supposed difference of function — it does not exist in every case — I have 
found that in Eudrilus the unpaired glands secrete calcareous particles entirely similar 
to the particles secreted by the paired gland, and so has Udb (4). There are so many 
instances among the Oligochaeta of organs being paired in one genus and unpaired in 
another, that a difference of this kind cannot be looked upon as of much importance ; 
the various parts of the generative apparatus are sometimes unpaired, though as a rule 
paired; the contrary occurs with the dorsal vessel which is as a rule unpaired, but 
occasionally paired ; even the brain is more or less completely divided into two halves. 
The actual fact as to the pau-ed or unpaired character of the glands is not therefore in 
my opinion a matter for serious consideration in deciding upon their homology. We 
may fairly regard them as structures which are serially homologous. 

These organs, the paired and the unpaired, consist of a diverticulum of the 
oesophagus, which is lined by epithelium continuous with that of the oesophagus ; 
as a rule this epithelium is rather different in appeai'ance from that whence it has 
been derived ; the minute structure of the glands has been studied in Lumbricus 
(CLAPAEiDE, 1), Pontoscolex (Peeriee, 5), Ocnerodrilus (Beddaed, 20, and Eisen, 1), 
Eudrilus (Beddaed, 62), Heliodrilus (Beddaed, 54), Alvania (Beddaed, 39), and in 
a few other types; Ocnerodrilus is the least complicated of these. In that genus the 
single pair of glands have a capacious lumen which is only moderately divided up by 
internal folds ; the cells lining the diverticula are ciliated, as are those of that part of 
the oesophagus at least which follows the apertures of the glands into it. Among the 
Eudrilidae the subdivision of the lumen by numerous anastomosing folds has reached its 
highest point ; and here as in many other species the epithelium is not ciliated ; in 
Pontoscolex the epithelium is remarkable for being a low flattened epithelium ; it is 

I 2 


elsewhere columnar. In the genus Fontoscolex and in most other genera each of the 
three or more pairs of glands opens by its own separate orifice into the oesophagus. 
Lwnibricus (including Allolobophora) is an exception to this statement ; in that genus, as 
was at first pointed out by Marshall and Hurst in their ' Practical Zoology,' the first 
pair of glands alone open into the lumen of the gut ; the two following glands of each 
side open into each other and into the lumen of the first pair ; there is thus but a single 
orifice into the gut for all three glands ; this difference has been emphasised by terming 
the first pair of glands oesophageal pouches in contradistinction to oesophageal glands. 
The descriptions of Marshall and Hurst have been confirmed by Kulagin (1). One 
other genus at any rate shows the same thing exactly ; I have found that in Microdrilus 
there are three pairs of glands of which the first pair only open into the gut; the 
remaining two of each side open into the first. A curiously analogous arrangement 
was found occasionally to exist in Polytoreutus ; Michaels SN discovered that in 
a specimen of Polytoreutus coerideus the last of the unpaired glands opened not 
directly into the oesophagus, but into the pouch in front. 

An example of the calciferous glands of this type, which are, it will be observed, 
not distinctive of any one family, is afforded by Octochaetus multiporus (see Plate V). 
In the young just ready to leave the cocoon the glands, which are most conveniently 
studied in that condition, owing to their small size, form merely a dilatation of the 
oesophagus. In a series of sections commencing at the head end of the worm the gland 
comes into view lying above the oesophagus and between the two dorsal vessels ; the 
latter immediately come to lie upon it. Still passing back, the gland gradually extends 
round the oesophagus opening into it at first in the dorsal middle line, afterwards 
along other lines. Ultimately the openings into the oesophagus become so frequent 
that the latter can be no longer distinguished from the suiTounding gland ; its somewhat 
more columnar epithelium passes gradually into the lower cubical epithelium of the 
gland. Only on the ventral surface can the oesophagus be said to exist apart from 
the gland for there are no ventral folds of the glandular membrane. The entire gland 
is covered with a layer of peritoneum of which the cells are not so long and pear-shaped 
as elsewhere. The lining epithelium is arranged in numerous folds some penetrating 
further towards the lumen than others and being generally thicker at their free edge 
owing to the fact that the cells are here more columnar. Peripherally the cells are 

The secretion produced by these glands is in the form of solid particles of calcareous 
matter; the actual form in which these concretions occur varies somewhat. In 
Zumbricus they are figured and described by Claparede as being 'small perfectly 
spherical bodies with a diameter of from a-6 micromm.'; in optical properties they 



Fig. 14. 

resemble, according to the same author, fat particles ; besides these round particles 
which constitute what is called by Glapaekdb ' Kalkmilch,' there are in the first pair of 
these glands rhomboedra, and larger concretions are figured but not specially described. 
These latter show radial striations. The large concretions as well as the small 
spherules are figured by Lankester. In Pontoscolex Peeeiee mentions only the fine 
spherules. In Microdrilus I have found rhomboedral crystals and also in Eudrilus. 
That these particles are formed of carbonate of lime seems to be proved by the fact 
that they dissolve in acetic acid, giving off carbonic acid. 

Besides the glands already described there ar^ other diverticula of the oesophagus 
which seem to be the equivalents of the calciferous glands, but are of different structure 
from those described and from each other. The Enchytraeidae have been known since 
the careful investigations of Vejdovsky (3) and Michaelsen (1-4) to possess glands 
appended to the anterior pai-t of the alimentary tract which have been termed by 
Michaelsen ' Chylustaschen,' and regarded as homologous with the similarly named 
structures in the Eudrilidae ; these occur in two varieties : — 

In Buchholzia there is a single dorsal pouch from which arises the dorsal blood- 
vessel ; this pouch communicates with the oeso- 
phagus by a wide orifice ; its interior is 
however not a simple lumen but is occupied 
by a mass of tubules or by a single much 
convoluted tubule, which has the appearance 
of a nephridial tubule, inasmuch as it has an 
intra-cellular lumen ; the interspaces between 
the tubules is occupied by blood spaces ; from 
these arises the dorsal vessel; this apparently 
single pottch is really composed of two closely 
applied pouches ; its lumen is not ciliated. 

In Fridericia leptodera there is a pair of 
diverticula with ciliated lining epithelium, 
and in Henlea ventriculosa two pairs of such 
diverticula ; the degree of folding of the 
lining epithelium varies somewhat. These structures are obviously related to the 
diverticula of earthworms, but it is not so certain whether the remarkable glands 
of Buchholzia are. 

The latter are to some extent paralleled in another genus, Gordiodrilus; in 
the four or five species belonging to this genus there opens into the oesophagus 
on the ventral side an ovoid pouch. This pouch is formed in the first place by 


(After Michaelsen.) 

1. Peritoneal layer. 2, 3. Tubules of gland 4. 
Dorsal vessel. 5. Lumen of intestine. 6. Ventral 
vessel. 7. Peritoneum. 8, 9. Muscular and epi- 
thelial layers of intestine, 

Vascular sinus of 



a diverticulum of the oesophagus ; round the epithelium, which is columnar, is a layer 
of tissue in which no cell boundaries could be detected ; abundant nuclei scattered 
throughout it appear to show that it is a mass of cells ; these cells are filled with oval 
or round particles like those which occur in the peritoneal cells in many parts of the 
body, particularly upon the nephridia. This mass of tissue is not separated from the 
epithelium by any membrane of any kind ; the epithelium rests directly upon it ; it 
is traversed in various directions by blood-vessels ; towards the blind end of the pouch 
this layer gets thinner ; at this end the lumen of the gland becomes intra-cellular ; the 

periphery of the gland in fact is 
Fig. 15. formed of a network of tubules 

exactly like nephridial tubes, but 
they are not ciliated, so far as 
I have been able to discover. It 
is a question as to whether the 
mass of cells covering the lining 
epithelium is peritoneal or is 
a specialization of the lining epi- 
thelium ; I believe the latter 
interpretation to be the right one ; 
as to the peripheral network of 
fine tubes I recur to the matter 
later. These glands or gland — for 
they may be paired or unpaired — 
are very diflferent in many points 
from the calciferous glands of 
earth-worms or from the pouches 
appended to the gut in the Enchytraeidae. They are perhaps best to be compared 
with a series of structures which characterise certain Eudrilidae. 

In the systematic part of this work I divide the Eudrilidae into two groups, in one 
of which there are calciferous glands and three unpaired pouches of the same structure ; 
in the remaining set the place of these is taken by a greater number of paired 
bodies, closely applied to the oesophagus. These glands were first described by myself 
in the genus Eudrlloides ; they have also been found in Stuhlmannia and a few 
other genera. They consist of a mass of tissue exactly like that which surrounds 
the lining epithelium of the pouch in Gordiodnlus ; this mass of tissue likewise 
surrounds a lumen which opens into the oesophagus (in Eudnloides at any rate) ; 
this lumen, however, is of very limited extent ; it does not nearly traverse the whole 


I. Lumen, 2. Nephriditun apparently in connexion with gland, 
3, 4, Septa, The blood-vessels and the vascular plexuses are shadeA, 


6f the glands, but ends very soon. The mass of cells surrounding it appears to 
be white in the dissected worm owing to the immense amount of minute spherical 
particles ; the cells are in places specialized, and get a certain resemblance to columnar 
epithelium ; this alteration takes place round a blood-vessel ; the effect produced is 
that of a gland tube cut across, the lumen of which is filled by blood ; the tissue 
in these modified tracts stains deeply with borax carmine ; elsewhere it is hardly 
at all stained by that re-agent ; this is of course due to the absence of secreted 
particles which are so abundant in the non-staining regions. I shall again recur 
to these glands in connexion with the blood glands of the Oligochaeta. Whatever 
the function of these glands may be, it does not appear to be exactly that of the 
calciferous glands of other Oligochaeta, for calcareous particles were never found in 
the lumen of the glands. They seem, however, to be in all probability serially 
homologous with calciferous glands ; I never found the two kinds of glands to 
co-exist in the same species ; and the glands now under consideration are clearly 
diverticula of the oesophagus as are the true calciferous glands. I believe that the 
mass of cells which surrounds the feebly developed epithelial lining is peritoneum, 
which has increased in amount pari passu with the gradual reduction of the glandular 
secreting surface, and has changed the function of the organ. Apart altogether from 
function, which is not now the question, these glands must from their position, and 
from the fact that they are diverticula of the oesophagus, be referred to the same 
category as the calciferous glands of other Oligochaeta. 

Intestine. — The oesophagus widens out to form what has been called the large 
intestine ; the exact segment at which this begins varies a good deal ; it is earlier or 
later as the case may be ; as a rule in the more simply organized forms it is more 
anterior than in the more complex species. The large intestine not only ditfers 
from the oesophagus by its greater calibre, but by various details of structure, not 
always, however, present. The most characteristic feature is the typhlosole ; the 
typhlosole is a median fold of the dorsal wall of the gut, which projects into its 
interior and diminishes the lumen while it increases the secretory surface. The 
degree to which the typhlosole is developed varies greatly ; in the lower Oligochaeta 
it is entirely absent ; it is also absent in a few of the more simple terricolous forms 
such as Ocnerodrilus. It has been often stated to be absent in Perichaetaj as far, 
however, as my experience goes, it is not absent in that genus, but very feebly 
developed, forming a slight fold which projects for but a short distance into the 
lumen of the gut. The complications in the typhlosole vary in different species ; 
in Octochaetus it is a fold which reaches nearly to the ventral surface of the 
gut and is trifid at the free edge ; an equally deep fold exists in Beinodrilus 



benhami, but it is not trifid. At a varying distance from the anus the typhlosole 
ceases; 1 have figured its abrupt ending, in the Acanthodrilid Odockaetus multi- 
porus. After this point the term rectum may be applied to the gut. 

The large intestine is generally a straight tube running vrithout a bend from 
commencement to termination ; in a few forms, however, it has a spiral arrangement ; 
this is seen in Plagiochaeta, and, according to Horst, in Pontoscolex corethrurus. 
In the course pf the tube it often exhibits a specialization into several regions; in 
Megascolex coeruleus this is especially evident — perhaps to some extent on account 
of the large size of the worm rendering the various regions more evident than 
they would be in a small species. In the first six segments the intestine is deeply 
pouched ; in the twenty-second segment the pouches became deeper still ; they extend 
to about the seventy-sixth segment; in several Oligochaeta the intestine commences 
in this way with a series of deepish pouches; for instance, in Urobenus brasiliensis 
(Benham, 3), where, however, it only extends from segment xvi. to xxv. In 
Megascolex coeruleus there are a series of glands appended to the intestine and lying 
beyond the pouches ; these are the ' kidney-shaped glands ' of my description of 
that Annelid ; they were also found by Bourne in his specimens ; the actual number 
of pairs of these glands appears to vary as both Bourne and I give different 
numbers. Their structure is quite simple ; they have the appearance of being formed 
by a much folded membrane ; it is a matter of some interest from a classificatory 
point of view that entirely similar glands occur in the genus Typhoeus belonging 
to the family Cryptodrilidae. Otherwise they are unknown in the Oligochaeta. The 
number of pairs is less in Typhoeus. Highly characteristic of the genus Perichaeta, 
but also, strange to say, found in the apparently remotely allied Urobenus, are a pair 
of caeca of the large intestine ; these occur nearly, if not absolutely, always in the 
twenty-sixth segment ; they are directed forwards, and occupy two or three segments. 
A few species of Perichaeta, for example P. sieboldi, have a mass of six or seven of 
these caeca arising close together and appearing to be formed by the branching of 
one caecum. In Urobenus the caeca are in the same segment. In the remarkable 
Cryptodrilid Millsonia there are more than thirty pairs of caeca precisely like those 
of Perichaeta. 

VI. Vascular System. 

The Oligochaeta agree with the Polychaeta in possessing a closed vascular 
system, i.e. one having no communication with the body-cavitj' (coelom). Unlike 
what is found in the Polychaeta there are no Oligochaeta known which have not 
a vascular system, though the complication of the vessels belonging to this system 


varies in the diflFerent groups ; they are more complex in the larger forms and on 
the whole less so in the smaller worms. 

§ I. Histology of blood-vessels. 

The blood-vessels of the Oligochaeta are tubular, rarely lacunar ; the larger 
vessels, some of which are contractile, have thicker or thinner walls, and give rise 
in the higher Oligochaeta to an extensive system of capillaries. In the larger 
vessels circular as well as longitudinal muscular fibres exist and the vessels are 
.lined by an epithelium and covered by the cells of the peritoneal investment. 
There is an increase in the elaboration of structure of the principal (dorsal) 
blood trunk, as we pass from the lower to the higher Oligochaeta. In Aeolosoma 
the dorsal vessel consists^ according to Vejdovsky, of no more than an exceedingly 
fine membrane, in which no structure, muscular or other, could be detected ; it is 
covered of course by the peritoneum. Very little more can be said of the Naids and 
Enchytraeidae ; there is simply a delicate membrane covered with peritoneum. In 
Ghaetogaster, however, which I reckon a Naid, it is possible to recognize faint longi- 
tudinal as well as transverse striae. Something similar is figured by Taubee (3, Tab. iii, 
fig. ii) for Rais elinguis, and Vejdovsky figures (24, Tab. iv, figs. 20, 21) and describes 
contractile cells (besides the peritoneal cells, with which they must not be confused) 
as constituting the walls of the dorsal vessel of Stylaria lacustris. One is inclined 
to suppose that the dorsal vessel of Aeolosoma, since it is contractile, will prove to 
possess muscular fibres. In the forms that have been hitherto considered there seems 
to be no definite lining of epithelium to the dorsal vessel. In the higher Oligochaeta, 
on the other hand, the epithelial lining consists often of very large and conspicuous 
cells. In Phreoryctes, for example, I have figured (18) a transverse and longitudinal 
section through one of the perioesophageal vessels (which have the same structure as 
the dorsal vessel) which show (Figs. 8, 9 of plate) the cubical cells lining the lumen 
of the vessel ; the strong circular and longitudinal muscular layers of the vessel are 
also shown. Vbjdovsky's figures (24, Taf. xiv, figs. 10-13) show the lining epithelium 
and at any rate a circular layer of muscles in Griodrilus lacuum; while Pbrriee 
has recorded a similar structure for Urochaeta. Naturally the circular ^ layer is much 
more important than the longitudinal,, and it has therefore been often the only one 
figured. The blood itself is usually red; yellower in Naids, and colourless in 
Aeolosoma^. The colour, as was first shown by Lankestek, is due to haemoglobin. 

' Both circular and longitudinal fibres are described in the text (p. ii8). 
" Lankester says that it is pink in Ae. quatemarium. 



The blood contains minute corpuscles suspended in it which are little more than the 
nuclei of the lining cells. 

The blood-vessels of the Oligochaeta are often provided with valves. These, 
however, are limited to the dorsal vessel and to the hearts. They do not appear 
to occur in the lower Oligochaeta — unless indeed the cardiac body of the Enchy- 
traeidae is derived from a fused series of valves such as occurs in Vertebrates — 
with the exception of Phreoryctes, where they have been described by Leydig (6) 
and by Timm in the dorsal vessel. In earthworms they are very general, 
probably universal. I have myself found them wherever I have looked for them. 
In the dorsal vessel they occur just where the vessel traverses the septa ; in 
the hearts they may occur all along (cf. e.g. Peerier, 5, PL XV, fig. 29) from its 
point of origin from the dorsal vessel up to the opening into the ventral vessel. 
The valves are essentially proliferations of the lining membrane of the blood- 
vessels, the cells forming them being large and granular. The blood gland of 
PhreodHlus may very possibly be regarded as homologous with a fused series of 
valves, unless the interpretation which I have suggested in describing it be accepted 
in preference. 

§ a. Main trunks. 

The principal longitudinal trunks in the Oligochaeta are five, all of which do 
not exist in the lower forms ; they will be now considered seriatim. 

The Dorsal vessel. — This, the principal of the longitudinal trunks, is present in 
all Oligochaeta and is invariably contractile ; it is not, however, always equally 
well developed; for in some forms (e.g. Aeolosoma) it is deficient posteriorly while 
in others (e.g. Chaetogaster cristallinus, Rhinodrilus ecuadoriensis, qq. v.) it terminates 
a little way before the anterior end of the body. In nearly all those forms where 
the dorsal vessel is fully developed, which include the great majority, it passes from 
one end of the body to the other upon the dorsal surface of the alimentary canal ; in 
Branchiura, however, and in Dero this is only true of the anterior section of the 
tube; posteriorly it comes to lie on the ventral surface of the body, near to the 
ventral vessel. The dorsal vessel is commonly separated by a little distance from 
the actual walls of the canal, but along the intestine it closely invests the gut ; it 
is not, however, in the higher Oligochaeta, covered by the intestinal peritoneum, 
but has a layer of peritoneal cells to itself. 

The way in which the dorsal vessel terminates anteriorly varies in different 
Oligochaeta. As regards the higher Oligochaeta we have two elaborate memoirs 


dealing respectively with Fontoscolex and Megascolex by Perrier and Bourne. The 
following account is deduced from these two memoirs. In Megascolex the dorsal 
vessel bifurcates anteriorly in the first segment of the body ; each of the two branches 
into which it divides is the equivalent of the succeeding dorso-tegumentary vessels, 
and like them forms a peripheral vascular network which only communicates 
indirectly with the ventral vessel. In Fontoscolex (Perrier) the dorsal vessel extends 
as far forward as the brain; there it bifurcates and joins the branches formed by 
the bifurcation of the ventral vessel, thus forming ' a vascular collar in front of 
the nervous collar.' In these genera, in Lumbricus, and in fact in all the higher 
Oligochaeta the dorsal vessel gets to be very much less in calibre anteriorly and 
communicates indirectly with the ventral longitudinal trunks ; in these Oligochaeta the 
vascular system as a whole is, as will be pointed out, excessively complicated as 
compared with the lower forms ; the fact therefore that in the lower forms the dorsal 
and ventral vessels communicate directly by vascular arches similar to those which 
occur in succeeding segments is not to be looked upon as a difference of importance, 
but merely as a result of the more highly-developed vascular system of the former. 
In the Tubificidae, for example, there is in the first segment of the body a vascular 
arch connecting the dorsal and the ventral vessels directly and quite similar to the 
following arches which put these vessels into communication in ensuing segments ; 
but these facts will be fully gone into in a subsequent section. Posteriorly the dorsal 
vessel appears gradually to fade away in some worms ; in others, as in Megascolex, 
for example, it terminates abruptly just after giving off the last pair of dorso- 
tegumentary trunks. 

A remarkable condition of the dorsal vessel was first described by myself in 
Megascolex coeruleus j in this worm the anterior part of the tube is partially divided 
into two halves which reunite at the septa : a large number of earthworms are now 
known which exhibit this peculiarity. It is, moreover, not a question of systematic 
position ; the most diverse families show this condition ; it has been found, for 
example, in OctocJiaetus viultiporus, Acanthodrilus novae-zelandiae, Microchaeta rappi, 
Teleudrilus ragazzii, etc. Three degrees of the division of the dorsal vessel exist; 
in Megascolex, for instance, the vessel is only double in the anterior part of the body ; 
in other types the dorsal vessel is double from end to end of the body ; here again 
there is a difference ; in Octochaetus multiporus and in Acanthodrilus annectens the 
vessel is completely double; there are two distinctly separate tubes running side 
by side on the dorsal surface of the gut ; in Acanthodrilus novae- zdandiae there 
are two such tubes in the middle region of each segment, but at the septa the two 
tubes fuse. The interest attaching to these facts lies mainly in that they seem to 

K 3 



Fig. 16. 

indicate the persistence of an embryonic condition ; for Vejdovsky showed that 
in Criodrilus the dorsal vessel was developed from two independent tubes, and he 

has recently extended this discovery to species of Lum- 
bricus (s. 1.). As a general rule the dorsal vessel of the 
Oligochaeta is of much the same diameter throughout ; 
but in a few forms it is locally dilated ; this is the case, 
for instance, with Microchaeta where there is heart-like 
swelling in the ninth segment. In various Enchytraeids 
too there is a dilatation of the dorsal vessel just where 
it arises from the peri-intestinal sinus. 

The Supra-intestinal vessel. — In addition to the dorsal 
vessel a good many Oligochaeta possess another vessel, also 
running along the dorsal side of the gut, which has been 
termed the ' Supra-intestinal ' vessel. It is not long since 
the presence or absence of this vessel would have been held 
to be at least partly distinctive of the two groups instituted 
by Claparede, the ' Limicolae' and the ' Terricolae/ Until 
the researches of Stolc the supra-intestinal vessel was 
considered to be confined to the Terricolae, althouffh not 
occurring in all the genera of that group. It is, however, 
now known to exist in a number of Tubificidae as . well 
as in the aberrant genus Phreodrilus. The supra-intestinal 
trunk is limited to the oesophageal region, where it takes 
the part with reference to the intestinal circulation that is 
elsewhere played by the dorsal vessel. In the oesophageal 
part of the alimentary canal the dorsal vessel comes to 
be some way removed from the gut, and it gives rise to 
a series of stout lateral branches which embrace the gut 
and join the ventral vessel without being connected in 
any way with the walls of the gut. 

It has been stated that the supra-intestinal trunk 
passes back along the intestine as the typhlosolar trunk ; 
but this appears to be very doubtful; the careful re- 
searches of BouEyE (4) into the circulation of Megascolex 
coeruleus do not support the existence of such a vessel ; 
and, considering its relations to the gut and to the body -wall in the oesophageal 
segments, it does not seem likely on a priori grounds that it does exist in the intestinal 
region, where indeed it would seem to be de trop. 



(After Stole.) 

I-IX. Segments. i-6. Lateral 
hearts. 7. Ventral vessel. 8. Pha- 
rynx. 9-14. Vessels supplying in- 
testinal network. 15. Sub-intestinal 
vessel. 16-19. Vessels supplying in- 
testinal network. 18, 21. Intestinal 
hearts. 22. Supra intestinal vessel. 
23. Dorsal vessel. 


This blood-vessel resembles the dorsal vessel in being also directly connected with 
the ventral vessel; among the Tubificidae such connexions occur, and in Branchiura 
there is a pair of hearts of this description coexisting in the same segment with a pair 
of hearts ai'ising from the dorsal vessel ; in the earthworms it is frequently the case, 
as I mention more in detail further on, that the same pair of hearts arise both from 
the supra-intestinal and the dorsal vessel. Another point in which this vessel resembles 
the dorsal is that it is occasionally double ; in Megascolex coeruleus this is the case, 
as has been described by both Bourne and myself. The supra-intestinal vessel is by 
no means present in all of the higher Oligochaeta ; it is present in most of the genera 
belonging to the two families Megascolicidae and Eudrilidae. It may perhaps be 
a question to which of the two dorsally-placed trunks the single vessel of, for example, 
the Naidomorpha corresponds. I have suggested that the intimate relations of the 
supra-intestinal to the alimentary canal indicate that the posterior part of the dorsal 
vessel is its homologue and that the anterior part of the dorsal vessel is a new structure 
in the higher Oligochaeta — the relations between the two being somewhat analogous 
to those which subsist between the vena cava posterior and the posterior cardinals in 
the Vertebrata. This view seems to be supported by the relations obtaining in the genus 
Phreodrilus. In that worm there are in the anterior region of the body two trunks 
upon the dorsal side of the alimentary tract which I homologize respectively with the 
dorsal and supra-intestinal of other Oligochaeta. These two vessels can be distinguished, 
not only by their position but also by their minute structure ; the vessel which is closest 
to the wall of the gut, lying in fact upon it, has thin walls and is full of blood after 
death ; the other has thicker walls and is not so full of blood after death ; moreover 
the latter has not the coating of chloragogen cells which occur in the former ; it is 
coated with flattened cells without pigment ; this vessel, which I believe is the dorsal 
vessel of the higher Oligochaeta, terminates at about the fifteenth segment ; behind 
this point there is only the supra-intestinal. The question, therefore, arises whether 
the vessel, which I have termed dorsal, is really the equivalent of the similarly 
named vessel in the Naids, etc. 

The Ventral vessel. — This vessel is present along the whole body in all Oligochaeta. 
It is invariably a single tube never showing any signs of duplication as does the 
dorsal vessel. The ventral vessel also differs from the dorsal vessel in that it is never 
contractile^. Another fact of importance is that it is the first blood-vessel to be 
developed ; this is at any rate the case with Rhynchelmis and Lwmhncus (several 
species) in which alone the origin of the vessel has been carefully studied. The ventral 

1 Phreoryctes is said to be an exception to the statement. 


vessel in its earliest stage is simply a sKght thickening of the splanchnopleure ; for 
some time it remains solid and only subsequently becomes hollow. In the view of 
some writers, for example Wilson, the ventral vessel is from the first hollow and its 
cavity is the remains of the segmentation cavity. 

In many of the lower Oligochaeta the ventral vessel terminates^ in front by 
bifurcating and ultimately joining the dorsal vessel by a loop round the gullet ; in 
Chaetogaster however it ends in a solid cellular cord which bends upward and ends 
freely in the neighbourhood of the cerebral ganglia. In the higher Oligochaeta the 
ventral vessel branches anteriorly and communicates with the dorsal vessel only through 
a capillary system. In two Tubificidae the ventral vessel terminates anteriorly in an 
altogether peculiar fashion. In Lophochaeta the ventral vessel appears to end in the 
eighth segment in a pair of hearts which communicate with the supra-intestinal trunk ; 
but from the angle formed by the bifurcation an extremely slender vessel passes forward 
and after receiving three branches from the sub-intestinal vessel joins that vessel in 
the seventh segment ; the single vessel formed by the fusion of the two terminates 
anteriorly in the usual way. In Bothrioneuron the narrow continuation of the 
ventral vessel commences in segment vii., and without receiving any branches from 
the sub-intestinal joins it in segment vi. These facts were described and figured by 
Stolc (3), whose figures are here reproduced (woodcuts, figs. i6, i8). 

The Sub -intestinal vessel. — I apply this term to the usually paired vessels 
called by Perkier and Bourne ' Intestino-tegumentary ' vessels, and by Benha.m 
' lateral longitudinal ' vessels. They are first seen in the family Tubificidae ; in 
Lophochaeta and in Bothrioneuron where their relation to the ventral vessel has 
already been referred to. In Lophochaeta the ventral vessel divides in the seventh 
segment into two trunks, one lying above the other : the uppermost of the two is 
closely applied to the ventral wall of the alimentary canal and appears to be intimately 
concerned with its blood supply ; it has much the same relations to it below that the 
supra-intestinal vessel has above ; the same vessel occurs in Bothrioneuron, but it 
begins earlier, in the sixth segment. In the higher Oligochaeta this vessel, or at any 
rate the vessel which I regard as its equivalent, is invariably double — with the possible 
exception of the Lumbricidae ; Howes in his ' Biological Atlas ' has figured in that 
worm a single sub-intestinal vessel which Jackson (' Forms of Animal Life ') considers 
to be non-existent, but to have been mistaken for the impression of the attachment 
of the mesentery. In many other earthworms there are a pair of these vessels which 
arise from the intestinal plexus ^ and run for a short distance closely attached to the 

1 In Benhamia schlegelii Horst iigures and describes these vessels as originating from the dorsal trunk 
just in front of the antepenultimate pair of hearts. 



Fig. 17- 

ventral surface of the gut; later however they leave the walls of the intestine and 

run freely suspended in the body cavity ; they end anteriorly in a network of capillaries. 

These vessels occur in Megascolex (Bodkne), Pontoscolex (Pekeiee) (woodcuts, figs. 

20, 21), Eudrilws (Bbddaed), and other forms. In Eudnlus 

the sub-intestinal vessel is partly double and partly single 

like the dorsal vessel of other earthworms ; it bifurcates 

in the neighbourhood of the calciferous pouches and is 

single between them. In another Eudrilid, Libyodrilus, 

I have traced the course of these vessels for a little 

distance ; they are double all the way and give off two 

pairs of branches in each of the segments through which 

I followed them ; one pair of branches go to the septum, 

the other to a muscle suspending the intestine ; numerous 

minute twigs connect them with the intestinal plexus. 

In Megascolex coeruleus the vessel in question only extends 

back to the thirteenth segment ; after this there is a small 

vessel in each segment running from intestine to septum 

which is its equivalent. 

The Sub-nervlan vessel. This is absent in the lower 
Oligochaeta and is by no means always present in the 
earthworms. It exists in Lumbncv^, Perichaeta, and 
some other genera, and lies, as its name implies, beneath 
the nerve cord ; it is even sometimes partially imbedded 
in the ventral body -wall. The presence or absence of this 
vessel served Clapaeede with one of his characters for 
separating the Limicolae from the Terricolae. 



(After Benham.) 

§ 3. Commissural vessels. 
The dorsal vessel in all Oligochaeta is connected with 
the ventral vessel by circular trunks which run round 
the alimentary canal. The simplest and most primitive 
arrangement of the vessels appears to be that which 
characterises the families Tubificidae, Phreoryctidae, and 
Lumbriculidae. In these worms there are at least one pair 

of such vessels in each segment of the body — the regular metamerism of the vascular 
system being thus very apparent. In the anterior region of the body one or more of the 
commissural vessels are specialized and become contractile, being then usually termed 

i, 4, 6. Perivisceral vessels. 2, 3, 7. 
Dorsal vessel. 5. Spermatheca. 8. 
Sperm-sacs. 9. Intestino-tegumen- 
tary vessels. 10. Ovary. 11, 12. Dorso- 
and ventro-teguinentary trunks. 



Fig. 18. 

' hearts.' In Tuhifex rivulonom, for example, the commissural vessels of segment viii 
are greatly dilated and contractile. In Limnodrilus and Clitelli there are two such 
contractile circles, but they are not always larger than the unmodified perivisceral 
trunks. In other Tubificids a larger number of the anterior periviscerals are modified 
into contractile hearts. Further details will be found under description of Tubificidae. 

In the Naidomorpha the direct connexions of the dorsal 
and ventral vessels only exist in a few of the anterior 
segments ; this also is the case with the Enchytraeidae 
and the Aphaneura. The smallest number of commissural 
vessels occurs in Pristina equiseta, where Bourne only 
found one. The fact that in the newly-formed bud of 
Uncinais littoralis there are commissural vessels in all 
the segments, most of which become afterwards lost, is an 
indication that the disappearance of all but the few anterior 
commissural vessels in the Naidomorpha and Enchytraeidae 
is due rather to degeneration than to the retention of any 
primitive character. In some Tubificids there is another 
class of commissural vessel in the anterior segments of 
the body ; one or more of the hearts instead of forming the 
dorsal and the ventral vessels unite the supra-intestinal 
with the ventral. The name ' intestinal hearts,' introduced 
by Pekriee for the corresponding structures in earthworms, 
may be applied to these. It is difficult to determine whether 
or not these trunks are serially homologous with the 
remaining hearts and with the peri-intestinal vessels of 
the segments in the posterior region of the body. 

In Bothrioneuron, Lophochaeta, and Phreodrilus, the 
intestinal hearts occur in segments posterior to those which 
contain the dorsal hearts ; but in Branchiura the eighth 
segment contains two perivisceral vessels, of which one 
seems to be the equivalent of the intestinal, the other of 
the dorsal heart. 

In the higher Oligochaeta the intestinal hearts are 
connected in many cases with both of the dorsally running blood-vessels ; so that one 
connection may be regarded as secondary. In these Oligochaeta, in the earthworms 
in fact, the commissural vessels are confined to the anterior segments of the body; 
but this specialization is not due, as it may be in the Naidomorpha, to degeneration ; 


(After Stole.) 

I. Ventral vessel. 2-6. Vessels 
joining intestinal network. 7. Slen- 
der posterior part of ventral vessel 
just after it joins. 12. Sub-intestinal 
vesseL 9, 13. Intestinal hearts. 15, 
16. Conunissural vessels. 17. Dorsal 


it appears to be due to the fact that the commissural vessels of the posterior segments 
have given rise to the integumental blood-plexus, and have become largely lost in 
the process ; there are, as I shall point out later, the beginnings of this plexus to 
be seen in the Tubificidae ; it attains to its highest development in the Oligochaeta 
terricola ; but in the emhvy oLumbricus, as Vejdovsky has shown (see woodcut, fig. 19), 
each segment has a pair of commissural vessels. The number and position of the 
hearts in the earthworms varies very considerably; it is rare, however, to find them 
extending behind the thirteenth segment. As a rule the last three or four, often two 
of these, are connected either with both the dorsal and supra-intestinal vessels or with 
the latter alone ; this kind of connexion, however, does not occur in the Lumbricidae, 
and apparently not in the Geoscolicidae ; it occurs only in the Megascolicidae and the 
Eudrilidae among the Megadrili ; the Moniligastridae agree in this character with the 
Geoscolicidae and the Lumbricidae. As a rule the intestinal hearts are much more 
dilated than the others, and they show in a more pronounced fashion the moniliform 
character which these organs often exhibit ; this is really due to the presence of valves 
along the course of the commissural vessels, which allow the blood to flow • in one 
direction only. 

§ 4. Peripheral circulation. 

In addition to the longitudinal trunks and to the commissural vessels which 
unite them, there is in all Oligochaeta a system of smaller vessels, which form 
plexuses and may be termed capillaries; it is convenient to divide these capillary 
networks into two series, the integumental and the intestinal. The former includes 
all the capillaries which ramify in the thickness of the body-wall, the septa, and 
in fact all the organs of the body except the alimentary system; the reasonableness 
of making this apparently artificial division is shown by the fact that in the lowest 
Oligochaeta there is no capillary system except that which supplies the intestinal 
walls, and that the two systems are quite distinct, being only indirectly connected 
in the higher forms. In the Aphaneura, the Enchytraeidae, and the Naidomorpha, 
the intestinal system of capillaries is the only one that is present. In the two 
former groups the dorsal vessel loses itself in this plexus ; it seems a little doubtful 
whether in the adults of any of these worms there really exists, as has beeii 
described, a blood-holding space surrounding the gut ; when the capillaries are 
gorged with blood, there would naturally be a tendency to the obliteration of the 
boundaries of the meshes of the network which would of course produce the 
impression of a continuous sinus. As to Aeolosoma, however, the figures of 
Vejdovsky show plainly that there is a plexus and not a sinus ; in the Enchytraeidae 



a sinus has been generally asserted to exist; but Michaelsen figures in Stercutus, 
for example, appearances which are much more in accord with the assumption that 
there is really a network : he speaks, however, in this case of a ' Darmblutsinus.' 
The same word is used in his revision of the family, and therefore containing his 
latest opinions on the matter, with which also Vejdovsey agrees. The existence 
of this sinus, if real, is of interest in connexion with the perienteric sinus in certain 
Chaetopods (for example Fahricia) ; the fact that in Aeolosoinia and the Enchytraeidae 
the dorsal vessel arises from this sinus becomes additionally interesting from Salensky's 
discovery that in Terebella, etc., the embryo has a dorsal vessel similarly connected 
with, indeed arising from, a perienteric blood-sinus. As to the Oligochaeta, embryology 
does not seem to indicate that this sinus is primitive ; in Rhynchel'mis Vejdovsky 
describes the dorsal vessel as arising from the perienteric plexus (not sinus). It 
will be noted, however, that the limited extent of the dotsal vessel and its origin 
out of the perienteric vessels is an embryonic character in the Aphaneura and 
Enchytraeidae. Whatever may be the case with the lowest Oligochaeta^ it is certain 
that in the Naidomoi'pha and in all the groups above them, there is not a plexus 
but a network of capillaries in the intestinal walls ; this has been figured by Stolc 
and others in the Naidomorpha, and by the same and also by others in the 
Tubificidae ; in these worms the network is fed from the dorsal vessel, and the 
blood returns into the ventral vessel ; the network at any rate has connections 
with both vessels. Among earthworms Sparganophilus has been said to possess 
a sinus by Benham ; I confess to being unwilling to accept this statement, for 
the reasons already stated, i. e. the difficulty of proving that the supposed sinus 
is not really a plexus with largely obliterated boundaries. The intestinal network 
of the higher Oligochaeta is described below in the genus Megascolex, in which it 
has been carefully studied by Bourne. 

The integumental blood plexus shows a gradual increase in complexity as we 
pass from the higher to the lower forms. It is, as already stated, entirely absent 
in the Enchytraeidae and the Aphaneura. The first traces are to be found in 
Limnodrilus, where the lateral commissural vessels give off (in Limnodrilus hoff- 
meisteri at any rate) a bunch of small vessels which seem to end caecally in the 
skin. In other Tubificidae the integumental system is much more developed ; in 
Ilyodrilus, for example, where the circulation has been carefully studied by SroLC, there 
is a complex network in the integument ; it appears from his figures (the text being 
in the Bohemian language is inaccessible to me) that in each segment of the body 
behind the first a special pair of vessels is given off from the dorsal vessel which 
supply the plexus, and that branches of the commissural vessels receive the blood from 



Fig. 19. 

the plexus and convey it to the ventral vessel. The integrity of the commissural vessels 

is, however, preserved ; in the higher Oligoehaeta the integrity of these vessels is lost 

(except in the embryo, see woodcut) ; they become largely dissolved into the network 

of capillaries. In the lower Oligoehaeta it is only the actual body- wall which sometimes 

possesses a capillary network, none of the other organs of the body (as a general rule) 

have such ; the sperm-sacs, for instance, are fed by specially elongated perienteric loops 

belonging to the segments in which they occur. The 

nephridia in Rhynchdniis are peculiar among the lower 

Oligoehaeta, in that they have similar loops attached to 

them and following their windings ; and in Tuhifex 

Nasse has described vessels in the nerve-cord. With these 

exceptions there is no peripheral vascular system other 

than the integumental network until we reach the true 

earthworms ; and here it is not always at the same pitch 

of development ; in the smaller forms, such as Ocnerodrilus 

the nephridia seem to be quite unprovided with vascular 


Epidermal capillaries. — It has been known for a long 
time that in the clitellar region the terminal branches of 
the vascular system push their way in among the cells, 
forming loops, which, however, stop shoi-t some little 
way below the surface. These capillaries are figured by 
Claparede and by others. As to the vascularity of the 
rest of the epidermis, it was, I believe, first pointed 
out by myself that this was the case with Megascolex 
coeruleus (2) and some species of Perichaeta (3). The 
extension of blood-capiEaries into the epidermis of Grio- 
drilus has been figured by KosA (12), and quite recently 
it has been discovered by Lenhossek (1) that the epidermis 
of Lumhricus — not only the clitellum — is also vascular. 
Even among the more delicate aquatic species this con- 
dition, very likely universal in the earthworms, is not 

unknown. Lininodnlus is furnished with apparently caecal vessels, which terminate 
between the epidermic cells. It has been stated by the Sabasins that in Perichaeta 
the capillaries actually reach the exterior, and open there ; but this appears on the face 
of it to be unlikely, and at any rate needs further proof. The statement was made 
incidentally in connexion with an alleged similar opening of blood-capillaries on to 

L 2 

(After Vejdovsky.) 

I. Aperture of 'head kidney.' 2. 
Brain. 3. Septal glands. 4, 5. Dorsal 
6. Intestine. 7. Ventral 



Fig. 20. 

the exterior in the Gymnophionae. The physiological meaning of ' intra-epithelial blood- 
capillaries,' as Lankester has termed these structures in the leech, whei-e he was the 
first to discover their existence, seems to be clear. One of the characteristics of the 
Oligochaeta is the rare occurrence of branchiae ; Dero, Chaetobranchus, Branchiura, 
Hesperodrilus branchiatus and the questionable Alvna are the only real (or reputed) 
Oligochaeta which form branchiae. The absence of these special organs is compensated 
for by the utilization of the entire skin as a branchial organ. The efiiciency of the 
skin as a branchial organ must be largely increased by the penetration of the 
capillaries into the outermost layers. 

Bourne was led by his researches into the vascular system of Megascolex coeruleus 

to a reasonable view of the course of the blood 
in that species which will probably, as he thinks, 
prove to be more widely characteristic. 

Excepting at the head end of the body, where 
the vascular system like other organs shows 
a cephaUzation, it is metamerically arranged. 
Moreover the fact that an earthworm . can live 
after a considerable part of its body has been 
shorn away indicates that each segment is to some 
extent independent in its circulatory mechanism. 

There is no doubt — all observers agree upon 
that point, which can be easily seen in the living 
worm— that the blood flows forwards in the 
dorsal vessel. In the cephalized region of the 
body the blood passes down into the hearts ; 
these contract from above downwards, just as 
the dorsal vessel contracts from behind forwards. 
In the intestinal region of the worm there are two sets of vessels connected with 
the dorsal vessel— the dorso-integumentary and the dorso-intestinal. It appears that 
blood flows from the latter into the dorsal vessel, and out of the dorsal vessel into 
the former. This view is not that universally accepted. In the anterior region of 
the body of course, the supra-intestinal vessel is supplied from the same source as 
is the dorsal vessel posteriorly. The ventral vessel of Megascolex communicates with 
the dorsal vessel anteriorly and directly by the hearts ; posteriorly it communicates 
with the same tube, indirectly by the ventro-tegumentary vessels. According to Bourne 
blood reaches the ventral vessel by the hearts ; of the truth of this there can be 
hardly any doubt ; he considers^ however, that the ventro-integumentary vessels take 



(After Perrier.) 

V. d. Dorsal vessel, v. s. Ventral, v. i. Sab- 
neural vessel. h. Intestinal plexus, v. t., b. 
Tegumental system of vessels, t. TypMosole. 



blood away from the ventral vessel to the parietes. 
There are thus two sets of vessels which convey blood 
to the integument — the dorso-integumentary and the 
ventro-integumentary. From the integument a series of 
intestino-integumentary vessels again returns the blood 
to the intestinal network, whence, as before stated, it 
reaches the dorsal vessel. 

Fig. 21. 

VII. Blood-Glands. 

Connected with the vascular system there are in 
several OHgochaeta structures which seem to be of 
a glandular nature. Such structures occur in the En- 
chytraeidae, the Lumbriculidae, the genus Phreodrilus, 
the Perichaetidae, and perhaps the Eudrilidae. 

It will be convenient to commence with a descrip- 
tion of these various organs in the groups mentioned, 
before discussing the morphology and the functions 
which they may perform. 

In the Enchytraeidae the dorsal vessel sometimes 
contains a cellular rod which has been named the 
'Herzkorper' by Michaelsen (woodcut, fig. 22). It 
was described first in the genus Mesenchytraeus 
(Michaelsen), where it consists of a solid rod of 
cells attached to the ventral side of the dorsal blood- 
vessel, and extending along its whole length; 'in M. 
mirabilis and M. primaevus it is thick with irregular, 
often strong swellings multicellular in section. In 
M. falciformis, M. beumeri, and M. flavidus, it is 
thinner, quite smooth, with only feeble swellings, and 
in section shows only a few cells.' The organ is figured 
by Michaelsen. In a later paper the same organ was 
described in Stercutus. It has been justly compared by 
him to the cardiac body of certain Polychaetes, such 
as Pectinaria belgica, and a few others. As to its 
function it was suggested by Michaelsen that it served 
to ease the contractions of the dorsal vessel ; he pointed 

— v.t 


OF vasculae system. 

(After Perrier.) 
V. d., V. m. Dorsal vessel. c. I. Con- 
tractile heart, c. i. Intestinal heart, -y. i 
Supra^intestinal vessel. v. I. Lateral 
vessel. V. 8. Ventral vessel, v. t. Peri-intes- 
tinal vessel, t.r., i.t. Smaller branches 
of ventral vessel. 



Fig. 22. 



(After Miohaelsen.) 

out that the flow of the blood forwards would be facilitated if the tube were 
completely closed posteriorly during contraction; the presence of this cardiac body 
would help to fill up the tube, and allow the lumen to be entirely obliterated without 

reaching the maximum degree of contractility of the dorsal 

The genus Lumbriculus and other Lumbriculidae have 
a series of caecal diverticula (woodcut, fig. 23) of the dorsal 
vessel, which are clothed with large chloragogen cells, and 
are contractile ; these were formerly mistaken for caeca of 
the gut itself, but there is no doubt that this was an error ; 
they serve as temporary reservoirs of the blood, which is 
presumably subjected during its sojourn in them to the 
action of the peritoneal cells which envelop them. 
Nothing of the kind exists in any worm that is not 
a Lumbriculid, though they are not found in the Lumbriculid genus Stylodrilus. 

A third kind of blood glands (woodcut, fig. 24) are to be met with in various species 
of the genus Perichaeta. They were originally described by Peebier (3), who thought 
them to be of the nature of salivary glands ; they occur in Perichaeta houlleti, for 
example, at the sides of the oesophagus and show a distinctly metameric arrangement, 

being grouped in accordance with the segments. These 
glands consist of a number of spherical acini, which have 
a solid appearance, and are composed of small rounded 
cells ; they have, however, as I (43) showed, no connexion 
whatever with the gut, and are simply dilatations upon 
the blood-vessels comparable to those dilatations which 
occur along the course of the nephridial tubes, than which, 
however, they are considerably larger ; a blood-vessel can 
be traced into them and out of them. They are covered 
externally by masses of pigment-holding peritoneal cells, 
and form with these fairly compact glandular masses. 
Something of the same kind appears to exist in at any 
rate one species of Acanthodrilus, A. rosae. 

In the aquatic Oligochaet genus Phreodrilus another 
variety of organ of possibly a similar physiological nature 
occurs ; in the tweKth and thirteenth segments of this worm is a coiled tube which 
puts the dorsal and the ventral vessel into communication ; it is so much coiled that 
I have not been able to ascertain its exact shape ; the interior of this vessel, which 



{After Claparfede.) 

I. Ventral blood-vessel. 2. Dorsal 
blood-vessel. 3-5. Branches of last 
witb caecal twigs. 



Fig. 24. 

is perhaps the homologue of the perienteric vessel of its segment, is largely filled by 
a mass of big vesicular cells containing granules. They are so numerous as to nearly 
occlude the lumen of the blood-vessel. 

The last variety of these organs are possibly the modified calciferous glands of 
certain genera of Eudrilidae, for instance StuMmannia. I have, however, described 
these in full under the description of the calciferous glands, and need not recur to 
the matter here. 

The nature of these various glands now requires consideration. There can, I think, 
be no morphological comparison between the organs in Perichaeta and those of the 
Lumbriculidae, on the one hand, with the other varieties of blood-glands ; nor is 
there any relation except perhaps a physio- 
logical one between the blood-glands of the 
two genera of worms mentioned; the other 
organs do appear to me to be probably mor- 
phologically connected. There is no doubt, 
I imagine, that the peculiar glands of certain 
Eudrilidae which I have described above 
are really the equivalents of the calciferous 
glands ; their lumen, although much reduced, 
communicates with that of the oesophagus ; 
and the segments which they occupy are those 
in which calciferous glands, when present, commonly lie. At the same time it appears 
to be at least probable that their function is a difierent one from that possessed by 
the calciferous glands ; I could find no evidence of a secretion of calcareous particles ; 
this merely negative evidence is not perhaps very strong, since I have often failed 
to discover any such particles in the lumina of glands which are undoubtedly calciferous 
glands. It is the structure which leads to the inference that the glands now under 
consideration are not functionally calciferous glands; this is most clearly marked in 
EudrUoides, and the structure in that genus offers a clue to what is the real function 
of the glands in question. The peculiar cells which make up the mass of the glands 
become altered in EudrUoides towards the periphery of the glands ; it gets to have 
a distinctly columnar arrangement ; but the columns of cells are disposed round a 
lumen which is filled with blood and which is a blood-vessel traceable into connexion 
with the other blood-vessels of the gland and the surrounding organs ; the structure 
seems to be irreconcileable with any other theory than that the glands in question 
have some secreting function in relation to the blood or eliminate effete matters 
from the blood; we have in fact a gland originally performing a function connected 


r. Dilatation upon blood-vessel filled with corpuscles. 
Small blood-vessel. 



12. Vascular sinus of 

with alimentation converted into a quite different physiological path, and one which 
must bear some relation to the vascular system. 

Now there is some evidence that the ' cardiac body ' of certain Enchytraeidae has 
had a similar origin. In Biochholziu the paired dorsal diverticula of the oesophagus, which 

are comparable probably to the calciferous 
glands of earthworms, are surrounded by 
a blood-sinus where they are attached to 
the gut ; this blood-sinus becomes further 
forward the dorsal vessel ; there is thus 
a diverticulum of the oesophagus, as it were 
thrust into the dorsal vessel ; it is to be 
pointed out in the first place that a dorsal 
diverticulum of the gut never coincides 
with a cardiac body ; Hoest was apparently 
the first who compared the cardiac body 
of the Chlorhaemidae with the gut diverti- 
culum of the Enchytraeidae ; it is not in 
my opinion a valid objection to this identi-! 
fication to point out, as Cunningham has done^, the absence of any present connexion 
between the cardiac body and the epithelium of the oesophagus ; such a communication 
may easily have become lost ; there are plenty of analogous instances. If we imagine 
the gut-diverticulum to lose its lumen a cardiac body is at once produced. We should 
therefore look upon those forms which have a cardiac body a3 being more modified 
than those in which there is a gut-diverticulum ; as in the genus Aeolosoma there is 
a trace at least of a cardiac body, better marked in the perhaps allied' ttenocZriZus; 
these genera are perhaps to be regarded as the descendants of genera in which there 
was a dorsal diverticulum of the oesophagus — an argument for the non-primitive 
character of Aeolosoma. It seems to me to be also possible that the blood- 
glands of Fhreodrilus are referable to the same characteristic diverticula of the 
oesophagus. The cells contained in the lateral vessels will be on this interpretation 
the remains of the calciferous glands ; the vessel which contains them will be the 
hypertrophied vessel originally supplying these hypothetical glands ; this may also 
help us in understanding the anomalous presence of an intestinal heart in Branchiura 
in the same segment as that which also contains a dorsal heart ; if the intestinal heart 
be regarded as the last trace of a vanished pair of calciferous glands the difiiculty 

(After Michaelseii.) 

I. Peritoneal layer. 2, 3. Tubules of gland 4. 
Dorsal vessel. 5. Lumen of intestine. 6. Ventral 
vessel. 7. Peritoneum. 8, 9. Muscular and epi- 
thelial layers of intestine, 

' On some points in the Anatomy of Polychaeta,' Q. J. Micr. Soi. vol. xxviii, p. 259, etc. 


of accounting for its existence vanishes also ; but this is, of course, purely 

There are various interesting analogies, if no more, between this series of glands 
in the Oligochaeta and certain structures in the Vertebrata. Weldon has traced 
the origin of the supra-renal body in the Marsipobranch Bdellostoma to a detached 
portion of the pronephros, and he has described and figured an originally secreting 
gland with blood in its lumina — a state of affairs closely paralleled by the condition 
which I have just described in Evdriloides. A closer resemblance still is offered by 
such glands as the thyroid and thymus; originally connected with the oesophagus, 
or with diverticula of the same, these glands become entirely detached from it, and 
have acquired some function in relation to the vascular system. It is also said 
that the spleen is developed as a diverticulum of the gut ; if this be proved to be 
the case the analogy is so close that we are almost justified in an actual comparison 
of the blood-glands of these Oligochaeta with the vertebrate spleen. The spleen is 
permeated by blood-vessels just as are the glands in the Oligochaeta, and its origin 
as a diverticulum of the mesenteron leaves no resemblance unaccounted for. At 
the very least the analogy is interesting and I draw attention to the resemblances 
for what they may be worth. 

VIII. Respikatoey Organs. 

DeaUng as this woi-k does with a group of worms which were termed by Cuviek 
' Anndlides s^tigferes abranches,' and which are constantly distinguished by the absence 
of branchiae from the Polychaeta the present chapter might seem to be superfluous. 
Nevertheless there are a few forms which possess special respiratory organs, and 
two in which these branchiae are much like those of certain Polychaeta. In the 
great majority of Oligochaeta there are no special respiratory organs— the generg,! 
body surface occupying the place of a lung or branchia^. Where the integument is 
thick there are always plexuses of blood capillaries in the integument, which bring 
the vascular system into close relations with the external medium and presumably 
allow of an exchange of gases. The blood, as has been already mentioned, is in 
all Oligochaeta, with the exception of Aeolosoma, Chaetogaster, and certain of the 
Enchytraeidae, tinged with Haemoglobin; we may fairly suppose that this substance 
plays the same part in respiration among these Annelids as in the Vertebrata. The 
efficiency of the skin as an organ of respiration in many Leeches is largely increased 

' The nephridia were once regarded as respiratory organs and as the equivalents of the tracheae of 
insects. The dorsal pores also were considered to perform a similar function. 



by the extension of the blood capiUaries to the epidermis itself; this interesting fact 

was first discovered by Lankestee. I myself extended this discovery (3) to 

Earthworms, and in the genera Megascolex, Perichaeta, Criodnlus, Moniligaster, &c., 

as I and others have shown, the epidermis of the body in general— not merely the 

clitellum— is vascular. As a general rule the lower Oligochaeta, in which the 

dermo-muscular tube, in correspondence with their small size, is actually— sometimes also 

relatively— thinner, than in the terrestrial worms, have no development of a capillary 

system in the skin. This rule, however, is not without exceptions. LimnodrUus 

Hoffmeisteri has tufts of blood capillaries which arise from the peri-intestinal trunks 

and enter the body wall, reaching even to the epidermis itself [Vejdovsky, 24, Taf. 

viii, figs. i6, 17]. More highly developed still is the integumental vascular network 

of Ilyodrilus (Stolc 3) and Branchiura (Beddaed 58). Many Tubificidae and Lum- 

briculidae, appear to respire chiefly at the posterior end of the body ; they live in the 

mud at the bottom of ponds, ditches, &c., with the head implanted and the tail 

waving about in the water. Now in the worms which have this habit, and for the 

matter of that in others too, the body gradually diminishes in calibre and in the 

thickness of its walls towards the anus ; hence the blood is necessarily brought into 

closer relation with the surrounding water. There is in these Oligochaeta, at least, 

a commencing localization of the respiratory function to the tail end, in accordance 

with the attenuation of the body in this region. And we find that when branchiae 

are developed they are in every instance but one confined to the caudal extremity. 

For a long time the only Oligoehaet known, which could be said to possess special 
respiratory organs was Dero. This genus of Naidae is remarkable for possessing at 
the end of the body an expanded hood formed by the widening out of the anus; 
from the inner surface of this arise two pairs of (usually) cylindrical processes. 
These processes, as well as the hood which covers them, are ciliated; they can only 
be moved by means of intrinsic muscular fibres. The principal descriptions and 
illustrations of this branchial area are those by Peebieii (10), and Bousfield (3)- 
In describing the organ I shall follow Peebier's description and illustrations of 
Bero ' ohtusa ' ( = really B. perrieri, Bousf.). The branchial processes themselves are 
covered with a ciliated epithelium and their cavities are largely occupied by spindle- 
shaped or stellate muscular fibres ^- The vascular supply is derived from a direct 
continuation of the dorsal and ventral vessels. The ventral vessel passes undivided 
to about the middle of the hood; here it divides into two trunks which run round 
the margin of the hood and give ofi" six branches — one to each of the branchial 

' In D. digilata there are also retractor muscles for each braiichia. 


processes, and one on each side passing obliquely across the hood ; they then bend 
upwards and become continuous with the dorsal vessel; each branchial capillary 
runs up the branchia to its tip and then bending upon itself runs down the opposite 
side of the branchiae ; these and the other branches fuse together and form the 
dorsal blood-vessel. 

In Dero obtusa the circulation in the branchiae is somewhat simpler; it was 
described and figured by d'Udekem (1). The ventral vessel divides as in D. perrieri 
but simply pursues a sinuous course, passing up and down each branchia; the right 
and left halves of the circle unite at the base of the branchial hood to form the 
dorsal vessel. Dero digitata has more complicated vascular loops than even D. 
perrieri. The arrangement of these has been worked out by Stolc (2). Each of 
the four branchial processes has two capillary loops instead of only one ; and there 
are two circular vessels, derived from the ventral vessel, which form complete circles 
and do not communicate with the dorsal vessel except by the recurrent loops from 
the branchiae. 

It seems clear from the habits of the worm and from the structure of these 
circumanal processes that they must be regarded as branchiae; the expanded hood 
which is itself in some species furnished with elongated processes is no doubt also 
branchial in function. The next Oligochaet, in order of discovery, which is 
branchiate, is Alma nilotica. In the systematic part of this work I discuss 
whether this form (recently re-described by Levinsen as Digitihranchus niloticus) be 
really an Oligochaet, The conclusion arrived at is that it is an Oligochaet. The 
branchiae, like those of Dero, are found at the posterior end of the body. It is very 
desirable that this form should be re-investigated, for at present there are no details 
concerning the structure of the supposed branchiae to hand. There is, however, 
hardly room for doubt as to the branchial nature of the short cylindrical processes 
which are found upon the last sixty or seventy segments of the body. They occur just 
to the dorsal side of each dorsal pair of setae and there are four or five on each side, 
sometimes simple and sometimes branched. Chaetobranchus semperi is a remarkable 
branchiate Naid, apparently known to Sempee, and lately described by Bo dene 
(1). I can confirm from personal observations Bouene's description of the branchial 
organs, which, unlike those of Bero, are confined to the anterior part of the body. 
The most anterior, after the first one or two pairs, are the longest ; and in those 
anterior ones are imbedded the long capilliform setae of the worm ; further back the 
setae are independent of the branchial processes. The inclusion of the setae within 
them is not unsuggestive of a comparison with parapodia of the Polychaeta. These 
branchia in Chaetobranchus are ciliated externally, they contain a capillary loop 

M a 


which runs in the axial cavity of the branchiae ; the walls are cellular and have 
no muscular fibres. 

In Branchiura the branchiae are not lateral in position as in the last genus, but 
dorsal and ventral. They contain a prolongation of the coelom, which, however, is 
shut off by a diaphragm from the general coelomic cavity. The branchiae are 
contractile owing to the presence of a layer of muscular fibres lying immediately 
beneath the epidermis ; the latter is not ciliated. Immediately beneath the epidermis 
is a blood-vessel on either side. The branchiae are limited to the posterior region 
of the body. The reason for this is probably that this Tubifieid, like others, rests 
with its head imbedded in the mud and its tail waving freely in the water. The 
last branchiate Oligochaet is also a Tubifieid, Hesperodrilus branchiatus, in which the 
branchiae are like those of Branchiura, but lateral instead of dorsal and ventral. 

IX. Eepeoductive System 1. 

The OHgochaeta like some other animals which are hermaphrodite possess a com- 
plicated series of organs related to the reproductive function. We can distinguish 
the essential organs, and those which are unessential and only concerned with 
impregnation or the liberation of the genital products. The essential organs are of 
course the ovaries and testes — the gonads ; these are the first part of the reproductive 
system to make their appearance in the embryo. Then there are the ducts which 
convey the sexual products to the exterior, the sperm-ducts and oviducts, and the 
sperm-sacs and egg-sacs where the sexual elements undergo development ; finally we 
have a series of organs which are concerned, in the mutual impregnation of the 
worms ; the glands and sometimes penial setae appended to the sperm-ducts, the 
spermathecae for the reception of the sperm during copulation, etc. The various 
organs essential and non-essential have fixed positions in the body of the worm ; one 
organ is found always in one segment, another in a second segment, in every species, 
the positions being characteristic for the species or the genus or family as the case 
may be. The table appended illustrates the varying position of the parts of the 
reproductive system in all the families of aquatic Oligochaeta and a few earthworms. 
The reproductive organs are segmentally arranged just as are most of the other 
organs of the body; but, as also is the case with other organs, the metameric 
arrangement is sometimes lost or obscured; the sperm-ducts for instance are not 
always confined to a segment nor are they framed of a series of metamerically 

' For asexual reproduction, see under Aeolosoma and Naids. 



arranged parts. The sperm-sacs, too, though often ranged with an absolute regularity, 
sometimes show the same kind of divergence from what we must consider to be the 
normal for a segmented worm. 

Plates I and II represent the genitalia of various Oligochaeta depicted diagram- 







Aeolosoma . . 

or 3, 4. 5 













or in 






, 8 












X, xi 




Tubiflcidae . . 


10, II 





Trichodrilus . . 

II, 12 





Glaparedilla . ) 

Stylodrilus . ) 






Phreatothrix . 

JI, 12 





Bhynchelmis . . 


ix, X 




Lumhriciilus . . 

lO, II, 13 


10, II 





Sutroa .... 



ix, X 




Phreoryctes . . 



X, xi 

xii, xiii 

II, 12 

12/13, 13/14 

Pdodrilus . . 



X, xi 


12, 12 


Perichaeta . 

5 (7)-8 (9) 


X, xi 




Megascokx . . 

5 (6, 7)-8/9 

13 (I4)-I7 

X, xi 





(7) 8, 9 

13-17 (19) 

X, xi 




§ I. Gonads. 

The gonads are developed from the peritoneal epithelium and are nearly always 
paired structures, probably they are really paired in origin in such forms as Aeolosoma 
where they appear to be single. Both male and female gonads are present in all 
Oligochaeta ; there is no instance known of an unisexual form. The ovaries are most 
usually a single pair but there are sometimes an additional pair; no more than two 
pairs of ovaries have ever been certainly made out^. The ovaries invariably agree 
in position with the testes, but they are of course situated in different segments ; 

' Except in abnormal specimens such as those described by Woodwabd (1, 2). 


they furthermore agree so closely with the testes that in the early stages it is 
impossible to distinguish the two kinds of gonads save only by the segment which 
they happen to occupy. Not only is there this close agreement in structure (in the 
immature condition) and in situation, but the shape of both gonads is at first 
identical ; the immature ovary like the immature testis is somewhat pear-shaped in 
outline, the broader end being attached to the septum ; later on the free end of the 
ovary, as is also the case with the testis, may become frayed out into a number of 
processes. The ovaries always lie behind the testes ; the only possible exception is 
in the genus Plutellus, where Pekriek (8) has stated that the ovaries are in front 
of the testes ; if, however, a worm described by Benham (8) as a Plutellus is really 
a member of this genus, there is nothing abnormal in the position of the female 
gonads. As a rule the gonads are all in consecutive segments, the ovaries following 
the testes, the rule indeed has very possibly no exception ; this at first sight appears 
to be an inaccurate statement ; for in Lumbricus, and indeed in all earthworms, the 
ovaries are in the thirteenth segment, while the last pair of testes, if there are two pairs, 
is in the eleventh segment ; there is thus a gap of a segment between the last testis and 
the ovary of its side of the body. In embryos of Lumbricus, however, as was shown 
by WooDWAED, and in embryos of Octochaetus, as has been demonstrated by myself 
(51), there is an additional pair of ovaries in the twelfth segment, which never comes 
to maturity and disappears early. There are, therefore, some grounds for believing 
that two is the typical number of pairs of these gonads in the Lumbricidae and 
Acanthodrilidae, and very likely in other terrestrial forms also. The fact that there 
are two pairs of testes seems to render this assumption probable at least. In the 
genus Phreoryctes the number of ovaries (as of testes) is normally two pairs. 
Another fact pointing to the same conclusion is the presence in more than one 
species of Perichaeta of two pairs of egg-sacs ; there is a pair in the fourteenth 
segment and another in many cases in the thirteenth segment ; the latter would appear 
to correspond to a missing pair of ovaries belonging to the twelfth segment. 

In fact both lines of argument appear to point to the primitive possession of two 
pairs of ovaries at any rate in the terrestrial Oligochaeta. The passage of ova from 
the gonad into the duct and still more into the egg-sac in the terrestrial Oligochaeta 
is not easy to understand; as to the former it must be remembered that in the living 
worm the distance is not great between the gonad and the large mouth of the 
funnel; it is possible that from time to time during the movements of the worm the 
distance is lessened; M'hen this takes place the oviducal funnel may approach so near 
to the gonad that the movement of its cilia may perhaps detach a perfectly ripe, 
ovum from the extremity of the ovary and direct its course into the funnel, and 


thence either directly to the exterior or into the mouth of the egg-sac which lies 
conveniently near to the funnel. Among the Eudrilidae the passage of ova is 
enormously facilitated by the fact that the ovaries are enclosed in special peritoneal 
sacs which are continuous with the efferent duct ; these sacs are late in their 
development, which seems to indicate their comparatively modern appearance; they 
were first discovered by myself in the genus Eudrilus (62); since that time they 
have been recognized in the majority of the Eudrflidae (which see for a more detailed 

Ova. — The parasitic and encysted Gregarines were at one time mistaken for the 
ova of Lumhricus — perhaps not altogether an unnatural error. The ova are now 
known in a large number of Oligochaeta; but it is in Rhynchelimis that they have 
been most thoroughly studied (by Vejdovsky). It is a remarkable fact that the 
Oligochaeta can be divided into two groups according to the character of their ova; 
in the aquatic Oligochaeta the eggs are large and contain an abundance of yolk ; in 
the terrestrial forms, on the other hand, the ova are of microscopic size and contain 
but little yolk. It is, it will be observed, the large forms which have small eggs 
and broadly speaking the small forms which have large ova. All the genera which 
were grouped by Clapaeedb within his group 'Limicolae' have large ova containing 
much yolk. Differences of size no doubt occur in both groups ; but in no earth- 
worm is the egg ever so large as it is in the aquatic worm with the smallest ova. 
The difference is a remarkable one ; it is almost, if not quite, as striking as that 
between the ova of a Mammal (not of course Prototherian) and a Frog. The large 
size of the ova was justly made use of by d'Udekem in his classification of the 

It is difficult to account for this striking difference; Vejdovsky (24) pointed out 
that there is a difference in the mode of nutrition in the ova in the two groups ; in 
the aquatic Oligochaeta the eggs are early detached from the ovary and undergo 
further development in the body cavity or in egg-sacs ; on the other hand, in 
earthworms the eggs reach maturity in the ovary which is furnished with abundant 
blood-vessels, which are wanting in the egg-sacs of the others ; this latter statement 
is not quite accurate — at least in one way; it is perfectly true that there is 
no development of blood-capillaries on the egg-sacs of those worms ; but special 
perivisceral vessels undergo an increase of length at the time that the egg-sacs are 
formed and accompany them. This cannot, however, be regarded as an explanation 
now; for in Eudrilus, and in the Eudrilidae generally, the eggs are apparently 
transferred at an early stage to the egg-sacs where they reach maturity; indeed, in 
some Eudrilidae the ovaries are even at an immature stage no longer to be found 


in an ovary; they have been bodily transferred to the egg-sac. In Libyodrilus this 
appears to be the case; I succeeded in finding the ovaries in a very young worm, 
but not in mature or even in nearly mature specimens ; no one has detected the 
ovary of Polytoreuttis, though its position has been probably fixed with accuracy; 
in StvMmannia, too, the mass of cells described by Michaelsen as an ovary may 
be one, but it has none of the characteristic appearances of germinal tissue; I am 
of opinion that this is to be explained on the assumption that all, or nearly all, of 
the germinal cells have been made over to the egg-sacs. 

The facts evidently require another explanation. It might be thought that there 
were difierences in the development of the embryos sufiicient to account for this. 
If, as one might perhaps infer from d'Udekbm's figures, the eggs when large and 
full of yolk completely filled the cocoon to the exclusion of any albumen, it would 
be at once apparent that the absence of a nutritive fluid necessitated other nutriment 
for the developing egg; but Vbjdovsky has specially described the albumen in the 
cocoon of Rhynchelmis — an Oligochaet with very large ova. A free larval stage in 
one group or the other might also get over the difficulty ; but there is none such ; 
in all the Oligochaeta whose development is known— not a very large series it must 
be admitted, but still a series comprising representatives of genera with large and 
genera with small ova, the young leave the cocoon at approximately the same age ; 
there are at least no striking difierences in this particular. 

Among the vertebrata it is always possible to trace some connexion between 
the abundance of yolk in the ovum and the needs of the embryo ; for example, the 
ova of Aonphioxus have little yolk and are minute in size ; this is correlated with 
the fact that the young are hatched in an immature condition ; in the mammals 
there are of course special provisions (the placenta) to prevent the otherwise early 
hatching of the embryo ; the frog which has much more yolk leaves the egg in a 
tolerably advanced stage of development; finally the bii-d and reptile is hatched in 
the practically adult condition. Nothing of the kind occurs in the Oligochaeta. 

I formerly attempted to show that there was truth in the idea that the nature 
of the ova is correlated with the habitat of the worm ; the ova of Allurus which is 
an aquatic form though related to the terrestrial genera are apparently larger than 
those of Lumbricus ; the difference, however, is not by any means an obvious one ; 
moreover, I since found that the nearest approach to the ova of the 'Limicolae' 
among the Terricolae was to be seen in Moniligaster ; here we have ova which 
are filled with large yolk spherules as large as those of, for instance, Tuhifex; 
Moniligaster is purely a terrestrial genus so far as we know. 

The mature egg of Rhynchelviis, which may be selected as the type of a large 


yolked ovum, is spherical ; it has a peripheral and extremely fine membrane, beneath 
which is a dense layer of protoplasm; connected with this is a protoplasmic net- 
work which ramifies through the entire egg and in the meshes of which are 
contained the yolk-spherules. The spherical nucleus is surrounded by a distinct 
doubly-contoured membrane. Around this membrane is a dense layer of protoplasm 
with a radial arrangement of its particles ; but the membrane seems to be imper- 
forate. The contents of the nucleus show an obscure meshwork in the interstices of 
which is a granular nucleoplasm ; there are as a rule two nucleoli, of a spherical 
contour ; the substance of these can be differentiated into two layers ; outside there 
is a radially striate coat within which is the granular core. The ovum is elaborately 
figured and described by Vejdovsky, to whose work reference must be made for 
further details (9). 

Among the Megadrili the most aberrant ova are those of the Eudrilidae. Like 
other Megadrili they are of small size owing to the small development of the 
yolk. As in Lumbricus, there is a membrane covering the egg externally; but 
this membrane is greatly developed in many Eudrilids. In Lumbricws it is 
an excessively fine membrane as in Khynchehnis ; but in Hyperiodrilus there 
appear to be two distinct membranes. The ovum is surrounded by a very 
thick, darkly staining membrane which is traversed by numerous pores ; beneath 
this is a fine membrane which I regard as the probable equivalent of the vitelline 
membrane of Lumbricus. The egg-protoplasm has a distinctly reticular arrangement, 
and the nucleus has also a membrane separating it off" from the surrounding 
protoplasm. The ova of Heliodrilus appear to possess much the same structure. 
In Eudrilus the ovum has a thick membrane, exactly like the thick outer membrane 
of the ovum of Hyperiodrilus in minute structure; but it has the remarkable 
peculiarity of being confined to one pole of the ovu;n; it does not extend right 
round. The structure of the ovum of this Annelid has been treated of by Hokst 
and myself. Hoest (8) speaks of this peculiar membrane as a protoplasmic mass 
formed of filaments which have the appearance of cilia. The membrane must, I 
think, in spite of the fact that it is only partial, be compared to that of Heliodrilus. 
The comparison seems to favour my opinion that the membrane in question is not 
a product of the ovum, but is produced by the modification of the follicular cells 
surrounding the growing egg. If it were a product of the egg protoplasm it would 
surely surround the entire circumference of the egg in Eudrilus no less than in 
Heliodrilus. I shall recur to this ' membrane ' in describing the egg-development. 

Development of the ova. — There appear to be three types of egg-development in 
the Oligochaeta. 


The more prevalent type is seen in all earthworms except certain Eudrilidae, 
and in a large number of the aquatic Oligochaeta; it is met with, for example, in 
the Lumbriculidae and most Tubificidae. Yejdovsky has described in great detail 
the facts for Rkynchelmis which I shall therefore take as an instance. The youngest 
eggs are indistinguishable from the mass of ovarian cells among which they lie in 
the egg-sacs ; the cells appear to be amoeboid in their youngest stages ; this is 
inferred from their frequent pear-shaped character and from the fact that they are 
not always in continuous contact. Any of these cells, it may be inferrred, may 
develop into ova. As the egg-cell grows its peculiar characters already described 
gradually differentiate themselves; the nucleolus is at first of course single and its 
bipartition has been observed; no share whatever appears to be taken in the 
development of the ovum by the surrounding cells ; at any rate no changes are 
noted in them by Yejdovsky. In lAimbricus the mature ovum has, what it has not 
in Rhynchelmis, a distinct follicular layer of flattened cells ; it is possible that these 
cells do bear a part in the maturation of the ovum ; but apart from this follicle no 
changes are observable in the remaining cells of the ovary which are not on the 
way to become ova. 

The second method of egg-development differs from that just described in 
the important fact that certain of the cells of the ovary do apparently take a 
share in the formation of the ovum by contributing to its nutrition ; this way 
of development has at present only been observed in Eudrilus by myself and 
HoEST. Our observations agree in all essentials. The development here, as in 
Rhynchelmis and many other worms, takes place in the egg-sacs. At first the 
young ova are seen lying among a quantity of indifferent cells; any of these, it is 
to be presumed, possess the capability of becoming ova ; later the cells in the 
immediate neighbourhood of the more mature ova gradually break down; the out- 
lines become obscured and the final stage reached is a mass of protoplasmic matter 
in which neither cells nor nuclei can be any longer recognized, and which has a 
fibrous appearance ; it is very possible, though I have no positive facts to go upon, 
that the peculiar membrane already referred to as surrounding one pole of the 
ovum is produced by this broken down mass of cells. The pores of the membrane 
in question are figured by HoRST as penetrating the vitelline membrane, and he 
thinks that they serve as the conduits of nourishment to the ovum. There is to 
my mind an undoubted resemblance in the mode of development of the ovum in 
Eudrilus to the formation of the Graafian follicle in the higher Mammalia; in the 
latter, the liquor folliculi is produced by the breaking down of cells of the follicle, 
at least partly; and this liquid may be fairly compared to the perhaps fluid 


protoplasmic mass referred to as surrounding the mature and nearly mature ova of 

The third mode of development of the ova in the Oligochaeta is especially 
interesting, as it appears to show great resemblances to the development of the 
spermatozoa in the same worms. It is characteristic of the Naids, the Enchytraeidae, 
the genera Ilyodrilus and Phreodrilus among the Tubificidae. It has been followed 
out by a good many observers; the eggs in these worms consist in their early 
stages of masses of cells, which become detached from the ovary, and either find 
their way into the egg-sacs or float freely in the body-cavity. Each mass consists 
of a peripheral row of cells surrounding a central mass of protoplasm, -vPhich has 
no nucleus and does not appear to be the equivalent of a cell. I have seen in 
Phreodrilus the whole mass surrounded by a layer of flattened cells. Later one of 
the cells increases in size and becomes a definite ovum ; when it breaks away the 
other cells may in their turn become ova ; in this case the development is almost 
exactly that of the spermatozoa, the only difference being that the peripheral row 
of cells are not simultaneously converted into sexual elements. In the early stages 
of the ovarian development of Ilyodrilus Stolc (3) represents the immature ova 
as not separated by distinct boundary lines from the central non-nucleated mass. 

The spermatozoa and their development. — The spermatozoa in the Oligochaeta are 
always filiform bodies and possess the power of free movement. As a general 
rule there is no thickening at one end ; this was only observed by Vejdovskt in 
the Lumbriculidae. The development of the spermatozoa has been studied by a 
large number of naturalists ; the reader is referred to Bloomfield and Nasse 
especially for details upon this subject. The principal facts appear to be the 
following: the sperm undergoes its development in the sperm-sacs in those worins 
(the majority) which possess these structures; the sperm mother-cells divide in 
such a way as to form a peripheral layer of cells, not at first marked off from 
a central, generally unnucleated mass of protoplasm ; this latter has been termed the 
cytophore or the sperm- blastophore ; this structure is simply the remains of the 
mother-cell, and possibly serves to nourish the growing spermatozoa; it is evidently 
to be compared to the central mass of protoplasm round which the ova of 
Phreodrilus, &c. are developed. 

§ a. Sperm-sacs. 

Nearly all the Oligochaeta possess sacs in which the sperm undergoes most of 
the stages of its development, and which are on that account termed sperm-sacs; 
this name is preferable, on the whole, to the earlier name of 'vesiculae seminales,' 

N a 


because that term suggests a comparison with the similarly named structures of 
vertebrates, which are a part of the efferent apparatus ; at fii-st the structures now 
known to perform the function of harbouring the sperm were regarded as testes ; in 
spite of the discovery by Herino of the real testes, the sperm-sacs were persistently 
termed the testes by almost everybody. From the fact that the sperm-sacs always 
appear to contain Gregarines, these organs were once regarded as ovaries — the 
Gregarines when encysted being mistaken for the ova; d'Udekem's discovery of the 
ovaries disposed of this view as well as of the alternative view due to VON SiEBOLD 
that the sperm-sacs were hermaphrodite glands. 

Their true nature was first pointed out by Heking who called them 'Samen- 
blasen.' The development (in Luvihricus) was first elucidated by Bergh, whose 
discoveries finally disposed of the theory that the organs were testes. The sperm- 
sacs originate from the intersegmental septa, as outgrowths of the same ; they 
contain from the first a cavity which is continuous with the cavity of the segment 
from the wall of which they arise ; this cavity is therefore coelomic ; later on it is 
divided up into a series of inter- communicating spaces by the complicated growth 
of the walls ; but none the less the internal cavity remains coelomic, and the 
epithelium found within it is peritoneum. The sperm-sacs are in fact coelomic sacs 
set apart for the maturation of the sperm. It is usual to distinguish between the 
paired sacs and median unpaired sacs which enclose, when present, the gonads, the 
ventral blood-vessel and even the nerve-cord, besides of course the funnels of the 
sperm-ducts. These communicate in the adult with the paired sacs ; the name 
sperm-reservoirs may be applied to the former, that of sperm-sacs may be reserved 
for the latter. As to the origin of the sperm-reservoirs there does not appear 
to be much positive information ; Bloomfield thinks that they are due to a 
coalescence between the paired sacs ; Beegh is not apparently inclined to back up 
this view. It can hardly be doubted that they originate from the septa as do the 
paired sacs ; they difier, however, in structure to a certain extent ; the cavity of the 
paired sacs, as has already been stated, is subdivided into numerous cavities ; that 
of the median sacs is not so divided ; the median sacs are by no means always 
present; they are absent, for example, in the genus Allolobophora ; they seem also 
to be absent in Acanthodrilus and in a number of other forms. In these genera it 
of course follows that the gonads and the sperm-duct funnels hang freely into the 
cavity of their respective segments. It is not always possible to refer a given 
structure to one of these two series of sacs ; the matter becomes easier if we agree 
to apply the term sperm-reservoirs to those sacs which enclose the sonads and the 
funnels, reserving the term sperm-sacs for diverticula of the septum which do not 



Fig. 26. 

enclose these organs. Looking at the matter in this way, the sacs in Moniligaster 
must be regarded as sperm-reservoirs, although they are paired; the sacs in question 
enclose the gonads and the funnels, and it will also be noticed that their cavity is 
undivided by trabeculae — another feature which characterizes the sperm-reservoirs 
of Lumbricus in contradistinction to the sperm-sacs. It appears from the results 
obtained by Bergh and Vejdovsky that the paired sperm-sacs are 
developed before the median sacs ; hence it would seem just to 
regard these sacs as older than the median sacs. It is conceivable 
that the median sperm-reservoirs of earthworms are more strictly 
comparable to the generally impaired and very voluminous sacs 
found in the aquatic Oligochaeta, than are the sperm-sacs. 

In only one family of Oligochaeta are the sperm-sacs nearly 
always absent^ this family is the Enchytraeidae ; the genus Mesen- 
chytraeus, however, has paired sperm-sacs, which originate from 
the septum bounding posteriorly the segment in which the male 
gonads lie. The number and arrangement of the sacs varies 
considerably in different genera, thus affording valuable characters 
which are sometimes of specific value only. The number of the 
sperm-sacs varies from one to four pairs. The aquatic forms have 
only a single pair or a single sac which usually extend through 
a large number of segments, enclosing the egg-sacs in many cases, 
as in Rhynchelmis figured by Vejdovsky (9). In the terrestrial 
worms the sperm- sacs are as a rule of comparatively small size; 
they rarely occupy more than a single segment. There are various 
exceptions to this rule, however; and these exceptions always 
concern species or genera in which the number of the sperm- 
sacs is reduced to one pair. The most remarkable instances are 
Polytoreiitus (Woodcut, fig. 36) and Trichochaeta ; in both of these 
genera there is one pair of sperm-sacs which extend backwards 
through twenty or thirty segments ; the extent of these sacs is 
only paralleled by certain aquatic genera such as Rhynchelmis. It 
is time that when the sperm-sacs are of such an extraordinary length they are thin ; 
but in spite of this the total space enclosed by them is greater than in the case of 
the genera where there are two or three pairs of sperm-sacs only, occupying a single 
segment each ; unfortunately our knowledge of the economy of these worms does not 
at present permit of any explanation of these remarkable divergences. 

Neuland has recently recurred to the earlier view respecting the sperm-sacs ; he considers 




I. Anterior end of 
sperm-sacs. 2. Dilated 
region of sperm-duct. 

3. Calciferous gland. 

4. Oviduct. 5. Spermi- 
ducal gland. 6. Sper- 
mathecal sac. 7. Pos- 
terior end of sperm-sac. 


that the sperm-sacs are to be regarded as a testis ; but in putting forward this somewhat belated 
theory, the author has forgotten the development of the organs in question. However these are 
the grounds upon which the identity of the sperm-sacs with testes is based. 

The testes as first described by Heeing were not always to be found in the situation where 
he described them ; bodies identical in every particular with these testes — so much so that one 
might have served as a model for the other— were found in the sperm-sacs (' Samenblasen- 
anhange '). Neuland also points out that younger spermatogonia were found in the extremities of 
the sperm-sacs than in the more proximal regions; the reverse ought to be the ease were the 
developing sperm-cells derived entirely from the testes (of Heeing) ; the nearer to the testes the 
less advanced ought to be the spermatogonia. Heeing has figured the mouth of the sperm-duct 
as nearer to the testes than to the mouth of the sperm-sacs ; why therefore, asks Neuland, does 
not the unripe sperm get into the latter? As is known this does not take place. Another matter 
raised by Neuland is the immense amount of sperm produced— an amount too excessive to have 
been produced only by the testes of Heeing. 

Even if it be true that germinal tissue is produced in the interior of the sperm-sacs — which as 
yet wants confirmation — it is not clear how this proves the contention that the sperm-sacs are 
testes ; in a loose sense of the word they might be called so ; but then so they might have been 
before, for the median sperm-reservoirs contain the true testes; no doubt the whole structure may 
be called a testis in the same sense that the testis of Astacus is so called, or even the testis of 
the vertebrate which contains besides the germinal tissue other structures. 

§ 3- Egg-sacs \ 

In the majority of Oligocliaeta the ova when ripe or nearly ripe are transferi'ed 
to certain sacs in which they remain for a time before being extruded from the 
body; these sacs are enormously developed in all the aquatic Oligochaeta, but appear 
to be rudimentary structures in most of the terrestrial forms. In Rhynchelmis, for 
example, the egg-sacs extend back as far as to the fifty-fourth segment, or even in 
more mature individuals to the sixty-seventh. In other aquatic families similar 
sacs are to be found; they are comparatively as large in the Tubificidae and 

Nasse has described their development in Tubifex: they originate as outgrowths 
of the dissepiment xi/xii ; into these sacs the sperm-sacs are pushed as they 
develop ; and thus in the mature worm we have two sacs one within the other, 
the outer of which contains ova and the inner sperm; the same is also the 
case with Rhynchelmis, where Vejdovsky (9) speaks of the egg-sacs and the 
sperm-sacs lying within and enclosed by a common membrane; it should be 
observed that these sacs are paired, lying one on each side of the intestine (see 
Vejdovsky 9, PI. Ill, fig. i o.s., fig. 2). But the egg-sacs are not always paired; in 
Stylaria proboscidea (cf. Taubeb 2, PL XIV, fig. i o.v.), for instance, and in other 

' Often called Receptacula ovorum. 


Naidomorpha the egg-sac is a single structure surrounded anteriorly by the sperm- 
sac; the same statement holds good for other Oligochaeta, indeed for the majority 
of the aquatic forms; in Tuhifex there is only a single sac, such too is the case 
with Mesenchytraeus. Vejdotsky (24) indeed seems to imply that unpaired sacs are 
the rule, for he states in a footnote to p. 137 of his 'System und Morphologic' that 
only in Rhynchelmis are the sperm-sacs, and, as we know, the egg-sacs, paii-ed ; the 
remark in the text to which this footnote is appended runs as follows :^' Sowohl 
die Samen- als Eiersacke sind unpaarig,' &c. The paired condition, however, seems 
to be usual in the Lumbriculidae, for it is met with in Sutroa as well as in 
Rhynchelmis; the formation of the egg-sacs in the aquatic Oligochaeta seems to be 
merely a pushing out of the septa, caused, perhaps, by purely mechanical reasons: 
i.e. the large size of the ova and the pressure which they must consequently exert 
as they are driven backwards by the development of those in front; I have figured 
in the genus Phreodrilus the first commencement of the egg-sac, where it is nothing 
more than a bulging of the septum (21 PI. II, fig. 31). The egg-sacs were generally 
termed ' ovaries ' just as the sperm-sacs were called erroneously ' testes ' ; the excuse 
for this error was in many cases the fact that the ovaries in the sexually mature 
worm had entirely broken up into clumps of developing ova, leaving no trace of 
their former existence. 

Among the teiTestrial genera it is only in the family Moniligastridae that the 
egg-sacs are at all large ; here they may extend through three or four segments ; 
they are always paired sacs and enclose the numerous ripe ova; as the Moniligas- 
tridae are the only earthworms with comparatively large ova loaded with abundant 
yolk, there almost appears to be some connexion between the large size of the 
egg-sacs and this fact ; for in the aquatic Oligochaeta, also, the ova are loaded with 
yolk spherules. 

The minute egg-sacs of Lumbricus were first discovered by Heking who wi"ote 
as follows : — ' Der innere Rand des Trichters reicht an den Parmkanal ; am oberen 
Rande ist ein kleines, durch das Septum in den vierzehnten Ring hineinragendes 
Knotchen angewachsen. . . . Am oberen Rande (der Tuba) stiilpt sich die Wand 
zu einem kurzen Fortsatz aus, der durch das Septum tritt und im vierzehnten 
Segmente mit einer blaschenfdrmigen Erweiterung endigt. Dieses ist das erwahnte 
Knotchen. Es zeigt eine sehr verschiedene Breite, durchschnittlich 0-5 mm., und 
ist von einem dichten Gefassnetz umstrickt, dessen einzelne Gefasse bisweilen einen 
Durchmesser von 0-03 mm. erreichen. Unter dem Mikroscop ist es schwierig, den 
Zusammenhang dieses Organes mit der Tuba nachzuweisen, well diese sich nicht von 
dem muskulbsen Septum ganzlich isoliren lasst, so dass jedes klare Bild unmdglich 


wird. Man kann indess jeden Zweifel dadurch heben, dass man bei massiger 
Vergrosserung die natiirliche Spitze eines feinen Haares von der Tuba aus in das 
Blaschen eindringt. Dies gelingt ohne Schwierigkeit und man sieht unter dam 
Mikroscope deutlich wie die Haarspitze an der hinteren Wand des Blaschens 
angelangt sicb umbiegt. In den meisten Fallen fand ich in demselben Eier, 1-5 
an der Zahl, von gleicber Beschaffenbeit und grosser, als die im Ovarialzipfel 
enthaltenen und es liegt nabe, es als einen kleinen Eibalter anzuseben, in dem 
sicb die Eier ansammeln, um dann gemeinscbaftlicb in eine Eikapsel entleert zu 
werden. Der grosse Gefassreichtum des Organes weist auf Absonderung einer 
Fliissigkeit bin, die vielleicbt den Transport der Eier durcb den Eileiter erleicbtert. 
Der letztere ist ein mit zablreicben Gefassen umstrickter Kanal, ausgekleidet von 
einem, &c.^ ' 

Tbe development of these egg-sacs in Lwmhncus bas been studied by Bergh (5); 
be bas sbown tbat tbey originate as thickenings of tbe septum xiii/xiv, which later 
become excavated and communicate with tbe cavity of tbe thirteenth segment by 
a small aperture ; their interior is broken up, by anastomosing trabeculae, into 
numerous compartments which lodge the ova ; the oviducal funnel just reaches tbe 
interior of the sperm egg-sac as is sbown diagrammatically in tbe sketch given 
by GoEHLiCH : this clearly facilitates tbe passage of eggs from the sac to tbe 
exterior. Tbe investigations of Bergh already referred to have plainly sbown tbat 
tbe egg-sacs are quite homologous structures to the sperm-sacs ; both have tbe same 
minute structure and originate in the same way. 

The cavities of the egg-sacs are of course portions of tbe coelom, and they are 
lined like tbe rest of tbe coelom by peritoneal epithelium ; tbe rest of their walls is 
muscular tissue. 

In other genera of earthworms egg-sacs are to be found ; but it is doubtful, in 
some cases at any rate, whether they have any function and are not rather to be 
regarded as rudimentary organs. Thus in Criodrilus egg-sacs exist; tbe ova found 
in tbe interior are according to Collin ^ (1) smaller than the ripe ova within tbe 
ovary itself; this may argue that the ova become degenerate in tbe egg- sacs or 
perhaps that this sac is a sort of forcing house for the ova — only tbe comparatively 
unripe ova finding their way thither to attain maturity; tbe fact that the egg-sacs 
have abundant blood-vessels in their walls, which indeed often renders them 

1 These egg-saos were passed over by both Lamkestek (8) and Clapakede (1) in their accounts of the 
structure of Lumiricus ; Horst, on the other hand, has described and figm-ed them (11). And they have 
been recognized by all subsequent observers. 

2 Who figures as an abnormality the egg-sac of one side of the body depending from septum xiii/xiv 
into the interior of the thirteenth segment. 


extremely conspicuous to the naked eye in spite of their minute size, seems to 
indicate that they have some function. In the large and important family of the 
Eudrilidae there appear to be invariably a pair of egg-sacs present ; but in every 
case the egg-sacs appear without doubt to be functional though they are no larger 
or but little larger than in Lumbricus ; in this family the passage of ova into the 
egg-sacs is nearly always facilitated by peritoneal sacs, which enclose the ovaries 
and form a closed duct leading to the egg-sacs ; I refer again to these in describing 
the Eudrilidae. In all the Eudrilidae the egg-sacs in mature individuals contain 
not merely ripe ova, but ova in various stages of development surrounded by 
masses of ovarian cells ; this state of affairs is similar to what is found in the 
lower Oligochaeta, where masses of cells are broken off from the ovary and find 
their way into the egg-sacs, where they undergo further development ; in these 
cases, therefore, there is a closer likeness as regards function between the egg-sacs 
and the sperm-sacs, for in both the genital products undergo at least the final 
stages of development ; whereas in Lumbricus only ripe ova without any attached 
cells lie in the egg-sacs. 

One pair of egg-sacs have be6n found in the same segment (the fourteenth) in 
the Acanthodrilidae, the Cryptodrilidae, Geoscolicidae, and certain Perichaetidae ; as 
a rule egg-sacs seem to be absent from the smaller and degenerate species ; thus 
they do not appear to exist in Ocnerodrilus or in Gordiodrilus. In many Peri- 
chaetidae the egg-sacs are interesting from the fact that there are two pairs of 
them ; I have pointed out that to the two pairs of testes correspond as a rule 
two pairs of sperm-sacs ; now there does not appear to be any earthworm in which 
there are normally two pairs of ovaries, though in many there are traces of a second 
pair, especially in development ^ ; we should expect, therefore, that originally, at any 
rate, two pairs of egg-sacs existed in correspondence to these two pairs of ovaries. 
And in the genus Perichaeta (s. s ) there are several examples of species in which 
there are two pairs of egg-sacs. Bergh (5) has remarked upon the presence of two 
pairs in a species nearly related to Hoest's Perichaeta hasselti ; one pair occupy the 
normal position, the other lie in the thirteenth segment just above the ovaries ; 
this is, it will be observed, precisely the position that the second pair ought to 
occupy, for the missing ovaries belong to the twelfth segment. I have found two 
pairs of egg-sacs in other species of Perichaeta similarly placed. In Perichaeta 
I have noticed that these egg-sacs are often rather larger than in Lumbricus, 
and that they are of an elongated form perhaps more like the sperm-sacs in 

' See WooDWAED (1, 2) for occasional presence of many ovaries. 


shape than is usually the case; in considering the possibility that Perichaeta is 
an archaic type of earthworm, the existence of two pairs of egg-sacs is of some 
importance. Bekgh found nothing in the egg-sacs of the species investigated by 
him but Psevdonavicellae and darkly pigmented bodies ; he comments upon the fact 
that the funnel of the oviduct does not open into the interior of the egg-sac as it 
does in Lumbricus. and suggests on this account that the bodies are functionless. 
They are certainly not always functionless in this group of earthworms ; I found in 
Biporochaeta intermedia a pair (one only) of egg-sacs in the usual position which 
were full of' developing ova, each surrounded by a mass of nutritive cells ; these 
egg-sacs had no communication with the oviducts, but the fact that they contained 
so many eggs, both fully mature and developing, seems to dispose of the view that, 
when there is no connexion with the oviduct, the egg-sacs are without function. 
In the species just referred to the egg-sacs contained abundant Gregarines which 
I have also met with in the egg-sacs of Evdrilus; this is another resemblance to 
the sperm-sacs, where these parasites are invariably to be seen. 

§ 4. Sperm-dv,cts. 

Special conduits for the semen are found in all Oligochaeta, with the sole 
exception of the genus Aeolosoma. In that worm, according to the recent researches^ 
of Stolc (1) true sperm-ducts do not exist; the nephridia of all the segments 
of the body conduct the spermatozoa to the exterior ; this was proved by direct 
observation ; although the spermatozoa may escape by any of the nephridia (some 
of the nephridia disappear wholly or in part during the period of sexual maturity), 
those of the sixth and neighbouring segments especially take upon themselves 
the function of sperm-ducts, and they are figured by Stolc as rather larger than 
the others. 

When true sperm-ducts are developed there are never more than two pairs of 
them^; and frequently only a single pair exist. Each sperm-duct consists of (1) a wide 
funnel-shaped opening into the coelom and (2) a tube more or less contorted which 
opens directly on to the exterior or through (3) a terminal chamber, the spermiducal 
gland, which will be described in the next section. 

Both the funnel and the tube are ciliated throughout. The funnel varies very 
much in form: In the simpler aquatic forms, and in some of the smaller terrestrial 
Oligochaeta, it is a flattened plate-shaped disc, with incurved and sometimes also 

' Pereier's statement that there are no sperm-ducts in Anteus requires confirmation. 
^ Three have been described in Glyphidrilus weberi; but this is probably a 'sport.' 


a recurved margin. In the Enchytraeidae it has a very peculiar form ; it is here 
barrel-shaped, with a lining of very thick glandular-looking cells, which encroach 
upon the lumen, reducing it to the smallest dimensions. 

Among earthworms the funnel is for the most part folded at the margins, whence 
the term 'ciliated rosette' which is often applied to it. The folding is often 
extremely complex. 

As a general rule the funnel of the sperm-duct opens directly into the general 
body cavity of the segment ; sometimes, as in Lumbricus. &c., the funnels are 
lodged in special sacs — shut off from the rest of the coelom — which contain the 
testes, and are in communication with the sperm-sacs. The funnels commonly lie 
close to the posterior septum of their segment, facing forwards and opposite to the 
testes which usually are attached to the anterior septum of the same segment ; 
a curious exception to this rule is seen in the Eudrilid genera Teleudrilus, Hyperio- 
drilus, and Heliodrilus — perhaps also in some others. In these genera the funnels 
depend from the anterior septum of their segment, and accordingly (see woodcut, 
fig. 36) the vas deferens perforates this particular septum twice on its way to the ex- 
ternal pore. In the Lumbriculidae something of the same kind occurs, owing to the fact 
that the male genital pore lies in front of though in the same segment as the posterior 
funnels ; these funnels face forwards and depend from the posterior septum of their 
segment as in the majority of the Oligochaeta ; the vas deferens passes backwards 
perforating this septum and then again perforates it on its way to the external orifice. 
The number of funnels appears nearly always to correspond to the number of the testes ; 
if there is only a single pair of testes there is onlj' a single pair of funnels and sperm- 
ducts; if there are two pairs of testes the number of funnels is^ also doubled. The 
position of the funnels is also in correspondence with that of the testes, that is to 
say as to the number of the segment which they occupy. Very rarely this is not 
the case. For example in Heliodrilus the testes lie a segment in front of the funnels, 
which must necessarily occur owing to the facts already mentioned about the 
position of the funnels. The duct ai'ising from the funne] is a tube with ciliated 
epithelial walls ; the cells composing it are more or less quadrangular in form and 
surround the lumen which is never intracellular. Outside the epithelium is a layer 
of peritoneum and in a few instances, e.g. in Eudrilus, a layer of muscles between 
the two. Very anomalous are the sperm-ducts of Fhreodrilus, where the windings 
of the duct, which has a caecal diverticulum, are largely within the peritoneal covering ; 
this is described more in detail below. The principal variations in the sperm-ducts 

' Not, however, apparently in many Lumbriculidae, where two pairs of funnels and one pair of testes. 



concern (i) theii- position in the body, (a) the point of opening on to the exterior, (3) the 
degree to -which the ducts of one side of the body are fused together. 

As a general rule the sperm-ducts lie in the body cavity ; this is always 
the case in the lower Oligochaeta where they are frequently much coiled ; even 
in the higher Oligochaeta (earthworms) where the ducts (except in Moniligaater) pass 
in an approximately straight course from funnel to pore, the sperm-ducts lie a little 
above the ventral parietes^ 

In Lumbricus (cf. Neuland, fig. 4) and other forms the sperm-ducts lie just 
within the peritoneum. 

The sperm-ducts lie deeper still within the tissues of the body-wall in a few 
other earthworms. I found this to be the case with Acanthodrilus annectens and 
have since recognized the same thing in Acanthodrilus paludosus and in a worm 
belonging to a totally different family, viz. Siphonogaster millsoni; in these examples 
as well as in Diplocardia communis (Gabman, 1) the sperm-ducts are imbedded deep 
within the longitudinal muscular layer. It is quite possible that in other earthworms 
whose sperm-ducts have not been seen, the reason for their invisibility in a dissection 
is due to their lying in this position. 

Finally the genus Sparganophilus (Fam. Geoscolicidae) is unique by reason of 
the fact that the sperm-ducts lie deeper still or more superficially in reality. ■ They 
are placed just beneath the epidermis; in the clitellar region the duct hes at first 
below the entire epithelium ; nearer to the pore it comes to be just underneath the 
superficial epidermis of the clitellum. 

The position of the external orifice differs greatly and is but rarely characteristic 
of a family. As a general rule the two sperm-di«3ts of each side unite to form a single 
tube with a single orifice. In Phreoryctes, however, the two sperm-ducts open on 
to as many consecutive segments ; they have absolutely no connexion with each 
other. An intermediate condition is to be seen in the genus Pelodrilus ; here there 
are two quite separate sperm-ducts which, however, open near to each other on the 
same segment. In the Acanthodrilidae there is a still more marked foreshadowing 
of a fusion between the two sperm-ducts, for they join just before the external 
orifice. In the majority of earthworms, for example in the Perichaetidae, the two 
sperm-ducts join as soon as they can, i. e. in the twelfth segment. 

§ 5. Oviducts. 

In the lower Oligochaeta there are no special conduits for the ova ; the genus 
Aeolosoma is provided with a pore upon the median ventral surface of the sixth 


segment; this pore has been figured by d'Udekem (2), and by Stolc (1), whose 
account, though establishing that of d'Udekem, is more recent, and being the result 
of more refined methods of investigation is important if only as a confirmation 
of the earlier description. 

Among the Naidomorpha similar pores appear to exist. It is among the 
Enchytraeidae that we first meet with indications at least of special tubes which 
conduct the ova to the exterior ; the structure of the oviducts in this group of the 
Oligochaeta is such as to suggest a degenerate condition. Buchholz controverted 
the opinion of d'Udekem that the sperm-ducts served as the conduits for the ova 
also, and suggested that there might be simple pores in the skin through which the 
ova made their way to the exterior ; such pores, however, were not found, though 
Buchholz believed that they were probably small, owing to the ease with which 
the ova could alter their size. Clapakede (3) described in Enchytraeus vermicularis 
a pair of orifices upon the twelfth segment (this position is erroneous) ; these orifices 
were figured (PL IT. fig. 8o) by Clapakede as existing on the same segment as that 
which bears the male pores ; this error is due to the fact that CLAPARiiDE was not 
aware that the setae were absent on the segment which bears the male pores ; hence 
he thought that the pair of setae behind the male pores belonged to that segment ; 
Vejdovskt (3) verified the existence of these pores by treating the living worm 
with a drop or two of Osmic acid which caused the pores in question to open widely 
and occasionally eggs were seen to pass out of them. They were, however, regarded 
as mere pores by Vejdovsky and by Michaelsen (4) in his account of Enchytraeus 
moebii. Michaelsen, however, spoke of ' trichterfdrmige Einsenkungen des Dissepi- 
ments 12/13 in das 13. segment.' And in his many subsequent papers upon this 
group of Oligochaeta he used the same expression. I have given (49) a figure 
of the oviducts of Pachydrilus ; they appear to consist of a few pear-shaped 
cells, not ciliated, which fringe the orifices. The degeneration of these structures 
is curiously paralleled by the degeneration of the spermathecal duct in Nemerto- 
drilus, where an orifice only has been left fringed with very similar cells on its 
coelomic side. 

The oviducts in the Tubificidae and Lumbriculidae are very similar to each other. 
The curious belief that in the former the oviducts form a sheath to the spermiducal 
gland I deal with later. The oviducts in this family, so far as they are known, are, 
as in the Lumbriculidae, short tubes which open into the coelom by a wide funnel. 
It is in the earthworms that the oviducts form long tubes, but not everywhere. 
Generally, indeed, the large funnel is followed by a very short tube leading to the 
exterior. In such cases, e.g. Lumbricus, Perichaeta, the oviduct is very little more 


differentiated than in the Lumbriculidae. But among the EudriJidae, for example in 
Lihyodrilus, it is of considerable length, and not unfrequently possesses a muscular 
sheath. In Alvania it has even a caecum lying close to it and bound up within 
the same sheath. The position of the oviduct or oviducal pore^ as the case may 
be, varies in different families. It is remarkable that in all Megadrili the oviducal 
ducts open on to the exterior in the fourteenth segment. Rhinodrilus proboscideus 
may be an exception, but the statement that the ovary lies in the seventeenth segment 
requires confirmation. In Lihyodrilus the oviducal pores appear to be on the fifteenth 
segment, but a dissection of the worm shows that the septum dividing the fourteenth 
and fifteenth segments lies behind the point of opening of the ducts. 

Considering the close agreement between the male and female gonads the 
differences between their ducts are perhaps more striking than the resemblances. 
There is the general agreement that both consist of ciliated tubes opening into the 
coelom by a wide funnel, often hidden within sperm-sacs or egg-sacs i. In both cases 
the ciliated tube consists of a single layer of cubical cells, which may be surrounded 
by a muscular layer in addition to the peritoneal coating. The caecum of the oviduct 
in Alvania may be compared to the caecum of the sperm-duct in Phreodrilus. In 
Phreorydes too, which appears to me to be undoubtedly an archaic type, the oviducts 
and sperm-ducts (there are two pairs of each) not only correspond in number and 
structure, but the last pair of sperm-ducts is shorter than the first pair and is 
therefore intermediate in length between it and the oviducts. 

In the Microdrili, indeed, the sperm-ducts and the oviducts agree in never occupying 
more than two segments, the funnel lies in one segment and the external pore is in the 
following one or rather between the two, and there is a relation between the position 
of one orifice and the other (see Table above, p. 85). 

In the Megadrili the oviducts invariably occupy only two segments, while the 
sperm-ducts nearly as invariably^ occupy more than two segments. There is, 
moreover, no ascertainable relation between the position of their pores, as the one 
varies, while the other remains fixedly constant. The coiling of the sperm-ducts 
is not paralleled by the oviducts ; and the latter are never connected with glandular 
structures like the spermiducal glands. An apparent exception to this last statement 
is shown by Eudrilus, where the oviduct opens in common with the spermathecal sac 
and a glandular diverticulum apparently belonging to the same. This connexion, 
however, is probably hardly comparable to the connexion between sperm-ducts and 
spermiducal glands, since it is so rare among the Eudrilids (where alone it is found). 

' Most completely in Eudrilidae. 

^ One exception is probably Teiragonurus. 



§ 6. Bevelopment and Homology of the reproductive ducts. 

The generative ducts of the Oligochaeta have for a long time been believed 

to have some connexion with the nephridia, but the precise nature of this relation 

has only quite recently been cleai-ed up; the oviducts are more like nephridia in 

the higher forms than are the sperm-ducts ; to begin with they occupy precisely 

the same number of segments as does a nephridium ; the funnel opens into one 

segment and on the segment behind this is the external pore. This is also the case 

with the sperm-ducts of the Microdrili but not of the Megadrili. Claparede found 

that the genital ducts in the ' Limicolae ' never coincided with nephridia and thus 

came to the conclusion that they were the modified equivalents of the latter. His 

observations turn out, however, to have been inaccurate ; for, although in the adults 

of the worms there are no nephridia in the segments which contain the genital-ducts, 

the nephridia are there in the immature worms, and only disappear on the appearance 

of the latter. The views of Claparede were extended to Lumbricus by Lankestee, 

who pointed out that there was some evidence of the primitive existence of two 

pairs of nephridia per segment in that worm, one series being complete the other 

represented only by the genita.1 ducts ; the intimate relation between the nephridio- 

pores and the orifices of the genital ducts and the setae on the other hand led to 

this view, which was subsequently strongly supported by Peeeier. This naturalist 

found that in some earthworms the nephridiopores were related to the dorsal instead 

of to the ventral setae as in Lumbricus, thus showing the persistence of the presumed 

second series of nephridia, the nephridia of Lumbricus being only partially persistent 

in the ffenital ducts of those worms of which Anteus was an instance. Later Peeriee 

found a worm (Plutellus) in which the nephridia alternated in position, now opening 

by the dorsal now by the ventral setae; in this case, therefore, the assumption was 

that both sets of nephridia partially persisted. The discovery of the occasional 

coincidence of a nephridium and a genital-duct at the same seta finally led Peerier 

to abandon the hypothesis. This difficulty was removed by my own discovery of 

the multiple nephridial pores of Octochaetus and other genera ; and during the progress 

of Perrieb's researches the discoveries of Balfour and Semper of the connexion 

between the excretory and genital systems in the Vertebrata of course strengthened 

the views which favoured the probability of a similar connexion in the Oligochaeta. 

Nevertheless facts seemed to be against any such homology. The development of 

Lumbricus showed, or appeared to show, the entire independence of the two sets 

of structures (see Beegh 5). On the other hand Stolc, from his investigations 

into the anatomy of the sexual organs of the genus Aeolosoma, supported the view; 


he found that in that Annelid the male-ducts were represented by slightly enlarged 
nephridia; an argument derived from a study of the embryology of an Enchytraeid 
led EoDLE to the same conclusion ; he states that the segments in which the male- 
ducts will appear have no nephridia, but that the male-ducts appear rather late and 
so suggest the idea of a pair of slightly modified nephridia which are delayed in 
appearance in connexion with their changed function. Spencee and I at one time 
argued against the connexion between the excretory and genital-ducts from the ground 
that in Perichaeta, which we believed to represent a highly archaic form, the genital 
ducts showed no trace of their supposed origin from nephridia, the latter being 
in a very primitive condition. Spencee also pointed out that the genital ducts have 
an intercellular lumen while the nephridia have an intracellular lumen ; this argument 
is, however, at best not a strong one, and, as has been mentioned, Vejdovsky says 
that the lumen in the nephridia is really also intercellular. The only positive 
evidence as to a connexion between the nephridia and the genital-ducts has been 
brought forward by myself. In Octochaetus multiporus the genital-ducts appear to 
be formed out of a part of the pronephridia, thus confirming the suggestion of Balfoue 
(Comp. Embryol. vol. ii. p. 617) that 'in the generative segments of the Oligochaeta 
the excretory organs had at first both an excretory and a generative function, and that 
as a secondary result of this double function each of them has become split into two 
parts a generative and an excretory.' The actual facts which I brought forward upon 
the development of the genital ducts are the following : at a comparatively late stage 
in the development of this worm, after the pronephridia have lost their distinctive 
character and have acquired numerous openings on to the exterior, the proximal 
part of the nephridium which consists of the remains of the funnel (the cilia have 
disappeared) and a straight tube leading to the body-wall separates off from the rest 
of the nephridium ; the funnel grows and re-acquires cilia ; the tube grows into the 
body- wall and becomes the genital duct; the genital ducts have precisely the same 
structure at first whether they are to become oviducts or sperm-ducts ; they are, 
moreover, only to be distinguished from the corresponding remains of the funnel 
and the first part of the nephridium in the preceding and succeeding segments by 
their larger size ; it should be mentioned also that there are at first traces of four 
genital ducts in correspondence with the four gonads. It is very remarkable that 
in the case of four pairs only of the pronephridia the commencing degeneration 
should be arrested and growth recommence ; the tube, at first hollow, becomes solid 
and then re-acquires a lumen ; an analogy to this state of affairs is, however, oflfered 
by the occlusion (temporary) of the lumen of the oesophagus in more than one 
Vertebrate. The disappearance followed by the reappearance of cilia upon the 


funnels is also not without its analogies ; particularly among the Protozoa where 
the vanishing of cilia so far from being an indication of degeneration is actually 
the prelude of renewed activity. In any case the facts appear to be as stated in 
Octochaetus. Another point to be emphasized in connexion with the development 
of the genital-ducts is their early appearance in Octochaetus as compared with 
Lumhricus. The facts which I have made out as to the development of the genital 
ducts in Octochaetus appear to absolutely contradict the facts established for Lumhricus 
— so much so, indeed, that it seems hardly credible that both can be correct. A little 
consideration, however, I think shows that there is not necessarily any contradiction. 
In Lumhricus the ducts appear after the nephridia have acquired their definite form ; 
there is every reason, therefore, why they should not show any actual connexion 
with them in development, since presuming their homology with the genital ducts 
of Octochaetus the latter are formed out of the remains of a part of the pronephridia. 
The genital ducts of Lumhricus are formed so late that they cannot be produced out 
of the pronephridia which have been in the meantime converted into the nephridia. 
Another point to be observed is that in Octochaetus the polynephric condition is to 
some extent established before the commencing differentiation of the sperm and 
oviducts. It is possible therefore to regard the existence of both in the same segment 
in Lumhricus as the last remnant of an ancestral condition where the nephridia were 
numerous in each segment. 

As tending to prove that in some ' meganephric ' worms at any rate the funnels 
of the genital-ducts are formed out of the funnels of the pronephridia I may say 
that I could find no nephridial funnel in the tenth, eleventh, and thirteenth segments 
of Alvania millsoni, but I did find them in the fifth to ninth and in the twelfth and 
fourteenth segments. In GordiodrUus there were none in the thirteenth at any rate. 

§ 7. Spermiducai Glands^. 

In some Oligochaeta the sperm-ducts open on to the exterior directly; in others, 
and these are the majority, the sperm-ducts open into a wide terminal chamber 
which itself opens exteriorly; this is the case with the LumbricuUdae, the Tubificidae, 
the Perichaetidae, many Cryptodrihdae, the EudriUdae, the Naidomorpha, the Chaeto- 
gastridae, the Moniligastridae, the Enchytraeidae ; in a limited number of genera of 
Cryptodrilidae and in all the Acanthodrilidae there are these glands, but the sperm- 
ducts open separately on to the exterior though in their immediate neighbourhood. 

1 I prefer this name, recently suggested by myself (80) to either atrium or prostate as it emphasizes 
the position and relation of the glands, and having been never used before has no preconceived meaning 
attached to it. 




The Lumbriculidae, the Phreoryctidae, the Geoscolicidae (for the most part) have no 
such glands at all; and in these worms the sperm-duet is with a few exceptions un- 
provided with any glandular apparatus at its terminal orifice. 

The simplest form of the terminal chamber is found, as might be expected, in 
the lower aquatic Oligochaeta ; in the Naids its structure is as follows : the sperm- 
ducts lead on each side of the body into a pear-shaped sac, which has been called 
the ' atrium ' and opens on to the exterior in the sixth segment ; this sac is lined by 
a glandular epithelium apparently without cilia; externally to this epithelium is 
a covering of peritoneal cells ; the ' atrium ' in Nais elinguis gradually passes into 
the sperm-duct; in Stylaria lacustris and in Bero there is an abrupt break between 
the two ; I cannot discover any positive statements as to whether the lining cells 
Fig. 47. ^'I'e or are not ciliated in the Naidomorpha ; there is no indication 

of any ciliation in the figures of Stolc (5) and Vejdovsky (24). 

Among the Lumbriculidae we meet with a terminal chamber 
which is modelled upon the same plan as that of the last family. 
Among the Enchytraeidae there is very commonly a homologous 
organ of an equally simple, though rather different structure. 
I follow here Michaelsen's account of ETichytraeus humicultor. 
It is a rounded or oval body (called by Michaelsen a ' penis '), 
lined with long cells. It communicates with the exterior by a 
short invagination of the latter, which is beset with groups of 
unicellular glands. The sperm-duct opens at the summit of the 
terminal gland which has a muscular layer outside the lining 
epithelium. ^ 

The Tubificidae have an ' atrium ' which is more complex 
than that of the last two families. In Tubifex itself, which 
may serve as a type of the family, the organ has been described 
by a large number of writers; it is an elongated sac receiving 
the sperm-duct at one end and opening on to the exterior at 
the other; the proximal part is ciliated; the distal region is 
not ciliated; the latter forms a protrusible penis which is 
described more fully in a separate section (see below). A 
remarkable feature of the atrium in the Tubificidae, with the 
exception of a few forms, is the presence of a glandular appendage variously termed 
'Cement gland' (Lankestee), ' Cementdriise ' (Vejdovsky), 'Prostate' (Eisen, &c.); 
this is a thick patch of pear-shaped glandular cells whose ducts (merely the pro- 
longation of the cells themselves) open into the lumen; Vejdovsky has shown that 


apparatus 01' 


(After Stole.) 

I. Funnel, a. Penis. 3, 

^Atrium.' 4. 'Prostate. 

5 Coiled sperm-duct. 



the origin of this body is an outgrowth of the lining epithelium of the atrium ; it is 
a curious fact that the covering of peritoneum upon the atrium ceases at the point 
where the ' prostate ' is attached, a fact which, if the development of these parts was 
not known, might give rise to the idea that the 'prostate' was nothing more nor 
less than a tract of peritoneum modified to serve a special function. Nasse speaks 

Fig. 39. 

Fig. 28. 



(After Stole.) 

I. Funnel, z. Sperm-duct. 3. Penis. 
4 Muscles for its retraction. 5. 

Glandular covering of 6. ' Atrium.' 



(After Stole.) 

I. Proximal region of 'atrium.' 2. 'Para- 
trium.' 3 Distal region of ' atrium.' 4. Penial 
setae. 5. External pore. 6. Muscles for pro- 
trusion of distal end of efferent apparatus. 
7. Muscle attached to sperm-duet (8). 9. Funnel. 

of a fine lumen to each group of cells in the prostate ; but it is fairly clear that 
the crland is solid: the cells of which it is composed appear to correspond with 
other unicellular glands found so commonly among earthworms ; for example, the 
capsulogenous glands of the Perichaetidae ; the 'atrium' being developmentally 
an involution from the outer layer of the body it is not remarkable to find 

p a 


that gland-cells ai-e developed in connexion with it just as gland-cells frequently 
underlie the epidermis. 

This form of atrium is found with but slight modifications in the majority of 
the genera of Tubificidae (woodcut, fig. 37) ; Ilyodrilus (woodcut, fig. 38) has, 
however, an atrium precisely similar to that of the Naidomorpha ; in Bothrio- 
neuron and in Branchiura the atrium is rather difierent, and the state of affairs 
characteristic of these two genera suggests another interpretation of the morphology 
of the parts in Tubifex; in both these last mentioned Tubificidae (woodcut, fig. 29) 
the vas deferens widens out into a terminal chamber which appears to be the 
atrium ; but this is provided with a single lateral diverticulum (fig. 29, 2), very much 
larger in Branchiura than in Bothrioneuron ; in Branchiura the lateral chamber 
has precisely the structure of the atrium in the Lumbriculidae. It is joined by the 
sperm-duct just where it passes into the muscular atrium ; there can be no doubt in 
my opinion of the relationship between Branchiura and Tubifex; hence we should 
expect to be able to furnish a comparison of the different parts of the efferent ducts 
in each ; two views seem to be possible ; either we must simply consider that the 
sperm-duct has come to open into the atrium some way from its extremity, or we 
must regard the sac-like appendix as the equivalent of the prostate of the typical 
Tubificidae ; in this event the solid character of the prostate in the latter will be 
a secondary matter ; Stolc (3) is apparently of opinion that the small lateral 
caecum attached to the atrium of Bothrioneuron (called 'paratrium' by Vejdovsky) 
is the homologue of the prostate of other Tubificidae, as he indicates both by the 
same letters in his plates. 

The genus Tdmatoclrilus has an 'atrium' which is intermediate in character 
between that of Tubifex or Psammoryctes and the more typical Tubificidae on the 
one hand, and Branchiura and the Lumbriculidae on the other: in that genus 
EiSEN described and illustrated the presence of a series of separate 'prostates' 
opening at intervals through gaps of the proper walls of the 'atrium' into its 
interior ; if these separate masses of glandular cells were combined into a continuous 
glandular structure we should have an 'atrium' exactly like that of Lumbriculidae, 
except for the absence of muscles. 

In the Megascolicidae the corresponding organ is constructed upon one or other 
of two types. In Acanthodrilus, for example, it is a tubular structure of varying 
length, occupying more or fewer segments, coiled or straight, which is divisible 
into two regions ; the part of the tube which leads to the exterior is of less 
calibre and has generally a glittering appearance; this region serves as a duct 
for the glandular secretion of the distal region; if is lined by a single layer of 


non-glandular epithelium and has thick muscular walls ; it is these which give the 
'nacreous' appearance already referred to; the glandular part of the tube is of 
greater calibre, and has an opaque white colour, and a rough exterior contrasting 
with the smooth external walls of the proximal part of the tube. This glandular 
section of the tube consists of two distinct layers of cells. The innermost coat is 
formed of a layer of not very deep columnar cells, often loaded with granules; 
below these are several layers, forming a stratum of considerable thickness, of pear- 
shaped cells with long processes, which penetrate between the columnar cells and 
therefore abut upon the lumen, being thus in a position to pour their contents 
directly into the lumen ; the entire structure of the lining epithelium is in fact very 
suggestive of that of the clitellum ; there is the same specialization of its epithelium 
into two sorts of cells. Outside the epithelium is the peritoneal membrane. This 
kind of gland is found not only in the genus Acanthodrilus, but also in very nearly 
all the members of the family Acanthodrilidae ; it also occurs in such Cryptodrilidae 
as have a tubular gland of a similar form to that of the Acanthodrilidae, with a few 
exceptions to be referred to, and finally in a few Perichaetidae. There are hardly 
even differencies of the minutest character whicli distinguish the gland in these 
various types ; sometimes the innermost epithelium is more plainly columnar and 
not so glandular as has been described ; but this is very possibly merely a matter 
of more or less activity in the secreting processes. Ocnerodrilus (including a few 
allied genera which I group together near this) and Kerria are the only exceptions, 
among the Megascolicidae which have a gland of a tubular form, to the description 
just given; in these genera the difference is in the fact that the lining of the entire 
tube is formed by a single layer of cells only. 

The second form of the spermiducal gland seen in the Megascolicidae characterizes 
the Perichaetidae and is to be found in a large number of Cryptodrilidae ; it occurs 
only in one Acanthodrilid. Here we have first of all the same differentiation of the 
gland into a glandular and a non-glandular portion ; the external duct is the non- 
glandular part and its length varies much in different species ; certain species of 
Perickaeta, for example, are characterized by this duct being curved into a horseshoe- 
shape, and diminishing in calibre towards its external aperture ; in others, on the 
contrary, it is short and straight ; in none is it absent ; it is, as in the case of the 
gland of the first kind, muscular, with a lining of columnar cells. In these worms 
the glandular part does not form a comparatively narrow tube of equal calibre 
throughout ; it has the appearance of a racemose gland much divided up into lobules ; 
the lobulation is sometimes so pronounced as to produce a very loose texture of the 
gland ; sometimes the organ is more compact ; it is also sometimes larger and some- 


times smaller ; in Perirhaeta taprobanae, for example, the spermiducal gland is so 
small as to be entirely contained within one segment ; in other species of Perichaeta 
it extends through a considerable number of segments. The peculiar appearance of the 
gland is produced by the branching of its lumen ; the tubes are lined by low columnar 
epithelium which does not appear to be ever markedly glandular in chai-acter; 
attached to the tubes are groups of pear-shaped cells massed into bundles, whose fine 
processes seem to open into the lumen between the non-glandular cells which line it. 
It is this division of the lumen coupled with the grouping of the glandular pear- 
shaped cells that gives its peculiar appearance to the spermiducal gland in the 
Perichaetidae, &c. The whole organ is covered with a fine covering of peritoneum. 
It will be evident from the figures illusti-ating the minute structure of the glands 
(woodcut, fig. 31) that there is no essential diflference between this type and the tubular; 
the difference lies in the fact that in the Perichaetidae the single tube has become 
branched and the glandular lining has become grouped instead of remaining a con- 
tinuous layer ; we find that genera very nearly allied in other particulars differ as to 
whether they possess a tubular or racemose spermiducal gland ; besides, as I have 
pointed out, there are among the Perichaetidae glands which seem to be intermediate 
between the two extremes ; the branching is much reduced and as a consequence the 
breaking up of the glandular cells into groups is not so marked. 

In the case of both kinds of glands the relations of the sperm-duct are pecu- 
liar; it never opens into the glandular part. As a rule the opening is into the 
muscular duct just at its commencement ; this rule has apparently no exceptions 
in the Perichaetidae (see, however, the remarks upon Perichaeta ceylonica, below) ; 
it is, however, not so common among those genera which have the tubular variety 
of the gland ; in Pontodrilus the sperm-duct has these relations ; but in no member 
of the family Acanthodrilidae has the sperm-duct any direct connexion with the terminal 
gland at all. In every species it even opens on to a segment distinct from that 
which bears the orifice of the these ; between this extreme and the other there are 
various intermediate stages ; thus in Microscolex novae-zelandiae the sperm-ducts 
open into the spermiducal gland just before the latter opens on to the exterior ; in 
Typhaeus the orifices are separate but are situated upon the same segment. There are 
as a rule but a single pair of glands in the Megascolicidae ; but exceptions are known ; 
thus with the exception of Acanthodrilus monocystis the Acanthodrilidae have always 
two pairs opening on to the seventeenth and eighteenth segments ; two pairs also 
characterize the genus Gordiodrilus. In the latter case they are in consecutive 
segments; a unique disposition of the glands is afforded in the aberrant species of 
Perichaeta — P. ceylonica; there are here two pairs, which are in the same segment 


upon which presumably the sperm-ducts open (the eighteenth) ; the two glands of 
each side are, however, not similar to each other ; one is of the tubular variety, the 
other like that of other Perichaetidae, i. e. ' racemose.' I was unable in examining 
this species to discover with which, if either, of the glands the sperm-duct was 

Developmentally it might appear that the glandular part of the spermiducal gland 
is distinct from the duct ; I have occasionally observed in immature Perichaetidae 
the duct alone present ; and there are several species (Perichaeta masaJcatae, &c.) 
in which the glandular appendix never seems to put in an appearance. 

The Eudrilidae have a characteristic spermiducal gland. It is superficially like 
that of the Acanthodrilidae but really dififers in a number of peculiarities, which 
together form one of the principal reasons for regarding the Eudrilidae as so 
distinct a family of Terricolae, and for removing it from the neighbourhood of the 
Cryptodrilidae. The gland has the same sausage-like form as in the Acanthodrilidae, 
but has almost always a nacreous glitter owing to its thick muscular coat. The 
main differences which distinguish the organ of the Eudrilidae from that of all 
Megascolicidae are : — 

(]) The usually thick muscular layer; 

(a) The opening of the sperm-ducts into the glandular part ; 

(3) The presence of a terminal muscular sac (bulbus). 

As a general rule the glands in the Eudrilidae are distinguishable, as they are 
in Acanthodrilus, &c., into a thicker glandular portion and a thinner duct, which is 
lined by a single layer of non-glandular cells. 

This is particularly well marked in the genus Eudrilus itself, where it has been 
figured by Peeriee, and by myself (see woodcut, fig. 30). 

In Eudrilus the muscular duct is very slender indeed as compared with the 
glandular tube from which it arises. 

In a few other cases, however, there is no such abrupt transition. In Nemerto- 
drilus, for example, the gland gradually dwindles towards the extremity which 
bears the external pore; at the same time there is an increase, though a slight one, 
in the thickness of the muscular coat ; and the lining epithelium becomes one layer. 
In Heliodrilus there is the same absence of any pronounced demarcation between 
the two sections of the tube, and the glandular lining becomes gradually one cell 
thick ; it retains, however, its glandular character, being formed of large oval cells 
between which are narrow packing cells ; but these disappear just before the opening 
into the terminal sac. 



Evdrilus and Lihyodrilus are among the genera whose spermiducal glands are 
furnished with very thick muscular walls ; the fibres are arranged in two layers ; 
the outermost, which is the thinnest, consists of longitudinally running fibres ; it is 
these which give the nacreous appearance to the organ ; inside the longitudinal is 
a thicker layer of circular fibres. In Nemertodnlus the muscular 
layer is greatly reduced, but it still consists of the two layers 
of fibres. 

The position of the opening of the sperm-ducts into the 
glands also varies considerably in difierent genera; they never, 
however, open either independently of them or into the terminal 
bulbus. The apparent opening is nearly always different from 
the real opening. For instance in Eudrilus (cf. woodcut, fig. 
30) the two vasa deferentia pierce the walls at a point which is 
roughly half-way along the gland ; microscopic sections, how- 
ever, show that the tubes run side by side between the muscular 
and epithelial walls of the gland to a point much nearer to its 
blind extremity, before opening into the himen. In Lihyodrilus 
the same thing occurs and the actual orifice is near to the very 
summit. Neniertodrilus occupies the opposite extreme in the 
series. The sperm- ducts here open into the glands nearer to 
the external pore. 

The third character of the spermiducal gland in these worms 
is also subject to some variation. The terminal bulbus is larger 
or smaller according to |he genus. It is very conspicuous in 
Evdrilus and in Pareudrilus. In the former of these two 
genera it is a rounded sac which appears on dissection like a 
hemispherical thickening of the body-wall. From its internal 
walls arise two processes which, together with the peculiar 
U-shaped gland connected with one of them, will be referred 
to under the description of the penis of the Oligochaeta. 

At the opposite extreme perhaps is Heliodrilus, where the 
bulbus is reduced in size to be merely a slight depression of 
the external skin into which both the glands open. This 
terminal sac is protrusible and is generally found more or less protruded in individuals 
killed with alcohol. 

Finally, it should be mentioned that in Eudriloides hrunneus, at any rate, tracts 
of epithelium in the glands are ciliated. 


■V. d. Sperm-ducts. pr. 
Spermiducal gland, a. Ap- 
pendices of terminal mus- 
cular sac. 


I select Kynotus michaelsenii as the type of the Geoscolicidae, as it has been 
very carefully described by EosA, though unfortunately his description is unaccom- 
panied by any illustrations. 

The terminal apparatus of the male efferent ducts in this Annelid consist on 
each side of the body of a long oval muscular sac extending through four segments ; 
posteriorly and externally a long tube arises from this which reaches back through 
another seven segments ; the terminal sac is attached to the parietes not only by its 
own walls, but by a moderately long muscular slip. The terminal sac has very 
thick walls, which as a consequence reduce the extent of the lumen ; its interior is 
divided by an incomplete septum into two chambers ; the upper of the two chambers 
is furnished with a shield- shaped thickening of the internal wall, on to which opens the 
glandular appendix to be presently described. The epithelium lining the ventral surface 
of the two chambers is like the external epidermis, but has rather more numerous 
glandular cells ; the dorsal chamber, on the other hand, is lined by a tall columnar 
epithelium without glandular cells like those of the lower chamber, but with 
glandular cells staining deeply with carmine ; the entire sac has a chitinous 
lining. The glandular appendix already spoken of opens on to the shield-shaped 
projection of the upper chamber; the canal which there opens has the same minute 
structure as that of the sac into which it opens ; this becomes further back a tube 
lined with two layers of epithelium as in Acanthodrilus, &c., ensheathed in a 
common peritoneal coat which does not follow the windings of the contained tube. 
The sperm-ducts penetrate the terminal sac near to its external orifice and pass 
along the thickness of its wall, finally opening into the lumen of the glandular 
appendix where it retains the non-glandular character of the terminal sac. Rosa 
considers that the organ in the Geoscolicidae is not the homologue of that of other 
Oligochaeta and terms it ' Pseudo-prostate.' 

It does not, however, appear to me to be possible to draw this hard and fast 
line between the Geoscolicidae and the Megascolicidae that Rosa wishes to draw; 
in the first place, as I have attempted to show, the Eudrilidae seem to agree more 
nearly with the Geoscolicidae than with the Megascolicidae ; in both of them the 
essential diflFerence is in the presence of a muscular terminal sac into which the 
spermiducal gland proper or the glands in the case of most Eudrilidae open. Rosa 
has, I think, overlooked the fact that in the genus Perichaeta something of the same 
kind also exists. In the systematic part of the present work I have laid some little 
stress as a specific character upon the fact that in some species the narrow muscular 
duet of the gland does not communicate directly with the exterior, but opens 
into the interior of a variously-sized sac ; this latter may be of large size, as in 




Perichaeta indica; or it may be almost absent, as in Ferichaeta lesperidum; 
probably, though I have no positive data upon the subject, it is extrusible and 

Fig. 3'- 


A. Svtroa. B. Monaigaater. C. Pontodrilus. D. Acanthodrilus. E. Eudrilus. 
F. FeriOmeta. G. TuUfex. p. Peritoneum. v. d. Vas deferens. b. Epithelial 
lining of spermidnoal gland, u. Glandular cells, m. Muscular layer sp. Spermatozoa. 

performs the function of a penis. In any case, whatever its function may be, it 
seems to be fairly comparable to the terminal muscular sac of the Eudrilidae and 


the Geoscolicidae. No other genus of Megascolicidae, so far as I am aware, shows 
the same terminal sac ; this fact is not without its importance ; it is so far an 
indication that the Perichaetidae, having preserved this structure, are nearer to the 
Geoscolicidae and the Eudrilidae in which it is better developed than any other 
sub-family of the MegascoUcidae. I use this as an argument in favour of the low 
position of the Perichaetidae in the series (see below). 

Another reason which led KosA to dissociate the 'pseudo-prostate' of the 
Geoscolicidae from the 'prostate' of other earthworms was the existence of special 
retractor muscles in the former. These, however, are not absent from the latter. 
In the genus Octochaetus, for example, there are numerous bands of muscular tissue 
which are inserted on to the bodj'-wall in the immediate neighbourhood of the 
male pores. A better example still is afforded by two species of Eudriloides (see 
Beddaed 84), where there is a complex system of retractor and protractor muscles 
attached to the ducts of the glands themselves. 

It may be useful to tabulate the principal varieties of the spermiducal glands 
which are illustrated in the accompanying diagram (woodcut, fig. 31). 

A. Glands with a double lining of cells. 

a. Tubular, e.g. Acanthodrilus. 

b. Racemose glands in which lumen has become branched and outer glandular 

layer of cells disposed in discrete tufts instead of forming a continuous 

e. g. Perichaeta. 
A'. Tubular glands with a double lining of cells covered externally by a layer of 
muscles ; terminal extremitj' modified into a copulatory apparatus, 
e.g. Eudrilus, Kynotus. 

B. a. Spherical or more elongated glands with two layers of cells separated by 
a muscular layer. 

e.g. Rhynchdmis, Moniligaster, Branchiura. 
b. Outer layer of cells grouped into separate masses. 

e. g. Sutroa, Telmatodrilus, Tubifex (one mass only present). 

C. Tubular glands with single lining of cells. 

e. s- Ocnerodrilus, Kerria. 
The structui-e which is here termed spermiducal gland has been called 'prostate' 
and 'atrium' by several; the distinction of terms implies a difference in homology 
between the different appendages of the sperm-duct, which aa-e by all called ' atrium ' 
in the aquatic Oligochaeta. Does this difference really exist, or are all the glandular 
sacs at the end of the sperm-duct homologous throughout the Oligochaeta? I myself 

Q 3 


incline to the latter view, and will here state my reasons for so doing. The fii'st 
reason which leads me to this opinion is entirely a priori ; this is, however, in the 
present case a more powerful reason than it might be supposed to be, and than it 
would be perhaps in some other cases. The Oligochaeta so evidently form a single, 
well-definable group that the structures of the terrestrial forms may be confidently 
expected to be represented in the aquatic genera; at any rate we know of no 
structures which are peculiar to one or the other except those now in question. 

No one, I imagine, will doubt that all the structures called ' atrium ' in the 
aquatic genera are homologous with one another. With inconsiderable exceptions 
these 'atria' receive the sperm-ducts; they have even been (erroneously) spoken of 
as an ' enlarged part of the sperm-duct.' In the Eudrilidae and in those Geoscolicidae, 
where they exist, the glandular tubes at the male pore also receive the sperm-ducts ; 
in Eudrilua itself, for example^ the sperm-ducts open into the middle of the 
glandular tube ; it can hardly be doubted that in this case we are dealing with 
a structure that does accurately correspond to the 'atrium' of the lower Oligochaeta. 
In the Acanthodrilidae, for instance, another arrangement occurs : here the structures 
which I have termed spermidueai glands open quite independently of the sperm-ducts, 
even on to a difierent segment; and yet the histology of the gland is in its main 
features quite like that of the Eudrilidae. The link between the two is aflforded by 
a series of genera ; in Pontodrilus the gland is not independent of the sperm-duct ; 
but the sperm-duct opens into the muscular duct and not, as in the Eudrilidae, 
into the glandular region ; in Typhaeus and in Microscolex the sperm-duct only 
communicates with the gland just before the opening of the latter on to the exterior; 
in Gordiodrilus the external pore of the gland is perfectly distinct from the pore 
of the sperm-duct but is placed upon the same segment; there are thus various 
intermediate conditions between the extremes. It can hardly be doubted that these 
latter genera possess organs which are homologous in spite of the rather different 
relation of the sperm-duct to them : in the Eudrilidae themselves the position of the 
opening of the sperm-duct into the gland varies in different genera. To insist 
upon a difference between the spermidueai gland of the Eudrilidae and that of the 
Acanthodrilidae, because in one there is a direct connexion between it and the 
sperm-duct, and in the other there is not, seems to me from a consideration of the 
intermediate stages to be absurd; and if we apply this argument to the aquatic 
Oligochaeta the absurdity is even more apparent. In the genus Branchiura which 
I have recently described, the 'atrium' is appended as a diverticulum to the sperm- 
duct ; the two open in common, but the ' atrium ' does not receive the sperm-duct 
at its summit, but at the point where it passes into a muscular duct which leads to 


the exterior ; in spite of this difference, who could deny that the structui-e which I 
have called ' atrium ' in Branchiura is the exact homologue of the ' atrium ' in other 
Tubificidae? To suppose that two structures so similar and yet morphologically 
different could exist in the genera of a limited family like that of the Tubificidae 
is to suppose too much, and to go counter to plain facts. Still, if Benham was 
right in calling the gland appended to the male duct in Pontodrilus ' prostate,' and 
the similar gland in EvArilus ' atrium,' it will be necessary to explain how it is 
that structures which are of different morphological import have come to possess an 
almost identical structure. And if the term 'prostate' is retained 'for those glands 
which either pour their secretion- into the sperm-duct or open independently to the 
exterior,' then the structure which I have called 'atrium' in Branchiura cannot be 
the homologue of the ' atrium ' in other Tubificidae and Lumbriculidae. This appears 
to me the reductio ad ab&urdum, and to dispose of the necessity for further argument. 
There has been, however, some little confusion as to the meaning to be attached to 
the term 'prostate,' independently of the facts ali-eady referred to. I have called the 
glandular investment of the ' atrium ' in the Lumbriculidae ' prostate,' and compared 
this glandular investment to the ' Cementdrusen ' of Tuhifex : it may be that I have, 
as Benham suggested, compared together structures which cannot be compared, since 
they are respectively epiblastic and mesoblastic in origin; in any case I did not 
seriously make that comparison after deliberation; I am now inclined to think that 
that comparison is after all the right one, provided only that embryology confirms 
it. The glandular investment of the ' atrium ' in the Lumbriculidae does not appear 
to be peritoneal ; we cannot, however, be certain, though it is so exactly like the 
glandular investment of the ' atrium ' in Moniligaster which is probably not peritoneal. 
I have seen sections through the ' atrium ' of Moniligaster in which the thick layer 
of cells investing it externally is covered by a thin layer of what is undoubted 
peritoneum ; and there is no doubt that the layer of cells in question opens by 
prolongations of the cells into the lumen of the gland — a state of affairs which is 
not suggestive of its being in reality peritoneum, as I at first believed it to be. 
Vejdovsky, too, has figured a similar prolongation of the cells of the glandular 
coating of the ' atrium ' in Ehynchelmis to join the lumen ^ ; the same thing undoubtedly 
exists in Sutroa ; in all these cases, therefore, I believe that we have to do with 
a second layer of epithelium of epiblastic origin; the absence, if it be ultimately 
proved, of a peritoneal layer in certain Lumbriculidae, and in some species of 
Moniligaster does not appear to be a matter of the greatest consequence, since in 

' Moreover, Benham has described a delicate membrane surrounding the pear-shaped cells which is 
probably the coelomic epithelium. 


the 'atrium' of Tuhifex the peritoneum is absent from the ' Cementdriise,' which has 
been shown by Vejdovsky to be an outgrowth of the lining membrane. I would 
therefore compare the 'Cementdriise' of Tuhifex to a portion of the glandular 
investment of the atrium in such forms as the Lumbriculidae ^ ; this view of the 
relations of the different parts brings matters into a far more satisfactory condition; 
and it helps us also to get a clearer insight into the meaning of the apparently 
great difference between the spermiducal gland of the Perichaetidae and the Acan- 
thodrilidae — in fact to compare more readily the ' lobate ' with the ' tubular ' form ; 
in the class of gland which I have termed 'lobate' the lumen is much branched, 
and the outer layer of glandular pear-shaped cells, instead of forming as it does in 
the Acanthodrilidae, and in other genera which have the tubular form, a continuous 
covering is broken up into masses of cells; now we get something very much like 
this in the genus Telmatodrilus among the Tubificidae, and in Sutroa among the 
Lumbriculidae ; but in neither of these genera is there any corresponding branching 
of the lumen of the gland itself. Those Perichaetidae in which the gland is the 
most compact, in which the branching is not so conspicuous, show the earlier stages 
in the conversion of a tubular into a lobate gland — for example, Megascolex 
newcotnhei, while Diplocardia is an almost ideal intermediate form ; there is in 
fact no difEculty in getting the one form of spermiducal gland out of the other. 

I have now to consider certain points in the histology of the glands which 
might seem to militate against a comparison between those of the aquatic and the 
terrestrial genera. In the former the cavity is lined by a single layer of cells often 
ciliated, which are separated by a layer of muscles from the outer layer of cells 
which have been spoken of by some as ' prostates.' In the terrestrial Oligochaeta, 
on the other hand, with a tubular spermiducal gland, the lining membrane is 
with a few exceptions always composed of two distinct layers ; its resemblance 
in fact to the clitellum has been often insisted upon ; I was inclined at one time to 
contrast these two forms^ and to connect the resemblance of the spermiducal gland 
of the higher Oligochaeta with the clitellum with its origin from the clitellar region 
of the body : I do not now think that the difference is a real one ; in such 
genera as Moniligaster it seems to me merely that the inner epithelial layer has 
been removed a little way from the other layer, so that a layer of muscle has come 
to intervene ; just as certain glandular cells, undoubtedly of epidermic origin, have 
come to be imbedded in the musculature of the body-wall; the two cases seem to 
me to be perfectly parallel. This view of the matter has been put forward by EosA 

' These views with regard lo tlie homologies of differently named structures are now those of Beniiam 
(25) as well as of myself (80). 


(4, p. 385) with regard to Desmogaster : he pointed out that in the worm the oute- 
layer of glandular cells were imbedded in a stratum of muscle, and that the atrium 
■was so far intermediate between that of Moniligaster and Acanthodrilus, &c. ; the 
muscles had not quite got to divide the two epithelial layers clearly from each 
other. Still, of course, this does not take away from the closer resemblance of the 
gland of Moniligaster to that of the Lumbriculidae than to that of the terrestrial 
earthworms. The occasional though rare ciliation of the lining epithelium in Eudrilidae 
removes all the differences that separate the spermiducal gland of the higher from 
that of the lower Oligochaeta. It will be seen therefore that the 'prostate' of 
Ferichaeta may be safely compared with the ' Cementdriise ' of Tubifex without doing 
any harm ; but that the insistance upon this homology must not be carried so far as 
to obscure the other obvious relations between the spermiducal glands of different 

A question which now requires consideration is the origin of the spermiducal 
glands ; are they to be regarded simply as dilatations on the sperm-duct or as 
separate structures which have come to have a relation to the sperm-ducts 1 I incline 
to the latter view. The chief reason which leads me to take up this position 
is the existence of supplementary glands which have no relation to the sperm- 
ducts ; in Bichogaster damonis the two segments following that upon which 
the sperm-ducts open are each furnished with a pair of tubular glands exactly like 
the spermiducal glands in structure, but rather smaller : it is also remarkable 
to find that these two pairs of glands open on to the exterior in exactly the 
same position as those of the seventeenth segment, and that the ventral setae of 
their segments, as of the seventeenth, are missing ; nothing in fact is wanting to 
complete their likeness to spermiducal glands, except that they have no direct 
relation to the sperm-ducts. The occasional presence of two pairs of glands to only 
a single pair of male pores, as we find in Acanthodrilus, Gordiodrilus, and Ferichaeta 
ceylonica, is not so remarkable, for in all these worms there are two pairs of distinct 
sperm-ducts, although they become one at the external pore. This view of the origin 
of the glands is the one held by RosA; in a paper upon the structure of Kynotus 
michaelsenii this author refers the spermiducal glands to the same category as certain 
glands found in that species in the segment following, and showing exactly the same 
minute structure. These glands are accompanied by penial setae. Rosa, it should be 
stated, is not of opinion that this view can be applied to the glands of other 
earth-worms ; he is only considering the Geoscolicidae, which according to him have 
' pseudo-prostates,' not comparable to the apparently similar glands in the Oligochaeta 
generally; this opinion of Rosa^s is not one that commends itself to me, and I have 


already attempted to show that the spermiducal glands in the Geoscolicidae are like those 
of the Eudrilidae more than any other group. In Microcheta benhami, also, there are a 
series of quite analogous glands which are referred by KosA to the same category as the 
spermiducal gland ; he also holds this view with respect to the posterior glands of 
Pontoscolex, Urobenus, and Brachydrilus ; as regards this latter comparison I may point 
out that the connexion of the glands of Pontoscolex- with the nephridia is curiously 
paralleled by the connexion of the spermiducal gland of ffeliodrilus with the nephridia 
of its segment ; this connexion, however, may not be more than accidental in the latter 
case. The existence of this relation, especially in Pontoscolex, suggests that perhaps 
Rosa has not gone quite far enough back in seeking for the origin of the glands. 
It may be that the glands, with which I think with Rosa that the spermiducal glands 
are homologous, are themselves derived from the nephridia ; as we now know that the 
sperm-ducts are homologous with nephridia, the connexion of the sperm-ducts with 
the spermiducal glands may be in this case comparable to the connexion of the nephridia 
of Pontoscolex with the posterior glands ; this, however, does away with the significance 
of the connexion between the two in Heliodrilus. The glands in Kynotus lying behind 
the spermiducal glands are provided with modified setae, and thus the resemblance 
to the spermiducal glands, as they usually are, is completed; we must not, perhaps, 
leave out of consideration the anteriorly situated glands, also provided with modified 
setae in a few species of Acanthodrilus (s. 1.), &c. ; these are very likely to be placed 
in the same category. It is possible that the glands which are so often found in the 
neighbourhood of the male pores and the spermathecal pores in the Perichaetidae 
should be also referred to the same series ; but it may be remembered that these 
latter have no lumen, and would therefore, have to be looked upon as much 
degenerated. StiU their frequently paired arrangement, corresponding to that of 
spermiducal glands, is an argument to be borne in mind. The spermatheeae also 
suggest the same origin, but I deal with their homologies under the heading 
' Spermatheeae ' (see below). 

Closely connected with the last question, and, of course, with the phylogeny 
of the Oligochaeta, is another question: what is the most primitive form of the 
spermiducal gland? If we accept their serial homology with the. copulatory glands 
it is evident that those spermiducal glands which are structurally most like the 
copulatory glands will have to stand at the base of the series. It should be noted 
in the first place that there seems to be a certain relation between the copulatory 
glands and the spermiducal glands in those few forms in which they coexist ; that is 
to say, differences in the structure of the copulatory glands are repeated in the 
spermiducal glands. In Kynotus both glands have a muscular covering; in 


Acanthodrilus both are without it; so too in Dichogaster damonis. The question, 
therefore, of the origin of the spermiducal glands is bound up with that of the 
copulatorjr glands, and must be deferred until we can determine which is the most 
primitive type of copulatory gland. It seems reasonable to suppose that the copulatory 
glands were originally modified tracts of the body- wall, which became invaginated and 
furnished with penial setae for their greater eflaciency ; the invagination, as it appears 
to me, might or might not involve the muscular layers of the body-wall ; so that very 
probably both circumstances have occurred, which would account for the corre- 
spondence, noted above, between the copulatory and spermiducal glands. In this 
case the spermiducal glands of the Geoscolicidae might be equally primitive with 
those of the Megascolicidae. Another point, which has to be taken into consideratioHj 
is the relationship of the sperm-duct to the spermiducal gland. The independence of 
the sperm-duct and the gland is, ex hypothesi, a primitive condition. This suggests 
that the Geoscolicidae, and particularly the Acanthodrilidae, Perichaeta ceylonica, 
and Dichogaster are primitive forms. The spermiducal gland of Moniligaster, which 
is characterized by the transference of the gland-cells to the outer side of the 
muscular layer, seems to be a later stage in the evolution of the organ than that 
found in the Eudrilidae. 

§ 8. Genital setae. 

In the neighbourhood of the male-pores there are in some OHgochaeta bundles 
of long modified setae which protrude through the orifices in question; to these 
structures the name of penial setae was first applied by Lankestek (1). In a very 
few species there are bundles of quite similar setae developed in the neighbourhood 
of the spermathecae, and in a few Geoscolicids, again, similar setae are found, not 
only at the male-pores, but on a greater or less number of segments in their 
immediate vicinity. Hobst has proposed to limit the name penial setae to those 
setae which occur in the neighbourhood of the male-pores, and to call copulatory 
setae those which are found in the vicinity of the spermathecae. It is perhaps 
more convenient to apply the term ' genitaL setae ' to all alike. 

These setae are nearly always found associated with glandular structures ; in the 
case of those which protrude through the male-pores, there are, of course, the 
spermiducal glands with which they are more particularly associated ; the sacs which 
contain them are often or at any rate have the appearance of being, diverticula 
of the spermiducal glands. The setae which occur in the neighbourhood of the 
spermathecae have special glands developed in connexion with them, which are 
described in the section devoted to the spermathecae. The special setae developed 



in various Geoscolicidae are nearly or always so accompanied tty glands. This 
obviously suggests some relation between the setae and the glands. It looks as 
if the long setae had the function of either conveying by capillary attraction the 
secretion of the glands or opening wider an aperture for the passage of the secretion. 
It seems likely that that is the function of those setae which are developed in the 
neighbourhood of the male-pores, perhaps more specially those in which the spermi- 
ducal glands and sperm-ducts open by a common pore ; they would facilitate the 
passing of the sperm and the secretion of the spermiducal glands. In the Acantho- 
drilidae, on the other hand, where the spermiducal glands open at some little 
distance from the sperm-ducts, the function would seem to be rather that of grappling 
the integument in the neighbourhood of the spermathecal pores in another individual ; 
here, however, it is possible that the sperm, when liberated from the male-pore may 
flow along the groove which connects this pore with those of the spermiducal glands, 
and may at these points mix with the secretion of the spermiducal glands, and be 
with it conveyed to the spermathecae of the other individual. Genital setae 
associated with the male-pores only occur in the Megascolicidae, the Eudrilidae, 
and (rarely) in the Geoscolicidae, Lumbriculides ; in fact, they only occur in those 
families of terrestrial Oligochaeta in which the male-pores are provided with spermi- 
ducal glands. They do not, however, often occur in the aquatic families which have 
also such glands. Though their range of occurrence is thus a wide one, they really 
are found in comparatively few species ; it is unnecessary here to give a list, which 
can be compiled from an inspection of the systematic part of this work. Naturally, 
too, the exact shape of the setae differs ; but in most cases the extremity is beset 
with spines, a state of affairs, however, which^ is by no means universal ; in many 
species of Acanthodrilus, for instance, the penial setae are quite smooth throughout. 
It is an interesting question as to whether these genital setae are new and 
special structures, or whether they are to be derived from the ordinary setae of the 
body which have become modified and converted to a new function. The second 
alternative is evidently the one which is most likely on a priori gi-ound to be true ; 
and, as a matter of fact, I believe it to be the true view of the origin of the setae. 
It will be noticed in the first place that when such modified setae are present the 
ordinary setae are wanting; they are present in the immature worms, but drop out 
when the genital setae make their appearance; and the genital setae as a rule 
occupy exactly the place of the missing ordinary setae. There are, too, not wanting 
transitional forms between the ordinary setae of the body and the modified genital 
setae ; I have pointed out in Odochaetus antarcticus that the genital setae are much 
less specialized than in the nearly allied Odochaetus multiporus. On the whole, 


therefore, it seems reasonable to regard the genital setae as being the modified 
representatives of the ordinary setae which should occupy their place. But another 
question requires answering: the bundles of genital setae generally contain more 
fuUy developed setae than would usually be present ; not, of course, in the genus 
Megascolex, but in such a genus as Acanthodrilus. I have suggested that this points 
to the former perichaetous condition, a number of setae having been retained for 
the purpose of serving this new function. Though this view has not met with 
acceptance, there are other considerations which seem to me to favour it. In 
Perionyx, as it appears to me, we have an early stage in the development of the 
genital setae preserved. In that genus there is (in some species) a row of the more 
ventrally placed setae, which are modified in structure, being beset with ridges at 
the distal end and being somewhat larger than the ordinary setae, which are smooth. 
These setae are sometimes imbedded in a groove into which opens the male-pore 
on each side ; if this groove were withdrawn, so as to convert it into a more strongly 
marked recess, it would follow that the setae would be crowded together, as is the 
case with the genital setae of the other genera, and that they would at the same time 
tend to become longer, in order to be capable of projecting out of the invagination 
in which they are imbedded. This suggestion is to my mind reinforced by the 
distribution of genital setae associated with the sperm-duct apertures; they occur 
in the Megascolicidae in bundles; the setae are numerous, though varying in 
number. Now it is at least arguable that all the worms belonging to this family 
are traceable to a form with a complete circle of setae. On the other hand, in the 
Eudrilidae the penial setae are not in bundles ; there is only a single seta so 
modified on each side of the body ; and this family is one in which perhaps 
the evidence of a descent from some form like Perichaeta is least arguable among 
the terrestrial Oligochaeta. It is true that the Geoscolicidae seem to be opposed to 
this way of regarding the matter. I am on other grounds disposed to connect them 
closely with the Eudrilidae, and this view of their affinities undoubtedly afibrds 
a way out of the difficulty. As will be seen later, I believe that the aquatic 
Oliofochaeta are not near to the terrestrial; and it will be noticed that in none of 
these that have paired setae are there bundles of genital setae. 

A curious fact has been observed in a few earthworms, for example in Benhainia 
annae, and in other Acanthodrilidae, and that is that the penial setae in a single 
bundle are of two kinds ; there are setae with an extremity marked by the presence 
of spines, and others in which the extremity is quite smooth ; that both kinds of 
setae are fuUy mature, that the one kind are not the immature kind of the other, 
seems to be proved by their being of the same size. 

K a 



§ 9. Penis. 

The presence of a copulatory organ, more or less analogous to the vertebrate 
penis, is not uncommon among the Oligochaeta. These organs may be single or 
paired, retractile or non-retractile. Sometimes they are directly connected with the 
male-ducts which open in them or on them ; more rarely they have no such close 

In the preceding section, dealing with the spermiducal glands of the Oligochaeta, 
I have described the terminal part of these glands, w^hich is almost always of 
a muscular nature; in many worms this part of the tube is partly protrusible, 

perhaps it is so in most; I have 
S- 3'- found that it certainly is so in Pen- 

chaeta houlleti, where a number of 
specimens killed with alcohol had the 
terminal part of the muscular ducts 
of the spermiducal glands everted ; 
I have not noticed the occurrence 
of this in any allied forms, but it 
very possibly takes place. In some 
Perichaetae the muscular duct of 
the spermiducal glands opens into a 
wide and rather thin-walled terminal 
chamber, which opens directly on to 
the exterior ; it is here, again, possible 
that this terminal chamber is pro- 
trusible ; but I have no facts at hand 
to prove or disprove the possibility. 
In most Eudrilidae the two spermi- 
ducal glands open into a single or 
paired terminal copulatory apparatus, 
which may be fairly termed a penis. 
This structure, as has already been 
pointed out, varies much in its development from genus to genus. It is very well 
developed in Eudrihis; in that Annelid the terminal apparatus consists of a widish 
chamber, opening directly on to the exterior; into this chamber open first the 
spermiducal gland, and secondly the peculiar horseshoe-shaped 'glands' which are 
figured in the accompanying woodcut (fig. 32). The spermiducal glands open on to 


pr. Spermiduoal gland. v. d. Sperm-duot. p. Penis, and 
c. Otisliion like pad in interior of ' Bnxsa oopnlatrix ' which has 
been cut open. 


a kind of penis {p), which projects into the interior of the terminal chamber. It 
seems quite reasonable to suppose that the chamber can be everted; ia which case 
the projection which bears the aperture of the spermiducal gland will play the part 
of a penis. 

In all Eudrilids there is something of the same kind, but the details differ ; the 
structures are described under the account of that family. Moniligaster — at any 
rate one species of that genus — has a penis which is a little different, though it 
agrees in being retractile. The muscular end of the spermiducal gland does not 
open directly on to the exterior, but into a terminal chamber whose walls are 
reflected round it ; we have, in fact, in this genus, an arrangement which is closely 
paralleled in the Tubificidae, and not very remote from the arrangement met with in 

In the genera Stylodrilus, Alluroides, Stuhlmannia, Alvania, Hypenodrihis, and 
Siphonogaster, there are penes of quite a different nature. In all these worms, which, 
it will be observed, belong to three different families, the penis or penes are non- 
retractile ; they are processes of the body-wall, which may or may not have an 
intimate relation to the aperture of the sperm-ducts. It will be necessary to describe 
them one by one. 

Of Stylodrilus the penis has been described by Vejdovsky and Benham. In 
Stylodrilus gahretae the two penes are figured by the former as longish, hollow 
processes of the body, whose walls are cellular ; they communicate directly with the 
spermiducal glands, of which they appear to be merely a continuation, but they 
are not retractile. 

The recently-described genus Alluroides, from East Africa, has a pair of penes 
which are clearly outgrowths of the body-wall ; they are placed, moreover, above the, 
opening of the male-ducts; they are peculiar in form, inasmuch as they are rather 
thick processes, hardly tapering at the extremity, which, in the preserved examples 
of the worm, are folded in an irregular fashion; the organ, though not retractile, 
seems to. be probably contractile, as the shape which it assumed in the preserved 
worms was hardly such as to ensure its usefulness as an organ for transferring 
sperm to the spermathecae. 

The penis of Stuhlmannia has been figured, as regards its naked eye characters, 
by MiCHAELSEN (6). Its varying position is not a little remarkable; but it always 
lies somewhere near to the male-pore, and is at any rate connected with that pore 
by a groove in the skin. In transverse sections of the body-wall (see Plate IV), the 
penis is seen to be asymmetrical in structure; it is roughly conical in outline; one 
side is covered by a thick glandular epithelium; the other side is covered by an 


unmodified epithelium, which is like the ordiaary epidermis. The interior of the 
organ is occupied by a quantity of laxly-aiTanged muscular tissue. In connexion 
with the penis there is developed a long sausage-shaped body of a nacreous aspect; 
this gland is covered by very thick muscular walls, and is lined by not particularly 
glandular epithelium. It narrows towards the external pore, which is situated on 
that side of the penis which is covered by thick epidermis. The penis does not 
always appear to be developed; I have examined a considerable number of sexually 
mature individuals in which I could find no traces of a penis. The penis here 
seems to be not much more than an outpushing of a part of the body -wall; the 
muscular sac connected with it may be the equivalent of the bursa copulatrix in 
Teleudrilus, divorced from its connexion with the rest of the terminal male efferent 
apparatus. In the genus Hyperiodrilus there are a pair of penes, which are connected 
by grooves with the male-pore; these penes are, as in Stuhlmannia, variable in 
position; they sometimes lie on the same segment as that which bears the male- 
pore, sometimes they are a segment in front, and in the latter case they are usually 
at least asymmetrical. In my description of Hyperiodrilus I described these structures 
as papillae ; I think that they are more comparable with the penis of Stuhlmannia, 
with which, indeed, I compared them. There are no details as to the minute structure. 
The penis of Siphonogaster is described under the genus. 

§ 10. Spermathecae. 

The spermathecae are very characteristic organs of the Ohgochaeta; it is only in 
a very few forms among those whose anatomy is well known that they are absent. 
They seem to be entirely unrepresented in the following : — 

Criodrilus lacuum. Lumbricus eiseni. 

Perichaeta acystis. AUolobophora con^tricta. 

Geoscolex maxi/nius. AUolobophora samarigera. 

Anteus gigas. Bothrioneuron vejdovskyanum. 

Siphonogaster millsoni. Bothrioneuron americanum. 

AU the remaining Oligochaeta (with possibly a few more exceptions) have a 
varying number of pairs of these organs. Eecent researches have brought to light 
the fact that the sperm-holding organs of the Oligochaeta are of two kinds, 
morphologically distinct. More generally the spermathecae are sacs which are most 
probably — though the actual origin has been traced in but few types — derivatives 
of the epidermis. In the family Eudrilidae these spermathecae are either more or 
less rudimentary or completely absent, their place being taken by sacs which are 
derived from the septa of the neighbouring segments and whose cavity is thus 



Fig. 3a. 

coelomic. These two sets of organs of different origin, but of similar function, will 
be discussed here separately. 

(i) Spermathecae derived from invaginations of epidermis. — This is the prevalent 
form of the organ. Spermathecae of this kind are spherical, oval, or more elongated 
pouches, with or without diverticula, varying in number and position, opening on to 
the exterior by a more or less pronounced duct and, except in rare cases, ending 
blindly at the other extremity. Like other organs developed from the epidermis, 
the spermathecae do not, for the most part, show any ciliation; in rare cases, which, 
however, may possibly be commoner than is thought, a ciliation does occur; it has 
been described, for example, in Tuhifex; and I have described a ciliation of the 
spermatheca of Acanthodrilus rosae (43). Cilia seem to be out of place in a sac 
communicating with the exterior and serving for 
the storage of sperm. 

MiCHAELSEN described, some years since, the 
remarkable fact that in Enchytraeus mobii the 
spermathecae instead of ending blindly opened 
into the gut ; he was able to trace the bundles of 
spermatozoa from the spermathecae into the lumen 
of the gut by staining methods ; he has since 
shown that many, indeed most, Enchytraeidae 
show the same remarkable connexion between 
the interior of the spermathecae and the lumen 
of the gut; I can entirely confirm this discovery 
for several species belonging to the family Enchy- 
traeidae ; sometimes the aperture is at the tip of the spermatheca, at other times on 
the side. More recently I have pointed out that in the genus Sutroa there is the 
same communication between the two organs ; and in all probability Rhynchelm,i8 
will be proved to agree with Sutroa. On a later page I shall have to direct 
attention to the fact that in Faradrilus, where the spermathecae are of the second 
type of these organs, the same thing occurs ; it seems, therefore, to be of functional, 
rather than morphological, importance. Possibly, as has been suggested by Michaelsen, 
this device ensures the getting rid of superfluous spermatozoa, which might otherwise 
decay and cause injury to the animal by so doing. On the other hand, a second 
suggestion might be made, which I give for what it may be worth; Whitman has 
recently made the very interesting discovery that impregnation among the 
Hirudinea may take place through the epidermis ; the spermatozoa apparently, in 
some cases, force their way through the body-wall and fertilize the ova lying within 

spermatheca of 
mesenchyteae0s beumeei. 

(After Midiaelsen.) 

Tlie upper extremity is cut off where the 
spermatheca opens into the gut. The two di- 
verticula contain sperm. 


the body ; he suggests that the same may be the case with the Oligochaeta, in 
many of which the spermatophores seem to be attached to the outside of the body 
'anywhere.' It is just possible that spermatozoa may be conveyed a short distance 
by the alimentary tract and then make their way out to fertilize the ova. In 
connexion with the above suggestion it may be noted that Vejdovsky could not 
find in a single case even one spermatozoon in the albumen of the cocoon. 

The above, however, are rare exceptions to the rule that the spermathecae end 
blindly in the body-cavity. Among the lower Oligochaeta the organs are generally 
simple pouches without any diverticula ; this is so with the Naidomorpha, the 
Tubificidae, the Lumbriculidae (except Rhynchel'nvis and Sutroa), the Phreoryctidae, 
and the Moniligastridae. Of the aquatic families the Enchytraeidae alone are usually 
provided with one or more diverticula ; the Lumbriculidae ai-e sometimes provided 
with diverticula — at least Sutroa and Rhynckelmis are. The higher Oligochaeta, as 
a rule, have diverticula ; this is the case with the families Perichaetidae (a few 
exceptions), the Acanthodrilidae (also a few exceptions), the Cryptodrilidae (like the 
others with a few exceptions). The members of the two families Lumbricidae and 
Geoscolicidae never possess diverticula. 

In the family Eudrilidae there are sometimes spermathecae met with ; this 
statement requires qualification to the extent that spermathecae of the type hitherto 
considered are sometimes present ; the majority of the members of that family have 
spermathecae of a totally different morphological nature, which are described fui-ther 
on. In those cases where spermathecae derivable from epidermic invaginations 
(probably, for that fact has not been actually proved) occur, these organs are 
invariably placed far back in the body, tho, most forward position being seen iq 
the genus Heliodrilus, where they are in the tenth segment; the spermathecae are 
always unpaired and of course open in the median ventral line. They never possess 
diverticula, and, indeed, appear to be of limited functional importance, as they have 
not been observed to contain sperm; in Heliodrilus, however, the spermatheca is of 
great length, extending from its external aperture on the tenth segment as far back 
as the fifteenth. 

When diverticula are present, they are of varying degrees of importance; they 
also vary in number; in the Perichaetidae there is, as a rule, only one diverticulum, 
which is often of considerable size ; two diverticula are found in various Acantho- 
drilidae and Cryptodrilidae; in Octochaetus multiporus there is a circle of small 
diverticula round the external aperture of the spermatheca. There are two facts of 
interest to be noted in connexion with the diverticula; in the first place they are 
of slightly different structure from the pouch of which they are appendages ; secondly; 



I'ig- 34- 

they contain, in mature worms, spermatozoa, absent in the pouches themselves (see 

accompanying woodcut, fig. 34). Acanthodrilus dissimilis may be used to illustrate 

the first point; in this species there are two diverticula to each spermatheca; they 

are lined with an epithelium which is largely converted into irregular masses of an 

amorphous appearance, in which are imbedded bundles of spermatozoa ; ifc looks very 

much as if the lining epithelium had become converted into a substance, the function 

of which was to hold firmly the spermatozoa until they were transferred to another 

individual. In other worms the diverticula are lined with 

a low quadrangular epithelium differing so far from the tall 

columnar cells which line the pouch ; this is the case with 

Perichaeta and other genera. In no case have I observed 

an absolute similarity between the epithelium of the pouch 

and of the diverticulum : in no case have I found any 

spermatozoa in the pouch itself; they were always in the 

diverticula. Eosa has indicated an exception to this ; no 

doubt it may easily happen accidentally that the sperm 

goes astray and finds its way into the wrong cavity ; the 

very fact that Rosa thought it worth while to mention the 

exception shows how widely prevalent is the rule. It is 

very curious that the spermatozoa should choose the narrow 

path leading to the interior of the spermathecal appendix 

rather than the broad and, one would have thought, easier 

road into the spermathecal pouch. It has been suggested 

by Ben HAM that when the male apparatus is everted in 

copulation (this can at any rate take place in some worms) 
and pushed into the spermatheca, the everted portion blocks 
the route to the spermatheca, but leaves free the aperture 
of the sperm-duct which is directed into the appendix. 
The question of the function of the various parts of the 
spermathecal apparatus will be discussed later (see below). 

Apart from the presence or absence of diverticula, the number and position of 
the spermathecae vary in different families of the Oligochaeta. They are sometimes 
further forward, and sometimes more posterior in situation. The most anterior 
position in which they occur is seen in the genus Aeolosoma; here the first pair 
of spermathecae may be as far forward as segment iii. In the Lumbriculidae and 
in some Geoscolicidae they are as far back as they are ever found — viz. in 
segments xiv-xvi. 





(After Horst.) 

I. DiverticuliLm containing 


The following table shows all the different segments which may be occupied by 
the spermathecae in the Oligochaeta: — 

Segment iii 
Segment iv 
Segment v 
Segment vi 
Segment vii 
Segment viii 
Segment ix 
Segment x 
Segment xi 
Segment xii 
Segment xiii 
Segment xiv 
Segment xv 
Segment xvi 















As a rule there is a certain relation between the position of these organs and that 
of the testes. Thus in Aeolosoma, where the testes are in the fifth segment, the 
spermathecae commence in the third; in the Naidomorpha they lie in the same 
segment as the testes, viz. the fifth; this is also the case with the Tubificidae ; 
both testes and spermathecae are in the tenth. In earthworms the spermathecae 
generally lie in front of the testes ; but in the Lumbricidae they are often in the 
same segments. No very definite relation between the position of the spermathecae 
and that of any other organ can be traced ; any attempt to trace such a relation is 
rendered difficult by numerous exceptions. It is only in certain Lumbriculidae and 
in Phreodrilus that the spermathecae are behind all the reproductive organs. They 
appear never to be placed behind the clitellum. 

As a rule each segment of those which contain them has only one pair of sperma- 
thecae; among the Geoscolicidae, however, a very large number of spermathecae 
sometimes occur in a single segment ; thus in Kynotus madagascariensis there are as 
many as fourteen pairs in one segmenb. Perichaeta sangirensis, and a few allied species, 
Allolobophora savignyi, are the only worms not belonging to the family Geoscolicidae in 
which numerous spermathecae take the place of the more usual pair. This great increase 
in numbers is accompanied by a reduction in size, more marked in the Geoscolicidae 
than in the Perichaeta. When the spermathecae are paired, the number of pairs 
varies from one to seven. One is the usual number in the lower forms, and is 


perhaps an indication of the simplification of these genera; we find only a single 
pair of spermathecae in the Tubificidae, the Enchytraeidae (with just one exception), 
the Naidomorpha, most of the Lumbriculidae ; Aeolosoma, the lowest of all the 
Oligochaeta, may have as many as three pairs in segments iii, iv, v. Among the 
higher Oligochaeta the largest number of pairs, seven, occurs in AUolobophora 
complanata ; a few Perichaetidae have five pairs, but on the whole two pairs is the 
prevailing number among earthworms ; the numerous exceptions prevent a comparison 
between this fact and the existence of two pairs of testes and sperm-ducts ; however, 
when a given species has two pairs of spermathecae and two pairs of male-ducts, 
and when a disappearance of one pair of the one set of organs in an allied form is 
accompanied by the disappearance of one pair in the other set of organs, there would 
seem to be some relation between the two cases; this has happened in Acanthodrilus 
monocystis ; the worm is closely allied to Acanthodrilus dissimilis (see below), in 
which there is always a double set of spermathecae and male-ducts ; in the former 
species one pair of each has vanished. Very rarely are the spermathecae median 
and unpaired ; this state of affairs exists in the Cryptodrilids grouped together by 
M1CHA.ELSEN into the genus Fletcher odrilus, in a few Eudrilids (e.g. Heliodrilus), and in 
the genus Sutroa; also apparently in the Tubificid Vermiculus. The fusion in the 
middle line of, we must suppose, originally separate and paired pouches is correlated 
with a similar fusion of the terminal apparatus of the male-ducts. 

There seems to be no doubt, after the investigations of Vejdovsky upon Tubifex, 
and of Beegh upon Lumbricus, that the spermathecae arise as ingrowths of the 
epidermis ; as to the appendices of the spermathecae, where they exist, it is not so 
certain; in Perichaeta it is common to find the appendix of large size, when the 
pouch itself is exceedingly small ; this looks as if the appendix were phylogenetically 
older than the pouch, and, if so, the term appendix should be dropped or appHed 
to the pouch itself. I failed to find in immature examples of an Acanthodrilus 
{A. falclandicus) any connexion between the appendix and the pouch; this almost 
suggests a different origin for the two. 

In the neighbourhood of the spermathecae, or appended to them, there are in 
a few Oligochaeta, pecuHarly modified setae, often accompanied by glands. They 
often bear the same kind of relation to the spermathecae that the spermiducal 
glands, with their penial setae, do to the sperm-ducts. Physiologically speaking, 
it is possible that the correspondence is even closer. 

Among earthworms these' structures have been described in the Acanthodrihds : 
in Acanthodrilus ungulatus, A. schmardae, and in Benhamia beddardi in their 
complete form. In the former species the ventral setae of the eighth segment are 

s 3 



Fig. 35- 

replaced by long ornamented setae, agreeing very closely in appearance with the 
penial setae (on the seventeenth and nineteenth segments) of the same worm. These 
setae are enclosed in a thin, muscular sac, on either side of which is ' a long some- 
what sausage-shaped glandular body, which communicates by a slender duct with 
the orifice through which the setae project on to the exterior.' The minute structure 
of these glands is unfortunately unknown. 

Among the aquatic Oligochaeta the only species 
with any appendages comparable to the above is 
Psanvmoryctes barbatus. In this Tubificid there 
is a sac (see fig. 3^) containing a single penial 
seta opening either into the spermatheca, just at 
its external orifice, or independently, but close 
to it. Two glandular sacs pour their secretion 
out on to the external surface by the same 

Of the origin of the spermathecae it is difficult 
to ofier any hypothesis that is convincing. Bergh 
regards them as quite new structures developed 
for the purpose that they have to perform ; it 
seems to be not proved that they have any 
connexion with the nephridia (see above). The 
occasional close relations in the way of position 
which the spermathecae have to some of the 
genital ducts, suggests a possible origin for them 
as diverticula of such ducts. In the genus Phreo- 
drilus I have described a long coiled diverticulum 
of the sperm-duct, which shows, at any rate, that 
such diverticula can exist ; 1 have suggested that 

A. The entire spermatheca. B. The distal , . i? i • ■ . 

extremity more highly magnified, i. Genital the meaning of this diverticulum may be that 
r"orm:tiv?rrXnitrsr'''^"'^'*'- i* i^- rudimentary second sperm-duct. In the 

Eudrilid Alvania the oviduct has a similar though 
smaller diverticulum ; if these diverticula became separated from the ducts of which 
they are outgrowths, and acquired an independent opening on to the exterior, they 
would be of the same character as the spermathecae. Beyond these scanty facts, 
however, there is no evidence of such an origin of the spermathecae, probable 
enough though it is on a priori grounds. In Eudrilus, of course, the spermatheca 
is a diverticulum of the oviduct, but this instance can hardly be made use of 

(After Stole.) 


since the speiinatheca of that worm is morphologically a different thing from the 
spermatheca of those types that have just been mentioned. 

There are some facts which tend to prove that the spermathecae are, like the 
spermiducal glands (see above), derived from copulatory glands ; the facts, however, 
are not so conclusive in this case as they are (in my opinion) in the other case. 
It must be noted, in the first place, that copulatory glands with modified setae occur 
in the neighbourhood of the spermathecae. They have just been mentioned. There are 
therefore no reasons against such an origin as there might perhaps be were the glands 
in question never present in the anterior region of the body, but always confined to 
the neighbourhood of the male-pores. Whether the glands in the anterior region 
of the body are exactly similar to those in the posterior region is not certain ; some 
facts tend to show that they are. In Bhynchelmis, for example (Ve.tdovskt 5), 
the albumen-gland is lined by two layers of cells, and recalls in this particular the 
spermiducal gland of the same species. The great similarity, too, between the setae 
which accompany the glands in Acanthodrilus ungidatus to the penial setae of the 
same species are facts which point in the same direction. 

The spermathecae are lined with a single layer of epidermis, and appear at first 
to differ very greatly from the glandular bodies with which it is sought to compare 
them. It must be borne in mind, however, that in certain Oligochaeta— in many 
Enchytraeidae — a smaller or greater part of the spermatheca is covered with a glandular 
layer, which seems to me to be strictly comparable to the glandular investment of 
the spermiducal gland. Even in the higher Oligochaeta this condition is not 
altogether unknown. Ferichaeta houlleti is characterized by the fact that there is 
appended to the spermathecae a pear-shaped body, described by Pekrier as 
a diverticulum, but shown by myself to be simply a mass of pear-shaped glandular 
cells. The relations of this to the spermatheca are very similar to those of the 
' Cementdriise ' to the 'atrium' in Tubifex^. 

There are two other points in which the spermathecae seem to show a resemblance 
to the spermiducal glands, and therefore to the copulatory glands, from which it is 
assumed that the latter have arisen by a slight modification. In Psammoryctes 
the long spermatheca has a diverticulum which lodges a penial seta. Its relations 
to the spermatheca are exactly similar to those which exist between the copulatory 
glands and their modified setae. The other resemblance is rather with the spermiducal 
glands, but it obviously amounts to much the same. In Moniligaster bahamens/s 
(Beddard 57) and in Moniligaster indicus (Benham 16) the spermathecae do 
not, as is ordinarily the case, open directly on to the exterior of the body; their 

I Cf. also the cap of glandular cells figured by me (28) in the spermatheca of Biaeliaeta Uttoralis. 


duct opens on to a papilla, which itself projects into the interior of a terminal 
sac. This is, of course, highly suggestive of the penis in the Tubificidae and in some 
others. I figured, but did not describe, this structure in my paper quoted above. 
On this hypothesis we have, of course, to account for the spermathecal diverticula, 
whose presence is so marked a feature of the Megascolicidae. An obvious suggestion 
is that they are the sacs which formerly gave rise to the copulatory setae. 

2. Spermathecae formed from mesoblastic structures. — The Eudrilidae are charac- 
terized by the possession of spermathecae of a different kind to those that have 
been hitherto considered. In a few genera, as already described, there are sperma- 
thecae of the ordinary kind, only differing from those of the majority of the 
Oligochaeta in being situated posteriorly. In most genera, however, the place of 
these organs is taken by sacs which are developed from the septa, and whose cavity 
is therefore coelomic. I describe the variations of these more particularly under the 
family Eudrilidae, inasmuch as they are peculiar to that family and appear to 
have no relation to any other organs found among other earthworms. These sacs 
are generally large and generally unpaired, paired only in Eudrilus, Pareudrilus, 
and Hemertodrilus j if true spermathecae are present, these sacs involve them ; 
if such spermathecae are not present, they open on their own account on to the 
exterior. Often they communicate with the sac that in the Eudrilidae generally 
envelops the ovaries, and they contain therefore both sperm and ova. That there 
should be organs performing the same function, but morphologically quite distinct, 
in a group which is so comparatively limited as that of the Oligochaeta, is perhaps 
the most remarkable fact in the economy of that group, and is not paralleled by any 
other structure. , 

I will take as an example of these structures the spermathecal sac of Libyodrilus 
(PI. iv. fig. i). When the worm is dissected, a large sac, extending through several 
segments (xiii-xviii), is seen to lie dorsally covering the gut. It is of a brownish- 
yellow colour, and is of an irregular elongated form, with numerous furrows on the 
surface which appear to indicate the possibility of distension. It has three pairs of 
diverticula reaching for some way down the sides of the gut. From the anterior of 
these arise the oviducts, one on each side. Anteriorly the sac bifurcates, embracing 
the oesophagus and reuniting below it. It then again divides so as to surround the 
nerve cord ; the two branches reunite below the nerve cord before opening by the 
single median pore on to the exterior on segment xiii. The oviduct does, not, as it 
appears to do, open into this spermathecal sac. A small roundish projection marks 
its attachment to the sac ; this round body is the egg-sac, and has the structure 
characteristic of that organ in other Oligochaeta. Its lumen does not communicate 


with the lumen of the spermathecal sac except indirectly. This communication is 
brought about by the oviduct, the funnel of which is drawn out into two tubes, 
one opening into the egg-sac, the other into the spermathecal one. In transverse 
sections the anterior wall of the spermathecal one is seen to be formed by the 
septum dividing segments xiii/xiv. The microscopic structure of the sac is as 
follows : it is a thickish muscular coat, the fibres of which run in various directions ; 
imbedded in this muscular coat is a structureless sheet, which has sharp edges 
and is highly refractive ; it is undoubtedly a membrane, and is probably elastic 
in nature, allowing for the distension of the sac with sperm, and its subsequent 
shrinkage when the sperm is evacuated. The sac is lined with cells of an irregular, 
somewhat elongated form, occasionally forming several pores ; these cells contain 
numerous granules and are very similar to peritoneal cells, in other parts of 
the body. 

I have been able to trace the development of this sac through several stages. 
In the youngest stage I found a sac below the nerve-cord freely opening into the 
body-cavity ; the walls of this sac were formed by the intersegmental septum xiii/xiv 
in front, and behind by a sheet of membrane, which traced backwards was seen to 
connect that septum with septum xiv/xv. The pouch was formed by the attachment 
of this membrane to the body-wall as well as to the septa, and was, therefore, freely 
open at both sides. The ovary was attached to this membrane, just where it joined 
septum xiii/xiv. In a later stage there was a more completely closed sac lying 
beneath the nerve-cord, extending for a short distance into the thickness of the 
body- wall; above this sac, divided to embrace the nerve-cord and dorsally to the 
nerve-cord, opened freely into the body-cavity; here its walls were obviously formed 
in front by septum xiii/xiv, and behind by the other membrane already referred to. 
The ovary lay between the two membranes. On the ventral surface of the body, 
just below the pouch, the epidermis had no gland-cells ; whether it is ultimately 
invaginated. to form the orifice of the spermathecal sac, or whether the sac bores its 
way to the exterior I am unable to say ; but in any case the cavity of the sac 
evidently begins as a coelomic space. 

§ II. Spermatophores. 

A considerable number of Oligochaeta form cases for the enclosure of packets of 
spermatozoa, which have been usually termed spermatophores. These were first 
discovered in the Lumbricidae, and from their being attached to the exterior of the 
body in the neighbourhood of the clitellum were regarded at one time as penes. 
Spermatophores, although met with in a good many Oligochaeta, are apparently by 


no means universal in the group ; neither do they occur in every genus belonging 
to those families in which they have been met with. At present they are known 
to exist in the following families only :— 

Naidomorpha? Geoscolicidae (Criodrilus). 

Tubificidae. Eudrilidae (Polytoreutus). 

Lumbricidae. Eclipidrilidae. 

In the Naidomorpha Lankester (9) has described spermatophores in B'ais 
serpentina. They are figured by him as long -coiled aggregations of spermatozoa 
more or less similar to the immature spermatophores of the Tubificidae, but entirely 
lacking the complicated structure which these bodies show when fully ripe. Vejdovskt 
(24, p. 153) could not, on the other hand, find any definite spermatophores in the 

Among the Tubificidae the spermatophores have been known for a long time. 
They wore figured by Duges (1, PI. VII, figs. 2, 4) in Tubifex {' Nais filiformis'), 
and termed in the description of the plate ' animalcule spermatique.' In the text it 
is suggested that they are too large to be probably so designated. Budge, though 
noticing the same structures, made no suggestion as to their nature. 

Lankestee has given a more detailed account of the spermatophore of Tubifex 
rivulorum in a subsequent paper (2), which is copiously illustrated. The sperma- 
tophore has a long, worm-like form, and generally a conical head with a collar just 
below; its peculiar form is due to the fact that it is moulded in the interior of the 
• spermatheca. The spermatophore is composed of a cementing matrix in which are 
imbedded the spermatozoa ; when fully formed it has rather a complicated structure. 
Centrally there is an axial band, apparently a canal, but filled with granular 
matter and darkly stained by carmine. External to this is a narrowish, highly- 
refringent band, and following this a broad band in which are imbedded the 
spermatozoa ; these are imbedded parallel to each other and run obliquely ; externally 
to this is another highly refringent layer, and beyond this project the ends of the 
spermatozoa, which are in constant vibratile motion. When the spermatophore is 
examined in transverse section, the spermatozoa are seen to have an imbricated 
arrangement which suggests that they have been fitted together by a twisting 

Fsammorydes has a spermatophore of a rather different form. Its most marked 
peculiarity is a series of recurved booklets at one extremity ; these are figured by 
VE.TDOVSKY (13, Taf. viii, figs. 11, 12), but not by Lankestee (2), who has also 
described the spermatophore of this worm. Otherwise the structure of the spermato- 


phore appears to be much as in Tuhifex. But the conical head is absent as it 
sometimes is in Tubifex. 

In Tuhifex hlanchardi, lately discovered by Ve-jdovsky (8), the ripe spermatophores 
have a distinct aperture at the anterior end, which, though apparently not always 
(see figs. 7, 9 of Vejdovsky's paper), communicate with the central canal. The free 
ends of the imbedded spermatozoa form, as in Limnodrilus, a spiral upon the outer 
surface of the spermatophore. 

A very remarkable kind of spermatophores occur in Bothrioneuron, where they 
have been described and figured by Stolo (3); as, however, his description is in 
the Bohemian language I have only been able to use the facts so far as they have 
been translated by Vejdovsky (8). These are invariably met with clustered round 
the male-pore, where they are attached to the skin. 

The genus is exceptional among the Tubificidae in possessing no spermathecae ; 
which organs in other Tubificids always contain the spermatophores. Their shape is 
unlike that of the spermatophores in the other genera and is more like that of the 
Lumbricidae. They consist of a narrow stalk and of a more swollen distal portion 
which contains the spermatozoa. In the Tubificidae the spermatophores appear to be 
derived from two sources ; the granular axis is the product of the secretion of the 
'Cement gland,' while the rest of it is produced by the epithelium of the sperma- 
thecae. This is the conclusion of Lankestee (2) and Vejdovsky ^ (24) ; but it must 
be admitted that the origin is not yet definitely established. In Bothrioneuron 
there are no spermathecae, and the structure termed by Vejdovsky paratrium must 
be the place where the case is formed. The difi'erent character of the spermatophores 
of this Tubificid correspond to the difi'erent place of their formation. 

Among the Lumbricidae spermatophores constantly occur. The most detailed 
account of these structures is to be found in Feaisse's paper upon the subject. 
They are small chitinous sacs, of an elongated, rounded form, with a more or less 
distinct stalk. The dilated end is filled with sperm, and Feaisse figures an aperture 
through which the sperm can escape, when the proper time arrives. There are 
trifling differences in form; the spermatophores vary in shape according to the 
species of worm. The origin — that is to say the place of formation — of these 
structures has been disputed ; Feaisse traces them to glands in the neighbourhood 
of the setae, on the first clitellar segments, which are the tubercula pubertatis. 
He figures, however, branching tubes in the thickness of the clitellar epithelium, 
composed of a single layer of columnar cells surrounding a lumen. Vejdovsky was 

' The 'moulding' of the head of the spermatophore to the shape of the distal end of the spermathecae 
in Tubifex rimlorum is, of course, a strong argument, 



unable to find these, and their presence, with the function attached to them by 
Feaisse, seems to me to be unlikely. It is possible that the tubes in question are 
the nephridial ducts, which we know, from Hubeecht's researches (see above), often 
open on to the exterior at some distance from that point where they first perforate 
the body- wall. A main objection to the formation of the spermatophores by the 
tubercula pubertatis is, according to Vejdovsky, their frequent occurrence at points 
somewhat remote from these tubercula. He (Vejdovsky) regards the spermathecae 
as the site of their formation. It is not, however, necessary to go into this 
suggestion at length, since spermatophores exist in Criodrilus, where there are 
neither spermathecae nor tubercula pubertatis. 

Rosa (4) has, therefore, suggested that the glandular protuberance on to which 
the sperm-ducts open in so many Lumbricidae is responsible for the secretion of the 
case of the spermatbphore. This suggestion is decidedly in accord with the facts. 
It explains, for example, the large size of the spermatophores in Criodrilus, where 
the glands in question are largely developed, as well as the fact that they have 
not been discovered, in spite of diligent search, in Allolobophora complanata, where 
the swelling at the termination of the sperm-duct is wanting. The suggestion also, 
according to Rosa, explains the position of the spermatophores ; for they are found 
'always on the segments which during copulation are placed in front of the male- 
pores.' In Allolobophora samarigera the spermatophores are, most exceptionally, 
found dorsally on the fifteenth or sixteenth segments. The pear-shaped spermatophore 
thins out on all sides into a thin and leaf-like lamella, reminding one in fact of 
a pine seed. 

Among exotic earthworms spermatophores have only been discovered (by myself 
39) in the Eudrilidae, and within that family only in the genus Polytoreutus. 
They have been found in more than one species of the genus, and are in all probability 
characteristic of it. The spermatophores of Polytoreutus are not at all like those 
of the Lumbricidae or of Criodrilus, and in fact are peculiar to the genus which 
they characterize, though somewhat like the immature spermatophores of Tubifex. 
They are very long, and commonly thicker at one end. They consist of a tolerably 
thick, transparent axis, to which the spermatozoa are attached, and within which 
they are but slightly imbedded. 

Finally, the spermatophores of Eclipidrilus may be briefly described. These are 
stated by Eisex (3) to be glassy transparent bodies ; they are sculptured externally 
with a raised spiral line ; the upper extremity of the spermatophore is dilated and 


§ 12. Clitellum. 

It is almost, if not quite, certain that the clitellum exists at the period of sexual 
maturity in all Oligochaeta. Doubtless a good many forms exist which have not 
been proved to possess a clitelJum, and the absence of this organ has been frequently 
used by some of the earlier writers as a mark of specific distinction ; but many of 
these cases have been shown to have be»n inaccurately described — a clitellum being 
really present in the fully mature worm. The quite recent discovery of the clitellum 
in the families Aphaneura and Moniligastridae has eliminated any doubt as to 
the presence of a clitellum in the only families in which it had hitherto eluded 
discovery. The Moniligastridae, indeed, were placed by Peeeier in a separate 
group, that of the ' Aclitellians,' supposed to be distinguished by the absence of 
this highly characteristic organ. The failure to find it in that group was probably 
due to its temporary presence, or to its not by any means strongly marked 

There is one genus of Oligochaeta, however, from which the clitellum may 
conceivably be really absent, or rather represented by a somewhat different 
structure. This genus is Siphonogaster. The male-pores of Siphonogaster open 
near to the extremity of long penial processes, which have been referred to on a 
previous page. The epidermis of these processes has quite the structure of the 
epidermis of the clitellum, and, in fully mature worms no trace of a clitellum 
occupying the normal position could be found. Unfortunately, as is the case with 
practically all the exotic earthworms, we have no such knowledge of the habits of 
the 'Yoruba worm' as would give a clue to the use of these long 'penial' processes. 
It is possible that their's may be the function of a clitellum. 

With this possible exception, then, the clitellum appears to be present in all 
Oligochaeta — in the aquatic no less than in the terrestrial forms. Its presence 
indicates, with more or less accuracy, the breeding period ; its function is that of 
producing the cocoon and when that function has ceased the clitellum disappears. 
The period of existence, however, seems to vary very much in difierent Oligochaeta. 
The aquatic Oligochaeta without exception, and probably the Moniligastridae also, 
have a fixed and definite breeding-time, during which alone the clitellum is 
developed. This period differs according to the species ; in the case of some the 
autumn, in the case of others the winter, is the time of sexual activity; this 
restriction of the breeding period accounts for the fact that in so many of the 
aquatic Oligochaeta the clitellum is unknown. Other influences, too, in addition 
to the season of the year, are doubtless at work in retarding or accelerating sexual 

T 2 


maturity and, in consequence, the development of the clitellum. I kept under 
continual observation a quantity of Aeolosoma tenehraruvi for more than a year 
without once seeing the clitellum or sexual organs. 

Among the terrestrial Oligochaeta, on the other hand, particularly among the 
exotic species, the clitellum seems to be more or less permanent, after they have 
reached their full size. It is quite possible that this permanency belongs to the 
species rather than to the individual ; but ;n any case it will be noted, on reference 
to the systematic pages of this work, how very few terrestrial Oligochaeta there 
are in which the clitellum is unknown. Considering that in so many cases the 
species are known from an examination of very few examples, collected at all times 
of the year, the assertion with which I have commenced this paragraph seems to 
be, to some extent at any rate, justified. 

The clitellum is a modified tract of epidermis, whose minute structure has 
been already dealt with. It generally happens that the muscular layers in this 
region of the body have undergone some diminution in thickness, in spite of which, 
however, the clitellum generally stands out above the general body surface. Its 
colour, too, is as a rule different from the rest of the body ; but not always. In the 
transparent aquatic Oligochaeta the clitellum, on account of its thickness, and on 
account of the enclosed genital products, is of an opaque white. As to minute structure 
the clitellum shows two principal modifications : in the aquatic Oligochaeta, without 
exception, and in ike genus Moniligaster there is only a single layer of glandular 
cells, the thickness of the clitellar epidermis being thus but little in excess of that of 
the general body-surface. In all other Oligochaeta, whose minute structure has been 
investigated, the clitellar epithelium is composed of several layers of cells. This 
important morphological difference is not at all due to size. Microdrilus and 
Microscolex, besides a quantity of other terrestrial species, are no larger than 
Moniligaster bahamensis, and not so large as Phreoryctes smithii. The difference 
in the structure of the clitellum may possibly be physiological ; the thinner layer 
of epidermis might secrete a thinner cocoon ; and a cocoon deposited in water or 
watery mud would run less risks of having its contents dried up than one deposited 
near the surface of the soil. As, however, Moniligaster seems to be a purely 
terrestrial genus, this explanation cannot yet be accepted. In the meantime, therefore, 
I would regard the difference in the histological structure of the clitellum as 
a morphological difference, acquired so remotely that it cannot at present be 
explained by environmental or other physiological causes. 

The clitellum of the Oligochaeta in naked-eye anatomical characters shows three 
principal variations. 


(i) In most of the aquatic families, in the Perichaetidae, many Cryptodrilidae, 
and Acanthodrilidae, in the Eudrilidae, one or two Geoscolicidae, and in the Monili- 
gastridae, the clitellum extends right round the body; that is to say, the peculiar 
modification of the epidermis is apparent throughout the entire circumference. 
A clitellum of this kind is sometimes spoken of as being 'complete.' EosA has 
suggested that the term cingulum be applied to this form of clitellum, expressing 
its girdle-like character, and that the term clitellum should be restricted to the 
variety of the organ next to be described. 

(a) In all Lumbricidae, nearly all Geoscolicidae, and in some worms belonging to 
the Megascolicidae, the clitellum has a saddle-shaped form ; this is due to the fact 
that the epidermis of the ventral surface, to a variable extent (variable according 
to the species or genus), has not been invaded by the glandular modification of the 
epidermis. The clitellum has therefore a saddle-like shape ^. Rosa attempted, at one 
time, to draw a hard and fast line between this kind of clitellum and the last. It 
is, however, not possible to draw such a line in a satisfactory way. The extremes, 
it is true, differ; the clitelluvi of Luvibricus is very different from the cingulum 
of Perichaeta ; but there are intermediate conditions. In some worms which appear 
to have a completely encircling clitellum, the glandular layer is distinctly thinner 
on the ventral surface ; in several species of Acanthodrilus the clitellum is complete 
anteriorly and incomplete posteriorly. In Pontoscolex corefheura the first one or 
two segments of the chtellum are complete ; then follow a few in which the median 
area free from glandular modification is narrow ; in the succeeding segments again 
this area becomes wider. The same differentiation of width in the ventral area 
occurs in Mhinodrilus gulielrtii. 

(3) Aeolosonia has a form of clitellum which is peculiar to that genus. The 
peculiarity consists in the fact that it is developed only on the ventral side of the body ; 
this form of clitellum is therefore the exact converse of the saddle-shaped clitellum 
just described. 

There is a great variation, among the Oligochaeta, in the segments which are 
occupied by the clitellum. This variation affects not only the number of segments 
included but their position ; and most useful specific, generic, or even family 
characters are obtainable from these variations. As a general rule, there is some 
relation between the position of the clitellum, and that of the other organs of reproduc- 
tion. Among the aquatic Oligochaeta there is a distinct relation between the position 
of the male-pore and that of the clitellum. In the Naidomorpha the clitellum is 
quite anterior in position, so too are the orifices of the sperm-ducts, which open 

' ' Simillimam ephippio quod anteriori arcu caret' (Willis). 


on to the clitellum. In the Tubificidae both have moved back. In the Lumbriculidae 
and the Enchytraeidae there is a similar relation between the male-pore and the 
clitellum. Among the higher Oligcchaeta there is, as a rule, no such relation 
apparent. The position of the male-pore, though it may coincide with one of the 
clitellar segments, does not show any alteration in position corresponding to alterations 
in position of the clitellum. The male-pore, for instance, in nearly all the Megascolicidae 
is upon the eighteenth segment, but the clitellum is most variable in extent, though 
it certainly does not fluctuate much as to the segment where it commences. The 
Geoscolicidae show greater variation than any other family of Oligochaeta in the 
position of the male-pore, and yet it is impossible to trace any clear connexion 
between the position of the male-pore and that of the clitellum. On the other hand, 
the Lumbricidae show the greatest amount of variation in the position of the clitellum, 
combined with a nearly invariable position (on the fifteenth segment) of the openings 
of the sperm-ducts. As to this family, however, it must be admitted that when 
{AUurus) the male-pore is advanced a couple of segments the clitellum also 
commences earlier. It may be noted, also, that in the series of variations exhibited 
by Perionyx excavatus (Beddard 41) an alteration in the position of the male-pore 
is accompanied by an alteration in the position of the clitellum. 

There do not seem to be any other relations between the clitellum and any other 
organs in the higher Oligochaeta — excepting, perhaps, that in no case are the sperma- 
thecae situated behind the clitellum. This is, of course, connected with the function of 
the clitellum in forming the cocoon, which is passed over the head, receiving the 
contents of the spermathecae on its way. 

It has been mentioned that the number of segments of which the clitellum is 
composed vary from species to species and from genus to genus. The smallest 
number of segments, viz. two, characterizes the Naidomorpha and Tubificidae, whereas 
we meet with the greatest number among the Geoscolicidae and Acanthodrilidae 
(twenty-seven in Trigaster lanlcesteri) ; every intermediate number is to be met with. 
Sometimes, as in the case of the Tubificidae, the number of segments is constant 
throughout the entire family ; in other cases — in Perichaeta at any rate, the number 
of clitellar segments (three) distinguishes, with very few exceptions, the genua. It 
is difiicult to detect any relation between the development of the clitellum and any 
structure in the worms. Why should one earthworm have a clitellum of barely 
three segments, and another have a clitellum of over thirty? The question is no 
easier to answer than that relating to the highly variable position of the organ in 
the genus AUolobophora. It is possible that there is some connexion between the 
extent of the clitellum and the phenomena of pairing among earthworms We know 


that in Allolobophora foetida, though not in all species of the genus, a mucous 
skin is thrown off by the clitellum of the worms, which binds them together more 
firmly when pairing. Nothing whatever, unfortunately, is known with regard to 
the pairing of exotic Oligochaeta. But it is important to notice that in many 
of them there are various organs whose function is, possibly, that of securing a 
firm contact between the individuals during copulation. The papillae of Penchaeta, 
which are sometimes present, not only near to the orifices of the sperm-ducts, but 
also in the neighbourhood of the spermathecae, very likely perform this function ; 
the penial setae of Acanthodrilus, Megascolex, &c., may play a similar role ; whereas 
in the Geoscolicidae there is generally an absence of any structures which might 
be supposed to act in this way. No doubt the modified clitellar setae of these 
worms and of the Lumbricidae have some such function, but their effect must be 
feeble when compared with that of the suckers of the Periehaetidae. Now it is in 
precisely these forms, in which accessory copulatory structures are highly developed, 
that the clitellum is restricted in extent, whereas in the Geoscolicidae the . clitellum is 
long, and accessory copulatory structures are commonly absent. It is possible, 
therefore, that the clitellum in these species performs a double ofiice, i.e. (i) that of 
producing the cocoon, and (2) that of secreting a ' skin ' to attach the two worms 
together during copulation. The suggestion is, however, only put forward in the 
most tentative fashion as it is not supported by all the facts so far as can be seen at 

§ 13. Genital papillae. 

There are apparently three kinds of papillae among the Oligochaeta, which have 
been more or less indiscriminately called 'genital papillae.' 

Genital papillae occur, and have been described more particularly, in the Peri- 
ehaetidae and the Lumbricidae (under the name of tubercula pubertatis) ; but they 
are also met with in other Oligochaeta ; in the Acanthodrilidae there are papillae in 
the neighbourhood of the male-pores and elsewhere, to which Michaelsen has applied 
the name of ' WoUustorgane ; ' and organs apparently most nearly corresponding to 
these have been described by myself in FelodrUus and in Fridericia. 

In the Periehaetidae the papillae occur in two situations ; they are found either in 
the neighbourhood of the male-pores or near to the spermathecal orifices, sometimes 
in both situations. They have proved most useful in the determination of species ; 
the variation exhibited by them is so great that it is very often possible to define 
a species by their number and position. To give a few examples : in Penchaeta 
forbesi there are series of pairs of papillae upon segments xvii, xix-xxi. In 


Perichaeta ■pokthuvia there are only two pairs, one on the segment in front of, and 
one on the segment behind, that which bears the male-pores. In Perichaeta indica 
there are no papillae near to the male-pores, but three pairs on the segments which 
bear the spermathecal orifices. Now in all these cases the papillae are little more 
than the external pores of glands, to which I have given (78) the name of 
' eapsulogenous glands ; ' it sometimes happens that a large number of these glands 
are associated together ; in that case there are a group of papillae ; the species 
Perichaeta aspergillum and Perichaeta bermudensis are partly characterized by the 
fact that in the neighbourhood of the male-pores, or near to these and to the 
spermathecal pores, are a group of orifices which belong to a mass of such glands. 
In Perichaeta, hilgendorfi, Beddabd (30), there is a median unpaired group of such 
orifices upon the eighth segment. 

The glands which are associated with these papillae are solid masses of pear-shaped 
cells, which are no doubt of epidermic origin ; they have long stalks, which penetrate 
the epidermis and reach the exterior; the stalks have a fibrous aspect, for they are 
simply the prolongations of the cells ; the glands in fact, in spite of their compound 
appeai'ance, are merely groups of unicellular glands ; very commonly the mass of 
gland-cells are bound together in a common sheath ; this is, for instance, the case 
with the species Perichaeta hilgendorfi just referred to ; I have figured in this 
species a mass of these glands which are enclosed in such a sheath ; it sometimes 
happens that these glands do not open directly on to the exterior ; the two species 
Perichaeta houlleti and P. d'udekenii (if they be really distinct) are each provided 
with one or two pear-shaped glands, attached to the stalk of the spermathecae, 
which have been termed diverticula of tlie spermathecae, but which, as I showed, 
are merely glands of a kind perfectly similar to the 'eapsulogenous glands' — 
different only in the fact that they do not open on to the exterior direct, but 
through the spermathecal duct ; so that when these species are described as having 
no papillae, it must be remembered that they have the glands which in other 
forms are associated with papillae. The function of these papillae is a matter of 
doubt ; more generally they have been held to be organs which allowed the worms 
to maintain a fii-m hold upon each other during coitus ; I have suggested that their 
function may be that of producing albumen for filling the cocoon ; and have in 
consequence compared them to the so-called eapsulogenous glands of the Lumbricidae — 
glands which occur in the neighbourhood of the spermathecae, but whose structure 
is at present not known. No direct observations are on record which would enable 
this question to be decided. 

Among the Lumbricidae the papillae are represented by the structures first called 


by EiSEN ' tubercula pubertatis.' These are swellings on some of the segments which 
form part of the clitellum: they appear, however, before the clitellum is developed; 
their structure is exactly like that of the clitellum, except that there are more 
unmodified hypodermic cells among them ; they are rather prominent, extending beyond 
the general body-surface, and they afford useful specific characters by their position 
and number; in a few forms, for example in the genus AUurus, the papillae form 
a continuous raised band, traversing all the segments upon which it is developed 
without a break ; it is more usual, however, for the tubercula to be paired, 
each pair being limited to a single segment. These organs differ morphologically 
from the papillae of the Perichaetidae, in that they are not associated with glands 
lying in the body-cavity ; it is possible, however, that the difference is not one of 
first-rate impoi-tance ; for the glands of the Perichaetidae are merely masses of 
glandular cells from the epidermis, which have got withdrawn into the body-cavity, 
but which correspond to the glandular cells which are found upon the tubercula 
pubertatis. I have dealt with the glands of a tubular character, which Feaisse 
has described as opening on to the exterior on the tubercula, in connexion with the 
formation of the spermatophores. 

In various Acanthodrilidae there are the papillae already spoken of as having 
been termed ' WoUustorgane ' by MiCHAEiiSEN; these have much the appearance of 
the papillae in certain Perichaetidae, but are found, on a microscopic examination, 
to be composed of elongated epidermic cells, and have in all probability rather 
a sensory than a glandular function ; they are often valuable mai-ks for the 
discrimination of species ; I have made use of them, for example, in distinguishing 
the two allied species Acanthodrilus georgianus and A. aqvMrwni dulciwm; their 
minute structure was first made known in the former species by Michaelsen. 
These papillae are very similar to papillae which I have described in the Phreoryctid 

§ 14. The Cocoon. 

All Oligochaeta, without — so far as is known — an exception, form cocoons in 
which the ova and the sperm ^ are deposited, with or without albumen, set apart 
from the nourishment of the embryos. The cocoon consists of a chitinous substance, 
and it is formed by the activity of the cliteUum; that this is the case has been 
proved by direct observation ; thus Vejdovsky (9) watched the secretion of the 
cocoon in Bhynchelmis from the clitellum. As all Oligochaeta possess this organ at 
the time when the genital products are ripe, there is no reasonable doubt but that 

' It does not appear certain that sperm always exists in the freshly-deposited cocoons. 



all Oligochaeta produce cocoons. The number of genera and species, however, in 
which they are known is very limited. Among the aquatic families of this country 
cocoons have been generally found ; the only exception up to the present is the 
genus Aeolosoma. In this genus cocoons were described by Maggi; but more 
recently I adduced evidence to show that these so-called cocoons were not really 
organs comparable to the cocoons of other Oligochaeta, but were cysts of a temporary 
nature, in which the adult worm enveloped itself; my observations were confirmed 
by Vejdovsky (2). Among earthworms cocoons are known in but few forms. 
For several species of Lumbricus, AUolobophora, and Allurus they have been 
described, especially by Duges and Vejdovsky (9). The cocoon of Crwdrilus is 
also known. It is among the exotic genera that our information is most defective. 
I described some years ago the large cocoons of the huge Megascolex coeruleus of 
Ceylon; more recently Fletcher has referred to the cocoons of a few Cryptodi'ilids, 
while Spencer (1) and Vejdovsky (4) have described and figured that of Ifega- 
ssolides australis. I have dealt with the cocoons of Octochaetus multiporus, and 
have seen also the cocoons of Acanthodrilus annectens. Benham (7) has given an 
account of the cocoon of his new genus Sparganophilus ; Fletcher (6) of Megascolex 
dorsalis. This, I believe, summarizes the scope of our present knowledge. 

In the Lumbricidae 'the shape of the cocoon is only characteristic for a few 
species,' remarks Vejdovsky; they are oval or barrel-shaped bodies, with a process 
at either end ; one of the two processes is commonly thinner than the other and is 
sometimes frayed out into a number of threads. 

In Lumbricus rubellus there is an obvious outer membrane, of a transparent 
appearance, which is prolonged beyond the end j3f the thinner process for a distance 
of lo mm. ; this membrane is soft, and minute particles of earth adhere to it ; later 
it gets to be less soft and is, therefore, not so clear. I have found that a double 
membrane is also characteristic of the cocoons of Octochaetus muLtiporus. Vejdovsky 
is of opinion that this outer membrane is formed in a diflerent situation from that 
where the horny inner membrane is formed: the latter is, he thinks, produced, as 
in the genus Rhynchelmis, from the epidermis of the segments which contain the 
gonads, the spermathecae, &c. ; the soft membrane he traces to the clitellum, for the 
reason that woi-ms were found with a mucous layer just thrown oflF from the 
clitellum. Direct observation, however, is wanting for the settlement of this point. 
There is variation in the colour of the cocoon of the Lumbricidae. In Allurus they 
are of a green colour ; a yellow colour characterizes the cocoons of AUolobophora 
putra, &c. ; AUolobophora octaedra has milk-white cocoons ; in Lwmbricus rubellus 
a dark brown or blackish tint is observable. The cocoon of Megascolex coeruleus is 


of a bottle-gi-een colour; one of the two specimens which I examined has three 
bands of a darker green at one end; the shape of the cocoon was much as in the 
Lumbricidae. In Octochaetus muUiporus and Acanthodrilus annectens the cocoon 
has also much the same form. A considerable number of species are illustrated by 
Vejdo-vskt; slight variations in the form can be observed in these drawings which 
are more readily noticed than to be described. Criodrilus, on the other hand, has 
a cocoon of a very remarkable form. It is extremely elongated and drawn out into 
a fine filament at either end. Hoffmeistek and Benham have figured it. In 
Benham's genus, Sparganophilus, there is a somewhat elongated cocoon, but nothing 
like the extreme attenuation which distinguishes from that of all other Oligochaeta 
the cocoon of Criodnlus. 

Among the aquatic Oligochaeta the cocoon is always of the same general form 
as in the higher Oligochaeta. 

The actual deposition of the cocoon has been observed by Vejdotsry in Rhyn- 
chdmis. There appears to be no doubt that in this worm it is formed by the 
cUtellum, and is a product of the hypodermal glands. 

The cocoon of the Oligochaeta contains a variable amount of albuminous fluid, 
and more or fewer eggs, from which a greater or less number of embryos arise ; the 
sperm, too, is voided into the cocoon so that fertilization takes place here. On the 
variations afforded by these characters d'Udekem founded (1) what is now, so 
far as concerns these facts, known to be an erroneous classification. He divided 
his 'Agemmes' into three families '(i) the Lombricins, (2) the Tubifex, (3) the 

In the first of these the egg is minute, there are many of them imbedded in a 
copious albumen. In the second the egg is voluminous, there are several of them in 
one cocoon but no albumen. In the third there is but one egg in a cocoon and no 
albumen ; in the last two divisions the yolk in the egg performs the function of the 
albumen. This classification has been since shown to be based upon inaccurate 
facts ; Vejdovsky has found that in the cocoon of Ehynchelmis, which belongs to 
the second group, there is albumen, while Michaelsen has described more than one 
Enchytraeid in which the cocoon contains a considerable number of embryos. 

The albumen within the cocoon is, of course, destined for the nutrition of the 
embryos ; its characters vary ; in some forms it is almost transparent and in others 
of a milky appearance ; the latter appearance is most general among the earthworms ; 
in RhynchdTnis and Allurus the albumen is a transparent fluid. 

The number of embryos which attain to maturity within the cocoon diff'ers 
in different forms ; in Lumbricus rubellus there are one or two. In Octochaetus 

u 2 


tnvlti'porus never more than one. In Allolobophora putra, cMorotica, and in some 
others, three is the number; Megascolex coeruleus has two embryos— at least this 
was so in two cocoons of this species which I examined. In Allolobophora foetida 
there are two to six embryos. In many of these cases there are probably more ova 
in the cocoon than the few which give rise to worms ; this is certainly the case 
with Allolobophora foetida, in which species Vejdovsky states that there are about 
twenty ova; those which do not produce mature embryos do not do so for the 
following reasons: either the ova are not capable of fertilization; or they are not 
fertilized ; the egg division takes an abnormal course in an unfavourable direction ; 
or the formation of ' twins ' ^ takes place. Similar variations in the number of 
embryos within a single cocoon occur in the family Enchytraeidae ; in some species 
there is but a single embryo ; in others as many as sixteen. 

The cocoons of the Oligochaeta are deposited in very various localities; Rhyn- 
chelmis attaches them to aquatic plants; those of Ilyodrilus are found in groups 
fixed to the roots of aquatic plants ; many earthworms deposit theirs superficially 
upon the soil; others, on the contrary, lay them deep down in the ground. Those 
of Megascolex coeruleus are laid in a deep burrow ; so also are those of Allolobophora 
trapezoides, according to Duges (6 inches deep). Octochaetus muUiporus selects a 
curious locality ; the cocoons of this species, described by me some time since, were 
found at the edge of a swamp, which in wet seasons is covered with several feet 
of water. 

X. Geographical Distribution 2. 

In considering the distribution of the group, ^nd in drawing conclusions therefrom, 
it is necessary to be very cautious in laying too much stress upon positive arguments. 
Negative arguments, i.e. the non-occurrence of a species in a given locality, are 
from some points of view more important. The reasons for this perhaps somewhat 
paradoxical conclusion are chiefly the accidental importation of worms from one 
country into another by man's agency. There is absolutely no doubt that this 
takes place ; I have myself received earthworms from various countries accidentally 
included in Wardian cases : again, various species of Perichaeta have been met with 
in this country and in America in hot-houses only. On the other hand, it is perhaps 
rash to go to the opposite extreme, and to explain every inconvenient fact in distri- 
bution by assuming accidental importation. The facts must be treated with discretion. 

1 It is not unusual, as has been described by Kleinenbers and Vejdovsky (0), for there to be double 
embryos, of which one only arrives at maturity. 

' For the details of distribution see the systematic part of this volume. 


The means of dispersal (other than through human agency) possessed by Oligochaeta 
are less than those which are possessed by many other animals. Salt water is 
fatal to, at any rate, many (although Pontodrilus, Glitellio, Vermiculus, and various 
Enchytraeids are, of course, exceptions) ; and even prolonged, though in many cases 
it has to be very prolonged, immersion in fresh water will drown some earthworms. 
Consequently, floating tree-trunks can hardly be safely made use of to explain the 
presence of a similar fauna on the opposite sides of a tract of sea, even if this sea 
were, through melting ice, comparatively fresh. 

The wide distribution of many land MoUusca, which, like the Oligochaeta, are 
incapable of much active migration on their own account, is believed to be at least 
partly explicable by their transference, as adults, or as eggs in the mud which 
clings to the feet of birds. The cocoons of the Oligochaeta seem in many cases 
suitable for carriage in this fashion^ — more especially perhaps those of the aquatic 
Oligochaeta, which are not only deposited in the situations frequented by wading 
birds, but are small in size. With earthworms the case is rather different, for 
the cocoons are deposited deep in the ground. Michaelsen has properly drawn 
attention to the fact that many Enchytraeidae have cocoons from which a large 
number of young emerge ; in such cases the transference of even a single cocoon 
might be sufficient to stock a new country. Among earthworms the number of 
embryos that reach maturity in a single cocoon is very limited. 

The conditions of life of Aeolosoma favour its wide migration. The practice of 
forming cysts by the mature worm would give it a double chance (for there is 
presumably a cocoon formed too, though nothing is known about the matter) of 
being carried away accidentally. 

Corresponding to the restricted capacities of earthworms for migration, we do not 
find, with a few exceptions to be noted immediately, that species are widely 
distributed. I may remark that I am compelled to practically leave aside the 
aquatic forms, owing to our comparatively small knowledge of them. The species of 
Allolohophora and Lwmhricus form the most conspicuous exception to the above 
statement ; but, as I point out later, there is every reason for considering their 
wide distribution to be due to human agency. The following species, not belonging 
to those families, have also a very wide range. 

Eudrilus eugeniae, Africa, America, New Zealand, Ceylon, &c. 

Pontoscolex corethrurus, America, Australia, India, &c., and a few Perichaeta 
(P. afflnis, P. indica, and P- houUeti). 

As to the second of these instances, it is noteworthy that the species can live 
on the sea-shore (as indeed its name denotes) ; hence there are possibly not such 


difficulties in the way of its transit across the sea, as there are apparently in the 
case of other species If the species quoted in the above list had all been met with 
in temperate regions, their wide range might possibly be accounted for by human 
agency ; and even though they are not so met with, it seems to me very possible 
that this is the reason fbr their being met with practically everywhere in the 
tropics. Their non-occurrence in Europe is possibly a matter of climate. It 
seems to me that the same arguments which I have used in order to prove that 
the exotic species of Lwmhricus are introduced appears to apply in the case of these 
also. It is almost too remarkable in the case of Pontoscolex, for example, that the 
examples from such widely separated regions of the earth as America, India, and 
Australia, should belong to the same species. Now these species are always amongst 
the most abundant in any gathering of worms from foreign parts ; this of itself 
shows how great would be their chance of accidental transference as compared with 
rarer species. But if this hypothesis be correct, how is it that other species, which 
are equally common in their native countries, are not also found in the same 
accidental way in other countries? For instance, Stuhlmannia variabilis, which is 
undoubtedly one of the species most frequently met with on the east coast of Africa. 
To this it is only possible to make the reply which, no doubt, largely begs the 
question, that some species are more fit for living under different conditions than 
others. It is a most remarkable fact about the Lumbricidae that they have the 
capacity of establishing themselves anywhere, and, moreover, of largely ousting 
the native inhabitants (a capacity, by-the-bye, which they share with their human 
fellow-creatures of their native region). In gatherings of worms from cultivated 
regions in New Zealand there were hardly any native species to be found ; the same 
was the case with gatherings from the seaboard of South America. Prof. Spencer 
has informed me that in Melbourne to get native species it is necessary to go well 
outside the town, the town gardens being filled with European species. 

In the following table is indicated the range of the genera of earthworms. 

1. Neotropical region (N). 

1. Geoscolex. 8. Tykonus. 15. Lumbeicus, N^, P, 0, E, A 

2. Khinodrilus. 9. Eudeilus, E, 0, A. 16. Poxtodeilus, P, 0. 

3. Anteus. 10. Peeichaeta, 0, A. 17. Cetptodrilus, 0, A. 

4. Pontoscolex, 0, A. ii. Acanthodriltjs, E, A. 18. Ocnekodeilus, N^. 

5. Diachaeta. 12. Trigaster. 19. Goediodrtlus, E. 

6. Onychochaeta. 13. Benhamia, E, O, P. 20. Trichochaeta. 

7. Urobenus. 14. Kerria. 21. Miceoscolex, P, A. 

22. Moniligastee, 0. 


2. Neardic {W). 
I. LuMBEicus, N, P, 0, E, A. 3. Diplocardia. 

3. Plutellus. 



3. Ethiopian (E). 








Benhamia, N, 0, P. 






Pekichaeta, N, P, 0, A. 






Megascolex, 0, A. 






Peeiontx, 0. 






















LUMBRICUS, N, W, P, 0, A, 





4. Palaearctic (P). 





.. 4. 









Benhamia, N, E, 0. 


Peeichaeta(1),N,E,0,A. 6. 


5. Oriental (0), 






Benhamia, N, E. 




Ceyptodeilus (?), A, 







Perichaeta, N, E, P, a. 






Megascolex, E, A. 




Eudeilus, N, E, a. 


Perionyx, E. 







1. Megascolex, O, E. 6. 

2. Perichaeta, 0, N, P, E. 7. 

3. Ceyptodeilus, N, 0. 8. 

4. Megascolides. 9. 


6. Australian (A). 

Miceoscolex, N, p. 
16. Diohogastee, E. 

II. Digaster. 
13. Perissogaster. 

13. Eudeilus, N, E, O. 

14. Pontoscolex, N, 0. 

15. LUMBEICUS, N, W, P, 0, E. 



It will be seen from the tables that the Neotropical and Ethiopian regions are 
richest in genera, and that they are at the same time richest in genera peculiar to 
each of them, and found nowhere else. The Ethiopian region is the richest in both ; 
and this fact gets additional weight from the consideration that Africa has been, perhaps, 
less explored than America and the West Indies from the present point of view. 

The Oriental and Australian regions are nearly equal to each other, both in the. 
number of genera and in the number of peculiar genera, and both of them are some 
way behind the Neotropical region in both respects. 

Then comes the Palaearctic region and finally the Nearctic. 

The above tables are too accurate in one respect to bring out the essential 
characters of the earthworm fauna of the different regions. In the Oriental region, 
for instance, Moniligaster and Perionyx are very characteristic forms, and are both 
represented by a number of species. If it were not for a single Moniligaster 
(M. bahamensis) in the Bahamas, and a single Perionyx (P sansibaricus) in the 
Ethiopian region, these two genera would be absolutely limited to the Oriental 
region. In the same way Gryptodrilus is confined to Australia, excepting for two 
doubtful instances. One of these is Gryptodrilus spatulifer, which, as I point out 
(below), is really perhaps in strictness not referable to that genus. The other 
exception is a Cryptodrilid, which I received myself some years ago from India, but 
whose identity with Gryptodrilus I should not like to positively assert. 

Megascolex ought really to be struck out of the list of Ethiopian earthworms, as 
it only just gets into that region with one species (Perichaeta madagascariensis = 
Megascolex armatus) in Madagascar. So, too, as already remarked, Perionyx. 

The following list indicates the characterisl^c genera of the several regions ; those 
which are abundant in individuals or species, or both, or are absolutely confined to 
the region in question, though not abundant. 




6. Urobenus. 





7. Tykonus. 





8. Trichochaeta. 





9. Perichaeta. 





10. Acanthodrilus. 





3. Annadrilus. 





4. Typhoeus. 





7. Desmogaster. 

8. Bilimba. 

9. Perichaeta. 
10. Megascolex. 

13. Benhamia. 

II. Perionyx. 
13. Pleionogaster. 

1. Megascolex. 
3. Perichaeta. 
3. Cryptodrilus. 

1. All Eudrilidae. 
a. Benhamia. 

3. Pygmaeodrilus. 

4. Gordiodrilus. 

5. Microchaeta. 

6. Kynotus. 

1. Plutellus. 


4. Megascolides. 

5. Acanthodrilus. 

6. Octochaetus. 

10. Digaster. 


a. Lumbricus. 


I. Lumbricus. 
a. Hormogaster. 

7. Deinodrilus. 

8. Plagiochaeta. 

9. Diporochaeta. 

7. Ilyogenia. 

8. Callidrilus. 

9. Millsonia. 

10. Nannodrilus. 

11. Siphonogaster. 

3. Diplocardia. 

3. Sparganophilus. 

4. Criodrilus. 

The genera can in reality (in my opinion) be sorted in a better way. It is not 
obvious from the above lists that there is a close resemblance between the Nearctic 
and Palaearctic regions; and yet there is, owing to the fact that Lumbricus (s. Z.) is 
the dominant genus in both. But it is a question how far this fact ought to lead 
one. I have pointed out (p. 150) that although Lumbricus is a genus of world-wide 
distribution, the exotic species are invariably identical with European forms ; were the 
former indigenous to those countries instead of being, as is probable, introduced by 
man's agency, they might be fairly expected to be not identical with European forms. 
To the case of N. America this argument cannot perhaps be so readily applied, on 
the grounds of its recent continuity with Asia. It must be borne in mind, however, 
that there are but few American species which are not also met with in Europe. 



I leave this question, however, undecided. It does, however, appear to be clear that 
the Australian region can have, as regards earthworms, no existence. To begin with, 
New Zealand is essentially different from Australia. The prevalent forms in Australia 
are Perichaetidae and Cryptodrilidae ; the most abundant worms of New Zealand are 
Acanthodrilidae. There are a few species of the two former families in New Zealand 
and three species of Acanthodrilus in Australia. The Australian continent shows, 
in fact, far closer resemblances to the Oriental region. In both the Cryptodrilidae 
and Perichaetidae are the prevailing forms. But the genera of the Oriental region 
are for the most part different, or, if identical, are more or less abundant than in 
Australia. Thus Perichaeta is the dominant Perichaetid in India and Malaya, and 
Megascolex in Australia, though both genera occur in both regions. As regards the 
Cryptodrilidae there are (with the doubtful exception of Cryptodrilus) no genera in 
common. I do not in fact propose to merge the Australian and Oriental regions ; 
but rather to dwell upon the resemblances in order to accentuate the differences 
between Australia and New Zealand. The latter islands do not, however, stand 
sufficiently alone to warrant the introduction of a special region (as was proposed 
by Mr. Huxley for other reasons). The characteristic earthworms of New Zealand 
are Acanthodrilidae — four genera, of which three are peculiar, and two of which 
(Deinodrilus and Plagiochaeta) are very weU marked; but the same family (the 
genus Acantkodrilus only) is equally characteristic of Patagonia and the adjacent 
islands. The only earthworms known from Marion and Kerguelen islands are also 
Acantkodrilus. These facts seem to me to be sufficiently important to require the 
■formation of an Antarctic region, circumpolar in extent. I am not inclined to put 
the Cape part of Africa in this region. Though Acanthodrilids occur there these 
are also abundant throughout tropical Africa. It seems very possible that a former 
greater extension northwards of the Antarctic continent may account for those 
Acanthodrilids; but the equally great or greater prevalence of Geoscolicids, and 
especially of Eudrilids, distinguish undoubtedly the Ethiopian region. 

The facts at present known about the distribution of earthworms lead me to 
divide the world into the following regions : — 

(i) Palaearctic (excluding Japan, but including Africa N. of Sahara); 

(2) Nearctic ; 

(3) Oriental ; 

(4) Australian (the continent of Australia only ?) ; 

(5) Antarctic (New Zealand^ islands of Antarctic ocean, Patagonia) ; 

(6) Ethiopian ; 

(7) Neotropical (including Central America and West Indies). 


I do not attempt the islands of the Pacific, concerning which there is not enough 

I have already referred to the resemblances between the Oriental and Australian 
regions. The latter is, as has been said, sharply marked off from the Antarctic ; nor 
is there a close resemblance between the Oriental and Ethiopian. The latter is, of 
course, especially characterized by the Eudrilidae, which, with the exception of the 
cosmopolitan Eudrilus, are restricted to that region. Some naturalists see a close 
affinity between the Neotropical and Ethiopian regions. 

The distribution of earthworms offers a certain amount of support to this con- 
tention. It will be observed that in both continents the family Geoscolicidae enters 
largely into the composition of the fauna, more, though, it is true, in the case of 
America than of Africa. It is also, as I point out later, possible to differentiate the 
old world from the new world Geoscolicids. It seems possible that these resemblances, 
as also the prevalence of Acanthodrilids in both continents, is due to the former 
extension northwards of the Antarctic continent. 



Before advancing the scheme of classification of the group which will be adopted 
in the present work, it is necessary to give a slight sketch of the principal previous 

Historical Survey. Linnaeus referred such members of the modern order Oligochaeta as he was 
acquainted with to two different orders of his class Vermes. Nereis lacustris (=Sti/laria lacustris) 
was placed among 'Vermes Mollusca,' Lumbricus among 'Vermes intestina.' 

In the thirteenth edition of the ' Systema Naturae,' Gmelin retained this classification and 
added to ' Vermes Mollusca ' the different species of Nais (described by 0. F. MullekI ; to ' Vermes 
intestina' were assigned species of the modern genera Tubifex, Lumiriculus, and l^nchyiraeus. 

CuviER included all the Oligochaeta, together with the Leeches, in one order, Abranchia, of the 
Anntdata ; a subdivision of the Abranchia, ' Abranches setigeres apodes ' included the two families 
Lumbricina and Naides. 

The first classification of the Oligochaeta which is of real value, being based upon an anatomical 
investigation, is that of Jules d'Udekem ; this classification he adhered to throughout. At first 
Milne Edwakds's name for the group ' Annelides setigeres abranches ' was retained ; later 
Savigny's term 'Lumbricina' was made use of. The group is divided into two, Monoica and 
Dioica. The Dioica include the Capitellidae ; the Monoica correspond to the Oligochaeta as now 
understood (i. e. with the exclusion of the Capitellidae) and are divided into ' Agemmes ' and 
' Gemmipares.' The 'Agemmes' contain three families (l) Lumbricidae, (2) Tubificidae, (3) Enchy- 
traeidae ; the ' Gemmipares ' only one family Naideae. The distribution of genera in these families 

X 2 


is as follows — (i) Lumbricinae (Lumbricus) ; (2) Tubificidae {Tiibifex, Lumbriculus, Rhynchelmis) ; 
(3) Enchytraeidae {Enchytraeus) ; (4) Naideae {Dero, Nais, Aeolosoma, Chaetog aster). 

The ' Greminipares ' are distinguished from the ' Agemmes ' by the fact that the sexual organs 
are always present, while in the 'Agemmes' the sexual organs only make their appearance at 
certain times. 

The Lumbricidae, Tubiiicidae and Enchytraeidae are separated by the following characters. 

CLAPAKi]DE distinguished the Oligochaeta from other Annelida and divided them into two groups. 

(i) Terricolae. With two ventral blood-vessels (one above, one below the nerve-cord) with 
' segmental organs ' in the segments which contain the spermathecae and reproductive ducts ; the 
clitellum developed behind the male sexual pores ; a vascular network on the segmental organs. 
Genera Lumbricus and probably Hypogaeon, and Criodrilus. 

(z) Limicolae. With one ventral blood-vessel (above the nerve-cord) without segmental organs 
in the genital segments; the male-pores opening on to the clitellar segment. No vascular network 
on the segmental organs. 

Genera Tubifex, Limnodrilus, Clitellio, Lumbriculus, Nemodrilus, Enchytraeus, Pachydrilus, Nais, 
Stylaria, Chaetogaster, Euaxes {=Bhynchelmis), Serpentina, Aeolosoma. 

The next classification of the group is that of G. Johnston. The division of Oligochaeta 
(termed by him Sooloces) included also the Capitellidae ; it was divided into two tribes : — 
I. Lumbricina. II. Naidina. 

The Lumbricina were again divided into two families : — 
(i) Lumbricidae, genera Lumbricus, Saenuris, Enchytraeus; 
(2) Littorales, Clitellio, Valla { = Capitella). 

The Naidina has only one family Naidea with genera Proto, Stylaria, Seipentina, Nais, Chaetogaster. 

The classification of Leon Vaillant is as follows : — 

Order Annelides lombkicines. 

I. Lumbricina. Setae simple. 

(i) Lumbricina propria. Setae isolated or grouped in pairs. Perichaeta, Megascolex, Pontoscolex, 
Hypogaeon, Echinodrilus, Lumbricus, Helodrilus, Criodrilus, Euaxes, Trichodrilus, Phreoryctes. 
(2) Enchytraeina. Setae three or four in number,in each bundle. Enchytraeus, Pachydrilus. 

II. Naidea. Setae bifid or hair-like, at least partly, rarely pectinated. 

(i) Naidea propria. Setae in four rows, exceptionally biserial, and then all hair-like. 
Heterochaeta, Stylodrilus, Lumbriculus, Clitellio, Tubifex, Nais, Aulophorus, Mesopachys, 
(2) Chaetogastrina. Setae biserial, never hair-like. Chaetogaster, Ctenodrilus. 
Benham (1) has proposed a classification which is really not different from that of d'Udekem. 
He divides the class into two sub-classes : — 

I. Naidomorpha. II. Lumbricomorpha. 

In the first sub-class are the families Aphaneura, Naididae, Chaetogastridae (and the genus 

It is defined thus :— ' Small worms of relatively few somites ; blood uncoloured ; male genital pores 
in front of the seventh somite, or in this somite. Asexual as well as sexual reproduction occurs.' 

The Lumbricomorpha are thus defined :—' Reproduction only by sexual process; no cephalization ; 
somites behind the peristomium all similar; and setae are similar throughout the body, except in 
special regions, e.g. on clitellum. Male genital pores behind the seventh somite. No eye spots 
(7 Helodrilus, Hopfmeistbe). 


This sub-class is divided into Order I— Microdrili {Lumbricomorpha minora,) ; Order II -Megadrili 
(L. majora). 

To these the terms ' -waterworms ' and ' earthworms ' may be applied. The only constant difference 
between the two orders is the absence (Microdrili), or presence (MegadriK) of a capillary network 
of blood-vessels on the nephridium. A number of characters usual to the two groups then follows. 

How far can the separation of the aquatic Oligochaeta, advocated by Clapaeede 
and others, be held now in the light of greater knowledge of the organization of 
the group? 

There are, undoubtedly, a certain number of points in which all these forms agree 
to differ from the terrestrial Oligochaeta; that is, when the structural features to be 
referred to are considered collectively. And there are furthermore a few points 
which at present are peculiar to the aquatic Oligochaeta. We -will commence with 
the latter ; among all the Oligochaeta which belong to Glapabede's ' Limicolae,' the 
ova are of large size and full of yolk; this holds good, without a single exception, 
from the smallest Enchytraeid up to so large a form as Fhreoryctes. Corresponding, 
no doubt, to this, the egg-sacs are of large size and extend through several 
segments ; both they and the sperm-sacs are very thin-walled, and their interior is 
entirely undivided by trabeculae ; in all, or at least very nearly all, of the aquatic 
Oligochaeta there is a head-pore, a structure which is, so far as our knowledge at 
present goes, quite unrepresented in the terrestrial Oligochaeta. The remaining point 
of difference concerns the structure of the body- wall; the longitudinal fibres consist 
of a single row of deep fibres only ; this, however, does not characterize Fhreoryctes, 
a genus which in other characters occupies an intermediate position. 

Besides these resemblances, all the aquatic Oligochaeta ^agree in a number of 
structural features which would, if it were not for the family Moniligastridae, 
distinguish them from all the teiTestrial Oligochaeta ; these are : — 

(i) The clitellum is only a single layer of cells thick. 

(2) It has a very anterior position, more so than in an.y terrestrial form. 

(3) The male- pores are also far forward. 

(4) The sperm-duct traverses only two segments. 

If we add the famUy Moniligastridae, it seems to me that a pertectly natural group 
of Oligochaeta will be the result ; this group will be capable of the following definition. 

Oligochaeta with a clitdlum covimencing not later than the tenth or eleventh 
segment, and condsting of only a single layer of cells; sperm-ducts only occupy 
two segments, the external pore being on the segment following that into which the 
funnel opens ; male-pore situated in front of the female-pore ■; eggs generally large, 
always provided with abundant yolk; egg-sacs large; spermiducal glands, when 


present, possess a muscular layer interposed between the inner epithelium and the 
glandular layer ; sexual maturity at a fixed period. 

In the above definition a number of characters are left out which do occur in all 
of the families which are contained in this group, but which also occur in the higher 
Oligochaeta; such are, for example, the invariably paired nephridia and the absence 
of a sub-nervian blood-vessel. The only genus among those not included in this group 
which presents any resemblance to it in any one of the points mentioned is Tetragonurus, 
in which the male-pores (presumably : its anatomy is not known) are only a segment 
behind that which contains the funnel. It appears to me, therefore, that the above 
definition is really sufiicient to separate ofi" a group of natural distinctness. And 
I propose to retain Benham's name of ' Microdrili ' for this group. 

The dilferences which Moniligaster shows from the other members of the group 
are not in my mind so great as the differences which it shows from other earth- 
worms ; it agrees with earthworms in having gizzards, in the thickness of the 
body-wall, and in the presence of specially thickened septa. But a little consideration 
will show that these characters are not by any means distinctive of earthworms ; 
the genus Pdodrilus is surely more akin to Phreoryctes than to any other family ; 
and it has thick body-walls and thickened septa. The presence of gizzards is 
really the only point in which Moniligaster is more like earthworms than any 
aquatic form ; but are we to decide a question of afiinity on so variable a structure 
as the gizzard, absent in so many earthworms, and possibly connected to some extent 
with the habitat of , the creature ? In considering this proposed grouping, too, 
a number of other structural features must be taken into consideration, which are 
important, though not distinctive of the grqjip. I have described, and so has 
Benham (16) in a different species, a protrusible penis in Moniligaster, which is 
exactly like that of the Tubificidae, and has not its exact counterpart in any genus 
of worms which do not belong to the Microdrili, as I here define them. All the 
Microdrili, with the exception of the Moniligastridae, have no nephridia in the 
anterior segments of the body ; but this fact cannot form the basis of a first 
grouping, since there are some undoubted members of the Oligochaeta which could 
not be placed in the Microdrili, though they show the same structural character : the 
genus Pontodrilus is a case in point. Nor can any stress be laid upon the fact that 
the nephridia of Moniligaster have a vascular network; this is, it is true, absent 
in the aquatic worms, but it is also absent in Gordiodrilus and some of its allies, 
which are not in other respects close to the aquatic families. In fact, there seems 
no way out of associating together closely the Moniligastridae and the Limicolae of 
Clapaeede, and making one family of them. Is it possible to pursue this ari-ange- 


ment further, and to make one group of the remaining Oligochaeta ? Rosa has associated 
all the terrestrial genera (including Moniligaster) into a group ' Terricolae.' It is 
difficult, however, to get together a number of characters of first-rate importance, in 
which they agree to differ from the Microdrili. The following list appears to me to 
embody all the characters which the diflferent famihes of earthworms have in common : — 
(i) The clitellum never commences before the twelfth segment. 

(2) Its minute structure difiers from that of the Microdrili, consisting always of 
two layers of cells. 

(3) The ova are invariably small, and with little yolk. 

(4) The sperm-ducts traverse two or more segments on their way to the exterior. 

(5) The egg-sacs are small. 

(6) The spermiducal glands, when present, have not a muscular layer interposed 
between the two layers of the lining epithelium. 

(7) Sexual maturity seems to be more or less continuous. 

(8) Oviducal pores invariably upon segment xiv, while ovaries are in segment xiii'. 
We might perhaps add the invariable three pairs of nerves given off in every 

segment from the nerve-cord. The number varies in the Microdrili. 

These characters are simply the reverse of those which I have used to define the 
Microdrili ; and they are, therefore, naturally in some cases negative characters, which 
are not always of such value as positive characters. For example, the fact that 
the sperm-ducts traverse more than one segment on their way to the exterior is 
a vague character, because the number of segments traversed is a variable one ; in 
some earthworms fewer, in others more, segments lie between the external and 
internal orifices of the ducts. The clitellum, again, is exceedingly variable in position, 
much more so than it is in the Microdrili. In spite of this, I consider that the 
characters which remain are enough to indicate a community of descent of all the 
different families which are included by EosA in his group Terricolae. 

The next question to be examined is the relation between these two groups. 
Can one, for example, be derived from the other, or are they both traceable only 
to a common stock? The facts, as to the influence of an aquatic life upon normally 
terrestrial forms, seem to tend to the conclusion that there is hardly any evidence of 
the characteristics of the Microdrili having been caused (except perhaps by a very 
prolonged exposure to those influences) by their aquatic life ; and, besides, those very 
characters are found in the terrestrial Moniligasters, which at least indicates that, if 
they are due to environment, it is a long time in acting, since Moniligaster is distinctly 

' Exceptions to this general statement which have been recorded (of. Rhinodrilus proioscideus) require 


a terrestrial form in habit, and bears no superficial resemblances to the aquatic 
Oligochaeta. All these facts seem to indicate a gulf between the Microdrili and the 
Megadrili (as I would call the earthworms, following Benham's nomenclature). 

Aeolosoma, as it appears to me, must undoubtedly form a group by itself. In 
many respects it is barely an Oligochaet. I refer particularly to the absence of 
specialized sperm-ducts, whose proper function is taken by at most slightly-modified 
nephridia; this state of affairs recalls the CapiteUidae and the Saccocirridae, in both 
of which families the generative ducts are apparently slightly- modified nephridia. 
The clitellum is so far unlike that organ in oth-er Oligochaeta (except certain 
Enchytraeidae) that it is only developed on the ventral surface of the body ; where 
the clitellum is deficient in other Oligochaeta, it is precisely here (on the ventral 
surface) that the deficiency occurs. The only characteristically Oligochaetous feature 
in the generative system seems to be the spermathecae. It is also the only 
Oligochaet in which there are a pair of ciliated pits at the sides of the head. 
These occur, as is well known, in some of the ' Flatworms ' and Nemertines ; they 
are also found in a few Polychaeta, for instance in Saccocirrus. It may be that 
a ciliated organ which I discovered (51) in the embryo of Octochaetus multlporus is 
referable to the same category ; this organ, however, was a prominence, not a 
depression, of the integument ; it had a bunch of particularly long cilia upon it. 

The ciliation of the prostomium is another absolutely distinctive character 
of Aeolosoma; this, again, while not occurring in any Oligochaet, is met with in 
Saccocirrus and Ctenodrilus, two genera of perhaps doubtful position, but more 
nearly related to the Polychaeta than to the Oligochaeta. 

As to the absence of the ventral nerve-co|-d, this may not be really so entirely 
absent as would appear from most of the published descriptions of the anatomy of 
the genus. To begin with, Vbjdovsky found the rudiments of a cord in the shape 
of more or less scattered cells; it is also possible that the ventral nerve-cord may 
be in connexion with the epidermis, as the cerebral ganglia certainly are. If the 
ventral nerve-chain be really absent, it might be put down to degeneration ; I should 
certainly be disposed to credit a process of degeneration with the almost complete 
absence of internal segmentation. A final peculiarity, very distinctive of the genus, is 
its habit of temporarily encysting itself in a chitinous cyst. 

These characters, taken together, seem to me to necessitate the placing of 
Aeolosoma in a group by itself, a group equivalent to either of the Microdrili and 
Megadrili ; it may even be that this separation is not wide enough ; there would be 
no great violence done to the systematic position of the genus if the group to which 
I refer it were regarded as the equivalent of both the Microdrili and the Megadrili. 


The group Aphaneura may be thus defined : 

Oligochaeta tvith a ditellum confined to the ventral surface, with no male-ducts, 
the female-ducts being replaced by a median pore, prostomium ciliated on the under 
surface, ciliated pits on the head, reproduction by budding as well as by the sexual 

As to the further division of the Microdrili, no one will dispute the justice of 
placing in separate families (i) the Enchytraeidae, {%) the Moniligastridae, (3) the 
Phreoryctidae. The remaining families of Naidomorpha, Tubificidae, and Lumbri- 
culidae, I associate into one superfamily, giving my reasons for doing so on a 
subsequent page ; this will form a fourth group, the Lumbriculides. 

The wide range of variation in structure exhibited by earthworms permits the 
group to be broken up into families. 

This has been done by Peeeiee (3), Vejdovsky (24\ Benham (1), and Rosa (20), but there is 
some difference in the classificatory systems of these authors. 

Peeeiee's classification ' depends primarily upon the relation of the male generative orifices to 
the clitellum ; according to the position of the male generative orifices, earthworms are divided 
into the following four groups : — 

(i) Anteclitellians : in which the male generative pores are in front 0/ the clitellum; 

(2) Intraclitellians : in which the male generative pores are within the clitellum ; 

(3) Postclitellians : in which the male generative pores are behind the clitellum ; 

(4) Aclitellians : in which there is no clitellum. 

The last division — that of the Aclitellians — cannot be defined as Peeeiee defined it, inasmuch as 
BouENE (2) has discovered a clitellum in several Indian species of the genus Moniligaster, of which 
genus alone the group consists. But the systematic position of Moniligaster has been already discussed. 

As Rosa (20) has pointed out, the classification of Peeeiee is artificial and must be abandoned 
for three reasons: 

(i) It places in different groups genera which have close affinities in other respects. An 
instance of this is Megascolex and Perichaeta ; the former genus is ' Intraulitellian ' the latter 
' PostcUtellian ' ; and yet Megascolex agrees with Perichaeta in such an important point as the 
possession of numerous setae arranged in a ring, round the middle of each segment. 

(2) It separates species of the same genus. 

In the genus Acanthodrilus the following species are ' Postclitellian ' : ^A. obtusus, A. ungulatvs ; 
the following are ' Intraclitellian ' : A. schlegelii, A. novae-zelandiae, A. dissimilis. 

(3) It separates even individuals of the same species. 

In some examples of Microscolex modestus, Rosa found the generative pores behind the clitellum, 
in others within. 

Peeeiee himself remarked that EudHlus seemed to be a transitional form between the Intra- 
and Postclitellian; the male sexual pores are within the clitellum, but the general organization of 
the worm was rather that of the Postclitellian; and this principally for the reason that the male 
reproductive ducts were furnished with prostate glands. In the present state of our knowledge, 

' This classification assumes that earthworms form a group 'Terricolae' equivalent to the 'Limicolae.' 



however, it is not possible to base any such wide distinctions on the presence or absence of 
'prostate' glands; since there are plenty of instances in which Intraclitellian male generative 
pores are associated with the presence of prostate glands. 

The classificatory scheme of Vejdovsky (24) is in some respects an improvement of that of 
Pekbiee. Benham carries in my opinion to too extreme a point the classification according to 
the excretory system originally proposed by myself. 

It does not appear to me that any of the schemes advanced express the total facts 
now known about the structure of the Oligochaeta. 

In the course of the following pages I have dealt with the mutual affinities of 
the different families to one another ; and for all details the reader is referred to 
these places. Summing up briefly the results put forward in detail, I may point out 
that the Acanthodrilidae, Perichaetidae, and Cryptodrilidae, form a natural assemblage 
which I make into a superfamily, Megascolides. In all the membei's of this group 
(with very few exceptions) there are spermiducal glands which are unconnected with 
the sperm-ducts, or into which the sperm-ducts only open near to their external 
orifice ; these glands, too, are entirely without an external muscular sheath. Sharply 
marked off by this character, the Megascolides are also the only Terricolae in which 
there is a ' plectonephric ' excretory system, and they are the only group in which the 
setae are often more than the ' typical ' eight. 

A group of equal importance is that of the Eudrilidae. In this group the 
sperm-ducts open into the spermiducal gland always, and far up it in the glandular 
part of the organ. This character does not, it is true, absolutely distinguish them 
from the next family, that of the Geoscolicidae j but in the Eudrilidae there are so 
many structures which are, without being universal, so widely spread and so peculiar 
that it is, I think, necessary to regard the family as a distinct one. Such characters 
are the presence of a system of coelomic sacs, involving the female organs of generation 
and forming a spermathecal sac, the existence in the integument of organs probably 
of a sensory function and of peculiar form ; the frequent network formed out of the 
nephridial ducts and lying in the skin. 

The Geoscolicidae and the Lumbricidae are the only remaining families of the 

Phylogenetic Aerangement of the Oligochaeta. 

The difficulty (greater in some cases than in others) of defining the various 
families of the Oligochaeta seems to show that there can be no question here of 
deriving the group from two or more stocks ; if we were limited to the types at 
the extreme outskirts of the group such as Aeoloscmia and Endrilus, this view might 


be tenable; but these, and other equally or nearly equally different types, are 
connected by so many intermediate forms, that the difficulty is not in uniting them 
into the same large group, but in finding adequate characters by which most of the 
families and genera can be differentiated. There is in fact no doubt, in my opinion, 
that the group Oligochaeta is a perfectly natural one. This being the case it is 
desirable to attempt a general sketch of the interrelationships of the several families, 
which shall indicate the most probable evolution of the class. In attempting a 
solution of this problem, no help is to be obtained from Palaeontology, and very 
little from Ontogeny; the development of only four types is known, viz. Lumbricus 
(various species), Acanthodrilus, Rhynchelmis, and Enchytraeus (s.l.). Many different 
opinions have been held as to the relative positions of the different families ; EoSA. 
has argued in favour of the archaic position of the Acanthodrilidae ; I have urged 
the claims of the Perichaetidae to occupy the lowest place in the series ; Benham 
has recently put forward the view that the terrestrial Oligochaeta are the most 
modern group, and that some family (not particularized) of the ' Lumbricomorpha 
minora' are to be looked upon as their ancestors. 

Eosa's opinion was mainly based upon the occasional doubling of the dorsal 
vessel in the Acanthodrilidae, which is known to be an embryonic trait, and upon 
the presence of eight nephridia per segment in Octochaetus Tnultiporus. Since then 
so many worms have been found to possess a ' plectonophric ' condition, and the 
double dorsal vessel has been met with so frequently, that more is wanted before 
the Acanthodrilidae can be accepted as the primitive family. My own arguments 
chiefly rested upon the structure of the nephridia in Perichaeta ; embryology has 
apparently shown that the plectonephric condition was preceded by paired tubes, 
though it is still not proved that the paired nephridia of existing genera are the 
precise equivalents of the embryonic nephridia of Octochaetus and Megascolex. 

As regards the earthworms, at any rate, it seems to me to be not so difficult 
a task to arrange them, if only the starting point could be agreed on; the aquatic 
families, on the other hand, are more puzzling; this is a good deal due, I should 
imagine, to the fact that our knowledge of them has by no means kept pace with 
our knowledge of the terrestrial forms; this is particularly to be regretted, in view 
of the fact, that all probability points to the origin of the terrestrial forms from 
aquatic and not vice versa; this cannot, I think, necessarily mean that we must 
look for the oldest type of Oligochaeta among the existing aquatic genera already 
known. Continuing our general resume' of the conditions which would on a priori 
grounds suggest an archaic position, we may refer to wide and discontinuous 
distribution; many undoubtedly ancient forms show this; a familiar example is 

Y a 


Peripatus ; so, too, are the Struthious birds, the marsupials, &c. There are but few 
Oligochaeta which fulfil this condition, and arguments based upon it would be 
unsafe; still it may be worth while to point out, without laying too much stress 
upon the facts, that Moniligaster found in India and the Bahamas is an example ; 
perhaps, too, Ocnerodrilus is one; the genus occurs in the West Indies, South and 
Central America on the one hand, and in tropical Africa on the other. Penchaeta 
is perhaps a better example still ; the impression gained from the details as to the 
distribution of this genus given in the present work may not appear at first sight 
to favour this statement ; the genus is. it is true, found in most parts of the world ; 
but it has, nevertheless, a somewhat discontinuous distribution ; it is excessively 
abundant in India and the neighbouring parts of Asia; this is, indeed, the head 
quarters ; it is rare at the very most in tropical Africa, and reappears, though it is 
not common, at the Cape. Three species of Perichaeta (s.s.) live in Australia, 
together with a number of Megascolex; it is very common in several islands of the 
West Indian group, rarer in South America. The distribution of the family does 
not suggest a migration from the head quarters, except in the case of the genus 
Megascolex; it is rather suggestive of a former universal range which has become 
here and there limited, and here and there perhaps increased in numbers where there has 
been but little competition. The distribution of worms in Australia offers particularly 
useful data ; if the general theory to account for the almost entire absence of 
placental mammals from this region of the globe be accepted, we should argue that 
the Geoscolicidae, which are totally absent from Australia ^, are a comparatively modern 
* type, while the Perichaetidae and the Cryptodrilidae, which are so abundant there, 
are more ancient ; this is quite in agreement wjih the other facts ; for there can be 
no doubt that the Cryptodrilidae are very nearly akin to the Perichaetidae. 

In considering the distribution of earthworms we are always fairly sure of our 
deductions, if we can only eliminate the interference of man ; the Perichaetidae 
and Crj ptodrilidae are too abundant in Australia to allow of our explaining their 
presence by artificial introduction ; they must obviously be indigenous to that 
country; the occasional occurrence of a species in a country far from its usual 
habitat, may, of course, be a fact of importance ; but, considering the possibility of 
accidental transport, it is not as a rule striking. I should be inclined, therefore, to 
lay some stress upon the probability of the archaic nature of the characteristic 
earthworms of Australia,. 

Apart from geographical distribution there appear to me to be other characters 
which tend to argue for the primitive position of the Perichaetidae among earthworms. 

' Pontoscolex coretJmirus (see p. 150) may be an accidental importation. 


Most of these characters will be dealt with in the succeeding pages ; it is only necessary 
here to briefly recapitulate. The characters upon which I lay stress are these — 

(i) The continuous circle of setae. 

(a) The (generally) undifferentiated character of the calciferous glands, which 
are often incompletely separated from the oesophagus. 

(3) The frequent presence of two pairs of egg-sacs. 

(4) The general existence of copulatory glands which are possibly the fore- 

runners of the spermiducal glands. 

(5) The existence of a series of small sacs along the septa, which may be the 

remains of a continuous series, of which some have been differentiated 
into the sperm- and egg-sacs. 

Some of these characters are shared with members of the family Acanthodrilidae 
and Cryptodrilidae ; but it is not an easy matter to distinguish these three families 
as has already been pointed out. Arguments of nearly — but in my opinion not 
quite — eqoial value might be put forward for placing either of the two remaining 
families of the Megascolides in the position of most primitive earthworm. The 
distribution of the Cryptodrilidae is obviously much like that of the Perichaetidae ; 
that of the Acanthodrilidae seems to be not sufiiciently world-wide to use it as an 
argument for the archaic nature of these worms. If all the Perichaetidae had the 
racemose form of the spermiducal glands, which I believe to be secondarily derivable 
from the tubular form characteristic of the Acanthodrilidae and many Cryptodrilidae, 
it would militate against their being regarded as the most primitive group. The 
Acanthodrilidae it will be remembered have occasionally (Plagiochaeta) continuous 
circles of setae ; the fact that the majority have lost this arrangement is, in my 
opinion, a further argument against regarding the Acanthodrilidae as at the base of 
the series of earthworms. It is no doubt an exceedingly difficult matter to adjust 
the conflicting claims of the Perichaetidae, Cryptodrilidae, and Acanthodi'ilidae, to 
the position of the most primitive family in the group. 

With regard to other earthworms, there would seem to be no doubt that the 
Eudrilidae and the Geoscolicidae are closely connected, and that the latter are 
intimately related to the Lumbricidae. . The arguments for these statements are put 
forward in the resumes which I give below of the structural characters of the several 
groups. The question now to be considered is the relation which they bear to the 
group Megascolides, and their relative position with regard to each other in the 
phylogenetic scheme. 

Rosa and others have placed the Eudrilidae, with some or all of the Cryptodrilidae, 
in one family ; this position seems to me to be quite untenable. The Eudrilidae diflfer 


from all Megascolides in a number of important points, of wliich the following are 
the most important: — 

(i) The spermathecae, if present, are in the neighbourhood of the female gonads ; 
they are always accompanied, and generally replaced, by coelomic sacs, 
which apparently perform the same function. 
(3) The sperm-ducts always open into the glandular part of the spermiducal 

(3) The spermiducal glands open by a terminal copulatory chamber. 

(4) Both paired and unpaired calciferous glands exist, or they are replaced 

by peculiar structures of probably a different function. 

(5) The ovaries are nearly always involved in coelomic sacs. 

(6) The nephridia are always paired, and sometimes form a network in the stin. 
There seems to be no doubt that the remarkable development of coelomic spaces, 

which is so characteristic a feature in the organization of the Eudrilidae, is a secondary 
development. But, on the other hand, the group shows certain primitive features. 
This is particularly the case with the spermiducal glands, which show, in the 
■possession of a terminal copulatory chamber, a close agreement with the copulatory 
glands from which they are, in my opinion, derivable. As a rule, there are no 
traces of these glands other than the spermiducal glands in the Eudrilidae ; but 
MiCHAELSEN has described and figured in Reithrodrilus minutus something which 
appears to be of the same nature. Mainly on account of the frequent presence of 
such glands in the Geoscolicidae, and the absence in them of coelomic spaces 
surrounding the genitalia, I should be disposed to regard that family as more 
primitive than the Eudrilidae. 

The relation of the Geoscolicidae to the Megascolides is not a very near one, 
and it is furthermore not an easy task to decide which of the three families of 
the Megascolides has most claims to have given rise to the Geoscolicidae. The 
Perichaetidae may, I think, be dismissed from consideration. The only positive 
point of resemblance is in the intestinal caeca of Urohenus, which not only are 
like those of Perichaeta, but occur in the same segment. In no Perichaetous worm 
are the setae ornamented ; this is one of the most general characters of the 
Geoscolicidae ; nor are there any segments in the anterior part of the body in 
Perichaeta free from setae ; this is another character often met with in the 
Geoscolicidae. It is among the Cryptodrilidae that all the characters distinctive of the 
Geoscolicidae occasionally occur; in Deodrilus we have ornamented setae and an absence 
of setae from the most anterior segments of the body. The long sperm-sacs of the 
Geoscolicidae (such as occur in Trichochaeta, Pontoscolex, &c.) have their counterpart 


in Typhaeus. The presence of additional glands behind the spermidacal glands, 
and corresponding to the copulatory glands of the Geoscolicidae, is met with in 
Dichogaster. A good many of these characters, it will be observed, particularly the 
last, are abo found in the Acanthodrilidae. In fact, it seems to be probable that 
the Geoscolicidae were given off from the main stem of the Megascolides, before the 
latter had become differentiated into the three families which now exist. As to 
the phylogenetic relationships of the different families of Microdrili among themselves, 
something appears to be clear, but a good deal is dark. I have associated together, 
and given my reasons for doing so, the Lumbriculidae, Tubificidae, and Naido- 
morpha. There appears to be a fairly straight line down from the Lumbriculidae, 
through Phreodrilus and the Tubificidae, to the Naidomorpha ; and there is a gradual 
change in structure which favours the view that that has been the course taken. 

The Lumbriculidae have the simplified muscular layers of the body-wall found in 
the forms lying below them, but the setae are all of the Lumbricid pattern, and are 
not by any means always cleft at the free extremity. They have retained the two 
pairs of sperm-ducts, and the spermiducal glands are less modified than in the 
majority of the Tubificidae. In Phreodrilus we get the first stage in the development 
of capilliform setae and the commencing reduction of one pair of spermducts. In 
the Tubificidae the bifid form of seta, occasional in the Lumbriculidae, has become 
definitely established. 

On the other side, it seems to me that the relations between the Lumbriculidae 
and the Moniligastridae and Phreoryctidae indicate the derivation of the former from 
a group more akin to the purely terrestrial Oligochaeta. The Moniligastridae and 
Phreoryctidae seem to be offshoots from the Lumbriculid stem, before the complicated 
structure of the body-wall necessary to a terrestrial Annelid had been lost. Granting 
the relationship of all these families, which I have argued above, their relative 
position seems to me to be that which I have just indicated. 

The question of the relations of the Enchytraeidae is a very difficult one. 
They are unique among the Microdrili in possessing dorsal pores and salivary 
glands, the existence of which features recall of course the Megadrili. ' Their aspect 
and habits are largely those of the earthworms,' remarks Vaillant (6, p. aa6). The 
considerable number of segments which separate the spermathecae and the testes is 
another character highly suggestive of the earthworms. 

On the other hand the Enchytraeidae present us with resemblances to the group 
of worms which lies at the opposite extreme of the Oligochaet series. They offer 
certain points of similarity to the genus Aeolosoma. These are: (i) pores instead of 
tubular oviducts; (a) a cardiac body in the dorsal vessel of certain species, of 


which traces exist in Aeolosomaj and (3) the development of the clitellum (in 
Mesenchytraeus at any rate) laterally and ventrally, but not dorsally; (4) origin of 
dorsal vessel from a perienteric sinus. 

Together with these resemblances to other groups of Oligochaeta, the Enchytraeidae 
are to be distinguished by certain pecuharities restricted to themselves. These are: 
(1) the peculiar form of the nephridia^; (2) the generally elongated and glandular 
funnel of the sperm-duct ; (3) the reduced size and simple structure of the spermiducal 
glands ; (4) the frequent presence of a double layer of longitudinal muscles. 

These four classes of facts in the organization of the Enchytraeidae appear to me 
to refer them to an isolated position between the two stems of origin of the Oligochaeta. 

The Enchytraeidae perhaps resemble Phreoryctes more than any other group of the 
higher Oligochaeta ; these resemblances, however, are not numerous, and are confined 
to a few species. The most striking is the existence in various species of Pachydrilus 
of the segmentally arranged lateral outgrowths of the nerve-cord ; structures similar to 
these appear to occur in Phreoryctes. Besides Phreoryctes, the only Oligochaet in which 
there are four setae per segment, implanted singly, is Enchytraeus monochaetos. 

The next point to be considered is the connexion between the MicrodrUi, the 
Megadrili, and the Aphaneura. In attempting to frame the broad outlines of a phylo- 
genetic scheme, such as is implied in the question here raised, regard must be taken 
of the relationships of the Oligochaeta as a whole to other groups of animals. 
A proper phylogenetic scheme can only be elaborated after considering these relations. 

The Oligochaeta are almost universally regarded as nearly akin to the Polychaeta ; 
the two groups are in fact generally placed within one larger group called the 
Chaetopoda. In so placing them it appears to me that important differences have 
been lost sight of, and too much stress has been laid upon the broader points 
of resemblance. What are the chief likenesses between the Polychaeta and the 
Oligochaeta? They are both regularly segmented worms, with a well -developed 
coelom, also arranged in accordance with the segmentation ; each segment has 
typically a pair of nephridia, which may in both cases be secondarily increased in 
number ; they have chitinous setae derived from the epidermis ; in both there is a 
closed vascular system, much complicated, and usually containing red blood, coloured 
so by haemoglobin; in both the gonads are proliferations of the coelomic epithelium, 
and are persistently unenclosed in special cavities, except in a few cases ; the ducts 
for the can-ying away of the genital products are formed out of metamorphosed 
nephridia; lastly, the nervous system is constructed on a similar plan. 

' I do not refer so mucli to the network formed by the lumen, which is to be found elsewhere, e.g. 
Ckaetogaster, as to the general form of the organ. ' 


It is undoubtedly the case that no other group than the Polychaeta comes nearer 
to the Oligochaeta in these essential features of organization. The Hirudinea, 
however, are quite as close, now that we know of the curious modifications of the 
reproductive organs in the Eudrilidae, and the frequently unpaired condition of the 
generative pores in that group and elsewhere ; a commencing division of the coelom, 
carried much further in the Hirudinea, is to be recognized in a few earthworms. 

The Gephyrea are further away still; yet in a few points they come nearer to 
the Oligochaeta than does any other group of worms ; for example, we have in 
Echiurus, at the tail end of the body, circles of setae (as in Perichaeta) ; and the 
'respiratory trees' are to some extent paralleled or foreshadowed in the anal 
nephridia of Octochaetus. 

The chief differences between the Polychaeta and the Oligochaeta are as 
follows : — 

(i) The testes and ovaries are limited in the Oligochaeta to a definite and 
small number of pairs ^, whereas in the other group they are much 
more abundant and are as a rule of one kind only in one individual. 

(2) The ducts for carrying away the genital products attain in the Oligochaeta 

to a specialization and variety quite unparalleled in the Polychaeta; 
in the latter (Capitellidae, Saccocirridae, Alciopidae), they are but 
slightly metamorphosed nephridia. 

(3) The nephridia of the Polychaeta are histologically very different from 

those of the Oligochaeta ; the only special resemblance between the two 
is in the nephridia of the embryo. And these nephridia it will be 
observed, recall those of the Platyhelminthes. 

(4) The setae of the two groups are different in form ; both, it is true, have 

long acicular setae; but when there is any peculiarity in form it is 

different in the two; and the long setae are closely paralleled by those 

of the Brachiopoda. 

These differences are not unimportant as it appears to me ; the question of the 

near relationship between the Oligochaeta and the Polychaeta has been prejudiced 

by the general term 'worm,' which has been applied to these and to so many other 

animals having even a more remote relationship. In the same way the Crustacea 

and the Insects have been confounded together owing to certain obvious similarities. 

The Insects are distinguished mainly by the possession of tracheae, which are totally 

different from any structures met with in the Crustacea; so the Oligochaeta are 

' See, however, above, where the very frequent occurrence of numerous gonads in Lumbricus is referred to. 



distinguished by the elitellum, and by other appendages of the reproductive system, 
which have not their counterpart in the Polychaeta ^. 

Only one group of Oligochaeta shows any particular resemblances to the Poly- 
chaeta, and that is the group represented by the genus Aeolosoma. There is no 
doubt that this Annelid does come very close in certain respects to some of the 
lower Polychaeta. It resembles them in the following points: — 

(i) In the presence of ciliated pits on the side of the head, found also in 

Saccocirrus, Ctenodrilus, and Polygordius, &c. 
(«) Ciliation of the prostomium, characteristic of Ctenodrilus, Saccocirrus, 
Protodrilus (general ciliation of body), &c. 

(3) Nervous system retains its primitive connexion with the skin. 

(4) Coelom divided up by irregular muscle-fibres, as in Saccocirrus. 

(5) Testicular products carried oif by very slightly metamorphosed nephridia; 

as in Saccocirrus, Capitellidae. 

(6) Setae often all acicular as in some Capitellidae. 

In other respects Aeolosoma is an undoubted Oligochaet. It has the distinctive 
elitellum (feebly developed, liowever) and spermathecae of that group. On the whole 
it appears to me necessary to place Aeolosoma near to the base of the Oligochaet 
series. I am not, however, at all certain whether it would not be more in accord 
with the anatomical facts at our disposal to place Aeolosoma among the Archian- 
nelida. If then Aeolosoma stands at the base of the series, is it necessary to derive 
the other Oligochaeta from this genus? Before attacking this problem it is requisite 
"to bear in mind that Aeolosoma, in spite of its primitive characters, is in some 
respects a degenerate worm ^. The absence of tjie ventral nerve-cord is undoubtedly 
a mark of degeneration ; so, too, is the almost complete absence of intersegmental 
septa ; there is at most a single septum present (see p. 178). The persistent connexion 
of the brain with the epidermis may be also rather due to degeneration than to the 
inheritance of an archaic character. The colourless blood, and lastly the small size 
are facts which appear to me to point in the same direction. These facts still 
further increase the difficulty of fixing the position of Aeolosoma in the Oligochaet 
series. The two families to which it appears to come nearest are the Naidomorpha 
and the Enchytraeidae. The resemblances, however, are not close to either. Both 

' EisiG considers the modification of the epidermis at the generative pores of the Capitellidae to be 
comparable to the Oligochaet elitellum. I should rather compare it to the glandular modification which 
surrounds the external aperture of the sperm-ducts in Lumbricus. 

' In this connexion I may be perhaps allowed to quote E. Meyek (Biol. Centralbl. x. p. sg6) to the 
effect that asexual reproduction only occurs in undoubtedly degenerate forms, or, in those where sexual 
products abound in the posterior, aa well as the anterior part of the body. 


the two groups mentioned agree with Aeolosoma in having no oviducts; in all three 
groups the ova are evacuated by pores or by a pore; but in the Enchytraeidae 
there seem to be traces of the former existence of a proper oviduct ; fringes of cells 
along the coelomic apertures of the oviducal pores seem to indicate the former 
existence of proper oviducts ; the resemblance, in fact, may be more apparent than 
real; as to the Naidomorpha we have but little information as to the exact nature 
of the oviducal pores. With the Enchytraeidae Aeolosoma agrees in the origin of the 
dorsal blood-vessel from a periintestinal sinus, and also in the presence of a cardiac 
body; traces of this seem to occur in Aeolosoma tenebrarum; the same structure is 
found in Ctenodrilus. Compared with the differences, the resemblances thus indicated 
are but slight ; there is really no group of Oligochaeta with which Aeolosoma has 
well-marked affinities ; I do not, therefore, think it unreasonable to place Aeolosoma 
far away from the other Oligochaeta, a derivative of the Annelid stock before it had 
thoroughly differentiated into the two groups, Polychaeta and Oligochaeta. I should 
leave it aside, in fact, in considering the origin of the remaining groups. 

As to these remaining groups, they are undoubtedly all nearly connected ; 
divisible though they are into Mici-odrili and Megadrili, there are no fundamental 
differences of structure. It is, in fact, perhaps a little doubtful into which division 
Moniligaster should go. We may, therefore, fairly consider them together with 
three possible conclusions in view: (i) that both Megadrili and Microdrili are 
derivable from a common stock ; (2) that the Megadrili have been derived from 
the Microdrili; and (3) that the reverse mode of development has occurred. 

First, as to structural characters which may fairly be regarded as indicating 
a low position; it is necessary to carefully eliminate those characters which may be 
due to degeneration; it would be absurd, for example, in my opinion, to consider 
Perichaeta hilgendorfi a primitive Perichaeta on account of the fact that it has no 
spermiducal gland, present in nearly all the other species of the family ; no doubt 
the presence of this gland is a secondary character; but in the particular case 
under consideration its absence must be secondary. One naturally turns to the 
reproductive organs for a clue to the difficulties of phylogeny. Considering that the 
reproductive ducts are developed from the nephridia of successive segments (in 
Octochaetus at any rate), it might be fairly supposed that the closer the correspondence 
between the ducts, the more primitive would be the characters of the worm ; the 
differences between the male and female ducts must, one would think, be secondary. 
Judged by this standard, Phreoryctes would have to be placed in the position of 
the most archaic form of Oligochaeta ; it is particularly worthy of note, in connexion 
with this matter, that the male-ducts of Phreoryctes, although longer than, and 

z 3 


therefore, to that extent, diflFerent from, the female-ducts, graduate into them; the 

anterior pair of ducts in fact are longer than the second pair; the latter, it is 

true, are longer than the oviducts, but still they form a transition to them. The 

absence of spermiducal glands here seems to me to be really a primitive character. 

There are other facts, too, in the structure of the reproductive organs which argue 

in favour of the primitive characters of the Phreoryctidae ; I have pointed out 

elsewhere that in the embryo Lwmh'icus and Octochaetus there are four pairs of 

gonads, of which only three come to maturity; in certain species of Perichaeta 

there are two pairs of egg-sacs, corresponding to the presumably ancestral two 

pairs of ovaries ; now in Phreoryctes, at least in P. smithii, there are actually two 

functional pairs of ovaries and oviducts. The development of Octochaetus seems 

also to distinctly favour the view that the primitive form of generative duct is that 

which only occupies two segments; the external pore being on the segment following 

that which contains the funnel. In the development of that Annelid the ducts in 

question ran straight to the body -wall, and there ended ; how the further growth 

was effected I had no facts to enable me to judge. Besides, apart from the actual 

facts of development, the necessity of assuming an early correspondence between 

the male and female-ducts would lead to the assumption that the short male-ducts 

were the most primitive ; there is no Oligochaet known in which the female-ducts 

occupy more than one segment ; whereas there is every possible variation in the 

number of segments occupied by the male-ducts; this of itself makes it probable 

that the female-ducts represent the earlier condition. All arguments, therefore, appear 

' to me to point to the conclusion that Phreoryctes is, in respect of its reproductive 

organs, the most primitive type. The vascular system of the Annelid is also in 

a primitive condition, though not more so than that of the Tubificidae. The 

development of Lumbricus (Vejdovsky) shows that a perivisceral trunk in each 

segment of the body is the primitive condition; apart altogether from the facts 

of development, this would seem to be on a priori grounds likely; we know, too, 

that the reduced number of the commissural vessels in the Naids is derived from 

a more primitive condition, in which there was a perivisceral arch in each segment 

(see below). In this respect Phreoryctes is primitive; it is true that in one species 

of the genus, at any rate, the periviscerals are not continuous round the body ; but 

they are in P. smithii. 

There is no type in fact, in my opinion, which has such good claims to occupy 
a low position among the Oligochaeta as Phreoryctes. It wiU be remembered also 
that this genus is one which was placed by Lankestee in a position intermediate 
between the ' Limicolae ' and the ' Terricolae ' of CLAPABiiDE ; it does undoubtedly 



combine the characters of those two groups, which are substantially the same as 
the Microdrili and the Megadrili of the scheme advocated here. It is also a form 
which, as regards habitat, is on the border-line between the two divisions ; it lives 
both in the water and on the land. I look upon Phreoryctes as representing, more 
nearly than any other existing form, the common type whence the Megadrili and the 
Microdrili have been derived. The recognition of this worm, as occupying the 
most primitive position in the two groups, seems to be at variance with the assump- 
tion supported above of the Perichaetidae being a primitive type on account of the 
continuous circles of setae. The question is whether it is conceivable that the 
perichaetous condition could have twice been reduced to the. condition of four groups 
of setae ; this may perhaps be a little difficult of conception ; but the necessary 
alternative is that from the eight setae per segment the perichaetous condition has 
been at least twice independently arrived at. The one hypothesis is to my mind 
at least as difficult as the other. In conclusion, the following scheme embodies the 
result of the foregoing remarks. 

Kg. 34. 

Aeanthodrilidae. Crjptodrilidae. 


Naidomorpha. Tiibifioidae. 









[Earthtvorms of Unrecognizable Position. — In the succeeding pages a number of species are dealt 
with, originally described under the generic name of Lumbricus, but really referable to other 
families or genera. A few forms remain which are undoubtedly earthworms, but which cannot 
be referred, even with probability, to any family. These species are the following : — 

Lumbrictis capensis^ (Kinbeeg), L. helenae (Kinberg), L. 'hortensiae (Kinberg), L. rubrofasciatus 
(Baird 3), L. vineti (Kinberg), L. apii (Kinberg), L. pampicola (Kinberg), L. tellus (Kinberg), 
L. iakitanus (Kinbebg), L. juliformis (Baird 3), L. guildingi (Baird 3). 

The genus Eurydame of Kinberg is regarded by Vaillant as synonymous with Titanus 
( = Geoacolex). As a matter of mere guess work, for Vaillant's suggestion is nothing else, I should 
rather prefer Aniens or Ehinodrilus. Eurydame insignis has the setae more separate posteriorly 
than anteriorly; it is a small worm from Panama 58 mm. in length. Segesypyle hanno, Kinberg 
is from Port Natal and is only 28 mm. in length. There are eight setae per segment, separated 
posteriorly ; anteriorly the ventral are closer than the dorsal. Is this possibly my Acanthodrilus 

Helodrilus oculafus, Hoffmeister. 

H. oculatus, W. HOPPMEISTEE, Die bis jetzt bek. Art. Regenw., p. 39. 

This is an extremely mysterious species, neglected by Rosa in his recent revision of the 
Lumbricidae, and, therefore, probably not believed bj' him to be a Lumbricid. Its most remarkable 
structural peculiarity is a pair of eye-spots on the buccal segment. There are four pairs of setae 
in each segment, which are straight instead of curved and said to be black ; the male-pores are 
"upon the fifteenth segment. The body is elongate and pink in colour ; the length at most 135 mm. 
It occurs on the sea shore in pools more or less dried up. 

Vaillant (6, p. 168) suggests that this worm is probably a Tubificid, on account of the 
presence of eye-spots, and on account of its habitat. The black setae are very suggestive of what 
I have myself observed in Tubifex rivulorum. But it does not seem to me that we are justified in 
relegating the genus to any family at present. 

1 Certainly not a Lumbricus. I have examined, by the kindness of Prof. Loven, Kinbero's type ; but 
such was the condition of the worm that I can only mention the fact that the gizzard is situated far 
forward, which is not a character of the Lumbricidae. 


Hypogaeon hirtum, Savigny. 

H. hirtum, J. C. Savigny, Syst. d. Annel., p. 104. 

The chief peculiarity of this species, which was met with in the environs of Philadelphia, 
is the nine parallel rows of setae. These setae are described as long, covered with spinelets, 
particularly upon the clitellum, where their distribution is described as confused. This suggests 
my genus Trichochaeta ; but this suggestion is not supported by the position of the clitellum 
(xxvi-xxxix), the male-pores (?) on xv and the spermathecal pores (?) on x, xi, xii; it is impossible 
even to guess at its position. 

Hypogaeon atys, Kinberg, might belong to any family; it has eight setae per segment and is 
32 mm. long. 

Another doubtful form, concerning whose position in the system nothing positive can be said, is : — 

Alma nilotica, Geube. 

Alma nilotica, Grube, Arch. f. Naturg., 1855, p. 129. 

Digitihranchus niloticus, Levinsen, Vidensk. Meddel., 1889, p. 321. 

I do not give any systematic description of this worm, since its position in the series is very 
doubtful. Gbtjbe, its original describer, does not give suflScient particulars to permit of a certain 
opinion as to its relations, "but compares it with Ehynchelmis (' Euaxes'). The illustration (Grube 
3, PI. V, fig. 11) represents a worm of the thickness and length of a moderately-sized earthworm, 
its measurements are given as 3-6 inches. During life it was, according to RilPPEL's notes, 
who was the discoverer, red coloured; it was found in the Nile mud. The most remarkable 
peculiarity of the worm is the possession of branchial processes upon the posterior segments of the 
body ; it is the presence of these structures which have led most writers to exclude it from the 
Oligochaeta altogether, or, at least, to consider its Oligochaetous affinities doubtful. As we are now 
acquainted with three undoubted Oligochaeta (viz. Chaetohranchus, Hesperodnlus, and Branchiura) with 
branchiae, there is no longer any reason, on these grounds, to refuse admittance to Alma among the 
Oligochaeta. Levinsen's Digitihranchus niloticus must, in my opinion, be regarded as generically, as 
well as specifically, identical with Alma nilotica; it was described from a fragment only, but so far 
as this fragment enables one to judge there are no differences from the worm described by Geube. 

The branchiae occupy the last sixty segments of the body (as in Branchiura), they lie just above 
the dorsal pair of setae ; they are small cylindrical processes with rounded extremities, and there 
are several of them (five or six) upon each side of the body; occasionally two or more arise from 
the same base. The setae are strictly paired, and according to Levinsen's figure (1, fig. 5), are 
exactly like the peculiar setae of Siphonogaster (see below). There is not much information in 
Grube's paper about the internal characters, but the vascular system seems to be, from what 
he says, comparable to that of an Oligochaet. The existence of a vascular system at all is, as 
Dr. EisiG has pointed out ', fatal to a comparison with the Capitellidae ; AlmM has been compared 
by SlMROTH^ vnth the Capitellid genus Mastobranchtts, to which it certainly bears not a little 
resemblance ; but neither Mastobranchus nor any other Capitellid possess a vascular system.] 

' 'Die Capitelliden.' Fauna u. Flora des Golfes von Neapel. 

2 Die Entstehung der Landthiere: Ein Blologlscher Versuoh, Leipsic, 1891, p. 236. 




As there is only one certainly known genus, viz. Aeolosoma, I do not attempt 
a definition of the family. 

Genus Aeolosoma, Ehrenberg. 

Syn. Aeolosoma, Ehrenberg. Aeolonais, Gervais. Ghaetodermos, Leidy. 

Dbpibtitioit. Freshwater Oligoehaeta of small size ; segmentation not marked by- 
regular septa ; prostomium ciliated ventrally ; setae in four bundles of one 
to six setae each, usually capilliform (only sometimes both capilliform and 
sigmoid). Clitellum developed only on ventral side of segments v-vii; testis 
single in v ; ovary single in vi ; no differentiated sperm-duets ; oviduct 
a ventral median pore in vi. Spermathecae in some or all of segments iii-iv. 
Ifervous system represented by cerebral ganglia only. Lateral ciliated pits 

Fig. 35- 

(After Lankester.) 
A chain of two individuals. 

This genus comprises a few species 
which are distributed over Europe, W. Asia, 
N. Africa, N. America, Central America, 
(New Granada), India, and Africa, in fresh 
water. They are all of small size, the 
largest species, A. tenehrarum, reaching a 
length of only lo mm. They are transparent 
worms, with a very ' loose and uncertain 
outline, the body flowing in various directions 
in a very strange liquid fashion.' The seg- 
mentation is not very marked externally, 
except by the implantation of the setae. 
When an AeolosoTna is fully extended the 
head segment is separated from the next by 
a constriction. The number of segments in 
an individual is small, not more than a dozen 
to fifteen. The prostomium is large, and 
is ciliated upon the under surface ; on either 
side are a pair of ciliated pits, which are 



Fig. 36. 

probably sense-organs ; they lie close to the brain, and in front of the mouth- 
opening. The setae are, in all the species of the genus except three, capilUform ; 
they are implanted in bundles, of which there are four to 
each segment ; each bundle contains a varying number of 
setae, of varying sizes, from two to nine ; the number of 
setae in the breadth is, in some cases at least, a cha- 
racter of specific importance. The only species in which 
there is any specialization of the setae are A. tenebrarum, 
A. niveum, and A. leidyi ; the posterior segments of 
these species have ./-shaped setae, which may or may 
not be bifid at tips. The body generally, and the pro- 
stomium especially, has minute, hair-like, rigid processes, 
which may be of a tactile nature^. 

The epiderTuis is covered by an excessively thin 
cuticle, only recognizable (according to Vejdovsky) after 
treatment by reagents. The cells of the epidermis are 
of two kinds ; there are more columnar cells, among 
which lie large, round, glandular cells. On the ventral 
surface of the prostomium the ciliated cells are figured 
by Leidy (6, H. ii. fig. 11) and Vejdovsky (24, PI. i. 
fig. 36) with a hexagonal contour. The gland-cells are 
of two kinds, coloured and colourless. The coloured 
epidermic gland-cells are highly characteristic of the 
genus ; they are found in every species of this genus, 
and Ctenodrilus, which is possibly an ally of Aeolosoma. 
Their colour varies according to the species ; they are 
red-brown in A. quaternariuTn, &c., olive green in 
A. tenebrarum, bright green ^ in A. varium, and bright 
green with a bluish tinge in A. headleyi. The colour is 
due to a large drop of oily substance within the cell ; 
that this drop is of a fatty nature appears to be shown 

by its black staining with osmic acid. The pigment itself j^^^ corpuscles, i.. second pair of 
is probably closely related, in spite of its difierence nephridia. 12. setae. 13. intestine. 

, . , 14. Coloured oU-globule in skin. 

of hue, in all species; I have shown (32, 74) that 

' Taillant (6) states that there are cilia over the general body-surface. This, however, is contrary to 
the statements of other authors and to my own observations ; the cilia are limited to the prostomium 
and to just behind mouth. 

^ The green colour suggests chlorophyll; Zachakias has recently suggested that the ' oil globules ' 

A a 


(After Vejdovsky.) 

:i. Cilia on under surface of pro- 
stomium, 3. Muscular cells. 4. Cili- 
ated pit, 5, Brain, 6. Pharj^nx, 
7. End of pbarynx, 8, First pair of 
nephridia, 9, Oesophagus, 10, Coe- 


in three species it is turned into a beautiful violet by the action of alkali ; it 
is dissolved out and destroyed by alcohol^. Colourless gland-cells appear to be 
particularly abundant in A. variegatuvi, and to be the only ones present in A. niveum. 
There can be no doubt that the coloured (and colourless) oil-globules form the contents 
of cells similar to those found in the epidermis of the Oligochaeta. Leidy (6), it is 
true, has spoken of them as nuclei either of the hexagonal cells of the prostomium 
or of the muscle-fibres of the body-wall ; but this is undoubtedly an incorrect opinion. 
The cells, however, which contain these bodies have, according to Vejdovsky, no 
nuclei ; so far as regards A. heviprichii I have been able (32) to confirm him, but not 
as regards A. tenehrarum (74), where the nuclei are not difficult to see. However, 
they are present in the developing cells of A. heynprichii; in such cells the coloured 
body first appears as a minute droplet, which gradually grows, pushing the nucleus 
to one side ; the nucleus seems to be itself converted into oil (Vejdovsky). 

The muscular layer of the body-wall is very delicate. Leidt (6, PI. ii. figs. 8, 9, 
12) figured a single layer only of circular fibres. Vejdovsky with difficulty, and then 
only occasionally, detected a delicate circular layer of fibres ; the longitudinal layer 
also exists, and I have figured both in A. headleyi (68, PI- xii. fig. 2 m, m'). 
Aeoloaoma is therefore in this respect a typical Oligochaet. The coelom is not 
divided by intersegmental septa into a series of segments ; one segment only, the 
first, is bounded by a distinct septum. Nevertheless, the body-cavity is traversed 
by muscular fibres which extend between the alimentary tract and the parietes ; 
and serve to keep the former in position. This statement applies no less to 
'A. quaternarnwi and hemprichii than to other species. The fibres in question 
are nucleated, the nucleus lying in an oval ^welling of the fibre at about its 
middle. They have been figured by Vejdovsky (24) and by myself (68). In 
the cavity of the prostomium, and in the section of the body-cavity round the 
pharynx, they are more numerous and closely set than elsewhere. There is 
occasionally a branching and anastomosis of the individual fibres. Similar uni- 
cellular muscular fibres (Beddard 68, PI. xii. fig. 3) are attached to the sacs of 
the setae, and serve to move them ; it is interesting to find that the muscles which 
connect together the seta-bundles of the same segment are histologically different; 
these strands (Beddaed 68, PI. xii. fig. 3 sm) consist of exceedingly fine fibrils, closely 
set together, and with no visible nuclei. These latter muscles appear to be absent 

are really symbiotic algae ; he observed them dividing. I believe that I have shown (74) that it is 
not chlorophyll. 

' In ^. tenebrarum a solution of iodine produces a remarkable effect ; a dense black colouration spreads 
over the outside of the oil-globule, which as rapidly disappears ; this effect was only seen in living cells. For 
further remarks on the pigment of Acolosoina I may refer to my two paper.% upon the subject (Nos. 32, 74). 


in certain species of Aeolosoma. The coelom ia of course lined by peritoneum, 
and contains a few round corpuscles, which vary in nunlber at different times. 
There are no dorsal-pores and no head-pore; the coelom, however, communicates 
with the exterior by the 

Nephridia. These organs are paired and metamerically arranged. They commence 
at earliest behind the first bundles of setae ; we might possibly, therefore, consider 
that they belong to the third segment, were it not for the position of the funnel ; 
the definite location of the nephridia is, however, rendered impossible by the 
absence of the septa separating the segments ; Vejdovsky considers that the first 
pair of nephridia belong to the second (i.e. the first setigerous) segment. In this 
case the entire nepbridium is confined to a segment, for the funnel does not 
perforate the only existing septum, that dividing segments i and ii. The fact 
that the external pore of the nepbridium (in A. headleyi at any rate) lies in 
advance of the funneP is a further confirmation of Vejdovsky's suggestion. In 
A, variegatv/m the first nepbridium is in segment iii. The nepbridium is attached 
by a delicate muscular band to the gut ; each is a much-coiled tube, with 
granular nucleated walls. Stolc has figured (1, PI. vii. fig. 4 a) a peritoneal 
covering of transparent vesicular cells ; the ciliated funnel is composed of very 
few cells ; it is hardly wider than the tube which follows. Just before the 
external pore, which lies ventrally in front of, and to the inside of, the ventral 
setae-bundles, the nepbridium dilates into a clear-walled vesicle, which Vejdovsky 
states to be contractile, and to be filled with a clear fluid. 

During sexual maturity some of the nephridia, as Dr. Stolc discovered (1), 
disappear; the others, particularly those of the sixth segment, convey the sperm to 
the exterior ; in the sixth segment one nepbridium only is figured by Stolc (1, 
PI. vii. figs, a, 3 e c h) ; this is larger than the rest, and the funnel, instead of facing 
backwards, as in the other nephridia, faces forwards. It also appears to be shorter. 

Vejdovsky discovered that in the buds of Aeolosoma the first segment is 
provided with a pair of provisional nephridia. These organs were chiefly studied in 
A. tenebrarum; they are most evident in the head segment, before the pharyngeal 
ingrowth has taken place, while the gut is still continuous from the parent worm 
to the bud; they lie closer to the gut on either side, and each consists of a thick- 
walled, hollow sac, of a pear-shaped outline ; the narrow end of this sac is attached to 
the dorsal body- wall, but no orifice leading to the exterior could be discovered ; nor 
is there any coelomic aperture. They disappear on the invagination of the pharynx. 

' Stolc, on the other hand (1), figures the funnel as in advance of the external pore; the relative 
position of the two, however, undergoes constant change owing to the movements of the worm. 

A a 2 


The alimentary canal, which is ciliated throughout, may be divided into 
a pharynx, an oesophagus, and an intestine ; the pharynx occupies the first segment 
of the body; in the young worm, budded off from the parent, the pharynx appears 
as an epiblastic involution; it is lined by cubical, ciliated cells, and has a layer 
of circular fibres. In A. tenebrarum the hinder margin of the pharynx gives rise, 
on either side, to a small, thick-walled diverticulum, to which the last of the muscular 
bands supporting the pharynx are attached. Following upon the pharynx is 
a narrow oesophagus, occupying two or three segments, and commonly furnished at 
its commencement with an oval dilation. The intestine is wide to begin with, and 
gradually narrows towards the terminally-placed anus. The structure of the intestine 
and oesophagus is exceedingly simple ; the walls of the tube consist of a cubical, 
ciliated epithelium, outside which is a peritoneal layer ; between the two, in the 
intestine, is interpolated a vascular network or sinus (see p. 73), and the dorsal blood- 
vessel lies underneath the peritoneal covering of the oesophagus. 

The vascular system is exceedingly simple. It consists merely of a dorsal and 
ventral trunk, which are united in the anterior part of the body ; the dorsal vessel is 
usually only recognizable in the oesophageal region, where it forms a pulsatory tube^ 
not, however, lying freely in the body-cavity, but beneath the oesophageal peritoneum ; 
anteriorly this vessel behind the brain divides into two trunks, which reunite to form 
the ventral vessel ; in a few examples of A. hemprichii Vejdovsky observed a pair 
of vessels given off from the cephalic ring, which, he believes, rejoin the dorsal 
vessel. Behind the oesophagus there is no dorsal vessel ; it ends here in a network 
of blood-vessels, ramifying in the walls of the intestine, beneath the peritoneal coat. 
The ventral vessel, on the other hand, lies , perfectly free in the body-cavity; 
anteriorly it is, as already mentioned, in communication with the dorsal-vessel by 
a pair of lateral commissures passing round the pharynx. In the intestinal region it 
gives off a number of trunks, sometimes regularly paired, sometimes not, which join the 
intestinal network. This intestinal network appears not to exist in either A. quater- 
nanum, A. variegatum, or A. headleyi; it is replaced in those species by a paired 
blood-sinus^. Proper walls to the blood-vessels have not been clearly demonstrated. 
Blood-corpuscles have, however, been found by Eichwald, Maggi ^, and Vejdovsky, 
but only in the dorsal vessel. In several species there is a row of somewhat 
fusiform cells in the dorsal vessel, which have yellowish fat-drops in their interior ; 
I have suggested (4) that these cells probably represent a rudimentary dorsal organ 

1 I say 'paired blood-sinus' in deference to Vejdovsky; in A. headleyi (see my figure 68, PI. fig. 6) 
the sinus did not seem to be paired, but to be continuous all round the intestine. 
' The corpuscles described by Magmji are, however, probably eoelomic. 


such as is found in the Enchytraeidae. The colour of the blood -plasma of 
A. hemprichii is stated by Lankester (6) to be of a pinkish colour. 

The nervous system consists of little else than the cerebral ganglion. The simplest 
cerebral ganglia are those of A. variegatum, where the fissure indicating its double 
nature is hardly visible ; in A. tenebraruni it is most complicated, with well-marked 
lateral lobes ; this species also alone shows any traces of a ventral nerve-cord, which 
is very short and is not connected with the brain. The latter is perfectly continuous 
with the epidermis, and is composed of cells dorsally and of nerve-fibres ventrally 
(A. variegatum) or of cells only ; it gives off numerous branches. Vejdovskt 
throws considerable doubt upon Maggt's description of a ventral cord in A. hemprichii, 
which is not accompanied by any figures. 

Generative organs. The generative organs of Aeolosoma .have been described by 
d'Udekem (2), Maggi, and more recently, as well as more fully, by Stolc (1). 
The reproductive organs are only occasionally developed, the usual mode of repro- 
duction being by budding and division. Curiously enough, sexual and asexual 
reproduction may be, according to d'Udekem (2) carried on pari passu. Maggi's 
figure of the ovaries and oviducal pores (fig. 8) is, it should be observed, an 
almost exact copy of that of d'Udekem (2, fig. i) ; in another figure, however, he sup- 
plements d'Udekem by illustrating the spermathecae, which were apparently not seen 
by d'Udekem. All these observers appear to have investigated the same species, viz. 
A. hemprichii and A. quaternarium. The testis is median and unpaired, and lies in 
the fifth segment ; the ovary occupies a corresponding position in the sixth segment. 
As has been already mentioned, there are no sperm-ducts, the nephridia, particularly 
that of the sixth segment, which is slightly different in structure from the rest, 
serving as conduits of the sperm. The ova, which are few and large, and apparently 
undergo amoeboid movements, escape by a large pore on the ventral surface of the 
sixth segment. The spermathecae are small oval sacs, one to three pairs occupying 
segments iii-v. 

At the epoch of sexual maturity a clitellum is formed, which is figured by Stolc 
as limited to segments v-vii; it is only developed on the ventral side of the 
body. It is not known whether a cocoon is formed ; the structures described by 
Eheenberg, and described and figured as developing eggs by Maggi, are probably, 
as Vejdovsky (24) suggested, encysted worms. At any rate, I have recently been 
able to show (27) that this annelid can encyst itself. The cysts are of variable 
thickness, usually spherical, but sometimes oval or of irregular form. I found them 
at the commencement of cold weather, at the end of autumn. 

The asexual propagation of Aeolosoma has been studied by Vejdovsky; the 


species selected was the large Aeolosoma fenebrarum. The most interesting fact 
in this process is the entire absence of the formation of any budding zone such as 
ejdsts in the Naidomorpha (see below). In Aeoloscmia there is simply a division 
of the parent organism into two ; the formation of the head of the newly-formed 
individual is ushered in by a dorsal thickening, which is the basis of the brain; 
the next event is the increase of the future head of the daughter zooid and the 
appearance of muscular strands. Very soon there appear a pair of provisional 
nephridia, such as seem always to occur in the head of the lower Oligochaeta when 
produced in this asexual way. It is important to observe that the pharynx is a new 
structure invaginated from the epidermis, and not produced by any modification 
of the existing section of the gut. This formation of the pharynx was also found 
to occur in Aeolosoma vanegatum (see Vejdovsky 18). 

Species of Aeolosoma. The division of the genus into species is a matter 
of no little difficulty, on account of the very imperfect descriptions which the older 
authors have given. Only six species can be at present regarded as well established ; 
four of these, viz. A. hemprichii, A. quaternarium, A. tenebrdrum, and A. variegatum, 
have been clearly discriminated by Vejdovsky (18 and 24) one, A. niveum, by 
Leydig (4) ; the remaining species, A. headleyi, I have described myself (68), and 
believe to be distinct from any of the above. The following is a complete list of 
all the species which may possibly be regarded as distinct. 

(i) Aeolosoma hemprichii, EsRENBEBa. (6) Aeolosoma headleyi, Beddaed. 

(a) „ quaternarium, Ehrenbeeg. (7) ,, niveum, Leydig. 

(3) „ variegatum, Vejdovsky. (8) „ leidyi, Ceagin. 

(4) „ aurigena, Eichwald. (^) „ ternarium, ScHMAEDA. 

(5) „ tenebrarum, Vejdovsky. (10) „ macrogaster, Schmaeda. 

The following species are somewhat uncertain in position and affinities. 

Two of the species described by Schmabda, viz. Aeolosoma ternarium and A. macrogaster, 
have been referred by Vaillant to a distinct genus, Pleurophlebs, on account of (i) absence of 
coloured globules in the skin, (2) the presence of two lateral blood-vessels. These two species 
appear Ukely to prove very interesting, but their structure requires, in my opinion, a re-examination 
before they can be admitted. 

Aeolosoma pictum of Schmabda (from New Granada) has red oil-globules and a spiral intestine ; 
it is very doubtfully distinct from one or other of the red-coloured species, A. quaternarium or 
A. hemprichii. 

Aeolosoma chlorostictum of Wood Mason is only known by a name ; he referred to it inci- 
dentally in describing an infusorian parasite. 

VoELTZKOW has recorded from Madagascar an Aeolosoma with 'golden-yellow' oil-drops,, but 
no further information is given. 


Aeolosoma stokesii of Cbagin with ' salmon-coloured nuclei ' is probably either A. quaternarium or 
A. hemprichii. 

A number of other names, which have been applied to supposed species, will be considered 
under the descriptions of the species with which they are probably identical. 

(i) Aeolosoma hemprichii, Ehrenbeeg. 

A. hemprichii, Ehrenbeeg, Symb. Phys. 1831. 

A. decorum, 

A. ehrenbergi, Oeested, Naturhist. Tidskr. 1842, p. 135. 

A. balsamoS Maggi, Mem. Soc. Ital. Sci. Nat. 1865, p. 9. 

A. quaternarium, Lankestee, Trans. Linn. Soc. 1867, p. 641. 

Chaetodemus multisetosus, Czeeniavsky, Bull. Soc. Nat. Mogc. 1880, p. 307. 

Definition. Head broader than the following segment. Setae nearly straight ; longer and 
shorter ones in the same lundle. Integumental glohules orange-red to dark crimson. 
SujM-aoesophageal ganglion divided posteriori)/ into two hy deep fissure. Nephridia 
commence in first setigerous segment. Intestine surrounded by a network of capillaries. 
Localities — England, Europe, N. Africa. 

If all the species which have received the above names are really one, it is clear 
that, as Vaillant has pointed out, the first of Ehrenbebg's two names has the 
priority, and that Oeested had no right to introduce the new name ' ehrenbergi.' 
The synonomy of the species however is difficult. I include A. balsamo in deference 
to the high authority of Prof. Ve.tdovsky, who was evidently influenced in this 
identification by the large size of the head as figured by Maggi. In the latter 
'species,' however, Maggi figures the setae of the lateral bundles as longer and more 
numerous than those of the ventral bundles. If this character were established it 
would undoubtedly be of specific value. Lankester's A. quaternariuTn is certainly 
identical with Vejoovsky's A. ehrenbergi, though it is regarded by Vaillant as 
synonymous with A. quaternarium of Ehrenbeeg. This is apparently because of 
the error (?) in Lankester's illustration, which shows the prostomium to be of equal 
or less diameter than the body. The comparatively large size of the worm, and the 
presence of nephridia in the first setigerous segment, as well as the form of the setae 
(straight), and the presence of an intestinal network prove that Lankester's A. 
quaternarium is wrongly identified by Vaillant. 

' Corrected to 'Balsamoi' by Czerniavsky. 


(2) Aeolosoma quaternarium, Eheenbekg. 

A. quaternarium, Ehrenbekg, Symb. Phys., 1831. 
A. venustum, Leidy, Journ. Acad. Nat. Sci. Philad. Vol. 11 (1850), p. 46. 
A. italicum, Magri, Mem. Soc. Ital. Sci. Nat. Vol. I (1865), p, 8. 
Chaetodemus quaternarius, Czeeniavsky, Bull. Soc. Nat. Mosc. 1880, p. 307. 

Deflnition. Head equal in breadth to following segments. Setae sliarply bent ; those in 
each bundle of the same length. Supraoesuphageal ganglion divided into two posteriorly 
by deep fissure. Integumental globules orange-red. Oesophageal segments without nephridia. 
Intestine surrounded by a paired blood-sinus. The worm can temporarily encyst itself. 
Localities — England, Europe, N. America. 

This is quite a distinct species from the last, as will be seen on comparing the 
definitions. A. italicum was incompletely, and possibly erroneously, described by 
Maggi. If correctly described in all particulars it should be referred to a distinct 
species, characterized by having only one pair of bundles to each segment instead 
of two pairs — each bundle containing only ten setae. The small prostomium and the 
small size of the worm are the chief reasons (I presume) which led Vbjdovsky to place 
it as a synonym of A. quaternarium. It is not clear why Vaillant regards A. italicum, 
as synonymous with A. hempricliii. In Maggi's paper the 'eggs' of A. italicum are 
described and figured. Vejdovsky suggested that these, being surrounded by a capsule, 
were cocoons, but remarked later (9) that it was not certain whether the structures 
in question weie really cocoons or encysted worms. That the latter suggestion was 
probably correct was shown by myself (27), and ]py him (2). The cysts are usually regu- 
larly spherical, rarely oval or of a more irregular rounded form. The cyst wall is of an 
appreciable thickness and is in some specimens thicker than in others. The cysts were 
found at the beginning of the cold weather. The formg,tion of cysts in A. hemprichii 
is perhaps hardly an additional argument for its identity with A. italicum, the only 
other Aeoloso7na in which they have been described. Leidy's A. venustum is only 
distinguished from A. quaternarium, according to Leidy himself, by its smaller size 
and longer setae. These characters can hardly be considered sufficient. 

(3) Aeolosoma variegatum, Vejdovsky. 

A. variegatum, Vejdovsky, SB. Bohm. Ges. 1886, p. 275. 

Deflnition. Head broader than following segments. Setae longer and shorter in the same 
bundle, sharply bent. Supraoesophageal ganglion only just divided by a shallow furrow 


posteriorly. Integurriental glolules colourless and bright-green. OesopJiageal segments 

without nephridia. Intestine surrounded hy a paired blood-sinus. Localities — Germany, 


This beautiful little Aeolosoma was first referred to by Vejdovsky, in his illustrated 

memoir (24), as probably identical with Leydig's A. niveum (p. 113, footnote), but 

afterwards (18) regarded as a new species and fully described with illustrations. Some 

examples sent from Cork by Prof. Hartog have allowed me to verify the chief points 

in Vejdovsky's paper. It may be the same as a species mentioned by Zachaeias (1). 

(4) Aeolosoma aurigena, Eichwald. 
Nais aurigena, EiCHWALD, Bull. Soc. Nat. Mosc. 1 847, p. 359. 
Definition. Two lines in length. Head broader than following segments. Integumental 
globules golden-yellow, lying close together in regular longitudinal lines. Three or four 
setae in each bundle. Locality — Russia. 
Whether this species is really distinct from A. tenebrarum is uncertain. Vejdovsky 
considers that it is chiefly on account of the regular distribution of the integumental 
globules. Eichwald's description is unfortunately defective in many important 
particulars which require to be known before the species can be properly defined. 
He describes and figures the dorsal vessel as running along the whole length of the 
body ; but probably the ventral vessel was mistaken for the dorsal. It apparently 
(so Vaillant thinks) inhabits the Baltic ; this may be an additional argument for 
its specific distinctness from A. tenebrarum, which has been hitherto only obtained 
from quite fr,esh water. 

(5) Aeolosoma tenebrarum, Vejdovsky. 
Aeolosoma tenebrarum, Vejdovsky, SB. Bohm.. Ges. 1879, p. 505 (footnote). 
Definition. Of considerable size (5-X0 mm.). Head broader than following segments, and pointed 
anteriorly. Setae capilliform and sigmoid ; supraoesphageal ganglion plainly double, with 
well-developed lateral lobes. Integumental globules pale-yellow to olive-green. Oesophageal 
segments with nephridia. Intestine with vascular network. Locality— Water from deep 
springs in Prague, England. 
This species is one of the largest of the genus. It has been fully described by 
Vejdovsky in two memoirs (17 and 24), and some details have been added by myself 
(74). Its structure is more complicated than that of any other species, and has already 
been for the most part described in the account of the generic characters of Aeolosoma 
<viven above. It is the only species except A. leidyi (if this is really distinct) and 



A. niveum in which there are two kinds of setae. The capilliform setae resemble those 
of other species, but the sigmoid setae are peculiar to this species and to A. leidyi and 
A. niveum ; they are like those of Naids, bifid at the free extremity ; these setae are 
found in the dorsal and ventral bundles, from the fourth setigerous segment onwards. 

It should be remarked that in the specimens described by myself under the name of 
A. tenebrarum the sigmoid setae of the posterior segments ave not bifid. I examined 
them with great care and with high powers, but could not detect the least notch at the 
extremity of the setae. The colour of the oil globules, too, does not agree with 
Vejdovsky's description or coloured figures of A. tenebrarum. These worms may 
therefore possibly belong to a distinct species. 

(6) Aeolosoma headleyi, Beddakd. 

A. headleyi, Beddakd, P. Z. S. 1888, p. 313. 

Definition. Of moderate size ; setae entirely capilliform. Integumental globules bright-green, 
occasionall)/ verging towards blue. First nephridium in first setigerous segment ; nine or ten 
pairs altogether. Habitat? [found in a tank at Zoological Gardens, London). 

This species evidently comes nearest to A. variegatum, but is to be distinguished 
by large number of nephridia. 

(7) Aeolosoma niveum, Leydig. 

A. niveum, Leydig, Arch. f. Anat. u. Phys. 1865, p. 360. 
A. laeteum, TiMM, Arb. Zool. Zoot. WUrzb. 1883, p. 155. 
? Chaetodemus panduratus, Leidy, P. Aca(J Nat. Sci. Philadelphia, V. 185a, p. 286. 

Definition. Prostomium not wider than folloiving segments. Setae sigmoid and capilliform. 

Integumental globules colourless ; only one pair of nephridia, at end of oesophageal region. 

This species was described and figured by Leydig (4). It is of very minute size, 
apparently smaller even than A. quaternarium and A. variegatum. Leydig thought 
that it was possibly the young of ' ^. decorum ' of Eheenbeeg, but pointed out that 
this was rendered less likely by the fact that Ehrenbeeg observed the red-coloured 
oU-drops of the latter species while yet within the egg. In all probability the ' ecrgs ' 
mentioned by Eheenbeeg are the cysts which I have described, and thus this 
argument falls to the ground. Vejdovsky thought that his A. variegatum might be 
the same. The same form was found in the Main by Timm, and mentioned under 
the name of 'Aeolosoma laeteum, Leydig' — doubtless a slip of the pen. I have 
contributed a few notes (25) upon this species, pointing out the existence of sigmoid 


setae. It is possibly identical with Leidy's Chaetodemus panduratus, also a transparent 
and colourless species. I could not myself detect even colourless oil-drops. 

(8) Aeolosoma leidyi, Cragin. 
A. leidyi, Cragin, Bull. Washburn Coll. Lab. vol. ii. p. 3. 
Deanition. Anterior segments with capilliform and sigmoid, posterior segments with sigmoid 
setae only. Integnmental glohules pale olive-green. 

As defined, this species can be distinguished from A. tenebrarum by the distribution 
of the sigmoid setae, which, in the last-mentioned species, are only found in the seta 
bundles of the posterior segments. 



Defiwitiow. Aquatic or terrestrial Oligoehaeta of slender form, often exceedingly- 
long. Setae in four rows of single setae or pairs, sigmoid. Testes in X, XI. 
Ovaries in XII, XIII, or XIII only. Sperm- ducts, two pairs opening 
separately, without spermiducal glands Spermathecae in front of testes, without 
diverticula. No genital setae. 

This family contains at present only two genera, viz. Phreoryctes, which has been 
long known, and Pelodrilus, recently described by myself. As these genera, although 
both conforming to the above definition, present numerous structural differences, it will 
be better to describe them separately instead of giving a general sketch of the anatomy 
of the family as has been done in other cases. 

Genus Phreoryctes, Hoffmeistee. 

Syn. Phreoryctes, Hoffmeistee. Haplotaxis, Hoffmeistbr. Nemodrilus, Clapaeiide. 

Georyctes, Schlotthatjbee. 

Defiwitioit. Setae in four rows of single setae or pairs. Clitellum XI-XIV. 
Frostomium divided by a transverse constriction. Testes in X, XI ; ovaries 
in XII, XIII. Vasa deferentia open separately on to XI and XII ; oviducts 
on to border line between segments XIl/XIII and XIII/XIV. Spermathecae 
two or three pairs in VII, VIII (IX). 

This genus was originally described by Hoffmeistee (3) under the name of 

B b a 


Haplotaxis, subsequently (1, p. 40) altered to Phreoryctes for the reason that Ilaplo- 
taxis had been previously used in Botany. Schlotthaubee altered the name to 
Georyctes for the reason that the worm is not exclusively an inhabitant of 'wells,' but 
is occasionally found in the soil. This change cannot of course be allowed. Claparede 
(2) applied the name Nemodrilus to a worm found by him in the Rhone, in ignorance 
of its identity with Phreoryctes'^. Accordingly the name Phreoryctes must stand, 
unless, indeed, it be thought that the same name may be allowed to an animal and 
a plant : but in any case the name Phreoryctes is now so well established that it 
seems hardly worth while to make a change. The genus Phreoryctes has been found 
in a good many parts of Europe (not in England), and in New Zealand, and in North 

The species or the genus are of considerable length and thinness (this is less the 
case with P. smithii), and the prostomiuTn is generally divided into two by 
a transverse constriction — a character also met with in the Capitellidae, which 
were at one time, not, however, for this reason, included in the Oligochaeta. The 
setae in the European Phreoryctes are of the typical Lumbricid pattern, sigmoid 
and not cleft at the free extremity. As a rule there are only four setae to each 
segment, which are implanted singly and at equal distances from each other ; each 
has commonly a nearly complete ' soie de remplacement ' beside it. Each of these 
setae, together with the incomplete seta lying beside it, represents the pair of setae 
in other Oligochaeta. The dorsal and ventral setae are usually of different sizes. 
A more normal arrangement of the setae is met with in P. smithii, where they are 
distinctly paired ; here, however, we still find the same difference in size between 
the dorsal and ventral pairs. The dorsal setae are largely obsolete in P. emissarius. 
The clitellum is not extensive ; it occupies segments xi-xiii inclusive and a portion 
of segment x. 

The ventral nerve-cord is provided in each segment with a peculiar appendage, 
first described by Leydig (6), and afterwards by Timm and Forbes ; these struc- 
tures, the ' ventral organs ' of Timm, are pyramidal masses of nucleated cells which 
support the nerve-cord as the ' chairs ' support the railway lines ; they are hollowed 
out where the nerve-cord rests upon them. These organs are regarded by Timm as 
of sensory function. Similar organs occur in the Enchytraeidae. 

The most remarkable feature about the vascular system is that (according to 
Clapar^de) the ventral vessel is contractile - a condition which is, so far as we are 
aware, unique among the Oligochaeta. 

The reproductive system is in several respects peculiar and interesting. The 

' This identity w-aa first pointed out by Vaillant (3, p. 249). 


existence oi four pairs of gonads was pointed out first of all by Letdig, and confirmed 
later by myself (69). I was able to show that these gonads are composed of two pairs 
of testes in x and xi, and two pairs of ovaries in xii, xiii. In no other Oligochaet 
are there more than one pair of ovaries. The ducts are interesting on account of 
their simplicity. Oviducts can only be distinguished from sperm-ducts by their 
position and smaller size. The funnel opens into one segment and the external 
aperture is upon the next. The second pair of sperm-ducts is shorter than the first 
pair, and thus offers a transition to the oviducts, which are shorter still. The 
spermathecae are two (P- smithii) or three (P. menkeanus) pairs, in segments vii, viii, 
and ix. They are simple pouches without diverticula of any kind. 

The number of species of this genus is at least four ; and they are all well character- 
ized. They may be recognized as follows : — 

Setae paired Comparatively stout worm. 

P. smithii. 
Dorsal setae longer than 

Prostomium divided by 

transverse constriction i Setae implanted 


ventral. P.filiformis. 

Ventral setae longer than 

dorsal. P. menkeanus. 

Prostomium not divided. Dorsal setae disappear posteriorly. P. emissarius. 

(i) Phreoryctes menkeanus, Hoffmeistee. 
p. menkeanus, Hopfmeistee, Die bis jetzt bek. Art. &c. 1845, p. 40. 
Haplotaxis menkeanus, HoFFMEisTEE, Arch, f Nat. 1843, p. 193. 
Georyetes menkei, Sohlotthaubee, Amt. Ber. Vers, deutsch. Naturf. zu Gottingen, 
1854, p. laa. 
Definition. Length, 600 mm. ; diameter, 1 mm. Number of segments, more than f)00. Ventral 
setae longer than dorsal, both implanted singly. Habitat — Europe. 
Besides above references see Leydig (6), Vbjdovsky (24): Giaed (2), Timm. 

(2) Phreorydtes filiformis, Glapae^ide. 
Nemodrilus filiformis, Clapaeede, M^m. Soc. Phys. Gen. 186a, p. 275. 
Phreoryctes heydeni, Noll, Arch. f. Nat. 1874, p. 360. 
Phreoryctes filiformis, Vejdovsky, SB. Bohm. Ges. 1875, P- '^9^- 
(?) Georyetes lichtensteini, Sohlotthaubee, Amt. Ber. Vers, deutsch. Naturf. 
Gottingen, 1854, p. laa. 
Definition. Length, 140 mm.; diameter, imm.; dorsal setae longer than ventral, implanted 

singly. Habitat — Europe, 
This species is confounded with the last by Vaillant, although both Noll and 


Vejdotsky have pointed out its differences from P. menkeanus. Noll has stated 
that the setae of the present species are shorter (a mm. instead of 3 mm.) and much more 
hooked than those of P. onenkeanus. Noll, however, figures and describes the ventral 
setae as longer than the dorsal. Vejdovsky states the opposite, and Vaillant 
mentions an individual in which the setae were about equal, which rather casts 
a doubt upon the distinctness of the two species, though the length is so very 

(3) Phreoryctes sraithii, Beddaed. 
p. smitMi, Beddaed, Ann. and Mag. Nat. Hist. 1888, p. 389. 
Definition. Comparatively dout worm. Setae paired, dorsal setae, of the posterior segments, 

longer than ventral. Two pairs of spermatheaae in VII, VIII. Hahitat — N. Zealand. 

There can be no doubt about the distinctness of this species; it is of a stouter 
build, not so long in proportion to its breadth as are the remaining species of the 
genus. In the paper cited above (and in Beddaed 18) will be found such details as 
are known about the species. The anatomical facts have been for the most part 
mentioned in the description of the characters of the genus. Other points, possibly 
of specific importance, may be now referred to. The dorsal setae, though of the same 
form as the ventral, are about three times their length; the shaft of the seta which 
is implanted in the body wall is curved, not straight, as in the other two species. 
The nephridia commence, in the sexually -mature worm, in the sixteenth segment, 
and their external orifice is in front of the ventral pairs of setae. The worm was 
collected by Mr. W. W. Smith chiefly in forest pools, where it lives in association 
with a species of Limnodrilus ; one example was discovered in marshy soil ; so that 
this species is equally at home in water and iif damp soil. 

(4) Phreoryctes emissarius, Foebes. 
p. emissarius, Foebes, Bull. Illinois State Lab. Vol. iii, 1890, p. 108. 
Definition. Long, slender species; prostomium not transversely furrowed. Setae implanted 
singly; dorsal setae disappear before middle of body, previously diminishing in size. 
Habitat — N. America. 

This species can be at once distinguished from the others by the complete absence 
of dorsal setae from all but the anterior seventy segments or so. This species is 
seven or eight inches in length and only -6-7 mm. in thickness, and consists of more 
than ^j^ segments 1. The nephridia, which were found from segment x onwards, 
open in front of the ventral setae ; the first six pairs, however, were more or 

' All the specimens examined by Mr. Fokbes were broken. 


rudimentary, and doubtless disappear when the sexual organs — concerning, which 
there is no information — appear. The cells of the funnels are ciliated, a number of 
cilia to each cell. A peculiar flask-shaped gland opens behind each seta. The 
oesophagus extends from segments ii-iv. It appears to be a subterranean species. 
See also Fokbes (2). 

Genus Pelodrilus, Beddard. 

DEFlBTlTioir. Small, slender worms. Setae paired. Clitellum XI-XIII. Sperm- 
ducts, two pairs opening on segment XII by four separate orifices. Ovaries 
in XIII ; oviducts opening on to border line between XII/XIII. Spermathecae 
one pair in VIII. 

This genus consists, at present, of only a single species recently described by myself. 
Although short and slender in form, and resembling a Lumbriculus, its affinities are 
clearly with Phreoryctes, from which genus, however, it shows certain important 
diflferences. The nephridia of the sexually-mature worm commence as far forward 
as the seventh segment, and are only absent in the eleventh and twelfth segments ; 
they are thus present in several of the genital segments. There are two pairs of 
sperm-ducts, the funnels of which open into the tenth and eleventh segments ; 
the tubes themselves are much coiled, but open on to the twelfth segment ; they 
are distinct from each other throughout their whole course, and there are four 
external pores ; these are situated two on each side of the body, one being in front 
of the other. The clitellum, contrary to the conditions which obtain in most aquatic 
Oligoehaeta, is undeveloped in the ventral region. The male pores, which lie a little 
to the anterior of the line of the ventral setae (these setae are absent from the twelfth 
segment), mark the ventral boundary line of the clitellum. A series of radiating 
muscular fibres are inserted near the male apertures and probably servo to protrude 
them during copulation. There is, contrary to what we find in Phreoryctes, only 
a single pair of ovaries present, which lie in the thirteenth segment. 

The genus Pelodrilus agrees with Phreoryctes in the following more or less 
important characters, those marked with an asterisk being the most important. 

(i) Setae sigmoid and paired (Phreoryctes smithii). 

(2) Testes in x, xi. 
*(3) Sperm-ducts two pairs, all four opening separately. 
*(4) Spermiducal gland absent. 

(5) Spermathecae anterior to testes. 
*(6) Hearts long, thin, and much convoluted. 


These points taken together ^ are such as to render the inclusion of Felodrilus 
in any other family but the Phreoryctidae impossible ; the only other alternative 
would be to create a separate family for its reception, which appears to be 

(i) Pelodrilus violaceus, Beddaed. 
Felodrilus violaceus, Beddard, Trans. Roy. Soc. Ed. Vol. xxxvi. Pt. ii. (1891) 
p. 301. 

Definition. Proglomium short. Nej)hridio-pores in front of ventral setae. Septal glands in 
V- VII, Chloragogen-cells covering alimentary tract commence in V. Sensory papillae two 
— one heJiind t/ie other — on segment X. Habitat — N. Zealand. 

As only one species of the genus Pelodrilus is known, the above generic and 
specific definitions will doubtless ultimately require revision. The species occurs in 
rich, wet soil, at a little distance from a swamp near Ashburton, New Zealand. In 
connexion with its terrestrial habit, the thickening of the integument and of the 
septa in the anterior part of the body (septa v/xi) is of interest. 


Definition. Large or small earthworms, with paired setae, eight per segment; 

clitellum X-XIII ; male pores X/XI or XI and XII ; sperm-ducts open into 

segment in front of that which bears the external pore ; spermiducal gland 

showing the same structure as in the Lumbrieulidae, with sometimes a pro- 

trusible penis, * 

This family of earthworms is in many ways exceedingly remarkable ; the 

genus Moniligaster — for a long time the only genus of the family — was originally 

made by Pekeier (3), its describer, into a separate group called AcliteUians, 

equivalent to the three other groups into which Pberier divided the terrestrial 

Oligochaeta ; this step was taken on account of the apparent absence of that 

characteristic organ the clitellum. A clitellum was, however, recognized by Bourne 

(2) in a large species, M. gravdis, some feet in length, where it would naturally 

be more obvious than in the small M. deshayesi ; the apparent absence of the clitellum 

is to be probably explained by the fact that it is only periodically developed, as in 

the aquatic Oligochaeta ; in addition to this possible explanation of the failure of most 

' The above characters hardly differentiate Pelodrilus from iMmbricus ; as, however, I am here comparing 
Pelodrilus with Phreorydes there is, perhaps, hardly any need to point out the more important facts which 
distinguish both from Lumbriais. 


observers to find the clitellum, may be added the fact that the clitellum, when 
developed, is not always very obvious when the worm is examined by a lens only, 
without subjecting it to microscopical investigation ; as I have pointed out, recently, 
(5) the clitellum in the genus Moniligaster is unique among ' earthworms,' by reason 
of the fact that it is only one cell thick, as in the aquatic Oligochaeta, without exception. 
This structural fact about the clitellum, coupled with its forward position, places this 
family in a very isolated position among the terrestrial Oligochaeta, and undoubtedly 
affines them to many among the purely aquatic forms. A second remarkable feature 
about this family, which now includes, besides the type genus Moniligaster, a second 
genus described by Rosa, Besmogaster, is in the relative position of the internal 
and external openings of the sperm-ducts ; in earthworms the funnels of these ducts 
are always some segments in front of the external pores ; as many as eight or nine 
segments in certain cases ; there is, it is true, a progressive series down to Allurus 
and perhaps Tetragonurus, where the external pores are upon the twelfth segment ; 
the funnels are presumably in the eleventh segment, though we cannot at present be 
sure as to this, for the genus has not been submitted to anatomical study ; in the 
aquatic forms, on the other hand, the funnels are never situated more than a single 
segment in front of that which bears the external pore ; and there are even cases where 
the same segment contains the funnel and bears externally the pore. This is pre- 
cisely what we find in the Moniligastridae ; the funnel is in the tenth segment, 
while the external pore is upon the groove separating segments x/xi. Besmogaster 
difiers from the genus Moniligaster in the doubling of the male organs ; here the two 
pairs of male pores are upon segments xi and xii ; the funnels of the sperm -ducts 
are placed in the segments in front of these. The sperm-ducts communicate with the 
exterior by means of spermiducal glands, four in Besmogaster, two in Moniligaster. 
In Moniligaster, at any rate, the glands resemble very closely those of the 
LumbricuUdae, and certain Tubifieidae such as Branchiura ; each consists of two 
parts — a glandular and a muscular ; the glandular section is an egg-shaped sac, 
which receives its sperm-duct at the extremity farthest away from the external 
pore; it consists of a lining epithelium of low columnar cells, surrounded by a 
layer of circularly-disposed muscle fibres, which are again surrounded by a mass 
of granular, pear-shaped cells, with long prolongations perforating the muscular 
layer and the lining epithelium so as to reach the lumen of the gland; the distal 
part of the organ is chiefly muscular, with a few gland-cells interspersed among 
the muscles, and is reflected to form a protrusible penis, very similar to that 
organ as met with in the family Tubifieidae. It is not quite clear from Rosa's 
account (11) how far the spermiducal gland of Besmogaster differs from that of 

c c 


Moniligastef ; a fuller description of its structure will be found on a preceding 
page (p. 119). The ovaries usually occupy an abnormal position — abnormal, that is 
to say, as compared with 'earthworms,' where they are invariably in the thirteenth 
segment ; this statement, however, only applies to certain species of Moniligaster, where 
the ovaries occupy segment xi ; in Besmogaster and in M. houteni and M. viridis 
their position is different ; they lie in xiii ; there is of course a con-esponding 
difference in the apertures of the oviducts. 

The Moniligastridae have very large egg-sacs — another point of resemblance 
to many ' Limicolae ; ' as a general rule the fact that these structures, which 
extend through several segments, are egg-sacs depends upon pi'obability ; their 
position is such as to lead to the inference that that is their nature ; in M. 
bahamensis, however, the egg-sacs were found by me (57) to contain numerous 
ova ; these ova were not particularly large, a little larger than is common among 
earthworms but nothing like the bulk of the enormous ova of the Tubificidae or 
Lumbriculidae. Nevertheless they were crammed with large yolk particles, so much 
so as to frequently obliterate the nucleus ; this fact may be an important point 
of similarity to aquatic Oligochaeta ; but it must be admitted that it may only 
point to the conclusion that the egg-sacs are not meant to store the ova but are 
merely the receptacles of useless ova, which there undergo degeneration. There are 
other facts which render this latter conclusion possible (see p. 96). In any case the 
large size of the receptacula ovorum is, as I have pointed out, and as EosA (11) 
has admitted, a resemblance to the freshwater Annelids. 

The fantily Moniligastridae is characteristically an old-world group ; it has been 
met with in India, Ceylon, Sumatra, Borneo, Buynah ; quite recently, however, I have 
described a species from the Bahamas. 

The affinities of the family are not so plain ; I have dwelt upon the resemblances 
to the aquatic Oligochaeta — a resemblance which Rosa has sought to minimize. In 
the following points the genus Moniligaster differs from earthworms and agrees with 
the lower Oligochaeta: — 

(i) The sperm-ducts only traverse one septum on their way to the exterior, 
and are much coiled, as in Pachydrilus for example. The doubling 
of the sperm-duets in Besmogaster is not important, for this often 
(2) The spermiducal gland and the penis are constructed upon the plan 
common to many ' Limicolae,' and different from that found in any 
'earthworm.' This does not altogether apply to Besmogaster (but see 
p. 119). 



(3) The forward position of the male-pores. 

(4) The minute structure of the clitellum and its forward position. 

(5) The large size of the egg-sacs, and the great abundance of yolk in 

the ova. 

Even if we follow EosA in placing but little stress upon the shifting forward of 
the clitellum and the male-pores, there remain a number of facts in the anatomy 
of the Moniligastridae which indicate differences from all other earthworms, and 
similarities to the lower Oligochaeta. 

Benham (1) places the family upon the direct line of evolution of the higher 
from the lower Oligochaeta, thus admitting their affinities to the latter. I have 
already discussed the way in which this evolution has most probably taken place, 
viz. in the reverse direction from that believed in by Benham ; but there still 
remains the question to which group of the higher Oligochaeta are the Monili- 
gastridae most nearly relfited. Eos a (20) suggests the Geoscolicidae and Lumbricidae; 
but is not able to bring forward many facts in support of this contention. The 
principal matter upon which he dwells, in comparing the family to the Geoscolicidae, 
is the intersegmental position of the male-pores; the presence also of caeca to the 
nephridia is referred to, though this is also met with among the Acanthodrilidae ; 
but it is thought by Rosa that there are also affinities in this latter direction. 
As regards the Lumbricidae he compares the Moniligastridae chiefly with Allurus and 
Tetragonurus ; this comparison is made on account of the forward position of the 
male-pores ; the posterior gizzards are also mentioned as a point of similarity to the 
Lumbricidae ; but we now know that more than one genus belonging to quite different 
families have the same character, viz. Pleionogaster, and Hyperiodrilus, &c., among 
the Eudrilidae. The forward position of the male-pores may well be, as Eosa thinks, 
a reason for associating the family Moniligastridae with the Lumbricidae ; but it is 
surely not quite fair to deny this point of likeness to the Lumbriculidae, &c., and to 
claim it for the Lumbricidae! On an earlier page of his paper dealing with 
Desmogaster (11), in which these matters are subjected to a renewed discussion, Eosa 
quotes the example of Buchholtzia appevdiculata to discount the strength of the 
arguments . derived from the forward position of the clitellum in Moniligaster. 

Other points of resemblance to the Geoscolicidae hardly exist ^ and it does not 
seem that the mere position of the male-pores is a likeness of first-rate importance ; 
the only conclusion to which we can, in the present state of our knowledge, come 
is that the Moniligastridae form an isolated group, with general affinities to the 

' Ornamented setae have been found in so many groups; but it must be admitted tliat they aie 
characteristic of the Geoscolicidae. In the present family they occur in M. houtmi. 

C C 3 


lower Oligochaeta, but with no particular resemblance to any one family of these 
or of the higher Oligochaeta. The family, as already mentioned, contains two genera, 
Moniligaster and Besmogaster. These two genera are at any rate generally recognized; 
it is, however, a matter for consideration whether Moniligaster should not be again 
subdivided into two genera; two species, viz. M. houteni and M. viridis (if they be 
really distinct) agree in the following points: — 
(i) The spermiducal glands are long. 
(2) The ovaries are in segment xiii. 

Probably M. deshayesi belongs to this group, though the incomplete information 
respecting these points which is given by Pberiee does not allow of any certainty in 
the matter. 

On the other hand, all the remaining species, of whose anatomy we have anything 
like a complete account, show the following assemblage of characters: — 

(1) The spermiducal glands are round or oval. 

(2) The ovaries are in segment xi. 

I am inclined to think that a few more characters could be added to these ; possibly 
the protrusible penis of M. bahamensis and M. indicus are among those characters. 

I do not, however, venture here to divide the genus Moniligaster, as accepted, 
into two genera, since we are as yet in ignorance of the characters of the considerable 
numbers of species, briefly recorded by Bourne (2), from India. 

Genus Moniligaster, Pekrier. 

DBPiWiTiOir. One pair of testes, sperm-saes, and sperm-ducts, the latter opening 
between X/XI ; ovaries in XI or XIII ; spermatheca usually in VIII, with long 
duct ; three to five gizzards ; a protrusible penis sometimes present. 

There has been a great deal of discussion as to the actual facts in the 
anatomy of this genus, let alone the conclusions to be drawn from the facts. 
The original description given by Perkier (3) was in some respects inaccurate, as 
has been subsequently shown by myself (19), Horst (1), and Eosa (11). Perkier 
thought that the worm possessed two pairs of male-ducts and pores; the mistake 
arose through his having confounded the spermathecae with an anterior pair of 
spermiducal glands. 

There has been also some confusion as regards the position of the vaiious organs 
— a confusion which is still, perhaps, hardly dissipated. For the observations of 
Horst (1), upon a large species of the genus from Sumatra, seem to show that the 


male-pores are a segment behind those of the other species; it is, however, not 
to my mind certain that there is not an error of one segment, and that M. 
houteni agrees with the remaining species of the genus in the position of the 
male-pores between segments x/xi. Rosa. (20), arguing from the preconceived 
idea that Moniligaster would be found to present the typical characters of earth- 
worms, suggested that Hobst's determination was probably right; but at that time 
I had fixed the position of the apertures in question as lying between segments ix/x, 
a position which I subsequently stated to be an error ; as a matter of fact, the 
setae of the first setigerous segment (the second) are so very small in M. harwdli 
that I at first overlooked them ; in M. hakamensis they seem to have vanished 
altogether. Rosa's own investigations upon M. beddardii (11), and those of Benham 
upon M. indicus, confirmed this location of the male-pores. 

The alimentary system of Moniligaster is remarkable for the fact that there are 
at least three gizzards, which lie in as many consecutive segments at the end of 
the oesophagus ; the oesophagus is rather wide in calibre, and has no trace of any 
calciferous pouches ; there does not even appear to be any vascular part of it, such 
as apparently does duty in so many earth-worms for the otherwise missing glands, 
except apparently in M. japonicus. 

The reprodvictive organs consist of (i) a single pair of testes in segment x, 
depending from the anterior wall of this segment ; they are enclosed within the 
(2) sperm-sacs ; the sperm-sacs in M. barwelli, lie partly in the ninth and partly 
in the tenth segment; in M. bahaviensis and, apparently, in M. houteni they 
are restricted to the tenth segment; in If. viridis they lie in the eleventh; 
their cavity is broken up by trabeculae in M. indicus, but not in the species 
investigated by myself, and they completely envelop the testes, as already men- 
tioned, and (3) the funnels of the sperm-ducts ; the latter belong, therefore, to 
the tenth segment, and lie in the same segment as that which bears the external 
pore of the sperm-ducts — a very unusual state of affairs ; the sperm-ducts are, when 
fully mature, long and much coiled ; their appearance, indeed, as I have already 
pointed out, suggests the sperm-ducts of the Enchytraeidae ; these coiled ducts lie 
partly in the tenth and partly in the ninth segments ; it may be that this peculiarity 
is confined to the species, M. bahrnnensis, in which I have described it; but it 
seems more likely that is not the case, but that the sperm-ducts, when the animal 
is fully mature, have the same great length. (4) The sperm-ducts open into 
a spermiducal gland, which consists of two parts ; the distal part is a sac which 
has very muscular walls, and of which the epithelium is reflected over a protrusible 
penis— precisely like that of the Tubificidae ; into this opens a sac with a lining of 


epithelium, surrounded by a layer of circular muscles, which is itself surrounded 
by a layer of large pyriform cells, whose prolongations penetrate the muscular and 
epithelial layers to communicate with the lumen of the organ; near to the summit 
of this glandular part of the organ opens the sperm-duct. (5) The ovaries are 
generally in the segment which follows that containing the testes; the oviducts 
lie opposite to them, and, therefore, open on the exterior between segments xi/xii ; 
a most characteristic feature of the genus (and of the closely allied Desmogaster), viz. 
the large egg-sacs has been already spoken of. (6) The spermathecae are a single 
pair -only ; they usually lie in the eighth segment ; each consists (except in 
M. Japonicus) of a small globular pouch communicating with the exterior by 
an extraordinarily long duct, which has well-marked muscular walls. It is not 
an easy task to discriminate the various species of Moniligaster which have been 
described ; altogether the following thirteen species have received names : — 

(i) Moniligaster deshayesi, E. P., Ceylon. 

(2) „ barwelli, F. E. B., Manila. 

(3) „ liouteni, HoRST, Sumatra. 

(4) „ japonicus, MiCH., Japan. 

(5) „ bahamensis, F. E. B., Bahamas. 

(6) „ indicus, Benham, India. 

(7) „ beddardii, EoSA, Burmah. 

(8) „ grandis. Bourne, India. 

(9) „ uniquus. Bourne, India. 

(10) „ sapphirinaoides. Bourne, India. 

(11) „ robustus, BOuRNE, India. 

(12) „ papillatus, Bouene, India. 

(13) „ ruber. Bourne, India. 

(14) „ minutus, Bourne, India. 

Seven of the species in the above list are too imperfectly known to admit of 
their being defined in a way sufficient to discriminate them from those that are known. 

Moniligaatef grandis appears to possess five gizzards (in segments xvii-xxi) ; the septum between 
segments ix/x is missing ; but Benham throws some doubt upon this fact, and explains it by the 
shifting of the septa, which does take place elsewhere. 

Moniligaster uniquus has gizzards in xv-xix and is a small species. 

Moniligaster sapphirinaoides has gizzards in the same segments as M. grandis, but appears to be 
a smaller form. 

Moniligaster robustus has the gizzards in an unusually anterior position, viz. in segments xi-xv. 

Moniligaster papillatus is chiefly defined by ' long tubular papillae in connection with the pores 



between somites x and xi'; I imagine that these papillae are in all probability the extruded 
muscular penis, in which the spermiducal gland of several species terminates. The gizzards are 
in xvi-xx. 

Moniligaster ruber has a more restricted gizzard than in any other species— it is only found in 
segments xiii, xiv ; but in segments x-xii there are ' soft-walled swellings of the intestine, looking 
like gizzard only not muscular.' 

Finally, Moniligaster minutus has a gizzard extending through segments xii-xiv only ; it has large 
egg-sacs reaching from the twelfth to the fifteenth segment. 

The species that are sufficiently known, for systematic purposes (including a form 
described for the first time in the present work) may be thus distinguished : — 








M. deshayesi . . 

150 mm. 

6& 13-22(5) 

in ix 

long and bent 


M. iarweUi . . 

28 mm. 

13-15 (3) 

in xiv 

in X (?) 



in xi 

M. indicus 

137.5 mm. 

in 13-16(4) 

in ix 

in ix 



in xi 

M. hahamensis . 

25 mm. 

13-15 (3) 

in ix 

in X 



in xi 

M. houteni . . 

T050 mm. 


in xi 

in xi 



in xiii 

M. japoniam . . 

28 mm. 

12, 13 (2) 


in X (?) 




M. riridis . . . 

350 mm. 

15-18 (4) 

in xi 

in xi 

long and bent 


in xiii 

(i) Moniligaster deshayesi, Peekieb. 
M. deshayesi, Peeriee, Nouv. Arch. Mus. 187a, p. 130. 

Definition. Length, 150 mm. ; breadth, 6 mm. Alimentary canal with a gizzard in VI and 

four gizzards in XIII-XXII ; hearts in fl-IX. Hah. — Ceylon. 

This species is the type of the genus ; but unfortunately it is not very clear if 

it be really diflferent from some of those described by Bouene. The existence of 

a gizzard in the sixth segment in addition to the gizzards which lie posteriorly 

is, so far as can be said at present, the distinguishing character of the species; but 

the absence of an anterior gizzard in all the remaining species of the genus renders 

the supposed presence of this gizzard in Moniligaster deshayesi perhaps just a little 

doubtful ; the specimen examined by Peeeiee was in a poor state of preservation ; 

hence it is possible that an error has been made. 

The description of this species is illustrated by a number of figures ; these are not all quite 
intelligible in the light of our present knowledge ; it is above all not clear as to the ' anterior 
pair ' of male efferent organs (Fig. 79) : the structures termed by Peebieb ' the glands of the seventh 
segment;' their minute structure is figured as well as described, and I am still disposed to think 
that my suggestion that these are the dilated extremities of the spermathecae, while the duct 


' entortille comme serait un Gordius ' is the long spermathecal duct, is correct ; the bodies termed 
by Pebeier ovaries are evidently the egg-sacs ; they are figured as containing ripe ova. 

Peeeier states that the spermathecal orifice is on a line with the ventral pairs of setae ; 
this is possibly an oversight. The spermatheca itself is furnished with a pair of large stalked 
glands. I have suggested that these are comparable to the diverticula of the spermatheca so 
common among earthworms. With this suggestion Benham agrees, but makes the additional 
remark that they correspond to the bilobed sac which he discovered at the end of the sperma- 
theca in Moniligaster indiaisj this appears to me to be a reasonable view to take. 

The sperm-duct is figured and described by Pereiee as possessing a quantity of little leaf-like 
bodies attached to it; these can be nothing else than the folds of the sperm-duct, which in this, 
as in other species of the genus, is extremely convolute. The spermiducal gland of this species is 
rather unlike that of others ; it is long and is bent upon itself near to the summit, the two parts 
running parallel ; but they are not, as in Moniligaster inridis, of equal length ; one part is not so 
much as one-half the length of the other ; the sperm-duct is represented as joining the gland at the 
point where it bends upon itself. 

(2) Moniligaster barwelli, Beddaed. 

M. barwelli, Beddaed, Ann. and Mag. Nat. Hist. 1886, p. 94. 
M. beddardii, Rosa, Ann. Mus. civ. Geneva (aa), Vol. ix, p. 379. 

Definition. Length, about ^omm.; alimentary canal with three gizzards in XIII-XV ; septa 
V/IX thickened ; sperm-sac lying half in segment IX and half in segment X. Hab. — 
Luzon, Burmah, Flores. 

This species is one of the few which are adequately characterised ; it was 
originally described by myself, and was at that time the second species of the genus, 
the only other being M. deshayesi. Hoest has pointed out that my M. harwdli 
is probably identical with Rosa's (11) M. beddardii. With this identification I fully 
agree ; there are no difierences that can be found in the descriptions given by Rosa 
and by myself of our two species. Rosa distinguished his M. beddardii from my 
M. barwelli in the following details : — In M. beddardii there are only three gizzards ; 
in my earlier description of M. barwelli I stated that there were four of these, but 
I subsequently corrected that statement. My figure of the sperm-funnel misled 
Rosa into thinking that these two organs differed in the two species. As to the 
colour which Rosa had used to distinguish the two, I am not of opinion that much 
weight can be attached to that diflference. 

I am doubtful whether to include in this species a small Moniligaster described 
in some detail by Hoest from the island of Flores. The specimen was only 18 mm. 
in length, and it appears to possess a protrusible penis; I should not, however, like 
to assert positively that M. barwelli does not possess this organ ; I have indeed 



suggested that the presence of a penis may prove characteristic of one group of this 
genus. As to the large size of the egg-sacs in Horst's species, that may be merely 
a question of more complete maturity. 

(3) Moniligaster houteni, Hokst. 
M. houtenii, Hoest, Notes from Leyd. Mus., IX, p. 97. 
Definition. Length, I. 50 M.; breadth, 18 mm. ; number of segments, 443: gizzards (4) in 

XV-XXII; septa, VI /X, XXIIT/? thickened; sfermidiwal glands open, XI/XII ; 

spermathecae in IX; ovaries in XIII. Ilab. — Sumatra. 
This species comes nearest to M. viridis; under the description of the latter 
most qi the points of resemblance as well as of difference are noted. M. houteni 
is quite the largest Moniligaster, and is indeed one of the largest of earthworms. 
Hoest has described and figured the setae as being ornamented with ridges after the 
fashion so very prevalent in the Geoscolicidae. It will be noticed from the definition 
that the aperture of the spermathecae has moved a segment further back in corre- 
spondence with that of the spermiducal glands ; but while the latter glands lie 
in the same segment as they do in M. viridis, the spermathecae are a segment 
further back; the sperm-sacs are in the eleventh segment; they are described as 
being large flat pale-brown organs, in which however only Gregarines were recognized ; 
thex-e is no doubt however that they are the sperm-sacs. It seems clear from Hokst's 
description that the two septa enclosing the ovaries join together peripherally as they 
do in M. viridis; this arrangement, which must facilitate the passing of the ova 
into either the oviduct or into the egg- sacs, is not commonly to be found among 

In a subsequent note upon this worm (17), Hoest reaffirms the position of the 
male pores, and states that they are in line with the dorsal setae, and not between 
the two pairs. 

(4) Moniligaster japonicus, Michaelsen. 
M. japonicus, Michaelsen, Arch. f. Nat. 189a, p. 233. 
Definition. Length, 28 mm.; breadth, 3 mm.; number of segments, 95; alimentary canal, with 
two gizzards in XII, XIII; spermathecae without a long duct. Hal. — 3apan. 
This species is clearly distinct from any of those that are well characterized ; it 
has been somewhat briefly described by Michaelsen, his description not being illustrated 
by any figures. The above definition contains, I think, all the points in which this 



species differs from others. The length of the egg-sacs may, however, be a further 
difference ; Michaelsen states that they occupy segments xii-xvi, with a little doubt. 
The spermathecae are a little further back than is usual in the genus ; they open 
between segments ix/x ; in having no stalk they differ from the spermatheca of 
other species. The ovaries also appear to be somewhat unusual in their size ; they 
reach from septum x/xi to xi/xii, and are only deficient above the oesophagus. In 
the large size of the ovaries the species resembles M. viridis. 

(5) Moniligaster bahamensis, Beddakd. 

M. bahamensis, Beddard, Proc. Zool. Soc. 189a, p. 690. 

Deflnition. Length, ^^ mm., three gizzards in XIII-XF ; septa V/IX thickened; sperm-sac 

lying entirely in X. Hal. — Bahamas. 

This species is mainly interesting on account of the fact that it comes from the 
New World, and is at present the only member of the family with that origin. It 
is a rather more slender species than M. harwdli, but otherwise seems to agree fairly 
closely with it. The apparent differences concern the spermatheca and the penis. 
In M. harwelli I did not find a protrusible sac connected with the spermathecae or 
with the spermiducal gland, such as exist in the present species and in M. indicus; 
but their apparent absence may, after all, be only a question of a less degree of 
maturity. It is mainly the habitat which leads me to distinguish this species from 
the other one mentioned. The same remark applies also to M. indicus, though there 
is here the further difference that there are four gizzards. I imagine that I have 
made an error in placing the spermathecal porgs in a position corresponding to the 
interval between the dorsal and ventral setae. 

This is the only species in which the structure of the clitellum and of the ripe 
ova has been seen. The ova are remarkable for the fact that they are filled with 
yolk spherules, though they are not markedly larger than those of other earthworms ; 
the structure of the clitellum agrees with that of the aquatic Oligochaeta in consisting 
of a single row of cells only. These points, however, are in all probability of generic 
or family importance, and have accordingly been treated of as such. I mention 
them here as a matter of history. 

(6) Moniligaster indicus, Benham. 
M. indicus, Benham, Q. J. M. S., Vol. xxxiv, p. 361. 
Deflnition. Length, i^y-^mm.j number of segments, 150 (about); gizzards in XIII-XVT; 


sejpta Vl/IX thickened slightly ; sperm-duct perforates body-wall on its way to spermiducal 
gland. Hab. — Nilgiris, India. 

This species is, thanks to Benham's researches, one of the best known. It comes 
near to M. harwelli and M. haJiamensis. Like that latter species, it has a protru- 
sible muscular sac, connected both with the spermatheca and with the spermiducal 
glands. The latter are bound down to the parietes by a few muscular strands — 
a state of affairs which recalls the similar glands in the Geoscolicidae and certain 

The sperm-duct runs for a part of its course in the thickness of the body-wall ; 
this occurs also in the species M. viridis, but has not been noticed elsewhere ; there 
are other worms in which the sperm-duct has this rather unusual course, which 
belong to so many different groups that it" cannot be of importance as indicative 
of any affinity (see p. loo). The interior of the sperm-sacs are divided up by trabe- 
culae ; though they appear to be suspended in the septum which divides segments x/xi, 
they are seen in sections to lie really in segment x, the wall of which bulges back 
at the place where the sperm-sac is attached to it. The muscular sac in which the 
spermathecal duct ends is remarkable for lying in two segments ; it is in conse- 
quence bilobed, but the spermathecal duct enters it exactly in the middle between 
the two lobes. 

(7) Moniligaster viridis (new species). 

Definition. Length, 350 wm.; breadth, 18 mm.; number of segments, 186; gizzards in 
XV~XVIII ; spermiducal glands long ; egg-sacs in XIV, not extending beyond this segment. 
Ilab. — Borneo, Fenrisen Hills, Sarawak. 
Of this new species I have been able through the kindness of Mr. Everett to 
examine a couple of specimens. It appears to be a perfectly distinct new species, 
coming nearest to M. houteni; it resembles that species in the fact that the ovaries 
are two segments further back than they are in all others. There are, however, 
several points in which M. viridis differs from M. hovieni ; the most obvious 
difference is perhaps in the fact that the spermiducal glands are bent upon them- 
selves so that the two ends are in actual contact ; the reason for this is that the 
sperm-duct after leaving the sperm-sac passes straight down to the ventral body- 
wall, and there perforates the muscular layers of the body-wall, only emerging to 
enter (after a short course) the summit of the speiTaiducal gland ; in M. houteni 
HOKST has figured (1, fig. i) the glands as lying straight, and he expressly compai-es 
them to those of Acanthodrilus, which are not bent upon themselves in the way 

D d 2 


that those of M. viridis are. In other respects it must be admitted that the present 
species comes very near to M. houteni. 

A remarkable point about it is its colour. What this may have been during life 
I do net know ; but after preservation in alcohol it is dark green above, with a very 
abrupt demarcation between this and the yellowish of the ventral surface. Benham 
has commented upon a similar plan of coloration in M. indicus. 

The dorsal vessel showed indications of being double in segment xi, a fact which 
is also noted by Hoest for M. houteni; the last pair of hearts were in the same 

The thickened septa show the same very peculiar arrangement that appears to 
characteiize M. houteni; there are a set of thickened septa in the anterior 
segments, and another set. behind the "gizzards. The septa in question are vii/xi 
and xix/xxv. The two septa enclosing the thirteenth segment meet laterally, so as 
to form a closed compartment. 

The sperm-sacs are large and rounded ; they lie in the same segment as that in 
which the spermiducal glands lie ; they are almost spherical in shape, with a long 
and narrow neck connecting them with the septum x/xi. The sperm-ducts are not 
much coiled ; they pass, as has already been described, from the interior of the 
sperm -sacs to the ventral body- wall, which they perforate, as in M. indicus; 
but emerge sooner to enter the spermiducal gland. The latter have an appearance 
like that of the corresponding glands of Diplocardia communis (see below) ; they 
are marked by rounded spots, which are in reality the masses of glandular cells 
which form the outer layer of the organ; the gland ends in the body- wall, without 
undergoing any apparent change in its structure. 

The ovaries are large, and lie in the thirteenth segment; they extend on to the 
dorsal surface of the gut, and are suggestive rather of a gland connected with the gut 
than of the ovaries, which a microscopical examination of their contents showed them 
to be. The egg-sacs are also large, but they do not extend even as far as the 
posterior wall of the fourteenth segment, to the anterior septum of which they 
are, of course, attached ; they open by a wide mouth into the cavity of the 
thirteenth segment, opposite to the ovaries ; projecting from this mouth was 
a curious structure, having the appearance of a mass of fine tubes, closely branching 
and anastomosing ; a good deal of the egg-sac itself was also lined with a similar 
mass of tubes, which are clearly 'blood-vessels ; the closeness and complexity of 
this tuft of vessels is, possibly, an indication that the egg-sacs have the capacity 
for growth, and that they may, when the worm is perfectly adult, have the same 
extent as in M. houteni. 


Genus Desmooaster, Rosa. 

Depiktitiobt. Two pairs of testes and efiFerent ducts opening XI/XII, XII/XIII; 
ovaries in XIII. 

This genus is at present only certainly known by one species, D. doriae. Its 
anatomy has been carefully worked out by Eosa (H). The principal differences 
from Moniligaster are indicated in the above definition of the genus; they mainly 
concern the doubling of the male apparatus. The sperm-ducts lie in two segments, 
the funnel lying in the segment in front of that which bears the pore belonging to it. 
The single spermatheca is, as in Moniligaster, to be found in segment viii. 

There is a peculiarity in the vascular system, which may, possibly, be of generic 
value; the two last pairs of hearts — those lying in segments x, xi — appear to be 
double. The outer vessel is the larger, and is sinuous in its course ; the inner 
vessel closely embraces the oesophagus. At its origin the outer vessel, which is the 
true heart, communicates with the inner by a short branch ; the inner vessel 
communicates, above the oesophagus, with its fellow of the opposite side, but has 
no direct communication with the dorsal vessel ; they end in two lateral vessels, which 
run along the body-wall. 

(i) Desmogaster doriae, Rosa. 

D. doriae, RosA, Ann. Mus. civ. Geneva (a a), IX, p. 369. 
Definition. Length 500 mm.; breadth 12 mm.; number of segments, 33O. Septae FIT/XI, 

greatly thickened ; last pair of hearts in XI ; ten, gizzards, com/mencing in XX. Hah. — 
Meteleo, Burmah (xooo-1400 M. in height). 

(3) Desmogaster horsti, Beddaed. 

D. sp., HoKST, Zool. Ergebn. Max Weber, III. p. 49. 
Definition. Length 140 mm.; number of segments nearly 3°° ! septa VII/X thickened; 
eight gizzards, commencing in XVII. Hab.— Sumatra, Mt. Sitigalay. 
This appears to me to be a distinct species, by reason of the fact that it has 
only eight, instead of ten, gizzards. Hoest also mentions that the sperm-ducts 
do not open into the spermiducal glands at their summit, as is the case with 
D. doriae. Hoest is inclined to think that the tenth septum is absent. 



This group includes the families Naididae, Tubificidae, and Lumbriculidae. 

Vaillant (6) united the two first of these families into one ; the chief distinction 
drawn between them by d'UdeKEM was the fact, that in the first asexual propagation 
by budding takes place — a phenomenon quite unknown in the Tubificidae. Vaillant 
is disinclined to allow that this is a distinction of first-rate importance, and I agree 
with him. In the first place, among the parallel group of the Polychaeta, the forms 
which multiply by budding are not in every case very widely separated from those 
which do not exhibit this mode of development ; in the second place, it is not 
easy to draw a hard and fast line between this budding and the breaking up of 
a Lumbriculus into pieces which are severally capable of independent growth. 

As to structure, it is difficult to fix upon characters which absolutely divide the 
two families. Ilyodrilus is a Tubificid in, perhaps, most of its characters ; the 
position, for instance, of the generative apertures is as in other Tubificidae ; the 
vascular system has the complications that are found in other members of that 
family, but are not found in the Naids. The spermiducal glands, however, are 
Naidiform. JJncinais is another annectent form ; it certainly shows the cephaiization 
so characteristic of, but not universal in, the Naids ; on the other hand, the sexual 
organs ^ are situated further back than in any other Naid in which their position 
is known; finally, the setae are entirely uncinate — a state of affairs not found in 
any other Naid, but often met with among the Tubificidae, e.g. in Limnodrilus. 

The Lumbriculidae are, at first sight, rather far removed from the Naids and the 
Tubificids ; but, at best, the separation cannot be a wide one ; compared, for 
example, to the characters which separate two such well-marked families as the 
Geoscolicidae and the Eudrilidae, the differences between the Lumbriculidae and 
the Tubificidae are by no means great. The main difference is the fact that in the 
Lumbriculidae there are two pairs of sperm-ducts ; this character is quite universal 
if we except certain dubious forms, to be dealt with immediately. Li Sutroa, 
however, which Eisen (2) has justly referred to the Lumbriculidae, one of the 
two pairs of sperm- ducts is decidedly smaller than the other, and seems to show 

' It is possible, however, that, as Eenham has pointed out, the 'testes and ovaries' of Uncinais are 
in reality sperm-sacs and egg-sacs ; in this ease the position of the generative organs may not be so nnlike 
that of other Kaids. 


some evidence of commencing disappearance ; a second stage in this disappearance 
is, possibly, exhibited by the genus Phreodrilus, a genus of rather doubtful affinities ; 
in this Annelid the single sperm-duct is furnished with a diverticulum which joins 
it some way before it opens into the terminal gland. This diverticulum is, indeed, not 
ciliated, and its epithelium is in some other particulars unlike that which line the 
sperm-ducts ; it is more like the epithelium which lines the spermatheca of the 
same worm, and I have suggested that we may possibly have here an indication 
of the formation of a spermatheca out of the second sperm-duct, which would be 
a remarkable and interesting example of a change of function. A third step is 
seen in Alluroides, a genus which, I think, is necessarily to be referred to the 
Lumbriculidae ; in this Annelid the sperm-ducts are a single pair only. Apart from 
the forms that have been considered, there are no others that appear to show marked 
intermediate characters between the Tubificidae and the Lumbriculidae ; but a con- 
sideration of these genera leaves very few points in which the Lumbriculidae are 
peculiar ; the only point is really the presence of a gland near to the spermathecae, 
which has been called the Albumen gland ; this gland has not hitherto had its 
counterpart in the family of the Tubificidae ; but quite recently a gland, which 
appears" to me to be strictly comparable, has been described in Emholocephcdus 
velutinvbs, a worm which, in all its other characters, must be undoubtedly referred to 
the Tubificidae, though presenting also certain points of affinity to the Naids. 

The only other point which, in the present state of our knowledge, seems to 
distinguish the Lumbriculidae from other worms (including the Tubificidae), is in 
the vascular caeca. This structural peculiarity is not quite universal ; for Stylodrilus 
has not these caecal appendages, and Vejdovsky (24) has spoken of it as a genus 
which, in many respects, recalls the Tubificidae. There are no other points in which 
the Lumbriculidae as a whole are to be distinguished from the Tubificidae ; no 
LumbricuKd has capillary setae, but then these setae are occasionally wanting in the 
Tubificidae — for instance, in the genera Clitellio and Limnodrilus ; moreover, the 
dorsal setae of Phreodrilus seem to be intermediate between ordinary capillary setae 
and the sigmoid setae of the Lumbriculidae. 


Depihition. Aquatic Oligochaeta of moderate size. Setae paired and /-shaped, 
sometimes with the free extremity bifid. The dorsal blood-vessel or the 


transverse vessels witli blind contractile appendages (exc. Stylodrilus ^). Two 
pairs of sperm-duets (exc. Alluroides) uniting to open by a single spermiducal 
gland on each side, which lies in front of oviducal pores. "So penial setae. 

It contains only eight well-characterized genera, viz. Lumbriculus, Rhynchel'niis, 
Trlchodrilus, Stylodrilus, Claparedilla, Phreatothrix, Edipidrilus, Alluroides, and 
Sutroa ; to these should, perhaps, be added Pseudolumbriculus. There are, however, 
several imperfectly described genera which may very possibly belong to the same family. 

Vejdovsky (24) places here Gkube's genera Bythonomus and Lycodrilus, more doubtfully 
CzEBNiAVSKY's genus ArcMeodrilus. They are, however, all ranged by Vejdovsky, as well as by 
Vaillant (6), as ' incertae sedis.' Bythonomus possesses blind appendages to the dorsal vessel, which 
character led Vejdovsky to associate it with the Lumbriculidae. The position of Czerniavsky's 
genus Archaeodrilus is more doubtful. There is nothing in the diagnosis of the genus or in the 
description of the two species to prove that the genus is a Lumbriculid. 

The two distinctive characters of the family, which are not found in any other 
family or genus, are the contractile appendages of the blood-vessels, and the 
arrangement of the vasa deferentia. 

As to the contractile blood-sacs, they have been described by most of those 
authors who have treated of the anatomy of the Lumbriculidae ; by Grube (9), (who 
mistook them for diverticula of the intestine), by Clapakede (2), Ratzel (3), 
Vejdovsky (24), Vaillant (6), Dieffenbach and Eisen (23, 5) ; they are figured 
by CLAPARi:DE (for Lumbriculus), Vejdovsky {Claparedilla, Phreatothrix, Rhynchelmis, 
and Lumbriculus), Eisen (Sutroa, Eclipidnlus), and Vaillant (Lumbriculus). Their 
, arrangement differs in various genera ; in all they consist of either simple, or branched, 
caecal diverticula of the dorsal vessel or of the perivisceral vessels ; they are covered, 
like the dorsal vessel, with chloragogen cells, loaded with dark granules — a fact which, 
as Dieffenbach has pointed out, probably led Grube to regard them as diverticula 
of the intestine. They are contractile, and, as they lie freely in the coelom, their 
position alters with each movement. I have already treated of the probable 
homologies of these organs. 

The sperm-ducts and the spermiducal glands of the Lumbriculidae differ from 
those of most other Oligochaeta. There are nearly always (not in Alluroides) two pairs 
of sperm-ducts which open into two consecutive segments, generally the ninth and 
tenth. The sperm-duct, which communicates with the anterior funnel, opens directly 
into the ciliated spermiducal gland, the orifice of which is placed upon the tenth 
segment; the second pair of funnels open into the tenth segment, and are closely 
attached to its posterior septum ; the sperm -duct traverses this septum directly it 

' And probably Alluroides. 





(after Claparfede.) 

I. Ventral blood-vessel. 2. Dorsal 
blood-vessel. 3-5. Branches of last 
with caecal twigs. 

leaves the funnel, and then, bending back again, passes through the septum, to open 
independently of the anterior vas deferens into the terminal gland. 

In Sutroa (see Eisen 2, 5, Beddaed 81) the plan 
upon which the male efferent apparatus is constructed is 
the same, but the spermiducal gland is enormously long. 
Eclipidrilus is altogether anomalous. 
Vejdovsky divides the family into two sub-families, 
(i) Trichodrihna and (a) Lumbriculina, characterized by 
the absence (in the former) or the presence (in the 
latter) of a paired or unpaired albumen-gland. Sutroa 
would, perhaps, be included in the second sub-family ; 
but it differs quite as much from either sub-family as 
the two sub-families do from each other; Eclipidrilus 
and Alluroides would, undoubtedly, need a fourth and 
fifth sub-family for their reception, if, indeed, it is 
justifiable to include them in the family Lumbriculidae at 
all. As it is, in my opinion, undesirable to break up so small a family (containing only 
twenty species) into four sub-families, I would prefer to separate it only into genera. 
The following table (p. aio) shows the resemblances and differences between those 
genera, excepting Eclipidrilus, whose structure is known. 

Another aberrant Lumbriculid is Alluroides; this genus, which is a native of 
tropical East Africa, was described by myself quite recently (84) ; it has the outward 
form of a Tubificid, but agrees with the Lumbriculidae in the following points: — 
(i) The setae are all S- shaped ; they are paired, 
(a) The spermiducal gland is clothed by a thick glandular layer. 
(3) The longitudinal muscle-layer of the body-wall is formed by a single row 
of deep plates. 
These three characters do not co-exist in any worm that does not belong to 
this family. In other points that are not distinctive of the Lumbriculidae, but 
which are characteristic of all the genera of that family, Alluroides agrees with 
other Lumbriculidae. These are: — 

(i) Ova of large size, and fiUed with yolk. 

(a) No glands appended to the alimentary tract except septal glands. 

(3) No gizzard. 

(4) Funnels of the sperm-duct a segment in advance of the spermiducal gland. 

(5) Ovaries in segment containing spermiducal gland. 

And in a number of other small points, without which the genus could not be 

E e 



assigned to the neighbourhood of the Lumbriculidae, but which are not of themselves 
sufficient to show its affinities to that family more than to other families of the 
aquatic Oligochaeta. 

The principal differences from the Lumbriculidae are the following: — 

(i) No vascular contractile caeca (these, however, are absent in Stylodrilus). 

(2) A single pair only of testes and sperm-ducts. 

(3) The position of the male-pores upon the thirteenth segment, and the 

related fact that the sperm-ducts pass through several segments on their 
way to these pores. 

(4) The position of the ovaries in the thirteenth segment, and of the oviducal 

pores upon the following segment. 
These characters do not unite this genus with the Tubificidae or with any other 
aquatic Oligochaeta ; they point, on the whole, to the terrestrial genera, and 
particularly, perhaps, to Allurus; hence the name Alluroides, which I have given 
to the genus. I have already discussed these characters in the section dealing 
with the general classification of the Oligochaeta. 







Lumbriculus . 


In X, xi, xii 

in viii 


in ix 

Claparedilla . 


in ix 

pear-shaped, opening on x 

two pairs 

TrichodrUus . 


in xi, xii 

pear-shaped, opening on x 

two pairs 

Stylodrilus . 


in ix 

pear-shaped, opening on x 

two pairs 

clitellum x-xii ; 
a pair of rigid penes 

Bhynchdmis . 


in Tiii, with 


long, opening on x 

two pairs 

in ix 

clitellum viii-xvi 

Phreatolhrix . 


in xi (and xii) 

opening on x 

two pairs 

Sutroa . . 


single in viii, 
with many 

long, opening on x 

two pairs, 

coelomic sacs on walls 

of X, xi ; clitellum 


Alluroides . 


in viii 

long, opening on xiii 

one pair 

clitellum xiii-xvi ; 
a pair of rigid penes 

The affinities of this family in one direction are not plain ; the two pairs of sperm- 
ducts, with the corresponding two pairs of testes, do not occur in any other group 
among the aquatic Oligochaeta ; there is, so far, a resemblance to the terrestrial 


genera, which is, however, not borne out in other particulars. It is true that 
among earth-worms it does not seem possible to lay much stress upon the existence 
of one or of two pairs of sperm-ducts ; but the existence of two pairs is so rare an 
occurrence among these aquatic forms that it seems to have more importance; hence 
it appears to me to be quite justifiable to found a family character upon it. 

Although it is difficult to institute any affinities between the Lumbriculidae and 
the higher Oligochaeta, their relationships to the lower forms are, perhaps, rather 
more obvious; as Vejdovskt (24) has said^ the genus Stylodrilus is in certain 
particulars not far removed from the Tubificidae; this genus has not the branched 
caecal appendages of the dorsal blood-vessel. A better intermediate genus or, rather, 
family, for I have placed it in a distinct family here, is Eclipidrilus ; that genus 
has the caecal appendages of the Lumbriculidae, but it has also spermatophores 
which are quite unknown in the Lumbriculidae ; I have also called attention to 
the structure of the spermiducal gland, so far as we know it, as affording evidence 
of a resemblance to the Tubificidae ; the position, also, of the oviducts in the genus 
Eclipidrilus is more like what is found in the Tubificidae than the position which 
characterizes the Lumbriculidae. 

The Lumbriculidae, however, also show affinities to the Naidomorpha ; this is 
chiefly seen in the nephridia; in both these families (and also in Ilyodrilus — a low 
form of the Tubificidae) there is a conspicuous brown mass immediately following 
the funnel of the nephridium; nothing of the same kind occurs in any other group 
of Oligochaeta. 

Ltjmbricultjs, Gkube. 

Lumbrioulus, Geube. non Clapaeedb. 
Lumbrieus, 0. F. Mullek (in part.). 
Saenuris, Johnston. 
? Acestus, Leidy. 
DEPiiTiTioii. Greatly elongated worms, aquatic in habit. Setae bifid at extremity. 
In each segment, after eighth, a pair of blind branched diverticula of dorsal 
vessel, and a pair of non-contractile perivisceral trunks uniting dorsal and 
ventral vessels. Spermiducal glands in VIII. Spermathecae in X, XI, XII. 
Oviducts in X, XI. Albumen-glands in IX. 
DiEFFENBACH has lately given strong reasons for believing that Gkube and CLAPARiiKE did 
not base their descriptions of Lumbriculus variegatus upon the same species, or even upon the same 
genus. Clapaebde himself remarked (2) upon the differences between the facts observed by 
himself, with reference to the structure of this worm, and the description given by Grube, but, 

E e 3 


considered that these discrepancies were due to errors on the part of Grube. Diepfenbach's 
paper was published about the same time as Vejdovsky's great work upon the Oligochaeta (24), 
and is, accordingly, not referred to there. Vejdovsky, however, has not passed over the divergent 
accounts given by Geubb and CLAPAEijDE, and has pointed out that CLAPAEiiDE's 'Lumhriculus' 
is really identical with Claparedilla; as this name, instituted by Vejdovsky in 1883, has the 
priority over Diepfenbach's name Pseudolunibriculus, it must clearly be retained. 

Our knowledge of the structure of this genus is principally due to Vejdovsky (24), Eatzel (3), 
and to DiEPFENBACH, and is restricted to that of the species Ltimhriculus variegatus. This knowledge 
is, unfortunately, far from being complete, as the sexual organs have been, as yet, only imperfectly 

The setae are disposed in pairs, and are bifid at their free extremity. 

There is, as Leydig first discovered, a pore upon the dorsal side of the prostomium, 
which is not always evident ; it puts the cavity of the prostomium into communication 
with the exterior. Dieffenbach states that in the neighbourhood of the pore there 
are cilia present ; and BuLOW (1) describes nerves which ai-ise from the cerebral 
ganglia and supply this region, which he regards as an organ of taste. 

The nervous system of LumbHculus is formed upon the same plan as that of 
other Oligochaeta ; the cerebral ganglia are figured by Vejdovsky (24, Taf. xii, p. 38), 
the cerebral ganglia and the commencement of the ventral cord by Ratzel (3, fig. 10, 
Taf. xlii.), and Leydig (7, Taf. iv. fig. 6). The brain (cerebral ganglia) consists of two 
ganglia, which are composed of nerve-cells anteriorly and fibres posteriorly; the two 
masses of cells are united by a narrow row of cells. The commissures passing round 
the oesophagus have no ganglionic cells ; where they join the ventral chain there 
is a group of cells in front of, and behind, the point of junction. The neuro-chord 
forks at the commissure, and extends alon^ the commissure for a very short way 
on each side. Vejdovsky represents the brain as more extensively cellular than does 
Ratzel, whose account has been followed above. According to Vejdovksy, the ventral 
nerve-chord gives off only a single pair of nerves in each segment, which almost 
immediately perforate the muscles of the parietes. Transverse sections show the 
presence of these ' giant fibres.' A pair of lateral ganglionated nerves are present. 

The sense-organs of Lumhriculus consist, according to Ratzel, of 'a sharply- 
marked white stripe in the median ventral line,' which gradually fades away 
posteriorly. This streak is composed of irregular spots, which consist of groups of 
cells in intimate connexion with the nerve-chord. Vejdovsky has denied the 
existence of this sense-organ; he describes, however, the cup-shaped tactile organs 
of which mention has already been made. 

The alimentary canal is, with the exception of the buccal cavity, ciliated 
throughout. It commences with a strong muscular pharynx, which can be everted 


and retracted by means of special muscles attached to the body -wall; it occupies 
segments ii-iv, the space in front belonging to the buccal cavity. The oesophagus 
occupies the next two segments, the intestine commencing in the seventh. The 
commencement of the intestine is marked by the chloragogen-cells which cover it 
throughout ; it is lined by columnar ciliated cells which are narrow at their base ; 
between them lie roundish cells, which appear to be young cells in course of growth ; 
outside the lining epithelium is a delicate muscle-layer, consisting only (Dieffenbach) 
of circularly-arranged fibres. An abundant plexus of excessively fine blood-vessels 
surrounds the gut. 

The vascular system consists of a dorsal and ventral blood-vessel, which are 
united by transverse trunks, and of a series of caecal appendages of the dorsal 
vessel. The dorsal vessel lies close to the dorsal wall of the gut, being covered by 
the chloragogen-cells ; in the first segment it divides into two the branches reuniting 
on the ventral side of the gut in the sixth segment. The dorsal and ventral vessel 
are united by pairs of non-contractile transverse vessels in every segment ; in the 
first eight segments (DiEFFiSNBACH), or fifteen (Katzel), these transversely running 
trunks give rise to branches which form a complicated and irregular network, 
continuous from segment to segment. At the ninth segment the network entirely 
ceases, and the simple vessels, which are at first much coiled, alone remain ; they 
lie in the posterior region of each segment. In front of each, commencing with 
the ninth segment, are a pair of lateral vessels, also arising from the dorsal trunk, 
which branch and end blindly ; at first the branches are less numerous ; afterwards 
each vessel has from eight to ten; they are attached to the parietes by slender 
contractile fibres ; they are themselves contractile. 

Vejdovsky's account differs somewhat from that of Dieffenbach — so much 
so that one is inclined to believe that the two naturalists studied different 

According to Vejdovsky, who examined young specimens (and starving, so that the 
vessels were specially clear, owing to the empty condition of the alimentary canal), 
each segment from the ninth to the sixteenth, has only a pair of non-branched, but 
contractile, vessels situated close to the posterior dissepiment ; from the seventeenth 
segment onwards there is no trace of these vessels. In the thirteenth segment the 
branched vessels commence, but in the thirteenth segment they are only short caecal 
appendages of the dorsal vessel ; the branching becomes quite obvious at about the 
seventeenth segment. 

The nephridia commence in the ninth segment; they are present, at least, in 
some of the genital segments, when the worm is matured. Where the funnel passes 


into the canal a granular, glandular-looking mass is present (cf. with Naids and 

The regeneration of the tail has been studied principally by v. BtiLOW (1, 2) and Ran- 
dolph (4). The principal point of interest in the actual process of regeneration concerns 
the formation of the mesoblast. This appears to be produced by the division of large 
peritoneal cells, which can be distinguished in the body-cavity of the adult worm ; 
the name 'neoblasts' is suggested by Randolph for these cells. This process of 
growth justifies a comparison between the regenerative process in this Annelid and 
the 'budding' of Naids (see below). In the latter worms the budding zone begins 
to be formed by a multiplication of peritoneal cells. The suggestion has been made 
that these neoblasts correspond to ova which are also, of course, peritoneal cells. 

Vaillant (6) allows five species to this genus, viz. : — 

(1) Lunibriculus variegatus, Geube. 

(2) Lumbrieulus limosus, Leidy. 

(3) Lunibriculus spiralis, Leidy. 

(4) Lunibriculus hyalinus, Leidy. 

(5) Lumbrieulus lacustris, Czeeniavsky. 

The last, however, with a query, and justly ; for in the description given there is not a particle 
of evidence that it is not, for example, a Clitellio. With regard to the two species, Lumbrieulus 
spiralis and L. hyalinus (first of all placed by Leidy in a separate genus, Acestus), there is, again, not 
the slightest evidence that they belong to this genus. The fact that there are from three to five or from 
three to eight setae in each bundle argues rather that they are referable to some Tubificid like 
Limnodrilus ; and that is the only possible character that can be drawn from the definition. 

Lumbrieulus limosus, on the other hand, appears at least a Lumbriculid, and it may be a Lum- 
brieulus ; but the pores on the tenth segment suggest male pores, in which case it is not a Lumbrieulus. 
Trhere are fifteen pairs of vascular caeca in each segment. 

Lumbrieulus variegatus, O. F. Mullee. 

Lumbrieulus variegatus, 0. F. MiJLiiEE, Verm, terrestr. vol. i. 3, 1774, p. 26. 
Lumbrieulus teres, Dalyell, Powers of the Creator, vol. ii. 1853, p. 140. 
Saenuris variegatus, Johnston, Cat. Worms, 1865, p. 6^. 

Deflnition. Length 80 mm.; number of segments, 300 y body dark green anteriorly ; 6-8 
caecal appendages of dorsal vessel in each segment posteriorly. Hob. — Europe. 

This, the earliest known species of Lumbriculid, is much more imperfectly known 
than most others ; the reproductive organs have not yet been properly described. 
The species has been studied by d'Udekem (1, 5), Buchholtz, Ratzel (3), Vej- 
DovsKr (24), BtJLOW (1, 2) and Randolph (4), in addition to the authorities 
quoted above. Figures are given by Bonnet (PL i. figs. 1-5), Grxjbe (9, PI. vii. 


figs. 2 a-d), Dalzell (PI. xviii, figs.. 10-12), Buchholtz (fig. 16), Katzel (3, PI. xlii. 
figs. 6, 10, II, 14, 19), Vejdovsky (24, PI. viii. fig. 68, pi. xii. figs. 16-32), BiJLOw 
(1, 2), Kandolph (4). 

Genus Rhynchelmis, Hoffmeistek. 

Syn. Euaxes, Gkubb. 
? Lyoodrilus, Grdbe. 

DEPiwiTioisr. Setae not bifld at tip ; clitellum VIII-XVI ; prostomium elongate ; testes 
in IX, X ; ovaries in XI ; sperm-sacs and egg-sacs paired, extending through a large 
number of segments ; spermiducal glands opening on to X, with a coating of 
glandular cells broken up into rounded masses ; spermathecae one pair opening 
on to VIII, each pouch with a single diverticulum ; a single median albumen- 
gland opens on to IX. 

The name Rhynchelmis is obviously the right one to apply to this genus ; but the synonym 
Euaxes has persisted for a long time, in spite of the fact that Gkube himself recognized the identity 
of his Euaxes fiUrostris with Hopfmeistee's (3) Rhynchelmis. d'Udekem (1), for example, has 
retained the name Euaxes, although he gave correctly the synonymy of the species fiUrostris ; so, too, 
Clapak^de (2) and Vaillant (3). In his later work (6) Vaillant has set this matter right, 
pointing out that Geube 'in 1851 preserved his own generic name of 1844, though giving as a synonym 
Rhynchelmis with the date 1843.' Vaillant (6) considers that Gkube's genus Lycodrilus must also be 
regarded as a synonym of Rhynchelmis ; this genus was created for a species, L. dyhowskii, from Lake 
Baikal ; from Geube's account it would appear to differ from Rhynchelmis by the presence of large 
bifid setae, replacing the setae on the ventral side of the body on segments ii-x. I am disposed, 
therefore, to agree with Vejdovsky (24) in not referring the worm to the present genus, but to regard 
it as 'incertae sedis.' Though using the name Lycodrilus as a synonym of Rhynchelmis, Vaillant 
prefaces the description of the species L. dyhowskii by a query. Our knowledge of the anatomy of this 
genus dates from the publication of Vejdovsky's important paper upon this worm (5). Neither 
HOPPMEISTEE nor Geube gave any satisfactory details upon the internal structure ; their memoirs 
were devoted almost exclusively to the external characters ; when Geube did venture upon aiiy 
references to structure, they were by no means successful; for example, he mistook the contractile 
appendages of the dorsal blood-vessel for caeca of the gut, which are non-existent. Some few 
additional details and figures were given by the same author in his work upon the Oligochaeta (8). The 
reproductive organs and the development have been recently described by Vejdovsky (24), some of 
his earlier statements being revised; the only other author who has written upon this genus is 
Kovalevsky, who investigated its development a good many years before the appearance 
of Vejdovsky's work upon the same subject. 

The genus Rhynchelinis is, so far as our present knowledge goes, confined to the 

fresh waters of Europe ; it has been met with in Bohemia, Russia, Belgium, and 

Germany ; I point out later that certain species, referred to this genus from other 

parts of the world, are probably not really referable to it. I have seen a specimen 


from some part of England ^, but cannot give any details. There is every probability 
that it is a native of this country. 

The most salient external character of the genus is the long prostomium ; the 
peculiar form of this is sufficient to prove that Hoffmeister and Grube were 
dealing with the same worm in their descriptions of Rkynchelmis and EiMxes; the 
same kind of prostomium occurs also in the nearly-related genus Sutroa from North 
America; but nothing of the kind is found in any other Lumbriculid; in Nais 
proboscidea, however, there is a prostomium which is of the same character. 

The nephridia are, of course, paired structures ; they commence in immature 
individuals in the seventh, in mature individuals in the twelfth, segment. The 
nephridia become enormously large in proportion to the worm ; they stretch so 
far back beyond their point of opening that, 'on a supei"ficial inspection each 
nephridium has the appearance of occupying several segments.' The nephridiopores 
are placed in front of the ventral setae. 

The vascular system of Rhynchelmis is described in some detail in Vejdovsky's 
original paper upon the anatomy of the worm ; a good figure of some additional 
particulars is to be found in the Entwickelungsgeschichtliche Untersuchungen 
(PL xxviii. figs. 7 and 8). The dorsal vessel is pulsatile; it communicates with the 
ventral vessel by a series of perivisceral trunks, a pair to each segment; after the 
eighth segment there are, in addition, a pair of vessels arising from the dorsal trunk, 
which do not end in the ventral vessel, but give oiF a number of contractile branches, 
as in other Lumbriculidae ; there are six or eight pairs of these branches, which 
were confused by Grube with diverticula of the gut; when the worms have 
attained to sexual maturity, the ninth, tenth, |tnd eleventh segments, are seen to 
contain each a pair of long vessels, giving off a rich network, which ramifies over the 
sperm-sacs and the other reproductive organs ; the intestine has a rich plexus, derived 
from the paired, non-contractile perivisceral trunks. The ventral vessel consists, in 
the first five segments of the body, of two separate halves, each half receiving the 
perivisceral trunk of its own side. 

The testes, at first overlooked by Vejdovsky, were subse(][uently (9) recognized 
by him as two pairs of gonads in the ninth and tenth segments ; as the worm gets to 
be mature, the testes disappear, their contents being transferred to the sperm-sacs ; 
the same thing happens to the ovaries, which lie in the eleventh segment. The 
spei-m-sacs have been already described in sufficient detail (p. 93). The spermathecae 
and the sperm-ducts open behind the ventral pair of setae of their segment; the 
albumen gland opens in the middle line, between these setae ; the oviducal pores are 

' I believe this specimen to be in the Oxford Museum. 


on the border-line between segments xi/xii, behind the ventral setae. The structure 
of the reproductive organs has already been described in treating of the characters of 
the family. 

(i) Rhynchelmis limosella, Hofpmeistee. 
R. limosella, Hoffmeistee, Arch. f. Nat. 1843, p. 19a. 
Euaxes fllirostris, Grube, Arch. f. Nat. 1844, p. 204. 
Definition. Length, lao mm.; number of segments, 150. Body more or less quadrangular. 
Prostommm long. Setae posteriorly reduced to one in each bundle ; cAloragogen-covering 
of gut commences in the seventh segment. Hob. — Euroj}e. 
Illustrations of this species and of its development are given in the following 
papers and books: Grube (9, PI. vii. figs, i a-d), Menge (PI. iii. pp. 14-17), 
KovALEVSKT (Pls. iii-v), Vejdovsky (5, Pis. xxi-xxiv), 24, PI. xii. figs. 33, 34, 
PI. xiii. figs, i-ii, PL XV. fig. 37, PI. xvi. figs. 1-16). 

(2) Rhynchelmis obtusirostris, Menge. 

Euaxes obtusirostris, Menge, Arch. f. Nat. 1845, p. 31. 

Definition. Length, ^^ mm.; number of segments, 100. Prostomium obtuse. Hal. — 
Belgium ; Germany. 

The principal reason for retaining this species (figured by Menge [PI. iii. figs. 1-13]) 
is that d'Udekem and Menge had the opportunity of studying both this species and the 
last, and were, therefore, presumably not likely to have made an error. Otherwise, 
as Vaillant says (6, p. 320), the characters are very incomplete. 

Genus Trichodbilus, Clapaeede. 

DEPuriTioiT. Setae simple ; dorsal vessel with 2-5 contractile perigastric branches, 
but with no caeca ; spermathecae in XI, XII ; male pores on X. 

This genus has been investigated by Clapaeede only. It comes very near to 
Phreatothrix, from which it differs principally in the fact that it has no caecal 
appendages of the dorsal blood-vessel. There is but one species — 

Trichodrilus allobrognm, Clapaeede. 
T. allobrogum, Clapaeede, Mdm. Soc. Phys. Gen. 1862, p. 267. 
Definition. Length, 25 mm.; about 70 segments. Prostomium 'conical, long, not separated by 
furrmv from buccal segment. Spermathecae opening behind ventral setae. Nephridia in VI 
and then again from XII onwards. Hah. — River Seine ( Geneva). 
The anatomical characters are fully illustrated in Ci.apaeede's memoir. 



Genus Phbeatothbix, Vejdoysky. 
Syn. Triehodrilus, Vejdovsky. 
Definition. Setae sigmoid, not uncinate; one pair of spermathecae in XI. In 
posterior region of body, five or six pairs of lateral contractile branches of dorsal 
vessel in each segment. 

This genus, first described by Vejdovsky as a species of Triehodrilus, was 
subsequently (24) recognized as a distinct genus. Our knowledge of its structure is 
entirely due to Vejdovsky. 

The reproductive organs are like those of Triehodrilus, except for the fact that 
there is only a single pair of spermathecae lying in the eleventh segment ; in the 
twelfth segment, however, a; rudimentary second pair make their appearance, but only 
acquire traces of a lumen, and finally completely degenerate. The external orifices of 
the spermathecae lie behind the ventral setae ^. There is no albumen-gland. 

The nephridia are very remarkable ; the first pair appears to belong to the eighth 
segment ; the funnels open into the coelom of the seventh segment, and the external 
pore of the organ is upon the eighth segment ; but the tube which connects these 
two openings extends back through the body as far as the fourteenth segment. The 
second pair of nephridia is similarly elongated. The funnels open into the thirteenth 
segment, and the external pores are upon the fourteenth segment. The tube is looped 
back as far as the twenty-first segment. Behind this, each pair of nephridia extends 
through three segments, although, as before, the internal and external apertures are 
upon successive segments. • 

The principal point of interest in the circulatory system is the condition of 
the characteristic appendages of the dorsal vessel. There are in the posterior 
segments of the body five or six pairs of these to each segment, which are not 
regular in their arrangement ; each of these branches terminates in a bifid extremity. 
The blood is described by Vejdovsky as being yellow rather than red. There are eight 
pairs of perivisceral vessels uniting the dorsal and ventral trunks in the anterior 
segments. The ventral vessel is forked in the fourth segment. In the young worm 
septal glands can be made out in the anterior segments. There are four pairs of 
these lying in segments iv-viii ; the glands of each side of the body are connected by 
a longitudinal strand, which is the duct, and which opens into the pharynx. They 
are not so plain in older worms. 

' In Vejdovsky's first account of the anatomy of this species (21) the setae of this segment are stated 
to be absent. 



The brain lies in the first and second segments; it has marked anterior and 
posterior paired lobes. Behind the pharynx a circle of nervous cells surrounds the 
oesophagus, as in Chaetogaster. 

Phreatothrix prageTisis is the only species well known. It may be that Euaxes 
haicaleThsis of Geube (4) is referable to the same genus, perhaps even to the 
same species. It possesses a pair of papillae on the tenth segment (clearly the male 
pores), and a pair of orifices on the eleventh segment, which I agree with Vaillant 
(6, p. aai) in considering to be spermathecal orifices. 

Phreatothrix pragensis, Vejdovsky. 

Trichodrilus pragensis, Vejdovsky, SB. Bohm. Ges. 1875, p. 196. 
Phreatothrix pragensis, Vejdovsky, Z. Wiss. Zool. 1876, p. 541. 

Definition. Length, 40 mm. j number of segments, 80. Prostomium twice as long as buccal 
segment, obtuse ; fve anterior segments hi-annulate. Hal. — Prague, in wells. 

To give any other characters in definition of this species would be merely, in 
the present state of our knowledge, to redefine the genus. 

Genus Clapabedilla, Vejdovsky. 

Syn. Lumbrieulus, Clapaeede. 

Pseudolumbriculus, Dibffenbach. 

Defiwition. Setae not bifid; both pairs of contractile perivisceral vessels with 
caecal appendages ; spermathecae in IX. 

I have already gone into the question of the probable identity of Vejdovsky's 
genus Claparedilla with Clapae^de's Lumbrieulus (see p. 311). The genus has been 
exclusively studied by the two writers whose names have just been mentioned. 

Clapaeede stated that the setae are ' simple, or indistinctly bifid ' ; but Vejdovsky 
found in the species studied by himself, and also in Clapaeede's species, that the 
setae did not show any bifurcation ; this slight point of difference, for, after all, 
Clapaeede is not very decisive upon the matter, should not, perhaps, be made too 
much of. The vascular system is very characteristic of the genus, but appears to 
differ a little in the two species ; the details will, therefore, be reserved until the 
species are described ; but, in the meantime, it may be pointed out as characteristic 
of the whole genus that both pairs of branches of the dorsal vessel are provided 

F f 2 


with caeeal diverticula. The nephi-idia commence in the seventh segment, but are 
wanting in the following segments; they recommence again in the thii'teenth. The 
only two known species of the genus are European. 

(0 Claparedilla meridionalis, Vejdovsky. 

C. meridionalis, Vejdovsky, SB. Bbhm. Ges. 1883, p. 236 (footnote). 
Lumbriculus variegatus, Claparede, Mem. Soc. Phys. Gen. 1863, p. 255. 
Non-Lumbriculus variegatus, Gkube et AucT. 
Pseudolumbriculus Claparedianus, Dieppenbach, Anat. u. Syst. Studien, p. 81. 

Definition. Prostomium as long as buccal segment; anterior perivisceral vessels in each 
segment covered with cJdoragogen-cells, dilated where they join ventral vessel, and giving 
off there four or five caeca ; posterior perivisceral vessels with pennate series of branches. 
Hab. — Geneva; Trieste. 

I follow Vejdovsky in identifying his Claparedilla with Clapakede's Lumbriculus 
variegatus; there are, however, certain differences in the descriptions given by these 
two writers, which are not easji, to reconcile with the theory that they were 
describing the same species. The question of the bifurcation or the non-bifurca- 
tion of the setae has been already referred to. Of more importance, perhaps, 
is the different account given of the ' vascular system ; according to Claparbde 
the two arches on either side are furnished with contractile caeca which arise only 
-on the posterior margin, in an identical position in the case of each ; on the other 
hand, Vejdovsky has figured and described the posterior of the two arches alone 
as furnished with a series (double, not single) of caeca arising along its whole 
length ; the anterior visceral arch swells into an ampulla on the ventral side of the 
body, just at its junction with the ventral vessel ; from this swelling are given off 
three or four caeeal appendages; this arch alone is covered with a coating of 
chloragogen-cells. CLAPAEioE describes the same condition as existing in this arch. 
The dorsal vessel is covered with a coating of the same brown cells from the fourth 
segment onwards, and the pigmented cells upon the alimentary tract commence 
in the sixth segment at its end (Clapabede). Vejdovsky states that the dorsal vessel 
is covered with the pigmented cells along its entire length. Dieffenbach's paper 
dealing with the non-identity of the species termed by both Clapak^de and Geube 
Lumbriculus variegatus, was published at about the same time as Vejdovsky's 
great work. His description of the reproductive organs of Pseudolumbriculus 
agree with those of Claparedilla (see under Lumbriculus). 


(2) Claparedilla lankesteri, Vejdovsky. 

C. lankesteri, Vejdovsky, Syst. u. Morph., 1884, p. 54. 

Iiumbriculus lankesteri, Vejdovsky, SB. Bohm. Ges. 1883, p. 226. 
Definition. Prostomium double as long as the buccal segment ; first five segments bi-annulate ; 
both pairs of perivisceral arches with pinnate series of caeca. Hah. — Podebrad, 

As this species is only known by a single immature example, it cannot be regarded 
as absolutely certain that it should be included in the same genus as the last; the 
difference in the arrangement of the contractile caeca of the perivisceral vessels is 
made use of in the definition of the species. It will be noticed that the arrangement 
which characterizes the present species is more like that of the last species as 
described by Vejdovsky. 

Genus Stylodriltjs, Claparbde. 

Depiw ITIOH". Setae bifid ; clitellum, X-XII ; spermathecae, a pair in IX ; a pair 

of non-retractile penes present on segment X, perforated by sperm-ducts ; 

spermiducal gland pear-shaped, with long duct. 

This genus has been anatomically described by Claparedb (2) and by Vejdovsky 

(24) ; more recently, Benham (9) has described a new species of which he has 

detailed the structure. A few details upon Stylodrilus have also been contributed 

by Eatzel (3). 

The vascular system consists of the usual dorsal and ventral vessels put into 
communication in each segment by two pairs of vessels, both of which, as weU as, of 
course, the dorsal vessel, are contractile ; one of the communicating vessels is called by 
CLAPABtDE the 'anse intestinale.' the other 'anse p^rivisc^rale.' The former lies 
anteriorly in the segment, instead of posteriorly, as in Claparedilla; it is imbedded 
in the peritoneal cells which cover the intestine, but is nevertheless easy to see on 
account of its large size ; neither of the trunks shows any caecal appendages, such 
as exist in other Lumbriculidae. . 

The nephridia are wanting in thei first six segments. They exist in the seventh 
and from the thirteenth onwards. The external pore is in front of the ventral setae. 

The most remarkable feature about the reproductive organs is the existence of 
a pair of penes which are non-retractile, but are perforated by the sperm-duct ; 
they exist in all three species of the genus, and lie behind the ventral setae ; the 


spermathecae open in a similar position behind the ventral setae of the ninth 

The three species of the genus are the following: — 

(i) Stylodrilus heringianus, Clapahede. 

S. heringianus, Clapakede, M^m. Soc. Phys. Gen. 1862, p. 263. 
Definition. Length, 30 mm.; number of segments, 80. Prostomium obtuse. Dorsal vessel 
without dilatation. Penis as long as half diameter of body. Hah. — Geneva ; Bohemia. 
ClapaeJide has figured all the important organs of the body in this species. 
A curious point is the existence of a large octaedral crystal in the spermathecae, 
which seems to distinguish this species from the next. Vejdovskt found this species 
in Bohemia. 

(2) Stylodrilus gabretae, Vejdovsky. 
S. gabretae, Vejdoysky, SB. Bohm. Ges. 1883, p. 325. 

Definition. Length, 40 mm. Prostomium conical. Dorsal vessel with dilatations ; hearts 
in VI and VII ; penis nearly equal in length to the diameter of the body. Hah. — 
This species is fully illiistrated ty Vejdovsky in his work upon the Oligochaeta. 

(3) Stylodrilus vejdovskii, Benham. 
S. vejdovskii, Benham, Quart. J. Micr. Sci. xxxiii. p. 309. 

Definition. Length, 25 mm.; dorsal vessel not dilated ; nephridia of first two pairs extend 
through three or four segments. Penis just greater than half diameter of body. Hob. — 
Biver Cherwell, England. 

This species (illustrated by Benham) is said to differ from the other two in that 
the setae are all notched. But, as these kinds of setae appear to be liable to wear, 
the difference is, perhaps, not real. The nephridia are obviously like those of 
Phreatothrix (see p. 218). 

Genus Sutboa, Eisen. 

Definition. Prostomium elongated; setae not bifid; olitellum VII/XV; spermi- 
ducal glands enormously elongate, and furnished with a coating of glandular 
cells, broken up into globular masses opening on to X ; the sperm-sacs 
surround the spermiducal glands ; a single spermatheca in VIII, with numerous 
branched and simple diverticula. Testes in X ; ovaries in XI ; ovidueal pores 
on XI/XII. 


This genus, which is in many respects a very remarkable one, is confined, so 
far as our present knowledge goes, to California. Its anatomy has been described 
by EisEN (2, 5) and by myself (81). The genus contains two species. 

Sutroa agrees with Rhynchelmis in its long and filiform prostomium; in many 
other points it differs from that genus. In the first place the spermatheca is single 
and median, and is furnished with numerous diverticula, which are both branched and 
single. The very peculiar shape of the spermatheca in Rhynohdmis—^ecnM&x, that is 
to say, as compared with other Lumbriculidae — suggests that one diverticulum is 
present in the spermatheca of that genus; but in any case the multitudinous caeca 
of Sutroa differentiate it from Rhynchelmis, even without going into any other 

There are, however, many other points of divergence between the two genera; it 
is not certain whether there is in Rhynchelmis any communication between the 
spermatheca and the lumen of the gut; Vejdovsky (5) speaks of 'einen Schlitz, 
welcher den Eindruck einer Offnung macht, und in der That bemerkt man, dass unter 
dem Druck des Deckglaschens auf das Organ nur auf dieser SteUe die Spermato- 
zoenbiindel herausgehen.' Eisen (5), too, mentioned a pore at the extremity of the 
spermatheca of >S'. alpestris. ' The object of such an opening,' he remarks, ' is not at 
present understood.' I have, however, myself shown (81) that this aperture 
communicates with the gut, though this discovei'y by no means explains the ' object ' 
of the pore; it is, however, not unknown in other Oligochaeta (see p. 127). 

A comparison of Vejdovskt's (5) original paper upon Rhynchelmis with Eisen's 
(2, 5) and my own contributions to the anatomy of Sutroa would indicate rather greater 
differences than those that actually exist. Vejdovsky's most recent contribution to 
the structure of Rhynchelmis is contained in his researches upon the development of 
that Annelid (9). From this it appears that the differences are not, after all, so great ; 
the spermiducal glands are, or rather may be, much greater in extent than is suggested 
in the earlier paper; Vejdovskt writes that they may extend as far back as to the 
thirtieth segment, and that they occupy the cavity of the sperm-sacs ; this is exactly 
what occurs in Sutroa, though the glands are by no means so extensive ; they reach 
back as far as the twentieth segment about, and are invested for the whole of their 
course by the sperm-sacs. A curious point about the sperm-ducts of Sutroa alpestris 
is that the anterior pair are much slighter than the posterior pair ; the funnel, too, 
is less developed ; one is inclined to come to the conclusion that the anterior pair of 
ducts is in course of disappearance, especially since there are no testes corresponding 
to them. So far, therefore, Sutroa seems to connect the Lumbriculidae with the 
Tubificidae, where there are, of course, only a single pair of testes and funnels ; but 


I have already gone into this matter in describing the characters of the family 
Lumbriculidae, and need not recur to it here. 

Attached to the anterior septum of the tenth and eleventh segments, and in the 
latter case close to the testes, are two pairs of bodies, called 'albumen-glands' by 
EiSEN. Their walls are delicate and muscular, and their cavity is subdivided by 
anastomosing trabeculae derived from the walls ; the interspaces thus formed contain 
loosely-packed cells, which suggest coelomic cells. I could find no duct leading to the 
exterior, and the fact that the sac on one side of the body contained one of the 
diverticula of the spermatheca suggests its coelomic character. I should be disposed to 
compare these sacs broadly to the sperm- sacs, septal sacs, &c. 

Sutroa rostrata, Eisen. 

S. rostrata, G. Eisen, Mem. Calif. Acad. Sci. (1888), vol. ii. No. i. 
Definition. Length 75 mm. ; ventral vessel forks in segment VIII ; only connected with 
dorsal vessel in prostominm. Cocoon not pointed at ends. Hob. — San Francisco^ in lake 
50 ft. above level of sea. 

Sutroa alpestris, Eisen. 
S. alpestris, G. Eisen, Zoe (189a), vol. ii. p. 333. 
Definition. Length about 40 mm. ; ventral vessel forked in segment VI, connected with dorsal 
vessel by perigastrics in each segment. Cocoon globular and pointed at ends. Hob. — 
California, Bonner Lake, 6,000 ft. 

Genus Alluroides, Beddard. 

Definitiobt. Setae simple ; male pores on XIII ; oviducal pores, XIII/XIV ; penes 
on XIII in front of male pores ; testes in X ; spermathecae in VIII ; no albumen 
gland ; no vascular caeca. 

This genus is at present known by one species only, which has been studied 
by myself (84). The chief points in its anatomy, besides those mentioned in the 
above description, are the following : — The clitellum is developed upon segments 
xiii-xvi. There are no special sperm-sacs, the two segments x, xi being filled 
with developing spermatozoa ; the ova lie in a considerable number of segments 
following that (the thirteenth) which contains the ovaries.)^ The spermathecae are 
without diverticula of any kind, and have very thick muscular walls, consisting of 
both circular and longitudinal fibres. The spermiducal glands are long, and have 



a thick investment of glandular cells, which are aggregated into groups; their ducts 
appear to traverse the muscular wall of the spermiducal gland ; the gland is ciliated 
at the end nearest to the external pore; just above the aperture of the gland, on 
each side of the body, is a process of the body-wall, which is, perhaps, comparable to 
the penis of the genus Stylodrilus; but it is not traversed by the canal of the 
male duct. The septal glands occupy the fifth to the ninth segment. The nephridia 
commence in the sixteenth segment ; they open in front of the second seta. 

AUuroides pordagei, Beddakd. 
A. pordagei, Beddaed, Q. J. M. S. vol. xxxvi. p. 244. 
Definition. Length about an inch; two pain of strong hearts in XII, XIII; thick septa 
IF/XII, that dividing X/XI being thinner . Hal. — Mombasa, East Africa; fresh water. 
The only two specimens known were collected by Mr. F. Finn in a swamp about 
four miles up country, opposite to Mombasa Island ; the worm has a delicate appearance, 
and appears to have no pigment in the skin. 

Appeindix to Lumbriculidae. 

Befim'ITIOII'. Aquatic Oligochaeta of moderate size ; setae paired, not bifld at 
extremity, of the usual Lumbricld pattern; the dorsal blood-vessel, with 
a series of slightly bifid contractile appendages in posterior segments; spermi- 
ducal glands opening on to X ; oviducts opening between X/XI ; a protrusible 
penis ; spermathecae in IX ; spermatophores are formed. 

I do not see any way out of the acceptance of Eisen's family of Eclipidrilidae 
for the very remarkable genus Eclipidrilus; Vejdovsky (24) refers it to the 
family Lumbriculidae, on account of the contractile appendages of the dorsal vessel 
and the characters of the setae. Vaillant (6), too, is disposed to follow this course, 
'at least provisionally.' 

It is, I think, undoubtedly desirable, as Vejdovsky says, that the genital organs 
should be subjected to renewed examination; but in the meantime it seems to me 
that we are in possession of sufficient knowledge concerning the worm to warrant 
its inclusion in a family of its own, which should be placed between that of the 
Lumbriculidae and the Tubificidae. 


The genus Edipidrilus does, it is true, agree with the Lumbriculidae in the 
characters mentioned by Vejdovsky. The setae are most unquestionably identical in 
their character with those of the Lumbriculids. When, however, we turn to the other 
organs of the body, we find divergences from the Lumbriculid type of structure. 
This is especially the case with the male efferent apparatus, which, however, 
evidently requires re-examination. The long spermiducal glands communicate with 
a retractile penis, apparently very similar to that of the Tubificidae ; the Lumbriculidae 
have no penis, with the exception of Stylodrilus and AUuroides, in which genera 
there are a pair of non-retractile penes, recalling those of certain Eudrilids. The 
spermiducal gland, too, is covered for part of its course with spirally-arranged 
muscular fibres, similar, as Eisen has remarked, to those of Camptodrilus. 

It is in the middle that the gland is invested with these spiral muscles. And 
here its diameter is much narrower than elsewhere, dividing the entire tube into 
two divisions ; in the lower part of the tube there are three apertures, placed one 
behind the other, by which its cavity communicates with the body-cavity; the walls 
of this tube are muscular, but the lining is, of course, epithelial. Freely suspended 
within this tube is another tube, which does not reach down so far as the first ; 
at the point where it ends, a little way in front of the commencement of the penis, 
it has a circular orifice. The long sperm-sacs occupy about the same segments as 
the remarkable efferent ducts, i.e. x-xv. There appear to be three pairs of gonads, 
which Eisen calls 'ovaries,' in segments ix, x, xi. The nephridia have a swollen 
glandular part immediately after the funnel, as in other Lumbriculidae, Naidomorpha, 
and Tubificidae. There seem to be no specially enlarged hearts ;. those of segment x 
are extended backwards, no doubt in relation to the development of the sperm-sacs. 
There is but one species, viz. Edipidrilus frigidus, from a spring in the Sierra Nevada 
of California, at an altitude of 10,000 feet. I attempt no definition of either species 
or genus. 


Definition. Aquatic Oligochaeta of small size and slender build, fresh-water or 
marine. Setae of three kinds, capilliform, pectinate, and uncinate, the former 
two kinds, when present, found only in the dorsal bundles. Dorsal and ventral 
blood-vessels, connected by perivisceral trunks, in every segment of the body. 
Testes in X, ovaries in XI, sperm-ducts always terminating in spermiducal gland, 


opening on to the eleventh segment. Oviducts open on to boundary-line XI/XII, 
Spermathecae one pair in X ^ Penial setae sometimes present. 

This family of Oligoctaeta is plainly derived from (or has given origin to) the 
Naidomorpha. Indeed, the two are united by Vaillant into one family, Naididae. 
The principal annectent genus is Ilyodrilus, formerly confounded with Tuhifetc, but 
now known, through the researches of Eisen (12) and Stolo (3), to be distinct. 
Ilyodrilus approaches the Naidomorpha principally by virtue of the structure of the 
nephridia and of the spermiducal glands, by the development of the oya, and by the 
relations of the intestinal vascular network. As in many Naids, a dark glandular 
swelling is developed upon the nephridium, immediately behind the septum; within 
this swelling the nephridial tube forms a network. 

The spermiducal gland differs from that of most other Tubificidae in having neither 
' prostale ' nor penis ; with it communicates a very short vas deferens, which is again 
a character found among the Naids. Ilyodrilus is, however, unmistakeably a Tubificid. 
The position of the sexual organs shows this, as well as the compUcated vascular system. 
Nevertheless, it serves to indicate the nearest affinities of the family in one direction. 
In two species of Hesperodrilus there is a slight cephalization, so characteristic of the 
Naidomorpha. Affinities with other families of Oligochaeta are not yet clearly apparent. 

The remarkable genus Fhreodrilus, perhaps, indicates a passage to the Lumbri- 
culidae. Its principal characters will be found below ; I separate it a little from the 
other Tubificidae. In my account of the anatomy (21) of this worm I pointed out 
that it ought to be regarded as the type of a new family ; I am not now disposed 
to separate it so widely. There is, however, it must be admitted, a considerable series 
of differences which distinguish the genus Fhreodrilus from the Tubificidae ; but 
rather greater differences distinguish the genus from other families of the aquatic 
Oligochaeta; and there are certain important resemblances to the Tubificidae, which 
are, perhaps, closer than any resemblances shown to the other grqups of Oligochaeta ; 
the chief resemblances to the Tubificidae are: — 
(i) Dorsal setae capilliform. 

(2) Spermiducal gland elongated without any thick glandular investment (but 

also without prostates). 

(3) Supra-intestinal vessel present. 

The two first characters, of course, are also those of the Naidomorpha. But, 
although I originally dwelt upon the likeness which Fhreodrilus shows to that 
family, I am not now so impressed with this view of its affinities. It has, for example, 
yet to be proved that Fhreodrilus is capable of multiplication by gemmation. The 

1 The position of the generative organs is abnormal only in Hesperodrilm {q. v.). 

Gg 2 


position of the reproductive organs is different to that of the Naids, though not, it is 
true, precisely that of the Tubificidae. 

Phreodrilus is also not far removed from the Lumbriculidae in many points. The 
chief resemblance is in the fact that the spermathecae open behind the male pores, 
a resemblance which is shared by Hesperodrilus. It is possible also that the periatrial 
sac is comparable to the periatrial sperm-sac seen in Sutroa; the sigmoid setae 
are something like those of the Lumbriculidae. Other features of resemblance are 
dealt with in the section dealing with the general characters of the Lumbriculides 
(see above). The genus Hesperodrilus, which has just been referred to, shows a slight 
resemblance to the Lumbriculidae in that the first pair of nephridia occupy a consider- 
able number of segments; the internal and external apertures of these lengthened 
nephridia are, however, as in the Lumbriculidae, in consecutive segments. We are at 
present evidently only very imperfectly acquainted with the extent of the range of 
structural variation in the family Tubificidae. 

It contains sixteen distinct genera, which are the following : — 

(i) Tubifex, Lamaeck. (9) Lophochaeta, Stolg. 

(2) Limnodrilus, CLAPAEiiDE. (10) Bothrioneuron, Stolc. 

(3) Peloscolex, Leiuy. (11) BrancMura, Beddakd. 

(4) Clitellio, Clapakede. (12) Psammoryetes, Vejdovsky. 

(5) Hemitubifex, El sen. (13) Hater ochaeta, Claparede. 

(6) Telmatodrilus, Eisen. (14) Embolocephalus, Eandolph. 

(7) Spirosperma, Eisen. (15) Vermiculus, GooDEiCH. 

(8) Ilyodrilus, Eisen. (16) Hesperodrilus, Beddaed. 

The Tubificidae thus contain a very large number of genera in proportion to the 
species; to anyone who has looked through Vaillant's work upon the Oligochaeta 
this statement will appear to be untrue ; but it appears to me very doubtful whether 
the large majority of the species allowed by that author are really tenable ; it will be 
seen, in the course of the following pages, that I have eliminated, and it is hoped with 
reason, a very considerable percentage of the described species ; out of the sixteen 
genera ten have only one definable species each. I am very doubtful how far it is 
really justifiable to separate Tubifex, Hemitubifex, Limnodrilus, Clitellio, Spirosperma, 
Camptodrilus (iacluded by Vejdovsky in Limnodrilus), and Psartwioryctes ; the 
latter is the most distinct, principally by reason of the peculiar setae attached to the 
spermathecae. There can be, on the other hand, no doubt about the distinctness of 
Telmatodrilus, Ilyodrilus, Lophochaeta, Heterochaeta, Bothrioneuron, Hesperodrilus, 
Embolocephalus, Vermiculus, and BrancMura. With regard to the first group, 
Tubifex blanchardi seems to differ more from Tubifex rivulorum than Tubifex 


rivulorum does from Limnodrilus. In considering the question of the generic 
separation of these forms, it must be borne in mind that the comparatively simple 
organization of the family leaves less room for differences than in the highly organized 
' Terricolae.' As all or nearly all the generic names have been in use for so long, and, 
as it is a less serious matter to multiply genera than to multiply species, I retain the 
names as they are at present. As to the species it is generally impossible to follow 
and compare descriptions by different authors. Minute points, emphasized by one, 
are neglected by another. "What is wanted is a careful study of the living worms by 
one naturalist who is able to devote a good deal of time to what is, after all, a very 
small matter. 

ElSEN proposed (12) to divide the genera known to him into two sub-families, one containing the 
genus Telmatodrilus alone, the others the remaining forms. 
They are distinguished by Eisen thus': — 
Subf. i. Telmatodrilus. 

Setae sigmoid, only bifurcate in young. 

Ventral nerve ganglia, distinctly paired, with numerous minute commissures. 
Five pairs of contractile hearts (not more dilated than others) in vi-x. 
Ventral vessel lateral in position approximated to dorsal. 
Spermiducal gland with numerous (ten) separate prostates. 
Subf. ii. Tubiflcini. 

Setae both capilliform and uncinate. 
Ventral nerve-ganglia not separated. 
One to two pairs of contractile hearts in viii, ix. 

Spermiducal gland with one large separate prostate, or without prostates. 
To these subfamilies Stolc added two others, viz. Ilyodrilinae and Bothrioneurinae. He defines 
these subfamilies and the Tubificinae as follows": — 
Subf. i. Ilyodrilinae. 

No penis. Penial setae present. Spermathccae without spermatophores. 
No prostates. Development of ova as in Naidomorpha. 
Subf. ii. Tubificinae. 

Penis present, but no penial setae. Prostates present and spermatophores. 
Development of ova as in higher Oligochaeta. 
Subf. iii. Bothrioneurinae. 

Penial setae present, but no penis. Male genital pore single and median. 
Prostates present. No spermatheca. 
If it were desirable to subdivide the family Tubificidae, my genus Branchiura would certainly form 
the type of a fifth subfamily. But it does not appear to me that much is to be gained by forming 
subfamilies in a family of so comparatively limited an extent as the Tubificidae. 

As regards the mutual affinities of the genera comprising this family, we may 

arrive at a comparison of the genera by the help of the accompanying table, which 

shows the principal anatomical resemblances and differences. 

' I abstract the two definitions from Eisen's two papers. 

^ As nearly as I can make out from the paper, which is in Czech. 












and uncinate 

? none-periv. 
loops of v-x 
gradually in- 
crease in calibre 





short and globular 
with naiTow duct 

Sothrionmron . 

uncinate only 

in Tii, viii 






or absent 

with continuous 

glandular covering, 

a diverticulum 


Telmatodrilus . 

uncinate only 

perivisceral loops 
of vi-x 





uncinate only 

in viii, ix, (and x) 

those of viii very 

large ^ 





with prostate 

CliteUio . . . 

uncinate only 

in viii, ix 



no prostate 

HemituUfex . . 

and uncinate 

with prostate 

Spirospen-ma . . 


uncinate, and 






with prostate 

TuUfex . . . 

and uncinate 

in viii mucli en- 
larged (intestinal) 




or absent 

with prostate 

Lophochaeta . . 


uncinate, and 


in ix (intestinal) 


? contractile 





with prostate 

Psammoryctes . 


uncinate, and 






' with prostate 

Heierochaeta . . 

uncinate and 

in vidi 



a prostate 

Vermicidus . . 


in X 


without glands 

Branchiura . . 

and uncinate 

v-x all con- 
tractile, last two 




a continuous 
glandular covering 

Hesperodrilus . 

and uncinate 






without prostate 

Emholocephdlus . 

and uncinate 


on X with 

a gland 

with prostate 

' In L. nmae-selandiae those of ix also are larger. 


The distinctively Tubificid characters appear to be (i) the presence of two kinds 
of setae, capilliform and uncinate, and the frequent modification of the last to form 
pectinate setae; (2) a long spermiducal gland, with a protrusible penis and a prostate; 
(3) a supra-intestinal, in addition to a dorsal, vessel ; (4) the contractility of at least 
one pair of the perivisceral trunks ; (5) absence of penial setae. 

I should, therefore, regard the genera Tubifex, Hemitubifex, Spirosperma, Psammo- 
ryctes, and Lophochaeta, as occupying a central position in the family. So far as can be 
seen at present, llyodrilus is the simplest form of Tubificid, connecting this family 
with the Naidomorpha. It may be, therefore, placed at the base of the genealogical tree 
which indicates the probable relationships of the different genera. Simplification of 
structure is doubtless often due to degeneration ; but the resemblance which llyodrilus 
shows to the Naidomorpha is not entirely due to the fact that it, like them, is a worm 
of simple organization. The simplification of the male efierent apparatus, absence of 
prostate and penis, might, indeed, be regarded from this point of view; but the 
connexion of the intestinal blood-plexus with the ventral vessel, the network formed 
by the nephridial tube immediately behind the funnel, and the presence of penial setae, 
are characters shared with the Naidomoi-pha which are by no means suggestive of 
degeneration; neither is the integumental blood-plexus. 

Limnodrilus and Clitellio are so far specialized genera, in that they have lost the 
capilliform setae ; the latter genus also has lost the distinctively Tubificid prostates. 
I should regard these forms as an offshoot, which has, however, progressed but a little 
way from one or other of the more typical genera. An attempt to fix the position of 
Telmatodrilus is more difficult. It shows certain resemblances to the very aberrant 
genus Branchiura, in the possession of several contractile perivisceral trunks, and in 
the integumental network. The last-mentioned character, as well as the absence of any 
specially dilated hearts, connects Telmatodrilus also with llyodrilus. Perhaps these 
characters indicate that Telmatodrilus has originated from the Tubificid stem a little 
before the acquirement of all the characteristic features of the group. Branchiura 
also must be looked upon (in my opinion) as a, comparatively speaking, primitive 
type. It has not lost the integumental network ; the perivisceral loops of five or six 
segments are contractile, the spermiducal gland is like that of llyodrilus, but the 
relations of the gland to the vas deferens indicate a specialization, as do also the 
branchiae of the hinder segments. The same arguments apply to Hesperodrilus. 

Bothrioneuron is also a much-specialized type, as is shown by the absence of 
spermathecae, and by the peculiar form of the prostates. 

The mutual relations of the different genera, as I believe them to be, are indicated 
by the accompanying phylogenetic scheme. 



Peloscolex is necessarily left out of consideration, as we do not know enough 
about it. 




Spirosperma - 







- Bothrionenron 





In all the Tubificidae the vas deferens of each side opens into a terminal chamber, 
■which I propose to call the spermiducal gland, abandoning the term atrium ; a portion of 
this is commonly exsertile, and forms a penis. The simplest form of the gland is seen 
in Ilyodrilus ; there is in /. coccineus (woodcut, fig. 39) a simple globular sac, which 
receives the sperm-duct ; the sac is lined with ciliated epithelium, outside of which 
is a feebly developed layer of muscle, covered again by a mass of large granular cells. 
This suddenly narrows into a duct which leads to the exterior ; it is not exsertile, but 
special muscles allow of the protrusion of this duct along with the epidermis of the 
body. Stolc figures (3, Tab. iii. fig. 1) a conical eminence of the body, well within 
which lies the duct. The resemblance of this chamber to that of the Naidomorpha 
and of the Lumbriculidae and Moniligastridae has already been pointed out. 

Branchiura , has an efferent apparatus which, besides possessing peculiarities of its 
own, is in some respects intermediate between Ilyodrilus and other Tubificidae. 

The terminal chamber itself is apparently much as in Ilyodrilus, only the lining 
epithelium does not show any ciliation, and the other layers are thicker. The principal 
difference is that the vas deferens joins the atrial duct just above the penis. This 
is also the case in Hesperodrilus ; this duct is long, and differentiated into two 
regions ; it has at first a flattened, ciliated epithelium ; this, changing abruptly, 
becomes taller, thrown into slight folds, and not ciliated. The whole duct is surrounded 
by muscles, and there is some evidence that the distal part of the tube is eversible. 
In any case the distal part of the tube, being of a different structure from the 
restj may be termed the penis. 

In all the remaining Tubificidae there is an extrusible penis, and the spermiducal 
gland is rather different in structure, being, moreover, in every case, except Clitellio 
and Hesperodrilus, furnished with one (Tubifex, Limnodrilus, &c.) or many (Telmato- 
drilus) prostates (' Cementdriisen,' ' Kittdriisen '). The simplest state of affairs is 
found in Limnodrilus, and, in fact, in all genera except Tuhifex. In Limnodrilus 
the terminal chamber is long, narrow, and somewhat pear-shaped, being swollen 
where it receives the sperm-duct. Its lining epithelium, like that of the vas deferens, 
is ciliated ; outside this is a layer of muscular fibres, and outside this, again, a peritoneal 
covering. At one side, near the upper extremity, is a lobate gland, the prostate, 
attached to it ; this gland, which has been variously termed ' prostate,' ' Kittdriise,' 
' Cementdriise,' ' v^sicule seminale,' is solid and compact, and composed of pear-shaped 
granular cells; where it is attached to the terminal chamber, the muscular and 
peritoneal layers of the latter disappear, so that the cells of the gland are in actual 
contact with the epithelium of the terminal chamber. This is well shown in 
Clapabede's figures (2, PI. i. figs, i and 4) of L. hoffmeisteri and udekemianus. 



Clapae^de, Nasse, and Dieppenbach speak of a cavity, or, rather, several cavities 
■within the lobes of the prostate, which unite to form the common duct of the gland 
which opens into the terminal chamber. The most detailed account of the lumina 
is given by Dieppenbach (for Tuhifex indeed, but the conditions do not differ in 
Limnodrilus), and runs as follows : — 

' Sie (die Kittdrlise) besteht aus einzelnen langlichen Drlisenlappen, die von der 
Miindungsstelle aus sieh facherformig ausbreiten und in deren Mitte ein feiner 
Kanal verlauft, dem die einzelnen, stark granulosen, mit grossem Kern und Kern- 
kbrperchen versehenen Driisenzellen aufsitzen. Ein Theil dieser Driisenlappen lagert 
dem Atrium auf, die Driisengange derselben vereinigen sieh mit den anderen zu 
gemeinsamer Miindung in das Atrium.' The figure, however (fig. 5 pr.), does not 
show these canals in by any means a convincing fashion, and there is no indication 
of them in the figures of any authors with whose papers I am acquainted. 

Vejdovsky states that they develop as an outgrowth of the lining of the terminal 
chamber, and are thus products of the epidermis. 

The distal extremity of the spermiducal gland is modified into a penis, enclosed 
by a sheath; the lining epithelium here is not ciliated. The penis itself is a direct 
continuation of the spermiducal gland. The sheath has an epithelial lining, and, 
outside of this, muscular walls, which have frequently a markedly spifal arrangement. 
The lining epithelium of the sheath secretes a chitinous layer, which encloses the 
penis. The development of this copulatory organ has been described by .Vejdovsky 
(24, p. 143), and that briefly^. He says 'that the cavity of the penis arises by 
a secondary invagination ; the sheath first appears, and then, after being invaginated, 
grows out again to form the penis. The penis and its sheath are therefore something 
superadded to the spermiducal gland, the original orifice of which would be at a in 
the figure.' 

This description refers to L. udeheniianus ; it applies, however, to the two other 
European Limnodrilus in every point, except that they have a double coating of 
spiral muscles, the outer coat reaching beyond the penis-sheath for some distance 
along the narrow terminal region of the tube. 

There are several other genera of Tubificidae in which the efferent apparatus 
only difiers in detail from Limnodrilus. In Spirosperma there is a shoi-t strongly 
chitinous penis, and only a very slight development of muscles round the penis-sheath. 
So, too, Lophochaeta (fig. 38), where the chitinous sheath of the penis is less strong. 
These two genera have been described by Stolc (3). 

' The account given by Eisen (12) is fuller, but substantially the same, apart from the curious views 
as to the 'oviduct,' for which see p. 238. 


In HetnitiMfex and Psammoryctes the upper end of the spermiducal gland, which 
is wider in most Tubificidae than the distal region, is more markedly dilated, and 
is constricted oft' from the rest to form a spherical chamber, receiving the vas deferens 
and prostate. This has been called the 'vesicula seminalin.' 

Bothrioneuro'ti (fig. 40) has a peculiar efferent apparatus. The terminal chamber 
is wide, and is continuous above with the spermiducal gland, which is enveloped in a 
glandular sheath ; about halfway down the prostate opens ; this prostate {paratrium 
Vejdovsky calls it) is rather different in structure from the prostate of other Tubificidae. 
It consists of two parts : (i) of a pear-shaped diverticulum of the terminal chamber, 
lined with tall cells ; (2) of a group of thicker cells, attached to the blind extremity 
of this. The two efferent ducts open (in B. vejdovskyanuvi) by a common median 
pore, a common enough arrangement in the family Eudrilidae, but rare in the aquatic 
families. Shortly before this common opening are (in B. veJclovsJcyanum) two diverticula, 
one containing penial setae. Bands of muscles are attached to the terminal part of 
the chamber, which is possibly protrusible. There is no penis as in the genera 
Tubifex, Spirospervia, &c. 

The terminal ■ efferent apparatus of Tubifex is like that of Limnodrilus ; but the 
penis has a more complicated structure which has been especially studied by Nasse, 
DiEFFENBACH, and Vejdovsky (24). The latter has described its development, which 
is as follows. The first appearance of the spermiducal gland is 'a spherical invagi- 
nation of the integument into which opens the vas deferens. This sac then lengthens, 
and the prostate arises from a thickening of the ciliated lining epithelium. The 
portion of the terminal chamber which Jies nearest to the skin then becomes, 
constricted off, and its epithelium breaks - up ' into numerous elliptical and bright 
elements, which completely fill the cavity of the sac, so that only a with difficulty 
visible canal remains to put the newly formed chamber into communication with 
the exterior.' The further development of these bodies was not traced; but the 
penis and its sheath is probably a product of this distal part of the terminal 
chamber. In the adult worm the penis has a very complicated structure. As in 
Limmodrilus the canal of the penis is a direct continuation of the atrial canal, 
and there is a penis- sheath lined with epithelium and covered with muscles 
externally; this seems to have undergone a second* folding, so that the retracted 
penis is enclosed by two sheaths; the inner of these secretes a thick, elastic, and, 
apparently, chitinous membrane, which is inserted on to the penis where the 
epithelium — the matrix of the chitinous membrane — becomes continuous with the 
epithelium of the penis itself. In Tubifex and in Spirosperma also the actual 
extremity of the penis is formed by a very thick mass of cells surrounding the 

H h 2 



almost obliterated lumen. None of these genera possess penial setae, which are only- 
found in Ilyodrilus and Bothrioneuron ; indeed the ventral setae of the eleventh 
segment, upon which the male pores are situated are altogether absent from the ■ 
sexually mature worm. It is not altogether easy to compare tho efferent apparatus 
of the various genera of Tubificidae among themselves. 

Fig. 40. 

Fig. 38. 


(After Stole.) 

I. Funnel. 2. Penis. 3. 
'Atrium.' 4. 'Prostate.' 
5. Coiled sperm-duct. 

Fig- 39- 



(After Stole.) 

I. Funnel. 2. Sperm-duct. 3. Penis. 
4. Muscles for its retraction. 5. 

Glandular covering of 6. 'Atrium.' 



(After Stole.) 

I. Proximal region of 'atrium.' 2. ' Para- 
trium.' 3. Distal region of ' atrium.' 4. Penial 
setae. 5. External pore. 6. Muscles for pro- 
trusion of distal end of efferent apparatus. 
7. Muscle attached to sperm-duct (8). g. Funnel. 

The simplest form is unquestionably that of Ilyodrilus, which leads to the Naids 
and Lumbriculidae. 

If it were not for Branchiura, the more specialized chamber of Tubifex, &c., 
might be regarded as produced by further diiferentiation of the simple globular sac 
of Ilyodrilus, particularly since, as Vejdovsky pointed out, the complex terminal 


chamber, with its appendages of Tubifex, originates as a simple sac, comparable 
to that in the lower forms. 

In Branchiura, the sac which clearly corresponds to the terminal chamber of 
Ilyodrilus, no longer receives the sperm-duct ; it has become merely a caecal 
diverticulum of the sperm-duct, which is specialized near to its termination into 
two sections, one ciliated, the other without cilia. The function of the vas 
deferens, with this terminal tube, is marked by a distinct narrowing of the vas deferens, 
and by a change in minute structure ; at the junction opens the glandular sac. 

It seems, therefore, to be possible that the tube lying between the end of the vas 
deferens and the exterior, and which corresponds to the duct of the globular sac in 
Ilyodrilus, is the equivalent of the entire copulatory apparatus in other Tubificidae. 
If this be so, it is then possible that the prostate of these Tubificidae is the homologue 
of the glandular sac of Branchiura, much reduced in importance and, except in 
Bothrioneuron, with a. nearly or entirely obliterated lumen. However, it is more 
probable that the cells of the prostate of Tubifex represent a part only of the 
peripheral glandular layer of Branchiura. 

As Benham (15) and I (80) have pointed out, the prostate in Tubifex is simply 
to be regarded as a portion of a continuous glandular covering such as exists in 
Branchiura and Ilyodrilus; an intermediate stage is offered by Telmatodrilus, in 
which genus the continuous glandular covering has already broken up into several 
separate glands. This probable explanation of the various conditions of the prostate 
in the Tubificidae does not do away with the possibility that the 'paratrium' of 
Bothrioneuron is the homologue of the large chamber in Branchiura; but it seems 
more probable that this ' paratrium ' is a new structure. 

In no Tubificid is there more than a single pair of spermathecae ; and these organs 
are always (except in Hesperodrilus, where they are in the eleventh) situated in the 
tenth segment, i. e. that which also contains the testes. The spermathecae never have 
any diverticula, except a rudimentary one in Hesperodrilus branchiatus, but in a few 
species there are glands appended to the base of the pouches. Such glands, which are 
probably to be regarded as of epidermic origin, comparable to the capsulogenous 
glands of the genus Perichaeta, are found in Hemituhifex insignia, and in Ilyodrilus 
sodalis and Psammorydes barbatus. In the latter worm the appendages are connected 
with a sac containing a single copulatory seta. The spermathecae of the Tubificidae 
differ somewhat in form in the different genera ; in Ilyodrilus they are globular 
and sessile; in the great majority of species there is a division of the spermatheca into 
a pouch and a duct leading to the exterior; in Lophochaeta, for example, there is 
a chano-e in the character of the lining epithelium which marks the commencement of 


the duct ; in many other forms this is still further emphasized by a constriction which 
divides the two regions of the spermatheca referred to ; this constriction is especially 
well seen in Clitellio, in which genus both Clapaeede and myself have figured it. 
In Hemitubifex insignis Eisen has figured the spermatheca as if there were cilia in 
the duct, but there is no mention of this in the text ; they may be intended to 
represent spermatozoa; cases where the spermathecae are ciliated are extremely 
rare, and, perhaps, a little doubtful. However, Nassb distinctly asserts the presence 
of cilia in the duct of the spermatheca of Tuhifex: 'Im Ausfuhrungsgange tragt das 
Epithel eine Cuticula und flimmert.' On the other hand, Vejdovsky has denied this 
ciliation, nor is' it figured by Stolc (3). ■ Between the peritoneal coat and the lining 
epithelium, two muscular layers are developed, particularly at the end of the organ 
nearest to the opening ; this region of the spermatheca can often, as in Tuhifex, be 
extruded. The spermathecae often, as in Clitellio and Hesperodrilus, extend through 
several segments. Bothrioneuron has no spermathecae at all. The spermatophores 
have been already described. 

The oviducts of the Tubificidae were first discovered by Stolc (4) ; this naturalist 
found them in the genera Ilyodrilus and Psam'moryctes ; a little later I found these 
organs in Clitellio and in Hemitubifex; later still Stolc figured the oviducts of 
Bothrioneuron ; finally, I have found them in the genera Branchiura and Hesperodrilus. 
There is now a considerable probability that in all Tubificidae there are a pair of 
oviducts opening by a comparatively large funnel into the eleventh segment (or 
twelfth in Hesperodrilus), and on to the exterior on the boundary line between this 
• segment and the following. 

Previously to the discovery of the true oviducts a :^ost curious view was prevalent as to the 
nature of the oviducts of the Tubificidae, which is of interest as an instance of the persistence 
of an error that had nothing in particular to recommend it ; the view in question was in fact 
neither probable nor ingenious. They were originally believed to be connected with the male 
efferent apparatus; d'Udekem (6) described the egg-sac as a 'matrix,' communicating, on the one 
hand, with the ovaries, and, on the other hand, opening into the 'cloaca,' which was the name 
given by him to what we now call the atria ; d'Udekem speaks of the ' matrix ' opening into the 
cloaca, but in the next paragraph he states, and his statement is illustrated by a figure, that the 
wall of this sac forms the outer wall of the cloaca ; hence the oviduct, according to this view, 
surrounds the atrium, which is thus invaginated in it. CLAPARiiDE (2) spoke of this description as 
being 'sans doute exacte '; but he limited his support of d'Udekem's statements to the enclosure of 
the atrium within the oviduct ; he denied any connection between the oviduct and the egg-sac. 

Vejdovsky (13) originally ranged himself on the side of d'Udekem and CLAPAKiiDE, and 
supported their views by his discovery of eggs passing out of the supposed oviducts ; later, however, 
he gave reasons for believing that these bodies were not ripe eggs at all, but immature egg-cells 
detached from the ovary by the compression of the cover-glass, and abandoned the belief that 
the oviduct of the Tubificidae was in the position assigned to it by his predecessors. Grranting 


that the oviduct was as described by d'Udekem and Clapaeede, Vejdovsky pointed out that 
it was difficult to understand the fixation of the penis; nor was he able to see the complicated 
layers round the penis, which were regarded by the authorities mentioned as representing the oviduct. 
EiSEN (12), indeed, who accepted the views in question, still further increased the complexity of 
the subject by distinguishing between Tubificids with a single and Tubificids with a double oviduct ; 
later, Eisen abandoned the distinction between the two kinds of oviducts, having discovered ' a minute 
penis-sheath ' in Telmatodrilus, which was the only representative of the former class. Lankesteb 
threw doubts upon the interpretation of a part of the penis as an oviduct ; and (also previously 
to the publication of Vejdovsky's great work) did Nasse. The latter asked, as did Vejdovsky, as 
to the whereabouts of the attachment of the penis to the body-wall, if it was invaginated in the 
oviduct ; he also considered that the end of the oviduct and the commencement of the penis were not 
clearly indicated by this way of supposing the oviducts to be outside the penis. Dieffenbach has 
devoted three or four pages of his studies upon the Limicolous Oligochaeta to a consideration of this 
question ; he believed that ' most zoologists who have dealt with this question have been led along 
a wrong path by an incorrect observation of d'Udekem"s.' Some of Dieffenbach's criticisms are 
reallj- beside the point ; he asks, for example, how it is that the ripe ova which lie in the sixteenth and 
seventeenth segments can press their way forwards to the eleventh ; they must get nearly as far, in any 
case, for the true oviducts are between segments xi/xii. Their large size, too, would be as great an 
obstacle to their passage out by the real oviducts as by the supposed oviducts ; Dieffbnbaoh thought 
that the exit of the ova was by a partial tearing of the skin, an occurrence which he observed more than 
once ; Vejdovsky, by treating the living worms, which were kept under observation the whole time 
with chemical reagents, noticed the ova to pass out between segments xi/xii ; he is entitled, therefore, 
to the credit of having first discovered that the supposed oviduct of d'Udekem, CliAPABiDE, and 
Eisen, performed no such function, but that the ova escaped from the body by the position just 

The form of the brain varies considerably in different genera of Tubificidae, and 
appears to be always complicated. The most distinctive feature is, perhaps, the 
anterior median process which is sometimes (as in the genus Ilyodrilus) a mass of 
cells continuous with the brain, and sometimes (as in Bothrioveuron) consists of 
a median nerve communicating with a small ganglion placed a little way in front 
of the brain (to be compared perhaps with the buccal ganglia of the MoUusca). 

The circulatory system, which has been principally investigated by Stolo (3), is 
more complicated than in either the Lumbriculidae or the Naidomorpha. In many 
features it recalls the cii'culatory system of earthworms. The dorsal vessel is 
contractile, and runs from end to end of the body, on the dorsa;l side of the 
alimentary tract. Branchiura is the sole exception to this rule, and a very curious 
exception; the dorsal vessel in this worm is only dorsal in position as far back 
as about the tenth segment, from this point to the posterior extremity, it lies 
below the intestine, side by side with, the ventral vessel. Anteriorly the dorsal 
vessel lies well above the oesophagus; in the intestinal region it is covered by the 
peritoneum of the intestine. Besides the dorsal vessel a good many Tubificidae 
have a supraintestinal vessel which has precisely the relations of the corresponding 


vessel in earthworms. It lies in close contact with the oesophagus, beneath the 
peritoneum covering this organ. A supraintestinal vessel is present in the genera 
Tubifex, Psammorydes, Spirosperma, Loplwchaeta, Limnodrilus, Bothrioneuron, and 
Branchiura; it is absent in Ilyodrilus; with regard to the remaining genera there 
is no information. The supraintestinal vessel, as in many, if not in all, earthworms, is 
limited to the oesophageal region ; it dies away gradually in front and behind. 

The ventral vessel is present in all Tubificidae, and is quite free from the wall of 
the alimentary tract. It is (apparently always) non-contractile. Most Tubificidae (the 
genera Bothrioneu,ron, Lophochaeta, Tubifex, Psammoryctes, Spirosperma, and Limno- 
drilus) possess also a subintestinal vessel, corresponding, on the ventral side of the 
oesophagus, to the supraintestinal on the dorsal side. In Bothrioneuron and 
Lophochaeta, at any rate, this vessel communicates in front with the ventral 
vessel. The subintestinal vessel may, perhaps, be the equivalent of the intestino- 
tegumentary trunks of earthworms. 

The dorsal and ventral vessels communicate, in nearly every segment of the 
body, by perivisceral arches. In the anterior segments of the body these are, 
as a rule, larger, and some of them are specially dilated, and constitute the so-called 
hearts. This, again, is a character met with in earthworms, as is also the contractility 
of some of the anterior perivisceral arches. In Telmatodrilus and Branchiura five 
or six pairs of these arches are contractile ; these, in Telmatodrilus, gradually increase 
in calibre in successive segments from before backwards ; in Branchiura those of 
ix and x are specially dilated. In the latter genus segment viii contains, in addition 
to the perivisceral loops connecting the dorsal and ventral vessels, a vascular arch, 
uniting the supraintestinal with the ventral vessel. In other Tubificidae there are one 
(Tubifex, Lophochaeta) or two (Limnodrilus, Bothrioneuron) contractile trunks, which 
pass round the oesophagus and unite the supraintestinal with the ventral vessel. The 
occurrence of these vessels is highly interesting. They evidently correspond to the 
'intestinal hearts' of earthworms (Pontodrilus, Perichaeta, &c.) which put into 
communication the supraintestinal and ventral vessels (being sometimes, moreover, 
also connected with the dorsa;l vessel) ; and it will be noted that, as in earthworms, 
they are the last of the series, and are specially dilated. 

In all Tubificidae the alimentary tract is surrounded by a network of blood- 
capillaries, which are derived, anteriorly from the supraintestinal, and posteriorly 
from the dorsal, vessel. In Ilyodrilus, as in the Naidomorpha, the intestinal 
network communicates with the ventral vessel by a pair of trunks in every segment. 
In other Tubificidae there seems to be no connexion between the intestinal network 
and the ventral vessel. 



Very characteristic of the genera Ilyodrilus, Branchiura, and apparently Telmato- 
drilus, is the existence of 

an integumental network 
of which also traces exist in 
Lininodrilus. This, again, 
is a character which brings 
the Tubificidae into relation 
with the higher Oligochaeta. 
For the details of the circu- 
latory system the reader is 
referred to the descriptions 
of the several genera. 

The two species, Bran- 
chiura sowerhii, and Hes- 
perodrilus hrancMatus are 
remarkable for the possession 
of branchial processes ; these 
are paired structures, either 
dorsal and ventral [Bran- 
chiura), or lateral (Hespero- 
drilus) in position. They 
are found only upon the 
posterior segments of the 
body. Their structure has 
been dealt with above. 

The Tubificidae may be 
divided into two sections, 
according as to whether the 
brown ovoid gland, so charac- 
teristic of the Naids, is or is 
not present. In the following 
genera it is present : — 


Fig. 42. 

Fig. 41. 




(After Stole.) 

I. Ventral vessel. 2-6. Vessels 
joining intestinal network. 7. Slen- 
der posterior part of ventral vessel 
just after it joins; 12. Snb-intestinal 
vessel. 9, 13. Intestinal hearts. 15, ■ 
16. Commissural vessels. 17. Dorsal 
vessel. , 



(After Stole.) 

I~IX, Segments. 1-6. Lateral 
hearts. 7. Ventral vessel. 8. Pha- 
rynx. 9-14. Vessels supplying in- 
testinal network. 15. Sub-intestinal 
vessel. 16-19. Vessels supplying in- 
testinal network. iS, 21. Intestinal 
hearts. 22, Supra-intestinal vessel. 
23. Dorsal vessel 


In all the others it is absent ; 

Stolc has figured these structures in the two genera mentioned, and his figures are 

I i 



repeated here (woodcut, fig. 43) ; within this brown ovoid body the lumen of the 

Fig. 43- 



After Stole. 

I. External pore 2. Vesicle. 3. Peritoneal coat. 
4, 5. Glandular body. 6. Funnel. 7. Ampulla. 
8. Muscular bands. 

nephridium divides into two channels, 
which give off a few branches ; these, 
however, are not represented as anastomos- 
ing ; the presence of this modified section 
of the nephridium afiines the two genera 
in question to the Naids and to the Lum- 
briculidae. Although this glandular body 
is only properly developed in the two 
genera mentioned above, there are traces 
of it, according to Stolc's figure (3, Tab. ii. 
fig. a), in, at any rate, one species of 
Livmodrilus, viz. L. claparedianus ; in 
this worm the lumen of the nephridium 
immediately following the funnel is wider 
than it is behind and in front of this 
point, and is at the same time somewhat 
tortuous in its course (woodcut, fig. 43, B 4) ; 
there is, however, no network. The two 
genera, Ilyodrilus and Bothrioneuron, are 
further to be distinguished by the fact 
that in neither of them is there an ampulla 
such as occurs in Limnodrilus and (even 
more prominently) in Spirosperma ferox 
(EiSEN, 12, PI. ii. fig. ag. + ) ; this ampulla 
is found in many earthworms. 

Genus Tubifex, Lamarck. 

*%)). Lumbricus, 0. F. Muller (in part.). 
Saenuris, Hofpmeister. 
Strephuris, Leidy. 
Blanonais, Geevais. 
Nais, Oken. 
Definition. Dorsal seta-bundles with capilliform, pectinate, and uncinate setae ; 
ventral bundles with uncinate setae only. Dilated hearts in the eighth segment. 
1^0 chitinous penis. 


This genus, which may be taken as the type of the family, has heen principally 
investigated by d'Udekem (6), Claparede (2, 3), Vejdovsky (24), Macintosh and 
Stolc (3) ; these investigations relate entirely to T. nvulorum, the best known 
species. The presence of capilliform setae in the dorsal bundles distinguish this 
genus from many other genera of Tubifieidae, such as Limnodrilus, which has been 
by some authors confused with it. Besides the capilliform and uncinate setae, the 
first fifteen segments of the body contain pectinate setae much like those of Psam- 
moryctes, but with the additional prongs less marked. Towards the end of the body 
the capillifonn setae of the dorsal bundles disappear. The ventral seta bundles 
contain only uncinate setae. The cerebral ganglia have well marked^ lateral lobes 
arising anterolaterally; posteriorly the brain is trifid with two larger lateral and one 
smaller central division. 

A second species of the genus has lately been described by Vejdovsky (8) ; this 
worm differs in several particulars from the type species of the genus ; it is indeed 
a little doubtful whether it is rightly included by Vejdovsky in the genus Tuhifex. 
The principal external difference is in the absence of capillary setae ; as, however, 
these are wanting in the posterior part of the body of Tuhifex rivulorum, the 
difference is not perhaps so great ; besides we know that Hemitubifex may, or may 
not possess capilliform setae in the dorsal bundles. The sexual orifices are stated by 
the describer of the species to be exactly as in Tuhifex rivulorum ; according to his 
figure, however, (8, PI. xv. figs. 2, 3) the male-pores are upon the tenth segment 
instead of the eleventh, as in T. rivulorwrri; there is clearly no slip of the pen here, 
for in the description of one of the figures (fig. 4), the author expressly mentions 
the tenth segment as being that upon which the spermiducal glands open. Another 
external difference from Tuhifex rivulorum is in the presence of 'penial' setae near 
to the orifices of the sperm-ducts and spermathecae ; these exist in many Tubifieidae, 
but not in Tuhifex rivuloruvi ^- As to the form of these setae Vejdovsky is unable to 
give any details ; the extremities were broken off. In other respects there are no 
differences of more than specific value between the two species. 

As to the possibility of Ilyodrilus sodalis being a Tuhifex, I refer to the matter 
later. It differs from the other species, if it is con-ectly referred to this genus, in 
having pulsating vessels, not dUated, in segments viii-x. 

' This description is from Vejdovsky. 

^ It must be remembered, however, that there is not certainly a difference here. Vejdovsky mentions that 
the sexual setae of this worm were broken. Now in TuUfex rimdorum there are the usual ventral setae at 
the male pores as at the female pores ; it is quite possible, therefore, that the setae of T. blanchardi are merely 
ordinary setae. ClapabJide (2, p. 23) refers to the existence of setae upon the segment bearing the male pores, 
which are placed a little in front of the setae. I have found them to persist in a species from New Zealand, 
which appears to differ in no way from T. rimdorum { = T. bonneti). 

1 i 2 


A very large number of species have been assigned to this genus, the names of 
most of which will be found as synonyms of species now known to belong to other 
genera, principally the littoral genus Clitellio ; Tuhifex marinus of Lamaeck ( = .£um- 
bricus tubicola of Mullek) is, according to Vatllant, a Clymene. 

The genus Tuhifex is met with in Europe, North America, and New Zealand. 

(i) Tubifex rivulorum, Lamarck. 

Iiumbrieus tubifex, 0. F. Mullek, Verm, terrestr. 1774, p. 27. 
Tubifex rivulorum, Lamarck, Hist. An. sans Vert. 1816, iii. p. 225. 
PStrephuris agilis, Leidy, Journ. Acad. Nat. Sci. Philad. 1850, p. 45. 
Nais tubifex, Oken, Lehrb. d. Naturg. 1815, Pt. i. p. 364. 
„ filiformis ', DuGES, Ann. Sci. Nat. 1838, p. 286. 
,, sanguinea, DoYiiKE, Mdm. Soc. Linn. Norm. 1856, p. 306. 
Tubifex bonneti, CLAPAEiDE, M^m. Soc. Phys. Gen. 1863, p. 330. 
Saenuris variegata, Hoffmeistee, De verm. quib. ad gen. Lumb. pert. Berlin, p. 9.. 
„ tubifex, Johnston, Cat. Worms, 1865, p. 64. 
„ sp. Gegenbaur, Z. wiss. Zool. 1852, p. 237. 
Blanonais filiformis, Gervais, Bull. Acad. Koy. Belg. 1838, p. 16. 
? Saenuris taurica, Czeeniavsky, Bull. Soc. Nat. Mosc. 1880, p. 333. 
? „ peeuliaris, Czeeniavsky, Bull. Soc. Nat. Mosc. ] 880, p. 333. 
? „ diversisetosa, Czerniavsky, Bull. Soc. Nat. Mosc. 1880, p. 334. 
? Tubifex eampanulatus, Etsen, Bih. Svensk. Akad. 1879, no. 16, p. 16. 

Definition. Brain concave in front with well-marked lateral lobes, posteriorly trifid with 

two longer lateral and one smaller central division. Setae of dorsal-bundles, capilliform, 

pectinate, and uncinate. 

As will be seen this species lias a longer list of synonyms appended than almost any other 
Oligoohaet. A strict adherence to the rules of Zoological nomenclature would necessitate the 
alteration of the commonly accepted name to Tubifex tubifex, if, that is to say, Lamaeck's Tubifex 
rivulorum be identical with 0. F. Muller's Lmnbricus tubifex''. On the assumption that there is 
more than one species of Tubifex in Europe, which is at present far from being proved, there are 
no positive reasons either for affirming or denying the identity of species described by the older 
writers with Tubifex riimlorum of d'Udekem, Vejdovsky and others ; there is, for example, nothing 
against their identification with Eisen's T. eampanulatus. It is perfectly clear, however, from 
0. F. MiJLLEK's description cited above that he is dealing with a Tubifex from his reference to 

' It is not at all clear to what species Williams (1) refers under this name in describing the vascular 

' CzERHiAVSKY (p. 330) does not allow Lamarck's Tubifex rivulorum to be the same as d'Udekem's, 
nor, in spite of the description of the tubes, Muller's L. tubifex. L. lineatus of the latter author is 
identified with T. rivulorum, d'Udekem. 


Bonnet, who figures a recognizable Tubifex, and from the sentence 'seta antice utrinque porrecta,' 
which I take to refer to the capilliform setae of anterior segments, and from his description of the 
tubes fabricated by the worm (also referred to in another work, 2, p. loo, note 28, and p. 102, note 29). 
No genus of Tubificidae, except Tubifex, has been stated to fabricate a tube (see, however, Emholo- 
cephalus). CLAPAKiiDE's species T. Bonneti is, as has been pointed out by Nasse, Vejdovsky, and 
others, identical with T. rivulorum ; the principal point of diiference used by Clapae^ide in 
distinguishing the species, viz the position of dilated contractile hearts in viii. instead of vii. is 
no doubt due to a wrong enumeration of the segments on the part of d'Udekem. Dieppenbach, 
though of opinion that the two 'species' are the same, uses the name T. Bonneti, which is hardly 
permissible. T. rivulorum, described by Macintosh, comptises, as he himself has pointed out, two 
■species; one of these I believe to be an Uyodrilus; the larger ('from the lakes' of Scotland) is, 
judging by the figure of the setae (pi. ix, fig. 3), identical with Tubifex rivulorum. Leidy's Strephuris 
agilis is doubtfully included by Vaillant ^mong the synonyms of T. rivulorum. Vejdovsky (24, 
p. 45) regards it as 'incertae sedis' and suggests its probable identity with one of the North 
American Tubificids described by Eisen (12). The position of the clitellum ('posterior to the 
ninth articulation'), and the form of the setae (figs. 6, 7) taken together argue that it is a 
Tubificid of some kind; the shortness of the oesophagus points in the same direction — but in 
Tubifex itself the oesophagus occupies only segments iii. and iv. (d'Udekem, Vejdovsky, Nasse), 
whereas in Strephuris it is said by Leidy to reach th« sixth segment. I therefore repeat Vaillant"s 
query in my list of synonyms. Vaillant includes among the synonyms Tlyodrilus coccineus, in spite 
of Stolc's papers upon the structure of this worm, one of which is included by Vaillant in his 
list of literature, and indeed partly abstracted (Vaillant, p. 349). I do not include /. coccineus 
among the synonyms of T. rivulorum. With regard to SaeHuris taurica, S. peculiaris, and S. diver- 
sisetosa (with two vars.), I am quite of the opinion of Vaillant that they do not differ from 
Tubifex rivulorum as they are described by Czeeniavsky. At the same time I may point out that 
there is nothing in the latter naturalist's definitions' which militate against their transference to 
the genus HemituUfex or Uyodrilus; T. diaphanus of Tatjbee, and T. longicauda of Kesslee are 
also possibly synonymous ; as regards the former it is stated that the uncinate setae are ' nonpalmate ' ; 
this probably refers to the pectinate setae of Psammoryctes umbelifer. If the slight additional processes 
found in T. rivulorum were absent from this species it is probably an Uyodrilus or HemituUfex. 

{%) Tubifex blanchardi, Vejdovsky. 
T. blanchardi, Ve.jdovsky, M^m. See. Zool. Fr. 1891, p. 596. 
Definition. Length about 35 mm.; number of segments, 6%. Setae of all bundles uncinate only. 
Hab. — Algeria. 
The setae of this worm, though they are everywhere uncinate, differ slightly in 
different regions of the body. The dorsal setae of the anteclitellian segment are 
equidentate, but there is occasionally a median denticle. The dorsal setae of the 
postclitellian segments have the superior denticle longer than the inferior ; the same 
is also the case with the ventral setae of the same segments, but the ventral setae 
of the anterior segments are exactly the reverse. Sexual setae accompany the 

' They principally relate to minute differences in the form of the uncinate setae. 


orifices of the spermathecae and the atria, but there is no description to hand of 
their forms. That they differ from the ordinary setae whose place they take seems 
probable ; in any case, those accompanying the male pores are arranged differently 
from the ordinary setae ; Vkjdoysky figures them (8, fig- 4) as lying one in front 
of the other. 

Genus Clitellio, Savigny. 

Lumbricus, 0. F. Mullee, &c. (in part.). 
Peloryctes, Leuckaet. 
Tubifex, d'Udekem (in part.). 
Pododrilus, Czerniaysky (in part.). 
? Psammobius, Levins en. 

Depinitioit. Uncinate seta only present. Dilated hearts in VIII, IX. No prostates. 
Spermatheca occupies four segments —X-XIII. 

In spite of Clapaeede's account of the structure of this genus and of Lininodrilus 
they are confounded by Vaillant (6), who leaves out from his generic definition, 
as well as from the description of the species, Gl. arenarius, the characteristic difference 
between the two genera, viz. the absence (Clitellio) or presence (Limnodrilua) of 
prostate glands appended to the effei'ent apparatus, which he does not consider of 
sufficient importance to be a generic character. In an earlier work (3) Vatllant had 
already taken up this position, which was accepted by Czeentavsky. My own investi- 
gations upon Clitellio (70) being entirely confirmatory of Clapaeede, besides adding 
a few details, emphasizing the necessity, I think, of keeping Clitellio and Limnodrilus 
separate. I showed, however, an additional point of resemblance between the two 
genera, i. e. the presence ' of two pairs of specially dilated ' hearts.' In the following 
table Limnodrilus and Clitellio are compared. It should be mentioned that CI. 
arenarius is the only species whose anatomy is known'. 


Setae. — Entirely uncinate. Ditto. 

Hearts. — Two pairs in vii, viii, both en- Ditto in viii, ix, last only enlarged (not 

larged, stretching through four segments in all species), limited to the tenth seg- 

when mature. ment. 

Integumental blood-capillaries. — Absent. Present. 

Vas deferens. — Comparatively short. Long. 

Spermiducal gland, — Without prostates. "With prostates. 

' 'Clitellio atcr' is not ii Clitellio (see below',. 


In a species of Limnodrllus from New Zealand there are two pairs of dilated 
hearts in precisely the same segments (viz. viii and ix) as those in which these 
hearts occur in Glitdlio. This lessens the gap between the two genera. It should, 
however, be mentioned here that the specimens of Liimnodrilus novae-zelandiae 
examined by me were sexually immature. It is therefore, perhaps, begging the 
question to refer them to the genus Limnodrilus, which I do on account of their 
fresh-water habitat. 

Vaillant (6) enumerates nine species of ClitelUo of which some are apparently not referable to 
this genus ; ClitelUo henedii is, as is pointed out elsewhere, a Hemituhifex ; Michaelsen (5) assigns 
ClitelUo linealus to the genus Pachydrilus. I discuss this view below. So, too, ClitelUo minutus. 

ClitelUo inaequalis, described by 0. F. Mtjllee as ' Lumbricus' inaegualis, and doubtfully referred 
by Geube (8, p. 104) to this genus is a difficult species to come to any conclusion about. All that 
MtJLLER says about it is — ' Papillis lateralibus simplicibus ; setis solitariis.' Vaillant justly remarks 
that these characters are 'trop succinctement donnes' to enable any conclusion to be drawn. The 
papillae obviously suggest Hemitubifex henedii. ^ 

ClitelUo neurosoma of Fkey and Leuokakt (p. 150) was referred by d'Udekem (1, p. 545, and 
5, p. 12) to the genus Lumbriculus — 'a cause du developpement des vaisseaux lateraux du aysteme 
circulatoire.' These were described by the authors who named the species as being very long and 
much convoluted, especially in the anterior segments. 

Clitellio tenuis was described by Leidy as a Lumbriculus (4, p. 148); Vejdovsky (24, p. 51) 
places it among the Lumbriculidae but as 'incertae sedis.' Vaillant (6, p. 421) doubtfully includes 
it in the genus ClitelUo. The position of the genital orifices (? male-pores) upon the tenth segment 
(ninth in Leidy's enumeration) appears to justify its being put in the family Tubificidae, when 
taken in connexion with the fact that the setae are uncinate and 3-6 in a bundle. The habitat 
as Vaillant remarks is also in favour of the correctness of this identification. 

ClitelUo irroratus of Veekill (p. 324 and 622), called also a ClitelUo by Vejdovsky (24, p. 45), 
queried by Vaillant (6, p. 422), is I imagine not a Clitellio ; as Vaillant points out the presence ■ 
of capilUform setae in the dorsal-bundles is against this identification ; the fact that these setae 
are not always present suggests the genus Hemitubifex. 

Clitellio dubius of Czebniavsky (p. 327) might be almost anything; it is only known from a 
fragment of the hinder end of body. There remains only— 

Clitellio arenarius, Savigny. 

Clitellio arenarius, Savigny, Syst. d. Anndl., 1830, p. 104. 

Lumbricus arenarius, 0. F. Mullee, Zool. Dan. Prodrom. 1776, p. 216. 

Lumbricus littoralis, Bkuguieee, Tabl. Encj'cl. 1791^- 

Peloryctes arenarius, Ledckaet, Arch. f. Nat. 1849, p. 161. 

Tubifex hyaUnus, d'Udekem, Bull. Ac. Roy. Belg. 1855, p. 544. 

1 Fide Vaillant. 


Definition. Length up to 30 ot. ; number of segments, 100/ setae 3-6 in each hinrlle. 
Hah. — Coasts of Europe. 
Other characters those of genus. 

LiMNODRILlTS, Clapabede. 

Syn. Tubifex, Budge (in part.). 

Clitellio, Vaillant (in part.). 
Camptodrilus, EiSEN. 
Pododrilus, Czekniaysky (in part.). 
PAeestus, Leidy. 
Lumtariculus, Leidy (in part.). 

DEFiiriTiOBr. I"resh-water Oligochaeta with'uncinate setae only. Contractile hearts 

in VIII Or in VIII and IX. Perivisceral loops in posterior segments of body give 

oflF branches wMcli penetrate body- wall. Penis with chitinous lining ; prostates 


The genus Limnodrilus was first clearly defined by Claparede (2). It had been 

confused by Budge and d'Udekem with Tubifex. Budge, for example, figures the 

chitinous penis of Limnodrilus in his paper dealing with Tubifex rivulorum, while 

d'Udekem, writing on the structure of the same worm, remarks upon the occasional 

absence of capillary setae ; this appears to show that he had before him a Limnodrilus, 

'for this is one of the principal difierences between Limnodrilus and Tubifex^. I do not 

agree with Vaillant's inclusion of Limnodrilus ■^^ithin the genus Clitellio ; the reasons 

for thus dissenting have been stated on p. 247. On the other hand it seems probable 

(as Vejdovsky and Vaillant believe) that Eisen's Camptodrilus is not really separable 

from Limnodrilus. It is distinguished by Eisen (12, p. H98) only on the grounds 

of the spiral arrangement of the fibres which surround the penis. ' In other respects,' 

he observes, ' this genus resembles Limnodrilus.' The question whether this be of 

sufiicient importance as a generic mark need not be discussed, since Vejdovsky 

(24, pp. 4S, 143) finds the same spiral arrangement in Limnodrilus. 

As to Pododrilus of Czeeniavsky— the defective description of this author renders it impossible 
to be certain as to its identity (as Vaillant thinks) with Limnodrilus. Pododrilus is not referred 
to by Vejdovsky. Czerniavsky's definition is as follows : — 

'Fasciculi setarum omnes uncinis furcatis (4-2 vel 5-1) formati anteriores in duphcaturis pro- 
minentibus ("von einer kleinen Hautfalte getragen") positi Limicolae marinae.' 

' Apart from Tubifex hlancharcU (see p. 246 above). 


This description, 'von einer kleinen Hautfalte getragen,' suggests, as Vaillant points out (6) 
p. 413), the parapodia of the Polychaeta. I have, however, seen in Limnodrilus a series of virart-like 
protuberances, which bear the setae and may be the same structures. They certainly suggest 
rudimentary parapodia; for this reason I am inclined to admit Pododrilus as a synonym, not of 
ClitelUo, but of LimnodriluK, though the marine habit of Pododnlus suggests rather that it should 
be included with the former genus. 

Acestus of Leidy was regarded by the author of the genus as but little different from 
Lumhrimlus. His definition of the genus is as follows : — 

' Body vermiform. Podal spines in 4 rows ; anteriorly 3 to 8 in each fasciculus, posteriorly in 
pairs; long sigmoid, bifurcated at extremity. Upper lip conoidal, inarticulate. AnnuH under 100. 
Blood red. Eyes, girdle, and muscular stomach none.' 

Imperfect though this definition is, it clearly shows that Acestus is not a Lumbriculus (at least 
auctorum; it is identical with Leidy's Lumbriculus), or a Lumbriculid of any kind known. In the 
Lumbriculidae there are only two setae to each bundle. Its probable identity with ClitelUo or 
Limnodrilus) was first pointed out by Czeeniavsky (p. 326), who also very justly ranged 
Lumbriculus (Leidy nee Geube et plurimorum auctorum) under the same heading. In this he is 
followed by Vejdovsky (24, p. 45), but not by Vaillant, who places Acestus under Lumbriculus, 
Gkube, a proceeding which has no justification except on the view that Leidy, who was acquainted 
with the real Lumbriculus, omitted from his definition facts which would show its identity with 
that genus.' 

The species of the genus Limnodrilus, so far as they are known, are invariably 
fresh water in habitat. The principal character of the genus, which distinguishes it 
from all other Tubificidae except ClitelUo, is the presence of uncinate setae only ; 
it is distinguished from ClitelUo by the presence of a prostate and by the great length 
of the narrow and much-coiled vas deferens. The chitinous penis, which is longer 
in some species than in others, distinguishes Limnodrilus from Tubifex. A character 
which is peculiar to Limnodrilus is the presence of vascular tufts penetrating the 
epidermis ; these are given off (in the posterior region of the body) from the peri- 
intestinal vessels ; a branch arising from each of these loops, near the middle, dilates 
into a sinus covered externally with pigmented cells ; from this sinus four or five 
capillaries penetrate the integument as far as the middle of the epidermis, and then 
appear to end blindly. 

The three European species, L. udeJcemianus, L. koffmeisteri, and L. claparedianus, 

can be readily distinguished, but it is not always so easy to make out the American 

species described by Eisen (in 7 and 12). The form of the brain and the comparative 

length of the chitinous penis sheath appear to be the most salient characters ; but, as in 

some cases, there is not an agreement between Eisen's figures and those of other authors 

(compare for example the figures of the brain of Spirosperma ferox, given by Eisen 

and Stolc respectively), a certain amount of doubt necessarily arises as to the value 

which can be legitimately attached to these characters. A revision of the genus would 

probably reduce the species considerably. 

K k 


The genus Limnodrilus as above defined contains a considerable number of species 
from Europe and from N. America. I have received examples also from New Zealand 
and from Hawai^. The following twelve species are perhaps recognizable : — 

Limnodrilus hofftneisteri, Clap., Europe. 









udekemianus, Clap., Europe. 
elaparedianus, Eatzel, Europe, 
ornatus, Eisen, California, 
steigerwaldi, Eisbn, California. 
montieola, Eisen, California, 
alpestris, Eisen, California, 
silvani, Eisen, California. 
spiralis, Eisen, California, 
igneus, EiSEN, California, 
corallinus, ElSEN, California, 
californicus, EiSEN, California. 

A number of others are not recognizable. 

Tubifex elongatus, d'Udekbm (Bull. Ac. Eoy. Belg. t. xxii. 2nd pt. p. 544) from the definition— 
' Teguments transparents ; corps mince tres allonge ; des crochets fourehus dans tous les faisceaux. 
Habite les eaux douces' — appears to be a Limnodrilus. It occurs in neighbourhood of Brussels and 
is to be distinguished from Tubifex riimlorum with which it lives by absence of capilliform setae. 

In spite of the impossibility of recognizing this species from the description, Taubeb (1, p. 71) 
records its occurrence in Denmark. Levinsen (2, p. 225) includes it under L. udekemianus, and 
L. hoffmeisteri, following CLAPAEiiDE (2, p. 71, footnote 2), who considers it to be perhaps 
•synonymous with one or other of these species. This view is also held by Vbjdovskt (24, p. 44). 

' Clitellio suchumicus,' Czekniavsky, Bull. Moso. 1880 (p. 328) is probably a Limnodrilus as it 
has uncinate setae (3 in anterior bundles, 2 behind) and to others, and occurs in fresh water along 
with Tubifex. It is not worth while to repeat the long Latin diagnosis given by Czeeniavsky, as 
this contains no facts of importance save those mentioned. Czeeniavsky remarks that 'it is met 
with in great numbers in shallow river of Suchum, under stones, particularly where it crossed a 
road.' The few setae in each bundle may perhaps distinguish the species. 

' Clitellio heteroseiosus,' Czeeniavsky (ibid. p. 328) is not defined in such a way as to render 
its identification possible, though there seems to be no doubt about its being a Limnodrilus. In 
possessing six or seven setae in each of anterior bundles it is a more normal Limnodrilus than 
the last. 

'Nais gigantea,' Kesslee, Trud. Eussk. Est. St. Petersb., 1868, is regarded by Vejdovsky as 
a Limnodrilics. 

' Saenuris abyssicola,' Smith and Veekill, Am. Journ. Sci. Arts, 3rd ser. vol. ii. p. 449. This 
woi-m is stated to consist of twenty-eight segments; 'anus terminal with three or four slight lobes"*; 

1 These were collected by Mr. E. C. L. Pekkiks ; I cannot differentiate them from the American 

' This is perhaps a little suggestive of the branchiae of Branchiura. I observed one specimen of this 
worm in which there were only two or three close to the anus. 


there are five or six setae in each fascicle ' simple acute, slightly curved ' ; in one specimen there 
were four minute ocelli. 

This species, which is obviously insuflSoiently described, is included by Vejdovsky among the 
Tubiflcidae, and is doubtfully regarded by Vaillant (6, p. 433) as a Limnodrilus. It is not stated, 
however, that the setae are uncinate, and therefore Michaelsen may possibly be right (5, p. 50) 
in placing it among the Enchytraeidae ; he remarks with regard to this and the following species 
'diese beiden Verril'schen Arten miissen den Enchytraeiden zugeordnet werden, da sie einfach 
zugespitzte Borsten haben.' 

' Saenuris limicola,' Smith and Veeeill, Am. Journ. Soi. Arts, 3rd ser. vol. ii. p. 450, consists of 
forty-four segments; 6-8 setae in each fascicle anteriorly, 4-5 posteriorly; setae 'long, slender, 
curved, and acute.' 

This species, again referred to the Tubiflcidae by Vejdovsky, and to Limnodrilus doubtfully by 
Vaillant (6, p. 434), is put by Michaelsen among the Enchytraeidae. In favour of this placing 
of the species and the last, in addition to the form of the setae emphasized by Michaelsen, is 
the fewness of the segments of which the body is composed. 

'Saenuris vagans' Johnston, Cat. Worms, B. M., pp. 65, 353. This species is doubtfully included 
by Vejdovsky (24, p. 46) as a synonym of Tuhifex rivulonim. Vaillant, as I think, with more 
probability regards it as possibly a Limnodrilus. So far as the definitions given by Johnston, and 
in the appendix to the 'Catalogue of non-parasitical worms' by Baied go, and this is naturally 
not very far, this is the only conclusion that could be safely arrived at. The setae are, however, 
described as 'acute at the outer extremity,' and there is no mention of a bifid tip. It is probably 
this which has led Michaelsen to include the species^ (5, p. 53) among the Enchytiaeidae, as 
well as its small size (6'"), and the fewness of the segments (50) of which the body is composed. 

Tyhifex uncinarius of DuiJES possibly belongs here, and T. deserticola and Limnodrilus bogdunovii 
of Geimm. 

In conclusion, I may briefly refer to the existence of this genus in Hawai, where 
it is represented by a species which shows the remarkable spiral disposition of the 
muscular fibres round the spermidueal gland. The specimens were obtained from 
a spring on a mountain, but I cannot distinguish any marked peculiarities which 
justify me in giving it a new name, though on the other hand I have not yet 
identified it with any of the known forms. Living material is essential for the 
proper description of these Tubificidp. 

(i) Limnodrilus claparedianus, Ratzel. 

L. claparedianus, Ratzel, Z. Wiss. Zool. 1868, p. 590. 

Camptodrilus spiralis, Eisen, Bih. K. Svensk. Akad., 1879, No. 16, p. aa. 

Camptodrilus californicus, Eisen, loc. cit., p. a4. 

Tubifex rivulorum, BuDGE, Arch. f. Nat. 1850, p. i (in part.). 

Clitellio (Limnodrilus) claparedianus, Vaillant, Annelds, p. 434. 

^ But no mention is made of the cleft extremity of the setae of T. rimloram by Johnston ; the neglect, 
therefore, to mention this fact with reference to the setae of 'Saenuris vagans' is not an argument that 
the setae indicate the Enchytraeid family. 

K k a 


Definition. Length about 50-70 mm. Brain almost square with squarish deft behind ; 
anterolateral lobes, each, divided into two, one directed forwards the other backwards. 
Pharynx reaches into fifth segment. Chitinous penis eight to ten times as long as 
broad. Hob. — Europe; California; San Francisco; and Sierra Nevada. 7,000. 

I associate Eisen's two species, CaTnptodrilus spiralis and C. californicus, with 
Ratzel's LimTwdrilus claparedianus in deference to the opinion of Vejdovsky ; but 
the definitions given by Eisen appear to show differences between L. spiralis and 
L. californicus. 

As regards the European forms L. claparedianus differs from the other two by 
the much greater length of the chitinous penis sheath ; the worm itself is also much 
longer. These facts were pointed out by its original describer, Ratzel, though the 
same facts about the penis sheath can be gathered from an inspection of Budge's 
plate. DiEFFENBACH (p. 98) puts down the proportions between the length and 
breadth of the penis as sometimes 30:1, and states that in the length of this sheath 
there are intermediate stages between L. hoffmeisteri and L. claparedianus. Diefpen- 
BACH is, however, not accurate in saying that these two species in other respects 
show an identical structure, for they differ, among other points, in the form of the 
brain — as may be seen by the figures illustrative of this organ given by Vejdovsky 
(24, Tab. viii, fig. 14), and Stolc (3, Tab. i, fig. 7). The pharynx too extends farther 
back in L. claparedianus than in L. hofimeisteri. 

(2) Limnodrilus hoffmeisteri, CLAPAEtDE. 

L hofaneisteri, Clapaeede, M^m. Soc. Phys. Gen. 186a, p. 248. 
Clitellio (Limnodrilus) hoffmeisteri, VailIjANT, Annelds, p. 424. 

Definition. Length about 35 *^''^- ! number of segments 95. Brain square, with shallow 
excavation posteriorly. Pharynx reaches into third segment. Nephridia with vesicular 
cells. Penis six to seven times as long as broad. Hab. — Europe. 

This species has been chiefly described and illustrated by Clapaeede and Vej- 
dovsky (24), who thinks that Doy^jee described it under the name of Nais sanguinea. 

(3) Limnodrilus iidekemianus, Clapaeede. 

Ii. udekemianus, Clapaeede, Mem. Soc. Phys. Gen., 1862, p. 243. 
Clitellio (Limnodrilus) udekemianus, Vaillant, Annel^s, p. 425. 

Definition. Length about 60 mm. Brain with median and lateral lobes posteriorly, with paired 
anterior and anterolateral lobes. Pharynx reaches into fifth segment. Penis about three 
times as long as broad. Hab. — Europe. 


(4) Limnodrilus ornatus, Eisen. 

L. ornatTiSj EiSEN, Bih. K. Svensk. Akad., No. 16, 1879, p. 17. 
Clltellio (Limnodrilus) ornatus, Vaillant, Annel^s, p. 436. 

Definition. Length about 30 mm. Brain with shallow concavity posteriorly. Spermathecae 
pear-shaped, sometimes constricted in middle. Penis about five times as long as broad. 
Nephridia without large globular peritoneal cells. Hah. — California ; San Joaquim river. 

The ' principal characteristic ' of this species is, according to Eisen, ' the star-like 
concretions round the upper end of the penis sheath.' His figure (12, PI. ix, 8 d cm) 
shows that these bodies are attached to the upper extremity of the penis just where 
it becomes continuous with the penis sheath (' Penisscheide,' Vejdovsky, not ' penis 
sheath,' Eisen). I do not understand what these structures are. Vejdovsky only 
doubtfully allows this species to be a Zimnodrilus. 

(5) Limnodrilus steigerwaldi, Eisen. 

L. steigerwaldi; Eisen, loc. cit., p. 18. 

Clitellio (Limnodrilus) steigerwaldi, Vaillant, loc. cit., p. 437. 

Definition, Length about 80 mm^. Brain with marked concavity posteriorly, wider in front 
than behind, and prolonged into lobes anteriorly. Spermathecae narrow and pear-shaped ; 
penis about as long as in last species. Nephridia without globular peritoneal cells, except 
just behind funnel. Hah. — California; Sierra Nevada. '], 000 ft. 

This species is to be distinguished from the last by its greater size and more 
complex brain. Vejdovsky again only doubtfully includes it in the genus 

(6) Limnodrilus monticola, Eisen. 

L. monticola, Eisen, loc. cit., p. 18. 

Clitellio (Limnodrilus) monticola, Vaillant, loc. cit., p. 437. 

Definition. Length about 30 mm. Brain much as in L. ornatus, but squarer. Spermathecae 
cylindrical, sometimes with slight constriction. Penis about eight as long as broad, 
truncated at extremity. Hab. — California; Sierra Nevada. 9,000/!. 

The truncated extremity of the chitinous sheath of the penis is, according to Eisen, 
the most characteristic mark of this species. It should be noted, however, that 
Diepfenbach (PI. ii, fig. i) figures a similar truncation in L. hoffmeisteri. 

' stated by Eisen in (12, p. 89s) to be 30 m. by an evident misprint. 


(7) Limnodrilus alpestris, Eisen. 

L. alpestriSj EtSEN, loc. cit., p. 19. 

Clitellio (Limnodrilus) alpestris, Vaillant, loc. cit., p. 428. 

Definition Length about 25 mm. Brain wider behind, sometimes trilobed. Nephridia with 
vesicular cells. Sjjermathecae wider at both extremities than in the middle, coiled at the 
upper- extremity. Chitinous penis sheath about eight times as long as broad, trumpet 
shaped at extremity. Hab.- — California; Sierra Nevada. '], 000 ft. 

The shape of the brain and of the extremity of the chitinous penis are stated by 
Eisen to be the distinguishing marks of the species. 

(8) Limnodrilus corallinus, Eisen. 

Camptodrilus corallinus, Eisen, loc. cit., p. 23. 

Clitellio (Limnodrilus) corallinus, Vaillant, loc. cit., p. 431. 

Definition. Length 25-70 mm. Brain nearly square, with square cleft behind ; anterolateral 
processes well developed. Nephridia in front of clitellum with, behind clitellum without, 
vesicular cells. Spermathecae wide and globular without narrow duct. Chitinous penis 
sheath eight times as long as broad. Hab. — California; Fresno county. 

The presence of vesicular cells surrounding the nephridia in the anterior segments, 

but not of the posterior segments, is a character which this species shares with 

L. hoffmeisteri ; the form of the brain in the two species is evidently not unlike, 

• though that of L. hoffmeisteri (Vejdovsky, 24, Taf. viii, fig. 14) want the deep and 

square posterior cleft. 

(9) Limnodrilus silvani, Eisen. 

L. silvani, Eisen, loc. cit., p. 19. 

Clitellio (Limnodrilus) silvani, Vaillant, loc. cit., p. 428. 

Definition. Length 180 mm. Brain wider than long, and wider behind than in front, some- 
times furnished with three posterior lobes. Nephridia without vesicular peritoneal cells. 
Spermathecae wider at extremities. Chitinous penis, only three or four times as long 
as broad. Hab, — San Francisco. 
This species is described by Eisen as occurring in two varieties ; one is very much 
longer than the other. In the above definition the longer form only defined ; the 
smaller worms (50 mm. in length) occur with the bigger ; their brain is longer than 
wide and never trilobed ; the spermatheca is straight. As intermediate forms are found 
(though rarely) Eisen unites them into one species. 


(lo) Limnodrilus igneus, Eisen. 

Camptodrilus igneus, Eisen, loc. cit., p. 23. 

Clitellio (Limnodrilus) igneus, Vaillant, loc. cit., p. 430. 

Definition. Length about 30 mm. Brain deeply cleft in front and behind ; the two 
anterior lateral processes of the brain marked by several protuberances, broader in front 
than behind. Nephridia with, vesicular peritoneal cells. Spermathecae wide distally ; 
expanded at its extremity. Penis at least ten times as long as broad. Hob. — San 

Genus Hesperodrilus, Beddaed. 

Defikttiobt. Dorsal seta bundles with capilliform setae only; ventral setae 
uncinate and simple, one of each, kind in a bundle. Clitellum XII -XIII ; 
male pores on XIII, spermathecae open on XIII. Ko prostate. 

This genus is at present known only from South America, whence I have described 
four species. The genus, though agreeing in general appearance with Tubifex and 
other Tubificidae (the resemblance is not so marked in the case of Hesperodrilus 
niger), is one of the most distinct of the genera of this family. As regards external 
characters the position of the clitellum and of the generative pores differs from all 
other Tubificidae — if we assign to that family, and except only, Phreodrilus. The 
segments occupied by these various structures are one further back than is usual. 
The setae of the ventral bundles are remarkable for the fact that in each bundle 
there is one uncinate seta and one whose extremity is not bifid. In two species, viz. 
H. albus and H. pellucidus, the dorsal setae are wanting upon the first setigerous 
segment, thus recalling the state of affairs so characteristic of the Naidomorpha. The 
iaternal structure is chiefly remarkable for the fact that the sperm-duct does not 
open into the summit of the spermiducal gland, but opens in common with this 
gland into the base of the protrusible penis. There are no distinct prostates ; nor 
is the absence of these compensated for by the presence of a thick glandular covering 
such as occurs in Branchiura. The spermathecae are abnormal in opening on to the 
exterior behind the male pores — a quite unique occurrence in the family excepting 
again only Phreodrilus. Coupled with the absence of prostates, there are no spermato- 
phores. Of the four species of this genus two come very close together, viz. H. albus 
and H. pellucidus. The species can be discriminated by the following scheme : — 
(i) Setae absent from dorsal region of first two segments : 

a. Spermiducal gland with narrow duct leading to penis. H. albus. 


h. Spermiducal gland without this duct. H. pellucidus. 
(2) Setae present upon all segments after the jBrst: 

a. Posterior end of body with branchiae. H. branchiatus. 

b. No branchiae. H. niger. 

(i) Hesperodrilus albus, Beddaed. 
H. albus, Beddaed, Ann. Mag. Nat. Hist. Feb. 1894, p. 209. 
Definition. Seiae of dorsal bundles commence in segment III. Spermiducal gland opens 

into penis by a moderately long, narrow, non-glandular duct. Hab. — Port Stanley, 

Falkland Islands, in a pond. 
This species, in the preserved condition, was quite white ; hence the specific name. 
It is about the size of a Tubifex. The intestine begins in xix. The first pair of 
nephridia, which occupy segments \i-x, open on to the exterior in vi. The condition 
of this first pair — and they are the same in the following species — recalls a character- 
istic feature of the Lumbriculid Phreatothrioc. The spermathecae are very peculiar — 
apart from the position of their external pore. They are very long, extending through 
five segments. The blind extremity is dilated into an oval sac ; in which part of the 
organ alone was sperm to be found. Then follows a narrow duct which runs in 
a fairly straight course, being in one specimen wound round the corresponding part 
of the spermatheca of the opposite side of the body. In the middle of this is a kind 
of trap formed by a sudden increase in the height of the lining epithelium, and diameter 
of the tube, which projects forward, and would tend to prevent the sperm from moving 
except in one direction, i. e. towards the blind end of the sac. After the narrow part 
of the tube it widens out into a terminal chaijiber, which opens on to the exterior 
in the unusual position already referred to. The spermiducal gland is not large; it 
ends in a thin tube lined by a non-glandular lining which opens into the penis. In 
one individual the clitellum occupied segments xi, xii, the normal segments for the 
Tubificidae. The entire spermiducal gland can be everted, and the penis thus formed 
is very long. There were ripe ova in segments xviii-xx, or in xxi, xxii. 

(2) Hesperodrilus pellucidus, Beddaed. 
H. pellucidus, Beddaed, loc. cit., p. aio. 
Definition. Setae of dorsal bundle commence in III. Setae of dorsal bundles shorter and 
more slender than in last species. No long duct, only a constriction separates spermiducal 
gland from, penis. Hab. — JJsclmia, S. America. 

This species exactly resembles the last, except in the points noted in the above 
diagnosis and in the pale brown colour of the preserved worms. 


(3) Hesperodrilus niger, Beddard. 

H. niger, Bbddakd, loc. cit., p. ao8. 

Definition. A stout species, black in colour. Dorsal setae begin upon first setigerozis segment. 
Hab. — Port Stanley, Falkland Islands. 

This worm has very much the appearance of a Lumbriculus. It is much stouter 
than either of those described. The black colour is due to pigment, which is chiefly 
deposited in the peritoneum dorsally^ extending into the musculature and also a little 
way along the septa. The dorsal setae are rather slender. It is a curious point 
about this worm that the spermiducal gland, though confined to the twelfth segment, 
pushes the septum xi/xii before it so that it comes to lie as it were partly in 
segment xi. This out-pushing of the septum in question begins when there are as 
yet no signs whatever of the gland itself. 

(4) Hesperodrilus branchiatus, Beddaed. 

H. branchiatus, Beddaed, loc. cit., p. 207. 

Definition. Last thirteen segments with a series of paired gills, attached a little below the 
lateral seta bundles. Spermathecae with a slight diverticulum at orifice. Hab. 
— Valdivia, Chili. 

This species is the only other Tubificid besides Branchiura, which has gills ; these 
will be found more particularly described above (p. 84}. The specimen which 
I examined consisted of fifty-three segments, and was about three-quarters of an inch 
long. I could find no proper sperm-sacs such as exist in the other species and in 
other Tubificids. The sperm lay loose in the cavity of segments vii-xi. 

Genus Hetebochaeta, CLAPARiiDE. 

DEFimTioir. Setae uncinate, except dorsal setae of V-XIII, which are mainly 

palmate. Dilated hearts in VIII. Spermiducal gland divisible into two 

regions ; the glandular region (vesicular) not dilated ; penis chitinous ; prostates 


This genus, originally described by CLAPAsiiDE (4), has been more fully dealt 

with by Benham (9). It is mainly to be characterized by the peculiar palmate 

setae which CLAPARiiDE erroneously referred to as cup-shaped. They are in reality 



better termed 'fan-shaped'; the free end is expanded, and apparently cleft into 
about seven 'teeth.' These are, however, united by a delicate membrane. As the 
absolute extremities of the teeth are bent over all in one direction, the shape of the 
setae is highly suggestive of that of a rake. There is but one species: — 

Heteroehaeta costata, CLAPAnfeDE. 
H. costata, Clapaejide, Beobacht. wirbeU. Thiere, 1863, p. 35. 

Definition. Length about 15 mm. ; nnmber of segments ahout 40. Ventral setae of anterior 
segment and dorsal setae of Il-IF have the prongs nearly equal ; in dorsal and ventral setae 
of posterior segments lower prong shorter than upper ; niimher of setae in a bundle, 4 (/J) 
to 14 (fill) dor sally ; in ventral bundles, i (XXV) to 5 (F). Hob. — Coasts of England 
and Belgium; Marine. 

There is a certain amount of variation in the arrangement of the setae. The 
palmate setae sometimes extend further. There has, however, been no detailed study 
of these variations. Setae of the pectinate variety occasionally occur, but, apparently, 
so rarely that their occurrence cannot be regarded as either a generic or a specific 
character. The species is fully illustrated in Benham's memoir, and a few figures are 
given by CLAPAEiiDE. 

Genus Peloscolex, Leidy. 

Depiwitioit. Each segment with a circle of prominent tubercles. Dorsal setae 
entirely capilliform, ventral setae uncinate. 

There is but one species, viz. : — 

Peloscolex variegatus, Leidt. 

p. variegatus, Leidy, P. Acad. Nat. Sci. Philadelphia, 1850, p. 134. 

Definition. Length 4 lines. Six to ten setae in each dorsal bundle. Two to three in ventral 
bundle. Hab. — N. America. 

It is quite clear that this worm is a Tubificid. The characters of the setae are not 
of themselves sufficient to prove this ; but the statement that the clitellum occupies the 
tenth segment, added to the characters of the setae, is not reconcileable with its location 
in any other known family. Vejdovsky (24, p. 45) includes it among the Tubificidae 


' incerfcae sedis.' I regard this genus and species as- distinct from Spirosperma ferox 
(with which, on account of the integumental tubercles, it might be confounded), 
for the reason that there are no setae but eapilliform setae in the dorsal bundles. 
This peculiarity has been stated to occur by Geube in his species ' Saenuris vdutina,' 
to which attention has been duly called by Vejdovsky, who considers further evidence 
desirable before the absence of uncinate setae from the dorsal bundles can be regarded 
as fully proved. It seems improbable that hoth Gbube and Leidt would have fallen 
into error about a point of this kind, and, furthermore, it seems probable that 
S. velutina is identical with EmbolocepJialus velutina (see below). Embolocephalus, 
however, has only two setae on each bundle ; this worm can also retract the pro- 
stomium and first segment, so that the body appears to commence with the first 
setigerous segment. It inhabits the depths of the Lakes of Geneva and Zurich. 
P. variegatus was found by Leidy in springs, the water of which was impregnated 
with iron, in the neighbourhood of Philadelphia ; there is no mention of any power 
of retracting the end of the body. I am disposed, therefore, to consider that the 
genus Peloscolex is a valid genus, distinguishable from nearly all other Tubificidae by 
the presence of eapilliform setae only in the dorsal fasciculus. 

It may be that ]}fais papillosa should be transferred from Spirosperma to the 
present genus. 

Genus Psammoryctes, Vejdovsky. 

Syn. Saenuris, Kessler (in part.). 
Tubifex, Lankestee (in part.). 
Archaeoryetes, Czeeniavsky. 

Depinitiom". Setae eapilliform, uncinate, palmate, and pectinate. Spermiducal 
glands, with a vesicula seminalis; prostate as in Tubifex; spermatheeae opening 
in common with a muscular sac containing a single long seta ; appended to 
this sac are two or four glands. 

The genus Psammoryctes has been investigated by Lankestee (2), Vejdovsky 
(13, 24), Peeeiee (7), and, more recently, by Stolc (3); Vejdovsky established 
its distinctness from Tuhifex, with which it had been confounded by Lankestee 
and Peeeiee ; the differences which this genus shows from Tuhifex were still 
further accentuated by Stolc, who described and figured the remarkable apparatus 
in connexion with the apertures of the spermatheeae. These have been already 
described (see p. 132). I quite agree with Vaillant in relegating Czeeniavsky's 
Archaeo7'yctes to this genus. The only difference alleged by Czeeniavsky is that 

L 1 2 


the setae of the dorsal bundles in the anterior segments are partly tridentate instead 
of pectinate — a difference which Vaillant justly considers as at most specific. 

Three species have been referred to this genus, viz. P. barbatus, P. remifer, 
and P. batillifer. About the first there is no question ; it is, perhaps, doubtful 
whether P. remifer should be placed in this genus. The peculiar setae of some 
of the anterior segments appear to difiier from those of the type-species as well 
as from those of any other Tubificid. It is obviously desirable that we should 
have more information about both P. reimifer and P. batillifer, before any definite 
statement can be made as to their systematic position. 

Psammoryctes barbatus (Geube). 

Saenuris barbata, Geube, Ein Ausflug nach Triest und Quamero, 1861, p. 75. 
S. (Naidina) umbellifera, Kesslek, Trud. Eussk. Est. St, Petersb., 1868, p. 107. 
Tubifex umbellifer, Lankestee, Ann. and Mag. Nat. Hist., Feb., 1871, p. 93. 
Psammoryctes umbellifer, Vejdovsky, SB. Bohm. Ges., 1875, p. 194. 
P. barbatus, Vejdovsky, SB. Bohm. Ges., 1883, p. 224. 

Definition. Length 40 mm. ; number of segments 90. Prostomium as long as buccal segment. 
Uncinate setae with equal prongs or with one longer than the other, the first kind found 
from eleventh segment onwards in both dorsal and ventral bundles, the other in ventral 
bundles of anterior segments ; no setae on XI ; palmate setae in II — X. Hab. — Europe. 

Geube's description was far too imperfect to admit of a recognition of the species; 
hence the synonymy. The identity of Geube's species was cleared up by Vejdovsky, 
who examined the type of 'Saenuris barbata.' This species is amply illustrated in 
the works of Lankestee (10), Vejdovsky (13, ^4), Stolc (3) ; see also Benham (9). 

Genus Hemitubifex, Eisen. 
Syn. Clitellio, CLAPARiiDB (in part.). 
Tubifex, d'Udekem (in part ). 
Peloryctes, Zengeb. 

DBPlBTiTioiir. Setae of two kinds, capilliform, and uncinate, the former only in 
dorsal bundles. A special chamber ('vesicula seminalis') constricted ofF from 
spermiducal gland above into which prostate opens. 

This genus comprises not only the species H. insignis, for the reception of which 
the genus was created by Eisen in 1870, but also, as I have shown (70), the marine 
worm 'Clitellio ater' of Claparede (3) ( = 7. benedii, d'Udekem). Possibly some 


others among the littoral Tubificidae, which have been but little studied, wiU be 
found to belong to the same genus. There are no points of special interest in the 
anatomy of this genus. It comes nearest perhaps to Tuhifex, from which it differs 
principally in the dilated extremity of the spermiducal gland, and in the chitinous 
penis. EiSEN placed the gfenus nearer to Psammioryctes, but Stolc's recent investi- 
gations upon this worm (3) have increased the differences between the two genera, 
which, however, agree in the division of the spermiducal gland into two chambers, 
and in the fact that there are glands appended to the spermathecae. It is very 
possible that both Ilyodrilus perrieri and Ilyodrilus fragilis should also be referred 
to the present genus. Vaillant (p. 395) has already made the suggestion with 
regard to Ilyodrilus fragilis. 

I have, however, pointed out (70, p- 487) that capilliform setae are sometimes 
present, and sometimes absent in worms belonging to a species, which must, I think, 
be CLAPABiiDE's G. ater, and d'Udekem's Tuhifex benedii. Thus the principal difficulty 
in the way of regarding these two worms as the same species is removed. Peloryctes 
inquUina of Zengeb is placed by Vaillant among the synonyms of C. arenarius, 
but also, evidently by an oversight, mentioned (on p. 435) as a distinct species of 
CUtellio. It cannot, I think, be distinguished from Clapabede's G. ater. Zengeb 
attempts to distinguish it mainly by the number of setae in the bundles (an 
eminently untrustworthy character except when applied with great caution) by the 
distribution of the papillae, and the presence of more than one pair of hearts. 

(i) Hemitubifex insignis, Eisen. 
H. insignis, G. EiSEN, Bih. K. Svensk. Akad., 1879, No. 16, p. 13. 
Definition. Length 25 mm. ; no integumental papillae. Spermathecae with glands at neck. 
Rah. — Sweden. 
For a fuller account of the anatomy of this species with illustrations, see Eisen (12)- 

(2) Hemitubifex benedii (d'Udekem). 

Tubifex benedii, d'Udekem, BuU. Acad. Roy. Belg. t. xxii., 2nd pt., 1855, p. 544. 
Clitellio ater, Clapae^de, M^m. Soc. Phys. Gen. 1862, p. 253. 
T. papillosus, ClaparJide, Beobacht. wirbell. Thiere, 1863, p. 25. 
Hemitubifex ater, Beddaed, P. Z. S. 1888, p. 485. 
Clitellio (Clitellio) benedii, Vaillant, Annel^s, p. 418. 
? M"ais pustulosa, Williams, Phil. Trans., 1858, p. 96. 
Peloryctes inquilina, Zengeb, Bull. Soc. Nat. Mosc, 1870, p. 221. 


Definition. Length ^^ mm. Integument covered with numerous greyish green papillae, which 
commence at middle of second segment and are not found upon clitellum. Hah. — Bea 
shores hf Europe. 

The synonymy of the species described as Glitellio ater by CLAPAEtDE is very 
difficult. There can be but little doubt, however, that it is synonymous with 
d'Udekem's Tubifex benedii as was first pointed out by Vaillant (3). I cannot 
myself see any valid reason for separating it from T. papillosus of Claparede, in 
spite of the fact that both species were described by the same writer. Claparede, 
indeed, says that the papillae of T. papillosus are flatter than those of C. ater; but 
the principal difierence upon which CLAPAKiiDE relied was evidently the presence of 
capilliform setae figured (4, Taf. iii, fig. 15) ia T. papillosus, which he speaks of on 
this account as 'ein echter Tubifex.' 

Genus SpirospebmA: Eisen. 

Nais, Kesslek (in part.). 

Tubifex, CLAPABi;DE, d'Udekem (in part.). 

Depiwition. Body covered with convex papillae. Setae capilliform, uncinate and 
pectinate. Prostate present; penis chitinous. Brain with median processes in 
front, prominent lateral lobes, with a square invagination posteriorly. 

The structure of this well-marked genus has been investigated by Eisen (12) and 
Stolo (3). The setae are like those of Psantgnoryctes, from which genus it differs 
in the absence of a specialized division of the spermiducal gland, and in the presence 
of the epidermal papillae which are evidently like those of Heniitubifex benedii. 
It is even possible that the two genera should be fused into one, the only difference 
being, apparently, the existence of pectinate setae in Spirosperma, which have not 
been described in Heniitubifex. The name Spirosperma was given to the genus by 
Eisen on account of the form of the spermatophore, which is coiled like a watch 
spring, or, as Eisen says, like the proboscis of a moth (12, PI. iii. fig. a i, 2 h). It 
appears, however, that the spermatophore has not always this shape, for Stolc 
figures one (3, Taf. iii. fig. 11) in which one half of the spermatophore is coiled 
spirally round the other half, and another (in fig. 10) which is bent into an S shape. 
The spermathecae are simple pouches swollen at the caecal extremity. The epidermal 
papillae are wanting upon the clitellum, which is stated by Eisen to occupy only 
one segment, the eleventh. 


Spirosperma papillosus (Kesslee). 

Nais papulosa, Kesslee, Trud. Kussk. Est. St. Petersb., 1868, p. 105. 
Spirosperma ferox, EiSBN, Bih. K. Svensk. Akad., 1879, No. 16, p. 10. 

Definition. Length i^ mm. Dorsal seiae capilliform and pectinate j ventral setae uncinate, 
some of them with two or three denticles. Nephridia without vesicular cells, no terminal 
dilated sac. Hah. — Europe. 

The probable identity of Eisen's Spirosperma ferox with Nais papillosa of 
Kesslee was pointed out by Eisen himself; and Vejdovsky (24, p. 45) suggested 
that Geube's ' Saenuris velutina ' may be identical. Both are placed as synonyms by 
Stolc. Nevertheless, according to Geube, Saenuris velutina has nothing but capilKform 
setae in the dorsal bundles, and the clitellum extends from segment ix-xii. Influenced, 
no doubt, by these differences in the descriptions of the two species, Vaillant does 
not unite them, but places Geube's worm in the genus Tuhifex. Benham has 
described and figured the webbed setae of this species (9). 

Genus Telmatodrilus, Eisen. 

Defiititiow. Setae all uncinate. Brain concave in front with a narrow process 
posteriorly. Circumoesophageal blood-vessels of segments VII-XI contractile. 
Integumental vascular system developed. Spermiducal gland furnished with 
numerous (8-IO) prostates; penis chitinous. 

This remarkable genus has been only seen and described by Eisen ; it consists of 
but one species. Eisen places it in a special subfamily, that of the Telmatodrilini j 
but a further splitting up of the Tubificidae seems hardly necessary. The most 
distinctive peculiarity of the genus is undoubtedly the numerous prostates; but 
except for this the eflferent apparatus shows no difference fx'om that of lAmnodrilus 
and Spirosperma. The contractility of the anterior perivisceral blood-trunks is shared 
by the genus Branchiura, but in Telmatodrilus only one of these pairs of vessels — 
that of the eleventh segment is larger than the rest. The approximation of the 
dorsal ventral vessels upon one side of the intestine, is also something like what is 
found in Branchiura; so too the integumental vascular network; in this respect, 
however, Ilyodrilus and, to a less extent, Limnodrilus, resemble Telmatodrilus. 
The spermathecae are a pair of simple oval pouches, which open between the dorsal 
and ventral setae bundles in the tenth segment — an anomalous position for the 


There is at present but one species known, viz. T. vejdovskii ; but as the external 
characters are similar to those of Limnodrilus and Clitellio, it is possible that some 
of the species doubtfully referred to these genera may belong to Telmatodrilus. 

Telmatodrilus vejdovskii, Eisen. 
T. vejdovskii, Eisen, Bih. K. Svensk. Akad., 1879, No. 16, p. 8. 

Definition. Length 3,5-50 mm. Intestine begins in XI ; pigmented peritoneal coating of intestine 
begins in XV. Nephridia with vesicular peritoneal cells and without dilated end sac. Hah. 
— California, at an altitude o/" 6,000- 10,000 /if. 

The worm is described by Eisen as being torpid in its habits, not active like 
many Tubificidae ; it lies in the mud as does Tubifex, &c. with the tail protruding. 
The setae are eight to fifteen in number in each bundle; in the adult they are 
sigmoid without terminal bifurcation, which is only apparent in young specimens. 
No spermatophores were found in the spermathecae. 

Genus Ilyodrilus, Eisen. 

DEFiifiTiow. Capilllform setae in dorsal bundles, uncinate setae with a web, 
penial setae, not very different from ordinary uncinate setae, in neighbourhood 
of male-pores. An integumental plexus of blood-vessels present. Nephridia 
with a dilated region after funnel. Spermiducal gland with thick covering of 
gland-cells ; no spermatophores ; ova develop as in Naids. 

It seems to me to be not a little doubtful whether this generic name Can be 
retained. The name was applied by Eisen (12) to a number of species of Tubificids 
found in California, which I cannot diiferentiate from the genera Tubifex, and 
Hemitubifex. To this genus (Ilyodrilus) Stolc (3) referred the species described by 
Vejdovsky (11) as Tubifex coccineus, and later (24) as T. rivulorum, var. coccineus. 
Stolc has given an elaborate account of the structure of 'Ilyodrilus' coccineus, 
which appears to me to prevent anyone from placing it in the same genus as any 
of the three species described by Eisen as Ilyodrilus. Eisen''s Ilyodrilus has small 
separate prostates, which are absent from I. coccineus; it has not penial setae which 
are present in I. coccineus, as was first pointed out by Vejdovsky (p. 156, footnote) ; 
the only resemblance of importance seems to be that in I. sodalis the eggs develop 
in the same way as in I. coccineus, and that in the American worms no spermato- 


phores are to be found i. Vaillant eliminates /. coccineus from the genus Ilyodrilus, 
restricting the genus to the three American species, remarking (6, p. 349) that the 
former is, according to Vejdovsky, merely a variety of T. rivulorum; this was 
certainly Vejdovsky's original opinion when writing his great monograph ; but in a 
footnote to a later page (p. 156) of that monograph, he distinctly refers to I. coccineus; 
by this time, no doubt, Stolc had completed his researches on the worm, proving 
its distinctness from Tubifex; hence the footnote, probably added as the work was 
passing through the press. Eestricting the genus, as I propose to do here, it will 
contain only one species — /. coccineus. 

Eisen's genus Ilyodrilus will not, I think, stand ; the three species of which it is 
composed are not, in my opinion, referable to the same genus so long as we allow 
the numerous genera adopted in the present work; the differential characters of the 
genus, as given in Eisen's work upon the Tubificidae, by no means apply to all the 
species; the only character that does so apply, and is of real importance, is the 
absence of spermatophores ; I should venture, however, to doubt the real absence of 
the spermatophores, since the three species all of them possess prostates; in all 
Tubificidae, with the exception of Teliniatodrilus^ which have prostates, there are 
spermatophores formed. Another character used by Eisen in the generic description 
is the absence of glands at the base of the spermathecae ; there are, however, such 
in 'Ilyodrilus' sodalis; can this species be put anywhere than in one of the genera 
Tubifex or Hemitubifex ? The glands at the base of the spermathecae ally it to the 
latter, from which the absence of a vesicula seminalis distinguishes it. In this it 
agrees with Tubifex, and in having a soft penis without chitinous sheath. As in 
both genera there are no pectinate setae; in the species under discussion the two 
prongs of the uncinate setae have fine denticulations along the inner margin, but 
these setae are hardly comparable to the pectinate setae of other genera. I am 
disposed to refer this species to the genus Tubifex, in spite of the presence of 
glandular appendages to the spermathecae. ' Ilyodrilus ' perrleri, unlike the last 
species, and unlike Tubifex, has a chitinous sheath to the penis ; there is a very 
faintly developed pectination of the uncinate setae ; this species seems to come 
nearest to Hemitubifex. '■Ilyodrilus' fragilis is the only remaining species of 
Eisen's genus ; it must go, I think, with the last into the genus Hemitubifex (as 
Vaillant has suggested). These three species seem to unite the genera Tubifex and 
Hemitubifex. Apart from these species the two genera in question are separated by 
very slender characters. In Tubifex the penis is soft, and there is no vesicula 

' Vejdovsky (24, Taf. iv. fig. 13) has figured the spermatophore of I', cocdnms, but it is to fee presumed that 
this is an error. 

M m 



seminalis ; in Hemituhifex the penis is chitinous, and there is a vesicula seminalis. 
The characters of these two genera, and of the three species of Ilyodrilus described 
by EiSEN, will be best shown in a table which follows : — 







Ilyodrilus sodalis 

uncinate setae with both 

prongs denticulate ; 

capilliform setae 

in dorsal bundles 

no hearts ; last 

perivisceral in x ; 

all contractile 


with bilobed 
gland at base 

no vesicula 

Ilyodrilus fmgilis . 

as above 

as above (?) 


no gland 

no vesicula 

Ilyodrilus perrieri . 

uncinate setae with web ; 

capilliform setae in 

dorsal bundles 

as above (?) 


no gland 

no vesicula 

Hemituhifex ater . . 

capilliform setae in 
dorsal bundles 



no gland 

a vesicula 

Hemilubifex insignis 

as above 



three glands 
at base 

a vesicula 

TuUfex rivulorum . 

as above 

hearts in viii 


no glands 

no vesicula 


Nephridia with 'brown gland.' 
Spermiducal glands, like that of 

Penial setae. 
Spermatheca globular and sessile. 


Nephridia without that gland. 
Spermiducal glands, as in Tubifex. 

No penial setae. 
Spermatheca with duct. 

There may be other points of diflFerence ; but in any case it seems to me that the 
first two are amply sufficient to distinguish the European Ilyodrilus from its supposed 
American congeners. 

Ilyodrilus coccineus, Vejdotsky. 
Tubifex coecineus, Vejdovsky, SB. Bohm. Ges. 1875, p. 193. 
T. rivulorum (in part.), Macintosh, Trans. Roy. Soc. Ed. 1870, p. 253. 
T. rivulorum var. coecineus, Vejdotsky, Syst. u. Morph., p. 46. 
Ilyodrilus coecineus, Vejdovsky, ibid. p. 150, footnote. 

Definition. £ram with a median process in front, and with two processes in the middle 
line behind at origin of visceral nerves. Spermathecae globular and sessile ; genital setae 
sometimes without notch. Hab. — Europe. 
The synonymy of this species is a little confused. There seems to be no doubt 


that, as Stolc has pointed out, Macintosh partially described this species in his 
paper upon the structure of Tuhifex. He distinguishes ' two species ' of Tubifex, the 
one which he considers to have most claims to be called Tubifex rivulorum, being 
in all probability identical with Ilyodrilus coccineus j the reasons for this identification 
are— (i) the shape of the setae ; Macintosh found no traces of the ' brush tip ' to 
the uncinate setae of this form such as characterize the true T. rivulorum; as he 
figures the latter kind of setae in the 'elongated form from the lakes' (really the 
true T. rivulorum), it seems clear that no mistake has been made, even if it was 
reasonable to doubt the accuracy of so experienced a worker ; (a) the integumental 
network is particularly referred to as existing in the shorter form of 'Tubifex'; 
(3) the peculiar brown glandular body lying upon the course of the nephridium is 
figured (PI. ix. fig. 18) for the form which is to be regarded as the same as Stolc's 
I. coccineus. Vejdovsky at first distinguished his T. coccineus from T. rivulorum; 
in his work upon the Oligochaeta in general he places (p. 46) T. coccineus among the 
synonyms of T. rivulorum, on the grounds that there are numerous transitions between 
this form and the worm described by various authors as T. rivulorum ; as, however, 
' T. rivulorum ' of some authors is now fairly certainly known to be no less 
than /. coccineus, this argument loses its force. Later (p. 146) in the same work 
Vejdovsky, in dealing with the egg development, which is after the Naid plan, again 
emphasizes the difierence between his T. coccineus and the common form, T. rivulorum : 
there seems to be no doubt that Ratzel's description of the egg development in 
' T. rivulorum ' referred to the present species, and that there was no dimorphism 
— as he thought —or pathological conditions — as thought Lankester and Nasse. Still 
later, in a footnote to p. 156, VEJnovsKY refers to the worm as /. coccineus without 
any explanation of his alteration of opinion, and in the section dealing with the 
relationships of the Oligochaeta the same author mentions the sexual setae (mentioned 
also in the place just referred to) as evidence of the relationships of 'Ilyodrilus' to 
the Naidomorpha. It is difficult to understand how Vaillant, with Stolc's paper 
upon Ilyodrilus before him, could have put it as a mere synonym of T. rivulorum ; 
his reasons for doing this are because Vejdovsky thought so at one time, before the 
species had been re-investigated by Stolc. It is the more remarkable that Vaillant 
should have overlooked the differences between the two species, as Ilyodrilus is so 
thoroughly intermediate between the Naids and the Tubificids, which Vaillant unites 
into one single family. To have emphasized its differences from Tubifex and resem- 
blances to Naids would have greatly strengthened Vaillant's position 

I take this opportunity of referring to Nicholson's Saenuris canadensis (see Vereill) merely 
for completeness' sake, as it is unidentifiable. 

M m 2 


Genus Bothrioneuron, Stolc. 

? Syn. Monopylephorus, Levinsen. 

DEPiETiTioif. Setae only uncinate, ventral setae of segment XI sometimes modified. 

Male pore single or paired. Spermiducal gland with a short diverticulum bearing 

the prostate. No spermathecae ; spermatophores fixed to the integument near 

to male pore. 

I believe that Levinsen's genus Monopylephorus is identical with Stolc's recently 

(3) described Bothnoneuron. Levinsen gives but few details about his genus ; he 

states, however, in the table of the generic characters of the Tubificidae (p. 223) that 

there are no capilliform setae, that the male pore is unpaired, and that the penis is 

without a chitinous sheath. In all these characters Monopylephorus agrees with 

Bothrioneuron, except perhaps in the presence of a penis ; in Bothrioneuron, however, 

Stolc has figured (Tav. iv, fig. 7) a short diverticulum of the spermiducal gland, near 

to the external aperture of the latter, which may possibly be everted at times and 

give the appearance of the penis of other Tubificidae. Vaillant (6) considers not 

only that Monopylephorus is synonymous with Glitellio, but that Levinsen's species, 

M. ruhroniveus, is identical with G. arenarius. It is true that the latter identification 

is prefaced by a query ; but I do not see any reason why Vaillant should suggest that 

Levinsen's genus should be considered ' doubtful.' 

The most marked characters of this genus are the absence of the spermathecae 
■ and the peculiar form of the spermiducal glands ; there is no other Tubificid in which 
spermathecae have not been found. The absence of these organs is correlated by Stolc 
with the remarkable form of the spermatophores ; these have an appearance in Stolc's 
figures which recall the corresponding organs in Lumbricidae ; they are stalked bodies 
invariably found attached in the neighbourhood of the male pore. The free end is 
wider and oval in form, and seems to contain the sperm. The two spermiducal glands 
open by a common pore, which is in the middle line of the eleventh segment ; each 
is divided into two regions, the two being separated by a constriction ; into the distal 
region opens the short and slender sperm-duct ; Stolc considers it to be a dilated 
part of the sperm-duct rather than a part of the spermiducal gland ; it differs from 
the succeeding section by the thicker coating of peritoneum, the thickening being 
due not to an increase in the number of cells, but in their greater depth ; the lumen, 
too, is much narrower than in the section of the terminal gland which follows, though 
the entii-e width of the tube is about the same ; the following section is lined by 
a higher cylindrical epithelium and is covered by a lower layer of peritoneal cells ; 


at about the middle point there is a short diverticulum, which Vejdovsky terms the 
' paratrium,' and on to the summit of which is grafted the prostate ; the cells lining 
the paratrium are larger than those which line the sac, of which it is a diverticulum, 
though they get to be very small before the opening of the paratrium into the latter ; 
the paratrium has presumably a very thin covering of peritoneal cells ; none are shown 
by Stolc in his figure. The prostate is composed of comparatively few cells. Near 
to the opening on to the exterior there is a short diverticulum to which muscles are 
attached ; this can possibly be extruded. 

(i) Botlirioneuron. vejdovskyanum, Stolc. 

Bothrionearon vejdovskyanum, Stolc, Abh. k. Bohm. Ges. 1888, p. 43. 
? Monopylephorus rubroniveus, Levinsen, Vid. Med. 1883, p. 225. 
? Clitellio arenarius (in part.), Vaillant, Annel^s, p. 415. 

Definition. Body covered with papillae. Penial setae with hooked extremity, helow which 
are two ridges with minute denticulations. Hah. — Denmark ; Bohemia. 

Levinsen regards as a possible synonym of this species d'Udekem's Tubifex hyalinus. 
I cannot see in d'Udekem's definition of the species any grounds for this assumption. 
I have elsewhere entered into the much more likely view that it is a synonym of 
Clitellio arenarius. 

(2) Bothrioneuron americanum, Beddard. 
B. americanum, Beddard, Ann. & Mag. Nat. Hist. Feb. 1894, p. 306. 
Definition. Length about one inch. Male pores paired. No penial setae. No spermato- 
phoi-es{?) Hal. — Buenos Ayres. 

Genus Lophochabta, Stolc. 

DEFiifiTiosr. Dorsal setae plumose and pectinate, ventral setae uncinate. One 
pair of intestinal hearts in IX, in II-VII slender periviseerals. Penis 

The most distinguishing character of this genus is in the form of some of the dorsal 
setae ; these are like those of Tubifex in being capUliform, but difier in that the seta 
is beset with fine branches, giving it the appearance of the hairs of many Crustacea ; 
this form of seta is unique among the Oligocbaeta, the only approach to it being found 
in Bohemilla, where one side of the seta is furnished with such processes. The structure 


of the reproductive organs does not materially differ from that of Tubifex or Limno- 
drilus ; the spermidueal gland is short and the sperm-duct rather long ; at the point 
where the solid prostate is attached to the former it (the spermidueal gland) is some- 
what dilated. There is only one species of the genus. 

Lophochaeta ignota, Stolc. 
L. ignota, Stolc, Abh. k. Bohm. Ges. 1888, p. 41. 
Definition. Pectinate setae of dorsal bundles, u-ith four prongs and a weh. CJdimous penis 
conical in form. Hah. — Bohemia. 
It is a curious fact about this species that no spermatophores have been found. 
The spermatheca are pear-shaped pouches, there being no constriction between the 
pouch and its duct, such as developed in Limnodrilus. 

Genus Beaktchiura, Beddaed. 

DEPiiiriTiOM'- Posterior segments with a series of contractile toranehial processes in 
dorsal and ventral median lines, a pair to each segment. Setae capilliform 
and uncinate, capilliform setae in dorsal bundles of anterior segments only. 
Two pairs of enlarged circular vessels (hearts) in IX and X. These and 
anterior perivisceral loops contractile. Dorsal vessel ventral in position in all 
segments after the twelfth. 

This very remarkable genus of Tubificidae was found by myself in the 'Victoria 
regia' tank in the Botanical Society's Gardens, Regent's Park, London, and described 
with illustrations in the Quarterly Journal of Microscopical Science. It agrees 
with Hemituhifex and some other genera in the form of the setae, but differs from all 
Tubificidae except Hesperodrilus branchiatus in possessing branchiae, and from all 
Oligochaeta in the form of those branchiae. They consist of a series of pairs of 
cylindrical processes arising from the middle dorsal and ventral lines ; each branchia 
has a capillary loop lying beneath the epidermis and a layer of muscles ; the cavity of 
the branchiae, which is shut off from the general body by a diaphragm, is lined by 
peritoneum and traversed by muscular strands. The branchiae move actively during 
life, and can be extended and retracted. The circulatory system, like that of Ilyodrilus 
and some other genera, has in addition to the dorsal and ventral vessels a supraintestinal 
trunk, and, as in Ilyodrilus and Teltnatodrilus, there is an integumental network of 
blood-vessels. All the perivisceral trunks of the segments in front of and including 
the tenth are contractile, those of the ninth and tenth segments being specially enlarged. 


Branchiura sowerbii, Beddaed. 
B. sowerbii, Beddakd, Q. J. M. S. vol. xxxiii (189a), p. 335. 
Definition. Length about a,^ mm. ; number of segments about 170. Branchiae upon last 
50-80 segments. Nephnd'm commence in segment XII. Ilab. — 1 {Found in Victoria 
regia tank in Botanical Society's Gardens, Regent's Park, London). 
This worm like most (? aU) other Oligochaeta can reproduce its tail ; I cut off the 
entire gill-bearing region of a specimen, and in nine days there were four pairs of 
gills, not on the regenerated tail, which had only one small gill, but on the stump 
left behind, which was unprovided with gills before amputation. 

Genus Vermiculus, Goodrich. 

Depinitioit. Setae entirely uncinate. Clitellum X-XIII. Sperm-duct pore, and 
spermathecal pore single. Sperm-ducts short and wide. 
This is a distinct genus, unless it be really identical with Monopylephorus ; I believe, 
however, that that genus is, as I have said, on p. a68, the same as Stolc's Bothrio- 
neuron ; in this case Vermiculus cannot well be identical with any form known at 
present. Its most remarkable peculiarity is partially shared with Monopylephorus; 
that is, the single genital orifice. In the present genus, however, there are spermathecae 
— wanting in the former. Both organs open together by a common median pore. 
The oviducts are unpaired and open as usual between segments xi/xii. The clitellum 
is unusually extensive for this family. The spermiducal gland is reduced to a widish 
cavity evidently formed as an invagination of the exterior of segment xi; into this 
open the two sperm-ducts, which are very remarkable in being short and wide with 
glandular walls ; they are not in the least coiled. 

Vermiculus pilosus, Goodeich. 

V. pUosus, Goodeich, Zool. An2., 1893, p. 474. 

Definition. Surface of body clothed with fine hairs. Setae 3-4 per bundle. Hearts in X. 
Hob. — Weymouth. Marine, 

The worm was met with on the sea shore at Weymouth ; the locality was below 
high tide mark, and there were found with it other Tubificids and Enchytraeidae ; 
to the naked eye it is said by its discoverer to be indistinguishable from Heterochaeta 
costata. The coelomic fluid is filled by a quantity of round corpuscles, which render 
the worm veiy opaque when examined microscopically. 


Genus Embolocephalus, Randolph. 
Syn. Saenuris, Geube (in part.). 
DEFtNiTioif. Setae of dorsal bundles capilliform alone or with uncinate setae 
also. Anterior segments retractile. Non-retractile sense-organs in rows round 
the segments. Live in a tube fabricated by themselves. Spermiducal glands as 
in Tuhifex. 
This genus was formed by Randolph (1) for Grubb's Saenuris velutina and 
a new species congeneric with it ; the two species thus - associated differ from each 
other in a number of points but agi'ee in the characters given in the above definition ; 
the most remarkable feature of the genus appears to be' the complete retractility of 
the first two or three segments of the body ; of less importance are the sense organs 
which Randolph compares to those of Slavina; they also resemble those of Spiro- 
sperma. The genus is doubtless, as its describer suggests, intermediate between the 
Tubificidae and the Naidomorpha ; Einholocephulus velutina has, as have some of the 
Naidomorpha, only capilliform setae in the dorsal bundles ; in the other species, how- 
ever, the setae are like those of other Tubificids. The most remarkable internal 
structure of the genus is found only in E. velutina ; in this worm are a pair of glands 
associated with the ventral setae of the tenth segment ; these glands have a lining of 
glandular cells, and attached to the outside is a mass exactly like the ' prostate ' 
of other Tubificids; in fact the whole gland exactly resembles the spermiducal gland 
in most other members of the family. There is no doubt in my mind that it is to 
be compared to the albumen gland of the LumbriCulidae ; and its presence thus cements 
more closely the union between the families Tubificidae and Lumbriculidae. 

(]) Embolocephaltis velutinu's (Geube). 

Saenuris velutina, Geube, JB. Schles. Ges. vat. Cult., 1878, p. 116. 
E. velutina, Randolph, Vierteljahrsch. nat. Ges. Zurich, 1892, p. 147. 

Definition. Length 50 mm. ; number of segments 70. Two rings of sense-papillae in each 
segment, one on a level with setae the other on a level with the septa. Dorsal setae all 
capilliform, two, three, or four to a lundle ; ventral bundle consists of two uncinate 
setae, not always with well-marked cleft at end. Brain concave in front with two 
posterior lobes. Clitellum X-XII. In segment X a pair of large glands connected with 
ventral setae. Hah. — Lakes of Ziirich and Geneva. 


Grube particularly mentions the absence of uncinate setae in the dorsal bundles 
of this worm, -which is one of its chief peculiarities, and enabled it to be identified 
later. The worm secretes a thick slimy coating in which a quantity of Bacteria and 
other foreign bodies exist, and which project here and there as papiUiform masses. 
The anatomy has beei^ carefully worked out, and fully illustrated by Randolph (2). 

(2) Embolocephalus plicatus, Randolph. 

E. plicatus, Randolph, loc. cit. p. 147. 

Deflnition. Length 40 mm. ; number of segments 50. Two rings of sense pajpillae on each 
segment equally far away from septa. Dorsal setae with uncinate setae among them. 
Glitellum X-XII. Hah. — Lake of Ziirich. 

There appears to be in this species no extensile proboscis such as exists in the 
last ; it is not quite clear whether this proboscis is merely the extruded pharjnix 
or buccal cavity, or whether it is a peculiar structure. The layer secreted by the 
worm is thinner in this species than in the last. There is no albumen gland, and 
there are uncinate setae in the dorsal bundles ; the retractility of the head exists, but 
the segments are not withdrawn as they are in the last species, but are simply closed 
up in each other like the rings of a concertina. Figures and a fuller description of 
the species are to be found in Randolph's paper upon the anatomy of this and the 
last species. 

Appendix to Tubificidae. 
Genus Phreodrilus, Beddaed. 

Depinition. Setae of two kinds, the dorsal capilliform, the ventral sigmoid ; testes 

in X, XI, fused to form one pair. Spermiduoal gland opening on to XII, 

sperm-duct much coiled, provided with an appendix of considerable length; 

spermathecae in XIII, long and coiled. 

This genus I place near the Tubificidae for the reasons stated above (p. 327). It 

contains at present only one species, which is found in subterranean wells in New 

Zealand. This was described by myself a few years ago. The characters of the setae 

at once distinguish the genus from any other genus of Tubificidae ; the dorsal setae 

N n 


are certainly capilliform, as in the majority of the genera, but they are really a little 
different in form from the capilliform setae of other Tubificidae. The ventral setae are 
quite different from those that occur in other genera ; there are two in each bundle, both 
sigmoid but urmotched at the free extremity ; one of the two setae of each bundle 
is decidedly larger than the other. The clitellum is more extensive than in other 
Tubificids ; it appears to occupy some four segments, commencing with xii or xiii. 
The alimentary tract is, with the exception of the buccal cavity, ciliated throughout ; 
the intestine begins, though the transition is not very strongly marked, in segment xiii. 

The blood system is like that of Lopkochaeta and some other genera in that there 
is a supra-intestinal vessel, which communicates with the ventral vessel by a single 
pair of ' intestinal hearts ' ; these latter are, however, apparently rather more than 
merely connexions between the two longitudinal trunks ; their characters have been 
already described elsewhere (p. 78). The testes lie in segments x, xi ; as one pair 
are attached to the posterior wall of the former segment and the second pair to the 
anterior wall of the latter, an appearance is presented of a single testis on each side 
of the body perforating the septum ; this appears actually to be the case, the germinal 
tissue being traceable through the septum ; the sperm-ducts are a single pair only ; 
they open by their funnels into the eleventh segment ; the sperm-duct itself is entirely 
enclosed in the following segment. The most remarkable fact about the sperm- duct is 
the presence of a diverticulum, already referred to as arising from it, just within the sac 
which surrounds the spermiducal gland ; the diverticulum is lined by an epithelium which 
is not ciliated ; its structure, in fact, is precisely that of the spermathecae. The gland 
is unusually long and very greatly coiled ; the terminal section appears to be protrusible 
and passes straight from the external orifice upon the twelfth segment: the greater 
part of the organ is contained within a sac which is formed by the splitting off of its 
peritoneal coat ; this sac also contains the greater part of the windings of the sperm-duct, 
and there are abundant bunches of spermatozoa lying loosely within its cavity ; the 
sac, in all probability, performs the function of a sperm-sac, but how the spermatozoa 
get into and out of it is a little difiicult to understand, unless they actually pass 
through the thin walls of the sac. The lining epithelium is not ciliated, and there is 
no prostate. There are a pair of ovaries in segment xii, and the oviducts open 
opposite to them. There are egg-sacs formed by a pushing back of the septum dividing 
segments xii /xiii. 

The single pair of spermathecae lie in the fourteenth segment ; they are long 
pouches dilated at the blind extremity, which in the fully mature worm is filled with 
sperm ; the position of these organs behind the other parts of the reproductive system 
is not found in any other Tubificid, but is known among the Lumbriculidae. The 



following table will show in a graphic form the resemblances to and differences 
from other Tubificidae shown by Phreodrilus. 






Phreodrilus . . . 
Other Tubificids. 




without prostate 
with prostate 

in xiv 
in X 

x-xii (or xiii) 

Phreodrilus subterraneus, Beudaed. 
p. subterraneus, Beddakd, Trans. Eoy. Soc. Ed., 1891, p. 273. 

Definition. Length about 50 mm.; body semitransparent. Setae one to each dorsal bnndle 
and two in each ventral. Hob. — New Zealand. 

This, the only species of its genus, has a peculiarly delicate appearance which 
differentiates it from any worm with which I have been able to compare it ; the first 
specimens which I received from Mr. Smith were obtained from a well ; other examples 
were afterwards met with in non-subterranean water. I dealt with certain points in 
the anatomy of the species in a preliminary account (24) which was later followed by 
a more detailed description. 


Defxhttiobt. Aquatic Oligochaeta of small size. Setae usually in four groups upon 
each segment, sigmoid, bifurcate, hastiform, and capilliform. Sexual reproduc- 
tion at fixed intervals, between which asexual reproduction by fission occurs. 
Sexual organs (only known in a few types) are situated far forward, commencing 
even in the fifth segment. 

This family of Oligochaeta comprises the following recognizable genera : — 

(i) Chaetogaster, Baek. 

(2) Amphichaeta, Taubeb. 

(3) Wais, 0. F. MtJLLEK. 

(4) BohemiUa, Vejdovsky. 

(5) Pristina, EHBENBEEa. 

(6) Uncinais, Czeeniavsky. 

(7) Chaetobranchus, BouENE. 

(8) Dero, Oken. 
N n a 


We have a fair knowledge of all these genera. In nearly all the genera of 
this family, the dorsal setae are wanting in a variable number (varying for 
the genus) of the anterior segments. This ' cephalization,' as it has been termed 
by Lankestek, is still more marked in Chaetogaster, where the ventral setae also 
of some of the anterior segments are wanting, a condition also seen in TJncinais. 
The setae themselves present a considerable range of variety in form, which is 
partly illustrated on p. 7 in woodcut, fig. % b. The two most usual kinds of setae 
are the capilliform and the uncinate ; the former, which are never found except in the 
dorsal bundles, are thicker at the point of origin, and taper gradually to the free 
extremity ; they are sometimes straight, sometimes slightly curved, and longer or 
shorter ; in Bohemilla alone these setae are provided with a row of lateral serrations. 
The uncinate setae are always found in the ventral bundles, and sometimes in the dorsal 
also. They are shorter and stouter than the others, sigmoid in shape, with a central 
swelling, and deeply cleft at the free extremity. These two kinds of setae are 
connected by transitional types, of which there are two types: (i) longish, straight 
setae, with a swelling at about the end of the second or third, and faintly bifid at 
the extremity (e.g. in dorsal bundles of middle segments of Nais elinguis) ; (a) shorter, 
straight setae, sharp at the extremity, near to which is a swelling (e.g. in dorsal 
bundles of hinder segments of Nais elinguis) ; (3) straight, but with a bent, 
sickle-shaped extremity (Ghaetobranchus). These two latter are termed by Bourne 
' spear-shaped setae.' Besides these, there are the genital setae, which have as yet been 
only described in a very few forms ; they are sigmoid in shape, and not cleft at the 
< extremity. The number of setae in each bundle is usually more than two. 

In this group of Oligochaeta alone (with the exception of Branchiura, Hesperodrilus 
branchiatus, and possibly of the doubtful Alma nilotica of EHBENBERG=Dig'iii6raTOcAws 
niloticus, Lev.) the body-wall is prolonged into branchial processes. These are, however, 
found in only two genera, Bero and Ghaetobranchus; in the former they consist of 
a set of processes surrounding the anus, in the latter of delicate outgrowths in which 
the dorsal setae are imbedded. A more minute description of these organs will be 
found below. 

The nervous system of the Naidomorpha is in advance of that of Aeolosoma, and 
conforms to the type characteristic of the Ohgochaeta in general. The cerebral 
ganglia are free of the epidermis, and lie in the first segment; they are manifestly 
composed of two halves, each of which is, as a rule, furnished with a lateral lobe. 
A circumpharyngeal commissure unites these ganglia with the ventral ganglionated 

The nephridia of the Naidomorpha often show a peculiarity which has been also 


recorded in Aeolosoma. The nephridia, instead of being paired, are occasionally only 
present to the number of one in each segment. In Nais heterochaeta, for example, this 
appears to be the invariable condition ; moreover, in that species the nephridium 
occupies the two segments following that in which the funnel lies ; hence, as Benham 
(15, p. 385) remarks, there is really one nephridium in place of four. Here, however, 
as in Phreatothrix, the external pore is invariably upon the segment immediately 
following that which carries the funnel. This absence of one of the two nephridia of 
a segment culminates in Uncinais littoralis. In this worm so careful an observer as 
Bourne remarks (5, p. 351) that he 'was unable, even after repeated examination, 
to discover any nephridia.' 

The nephridia in the Naidomorpha are not present in the anterior segments of 
the body ; the segment in which they do commence varies according to the species, 
and there is frequently an omission of a segment or segments in the series ; as, for 
instance : — 

Bohemilla cofnata, viii, xi, xvi, xviii, xx, xxi. 

Pristina equiseta \ . „ 

Pristina breviseta ) 

Nais heterochaeta, vii, &c. (with some variation in individuals). 

Nais elinguis, viii, &c. 

That part of the nephridial tube which follows the funnel, but lies in the following 
segment, is nearly always covered with a brown glandular investment, also found 
in some of the Lumbriculidae and Tubificidae ; this appears to be absent in 
N. lacustris. 

Among the Naidomorpha the vascular system shows various modifications. In 
some it is more on a level with that oi Aeolosoma; in others it is distinctly like that 
of the Tubificidae. In the latter group, which is represented by N. josinae, the dorsal 
and ventral trunks are connected by a perienteric arch in all the segments of the 
body; in the former group it is only in the anterior segments of the body that this 
communication occurs directly. This being so, it is -possible to divide the Naido- 
moi-pha into two groups, of which one is nearer to the higher Oligochaeta. It is 
a matter of obvious interest to inquire whether the resemblance to the higher 
Oligochaeta is secondary — whether, in fact, the species which show this particular 
likeness are above or below the others in the scale. Fortunately, we are in 
possession of evidence which enables this question to be decided with a great 
amount of probability. Bouene has observed that in the buds of Uncinais 
littoralis there are a series of vascular arches, one to each segment; but that on the 
full-grown worm it is only the head-segments which are thus provided. 


The vascular system in the genus Ifais itself has been principally studied by 
Stolc and Vejdotsky. It is only in N. josinae that the dorsal vessel has a series 
of lateral arches in all the segments of the body ; in the other species there 
are only a few such vessels in the anterior segments. In N. elinguis, for example, 
there are three pairs of lateral vessels ; these are quite simple ; a fourth is sometimes 
present; three pairs of lateral vessels also occur in JSf. barbata, K lacvMris, and 
Naidiutn luteum ; in N. lacustris these vessels are in segments iv, v, vi ; in Naidium, 
in the same segments ; Pristina longiseta has four pairs of equally simplfe perienteric 
vessels in segments v, vi, vii, viii ; Pristina offers a further degree of complication in 
that the last pair are dilated. 

iV. seri^entina exhibits another kind of complication, a first step, which is after- 
wards more fully developed ; there are three pairs of trunks as in most other Naids, 
but the first of these are dichotomously divided ; in Bohemilla the same thing occurs, 
but the first pair of vessels is not only once dichotomously divided, but again. 
In iV. heterochaeta all of the lateral vessels give off branches which join the successive 
arches, and often make more than one connexion with the ventral vessel ; this wiU 
be found more fully described below, under the description of that species ; finally, in 
N. Josinae the anterior lateral vessels are hardly recognizable, as they form an elaborate 
network. In Uncinais littorulis there are still more pairs of lateral vessels; five 
pairs in all are figured by Boukne (5) in addition to the terminal bifurcation of the 
dorsal vessel to join the ventral vessel ; of these the first two and the terminal 
bifurcation anastomose by several branches ; these lateral vessels end in the seventh 
.segment. In Ghaetogaster the dorsal and ventral vessels are only united by one pair 
of lateral vessels ; the communication elsewhere is effected by paired branches from 
the network upon the intestine. 

It has been already mentioned that I^. Josinae has a pair of commissural vessels in 
every segment of the body; the same state of affairs is met with in the genus Bero. 
The circulatory system of this Annelid had been studied by Peeriee, Stolc, and 
by myself. As compared with the other Naidomorpha, Dero is very highly organized 
in its vascular system ; a variable number of the anterior lateral vessels are contractile ; 
I found that six were so in Dero perrieri ; four (BouSFiEiiD) is a more usual number. 
Stolc has described a network formed by the first few lateral vessels in Dero (? sp.), 
similar to that of N. josinae. 

In both Nais and Dero the intestine is invested by a close network of capillaries ; 
this network takes its origin from the dorsal vessel, and communicates with the 
ventral vessel by a series of median unpaired vessels, one to each segment. 

The reproductive organs in the Naidomorpha have been examined in a few species 


only ; but so far as the facts go they show a nearly absolute uniformity. The 
following are the species in which the sexual forms have been studied : — Nais elinguis, 
JV. lacubtris, K serpentina, Bero multibranchiata, and Dero sp. In all of these the 
reproductive organs are very far forward ; I do not include Uncinais littoralis in the 
list because there are no data in the account given by Bourne of the ducts, and it 
is not certain whether the organs which he describes as testes and ovaries are not 
the sperm-sacs and ovisacs respectively. 

The clitellum in Dero occupies segments v, vi, vii ^. The ventral setae are wanting 
on the sixth segment which bears the orifices of the sperm-ducts. In the species of 
the genus Nais enumerated above there are special genital setae on the segments named, 
which replace the ordinary setae. In all the Naidomorpha there is a single pair of 
spermathecae in segment v ; in the same segment lie the testes ; the sperm-duct is 
a short tube which expands into the spermiducal gland ; in N. elinguis the transition 
between the sperm-duct and the spermiducal gland is so gradual that both structures 
appear to be one. The ovaries are in the next segment to that which contains the 
testes ; the ripe ova escape by means of a pair of slits between the sixth and seventh 
segments. The sperm-sac appears to be unpaired. 

The Naidomorpha are characterized by the asexual mode of propagation as well 
as by the apparently rarer sexual process. This process has been studied by a large 
number of those observers who have dealt with the species belonging to this family ; 
more particularly to be mentioned in this connection are Leuckakt, Max Sohulze, 
Peeriee, Bouene, and Semper. Not having myself investigated the subject I do 
not attempt any criticism of these various observers, but content myself with a brief 
description of how the process takes place in the group according to the paper of 
Bouene upon the matter. I have furthermore added a short account of a recent 
memoir by Kallstenius which deals with the same process in the genus Amphichaeta. 
It is important to notice that the asexual method of reproduction in the Naidomorpha 
follows a more complex course than in Aeolosoma; in the latter it is merely division ; 
in the Naidomorpha there is a growth of fresh segments in the middle of the body 
of the parent worm. I have elsewhere (below) called attention to an appearance often 
presented by the earthworm Pontoscolex, which seems to be not unsuggestive of a trace 
of a similar process in the higher Oligochaeta. 

Budding in the Naidiforni Oligochaeta takes place, according to Bouene (5), in 
the following way : — 

1 Stieken says vii-xii for D. multibranchiata, but the epidermis showed no modification ; this part of the 
body was swollen (perhaps only by genital products?). 


A worm divides at a given spot (probably characteristic for the species) i; at this 
point growth takes place and a number of new segments are formed which are in 
two series, one on either side of the line of division already referred to. The anterior 
set consists of an indefinite number of segments and form the tail of the anterior of 
the two worms (produced by division of the original worm). The posterior set consists 
of a definite (and characteristic) number of segments which will form the head of the 
second worm. 

Kallstenius has described the same process in Amphichaeta. The worm which 
is not budding consists of seven segments. The first step towards asexual multipli- 
cation is the production of an eighth segment which comes to be separated from the 
seventh segment by a budding zone. This eighth segment becomes the fourth of 
the daughter bud, the three anterior being formed from the budding zone and those 
behind from the original segment— the eighth of the parent. The two individuals 
then come apart, leaving a budding zone and an' eighth segment upon the parent, 
which gives rise to further division. 

The process may be graphically represented thus : — 



— 8 

6 — 7 — X 


6 — 7 — X 1 X 


-7 — X 1 IX 2x 3x 

In considering the genus Aeolosoma I have pointed out the characters of certain 
species which form a transition to the Naidomorpha. Voeltzkow has observed in 
Madagascar a form like N. lacustris, which has in the integument numerous colourless 
oil-drops ; these are said to present almost the appearance of a black pigmentation. 
Chaetogaster, too, is said to have colourless oil-drops in the integument, and both these 
forms so far approach Aeolosoma, particularly A. niveum. 

There is one genus of wliicli I do not give a deliberate description with synonyms, &c., as its 
systematic position cannot at present be fixed with certainty. 

The position of Vetrovermis hyalinus is uncertain. It is described by Imhop as occurring in 
several lakes in Switzerland. The body is not ciliated, and there are only one pair of setae 
bundles in each segment. 'The setae are thin, quite straight up to either extremity, where they 
are easily bent from side to side ; at the free extremity they are bifid.' There is a swelling at the 
end of the first third of the seta. It multiplies by division, and sexual organs were not met with. 

' Not so, according to Benham (9), at least not for A', barhata and JV. lacusiris. 


The body consists of four segments. The ventral nerve-cord has a large number of ganglionic 
swellings, not always sharply marked oif from each other. 

Imhop regards it as nearly allied to Ctenodrilus, but remarks that the nervous system lies freely 
in the coelom, and is not as in that genus imbedded in the body-wall. It differs, however, from 
Ctenodrilus in having only two series of setae, and in this same character agrees with Chaetogaster. 
I may point out also that the arrangement of the nervous system (more ganglia than apparent 
segments) seems to agree with that of Chaetogaster. 

Genus Nais, 0. F. Muller. 
Syn. Opsonais, Geevais. 
Slavina, Vejdovsky. 
Ophidonais, Gekvats. 
Serpentina, Oersted. 
Stylaria, Lamabok, &c. 
Stylinais, Gervais. 
ITereis, Linnaeus (in part). 

DEFlWiTiOlsr. Dorsal setae commence in segment VI, capilliform only, or capilliform 
and uncinate, or straight with bifurcate end. Eyes present. 

I believe that we cannot separate the five 'genera' included in the above list of 
synonyms. Bourne allows all of them, with the exception of course of Opsonais; 
Vatllant, on the other, hand, merges Slavina in Nais, and Bousfibld includes 
Ophidonais in Slavina. All the various species, which I here include under one 
generic title, agree in the important fact that the first five segments are cephalized — 
that the dorsal setae do not commence until the sixth segment. It is true that two 
other undoubtedly distinct genera, viz. Bero and Ripistes, show the same character ; 
but the former of these is distinguished by the remarkable branchial apparatus at the 
tail end of the body, while Mipistes shows a more advanced degree of cephalization 
in that the ventral setae also are missing from segments iv, v. The genus Slavina 
of Vejdovsky only differs from Nais in two points : in the fact that the dorsal setae 
of the first bundles are much longer than those which follow, and in the presence 
of the cutaneous sensory papillae. As regards the first point, it may be mentioned 
that similar difierences occur in the species of the genus Pristina ; the sensory 
papillae seem to be wanting in S. gracilis of Leidy, which Vejdovsky associates with 
his genus Slavina; in refusing to regard these structures as of generic value I am 
in accord with Stolc and Bourne, who blame Bousfield for allowing this character 
to have led him to associate together Slavina and Ophidonais. At anj' rate, if these 



papillae can occur in species which differ in other comparatively important ways, 
they cannot be of such use for the purposes of generic definition as Vbjdovsky would 

have it. 

As to the identity of Ophidonais with Slavina, which Bousfield argues, and which 
Stolc and Bouknb dispute, I think it must be upheld. The only difference between 
the genera is in the setae; in Ophidonais the dorsal setae are all short and straight 
with a cleft extremity. If the dorsal setae of Ophidonais were peculiar to the single 
species assigned to that genus, this character would be unquestionably of some impor- 
tance ; but, as a matter of fact, the two forms are the two extremes met with in the 
genus Fais as defined here ; for Nais elinguis is exactly intermediate, having both 
kinds of setae in the dorsal bundle. Finally, as to Stylaria: really the only dis- 
tinguishing feature of this genus is the long prostomium; this structure has led to 
its being confounded with Ripistes (q. v.), with which it has not much in common. 
I do not believe that this character is sufficient to distinguish it as a genus distinct 
from Ifais. There is also a Pristina with a long proboscis that does not otherwise 
differ from its congeners — an argument from analogy. Minor has urged differences 
in the way of asexual reproduction; but as Vaillant has pointed out, he did not 
always find a distinction between the ' gemmiparity ' of Stylaria and the ' scissiparity ' 
of Nais. 

As thus constituted the genus Nais will include the following well-characterized 
species : — 

(i) Nais barbata, MiJLLEK. (6) Nais appendiculata, d'Udekem. 

(a) „ elinguis, Mullek. • (7) „ lurida, TiMM. 

(3) „ laeustris, Linn6. (8) „ josinae, Vejdovsky. 

v'(4) „ serpentina, Mollee. ^, (9) „ heteroehaeta, Benham. 

(5) » gracilis, Leidy. j (10) „ reckei, Floeeicke. 

In addition to these there are a number of worms which are very possibly members 
of this genus, but of which no sufficient descriptions exist. These species may be first 

The Nais fusea of Carter (Ann. and Mag. Nat. Hist. (3), ii. p. 20), from its habitat might be 
supposed to be distinct had it not been shown that these smaller OUgochaeta have often a wide 
range. Its chief difference from Nais elinguis is, as Vaillant has indicated, the absence of eyes; 
pending further information it may there be regarded as doubtfully distinct. 

A number of other species have been provisionally referred to this genus by Vaillant ; some 
of these, however, are clearly not referable to the genus, while others are insufficiently characterized 
to permit of their generic position to be stated with any approach to accuracy. They will be found 
cited on other pages. 


I am inclined to think that Flobeicke's genus Caecaria will have to be incorporated with 
Nais. He regards it as intermediate between Stylaria and Pristina. It has no eyes, a long prostomium, 
and the dorsal setae begin on the sixth segment. In Caecaria rara the second, third, and fourth 
dorsal bundles are much shorter than the others. In C. siUsiaca only the first bundle is shorter. In 
C. brevirostris all are equal. Pending farther information, I leave the species as incertae sedis. 

N. clamcomis of Saes (Beskrivelser eg iagttagelser, &c., Bergen, 1835), is apparently a Tomopteris. 

N. Upunetata (delle Chiaje), N. picta (Dujaedin), N. quadricuspidata (Fabeicius) are also 

N. Carolina (Blanchaed) is quite unrecognizable. 

N. scotica is referred to below. 

(1) Nais barbata, 0. F. Mullee. 

? N. barbata, 0. F. Mullee, Verm, terrestr., 1774, vol. i, pt. ii, p. 23. 
I Opsonais obtusa, Geevais, Bull. Ac. Roy. Belg., 1838, p. 17. 

If. elinguis, Dieffenbach, Anat. u. Syst. Studien, 1885, p. 98 (in part.). 

Deflmtion. Dorsal setae bundles with capllliform setae only, which are of two sizes, four to 
eight in each bundle. Eyes present. 

The brief diagnosis of this species given by 0. F. Mullee would not be enough to identify 
even the genus to which the worm belongs, did he not refer to his account of the bearded Naid 
(die bartige Naide) published with a plate in an earlier work (2). The drawings referred to 
indicate a Nais or a Bohemilla, more probably the former, since the much smaller eyes of the latter 
genus would perhaps have escaped Mtjllee's attention. There is, however, nothing in the figure 
to enable one to determine with certainty the identity of this ' bartige Naide ' with the species 
subsequently called Nais harbata by Taubee, Vejoovsky, and others. The only distinction between 
the 'bearded Naid' and the 'tongueless Naid' mentioned by Mullee is the 'beard' of the former, 
which is evidently the ventral setae of the four first setigerous segments closely crowded together', 
for MtJLLEE says (2, p. 81) 'unter dem Kopf erscheint ein Elumpen kurzer Borsten gleich einem 
Barte,' and is, therefore, not a distinction. Strictly speaking, the name ' harbata ' ought perhaps to 
be dropped, and replaced by Gervais' 'obtusa,' but as the former name is now so well established, 
such a change is inadvisable. 

This species has been greatly confused with the nearly allied 0. elinguis. d'Udeeem 
distinguished them only by the presence or absence of the glandular ventricle. This, 
however, is no difference ; the structure in question appears to be present in both ^. 
Taubee (p. 73) first indicated the only real difference between the two species 

' DiErrENBACH (p. 104) evidently did not so interpret the structure in question when he wrote 'eine 
Nais barbata . . . mit besonderen kleinen Haaren am Munde, habe ich nie fiuden kounen.' Vaillakt's figure 
(6, PI. xxii. fig. 14), by exaggerating the tactile hairs of the prostomium, might lead to the inference that 
these were the ' beard.' 

' Tejdovskt, at least, does not refer to its absence in 0. barbata; but DiErrEHBACH says it is occasionally 
absent in his N. elinguis ( = 0. elinguis + 0. barbata). 



which concerns the seta, and is mentioned in the diagnosis of the two worms 
given here. Another difference referred to by Taubee is that in 0. barbata 
the ova are deposited in October and November^. As Taubee pointed out, Sempee 
also confused these two species. So, too, Dieefenbach (p. 99) who regarded the 
individuals with bifurcate setae in the dorsal bundles as a variety of I^. elinguis. The 
species is figured by Mullee, Vaillant (6, PI. xxii, figs. 14, 15). Floeeicke has 
applied the name of JSf. greeffi, to a worm which I regard as a variety of the present 
species ; the ventral setae are, as in N. elinguis, of equal length throughout the body ; 
whereas in N. barbata those upon the first four segments are longer. 

(a) Nais elinguis, 0. F. Mullee. 

J W. elinguis, O. F. Mullee, Verm, terrestr. vol. i, pt. ii, 1774, p. 22,. 
Opsonais elinguis, Geevais, Bull. Ac. Eoy. Belg., 1838, p. 17. 
N. rivulosa, Leidy, Journ. Acad. Nat. Sci. Philad., 1850, p. 43. 

Definition. Dorsal seta bundles with capilliform and hastiform setae, one to three in each 
bundle. Ei/es present. Hob. — Europe; N. America. 

I think that Leidy's N. rivulosa cannot be differentiated from N. elinguis. This 
is also the opinion of Vaillant (6, p. 371) and d'Udekem (5, p. 2,0). Vaillant, 
however, suggests as a possible difference the greater slenderness of the ventral setae 
in the American species. 

(3) Nais lacustris (Linnaeus). 

Nereis lacustris, LiNNAEUS, Syst. Nat., i J67, p. 1085. 

Nais proboseldea, 0. F. Mullee, Verm, terrestr. vol. i, pt. ii, 1774, p. ai. 

Stylaria proboscidea, Eheenbeeg und Hempeich, Symb. Phys., 1831. 

Stylaria paludosa, Lamaeck, Hist. An. sans Vert., 181 6, iii, p. 234. 
, Stylinais proboscidea, Geevais, Bull. Ac. Eoy. Belg., 1838, p. 18. 
"^Stylaria lacustris, Johnston, Cat. Worms, 1865, p. 70. 
? Stylaria phyladelphiana, CZBENIATSKY, Bull. Soc. Nat. Mosc, 1880, p. 309. 
? Stylaria fossularis, Leidy, P. Acad. Nat. Sci. Philadelphia, 185a, p. 287. 

non-TS. lacustris, Dalybll, Powers of the Creator, ii, 1853, P- 13° ^• 

Definition. Length, i^m.; number of segments, ^S- Prostomium much elongated. Dorsal 
seta bundles with capilliform setae only. Eyes present. Hab.— Europe and America. 

As Vejdovsky (24, p. 29) has pointed out, the ova of 0. dmguis, though usually laid in June, are 
sometimes deposited in autumn, so that this difference is not absolute. 
^ Chaetogaster. 


This species is referred by recent systematists (Vejdovsky, Bourne, and others) to 
a distinct genus. I prefer to follow d'Udekem and Levinsen in regarding it as 
congeneric with iV. elinguis, &c. If any species of this genus is to be removed from 
the others, it should be O.Josinae. The only character upon which a generic definition 
could be based is the long prostomium. Were we acquainted with a considerable 
number of species showing this character, it might be advisable to give it generic 
value. In the meantime, as there is only one species certainly known, it seems more 
consistent to lay stress upon the similarity in the cephalization with other species. 

This Annelid is one of tlie best known. It has been studied by Bonnet, Tremble y, Mullee, 
GtEUITHUISEN, Taubee (2, 8), Vejdovsky (24), and others. 

There is absolutely nothing in the very imperfect description given by Leidy of his two species 
' Stylaria paludosa,' and ' Stylaria fosfularis' to enable one to understand on what grounds they are 
separated from ' Stylaria lamistris.' 

The worm measures IQ-15 mm. in length; the prostomium is long and slender; 
its length is put down by Mullee as equal to ten segments of the body. The setae 
of the dorsal bundles (two or three to each bundle) are long and capilliform ; the 
ventral setae (four to six per bundle) are sigmoid and cleft at the extremity ; there 
is sometimes a median swelling. Bourne ' occasionally observed specimens in which 
dorsal setae were present in the one or two most posteriorly placed of the usually 
cephalized segments.' 

The species is illustrated by Blainville (PI. xxiv, fig. 3), Cuvier (PI. xxi, 
figs. 2, 3 a), Dal yell (PI. xvii, figs. 6, 7), d'Udekem (4, PI. iii, figs. 17-ai), Vejdovsky 
(24, PI. iii, fig. ay, K- iv, figs. 1-24, 26-31). 

(4) Nais serpentina, 0. F. Mullek. 

H". serpentina, O. F. MiJLLER, Verm, terrestr., vol. i, pt. ii, 1774, p. 20, 
t\ Ophidonais serpentina, Gbevais, Bull, Ac. Roy. Belg. 1838, p. 19. 

Serpentina quadristriata, Oersted, Nat. Tidsskr. 1842, p. 134. 
? Ophidonais vermicularis, Gervais, Bull. Ac. Roy. Belg. 1838, p. 19. 

Slavina serpentina, Bousfield, J. Linn. Soc. 1886, p. 268. 

Definition. Dorsal setae short, straight, and hifurcate. Ventral setae of first four bundles 
with a median thickening; on setae of posterior segments the swelling nearer the fixed 
end. Three or four pigmented bands on segments II- V. Hob. — Europe. 

This species, described by Rosel and 0. F. Mullek (2) (by the latter under the 
name of 'Die geschlangelte Naide') in the last century, has been chiefly studied by 


Lankestee and Vejdovsky. It is the largest of European Naids, measuring, according 
to the last-named observer, 1-5 to 2 centimetres in length. A chain of four Zooids 
measures as much as 3 cm. 

An illustration showing the natural size of the woim is given by 0. F. Mullee (2, Tab. iv, 
fig. i). In a magnified representation (2, Tab. iv, fig. 2) the characteristic transverse bands of 
pigment, which led to Oeested's specific name of ' giiadristriata' are shown. A drawing byn'TlDEKEM, 
reproduced by Vaillant in the third volume of the ' Histoire des Anneles' (PL xxiii, fig. 12), 
shows the same feature, also recognizable in the first published illustration of the worm — that by 
RbSEL. Ophidonais seipentina is not only larger, but more opaque than other Naids, and is, therefore, 
less easily studied. It is, therefore, doubtful whether Williams (1) really investigated this species, 
since he speaks of 'the softness and transparence of the integuments.' 

The pharynx occupies the first four segments, being somewhat constricted in the 
middle and dilated at the two extremities. The oesophagus has no glandular swelling 
but passes (in the ninth segment) into the intestine ; the intestine as well as the last 
three segments of the oesophagus, are covered with brown peritoneal cells. The dorsal 
blood-vessel communicates with the ventral by three pairs of vascular arches as well as 
at the anterior extremity, where it divides into two branches which reunite to form 
the ventral vessel. The two last pairs of arches, belonging to segments ill and iv, are 
quite simple ; the first pair branch dichotomously, the branches join either one of the 
two halves of the dorsal vessel or the single ventral trunk. This branching is evidently 
a simplification of the conditions which occur in N. josinae (p. 288). This species has 
been found with fully developed sexual organs in June (Lankestee), September, and 
October (Vejdovsky). 

Gebvais (pp. 1 9, 20) describes both 0. vermiciilaris', and 0. serpentina, but declares that it is 
impossible ' bien caraoteriser deux especes de ces animaBx ' ; above Gebvais remarks upon having 
taken 'plusieurs fois des Nais vermicularis ou serpentina, a Paris.' I presume, therefore, that I am 
right in adding 0. vermicularis to the list of synonyms. 

The species is figured by Mullee, Beuguieee (PI. liii, figs. 1-4 [after MuLlee]), 
Vejdovsky (24, PI. iii, figs. 14-16), Vaillant (6, PI. xxiii, fig. 12), Oeested (3, PI. iii, 
fig. 3), and BousFiEi;D (1, PI. xxxiii, fig. 5). 

(5) Nais gracilis, Leidy. 

N. gracilis, Leidy, Journ. Acad. Nat. Sci. Philad. 1850, p. 43. 
. Slavina gracilis, Vejdovsky, Syst. u. Morph. 1884, p. 30. 
Definition. Length, 10 mm.; number of segments, 50; dorsal setae capUUform only ; those of 

' 'Hais vermicyiaris audorum,' Geevais. 


segment VI longer than the rest ; dorsal bundles of this segment contain three setae each ; 

in the following segments only one dorsal seta on each side. No glandular ventricle. 

Eyes present. Hob. — Neighbourhood of Philadelphia. 
This species evidently belongs, by virtue of the specially elongated setae of the 
sixth segment, to Vejdoysky-'s genus Slavina, which, as it appears to me, Vaillant 
was right in joining with 'Nais' {=:Opsonais). N. gracilis, however, is probably not 
identical with N. appendiculata, for Leidy makes no mention of the sensory papillae 
which, if present, would hardly have escaped his attention. Furthermore, it seems to 
differ in the limitation of the number of the dorsal setae, and in the absence of 
a glandular ventricle. 

(6) Nais appendiculata, d'Udekem. 

N. appendiculata, d'Udekem, Bull. Acad. Roy. Belg. t. xxii, pt. ii. 1855, 

P- 552- 
^ Slavina appendiculata, Vejdovsky, SB. Bohm. Ges. 1883, p. 219. 
?N. escherosa, Geuithuisbn, Nova Act. Ac. Nat. cur. 1828, p. 409. 

Definition. Dorsal setae cainlliform only, those of sixth segment very much longer than those 
of other segments. A row of sensory papillae upon all the segments of the body. Hah. — 

There can be no doubt that Slavina appendiculata of Vejdovsky is the same species 
as iV. appendiculata of d'Udekem. The characteristic sensory papillae are referred to 
by d'Udekem as 'appendices couverts de spicules,' and figured (fig. 3 of his Plate 
illustrating Memoir). It has been also thought to be identical with the N. lurida of 
TiMM, with which, however, it cannot be confounded. 

The worm is figured, in addition to the above quoted authorities by Vejdovsky (24, 
PI. iii, fig. 17-26), BousFiELD 1 (PI. xxxiii, figs, a, 3, 4, copies from Vejdovsky and 

(7) Nais lurida, Timm. 

N. Ixirida, Timm, Arb. Zool. Zoot. Wiirzb. 1883, p. 153. 
Slavina lurida, Bousfield, J. Linn. Soc. 1886, p. 268. 
Definition. Number of segments, 40; first pair of dorsal setae very long ; not more than two in 
each bundle. Two rows of infegumental sense bodies on each segment. Hah. — Europe. 
Bousfield has re-examined this species, and shown it to be quite distinct from 


iV. appendiculuta, with which it has been confounded by Vaillant. A ventral view 
of the anterior segments is given by Bousfield. 

(8) Nais josinae, Vejdovsky. 
N. josinaej Vejdovsky, SB. Bohm. Ges. 1883, p. 218. 

Definition. Lengthy 8 mm. Dorsal setae bundles, consisting of loth capilliform and uncinate 
setae ; ventral setae uncinate only. In anterior segments the dorsal vessel forms a network. 
Eyes absent. Hab. — Teufelsee, Bohemia. 

This species has been carefully studied by Vejdovsky (24) ; it is impossible to say 
whether it is or is not identical with N.fusca of Caetee and N. scotica of Johnston. 
The worm is described by Vejdovsky as resembling a Tubifex on account of its size 
(6-8 mm. in length, 15 mm. when dividing). The setae at once distinguish it from 
0. elinguis, for the bifurcate setae of the dorsal bundles are sigmoid, and not straight as 
in that species. The dorsal vessel anteriorly becomes reduced in calibre ; it gives off, in 
the first five segments, five or six pairs of lateral branches, which form an irregular 
network ; in the posterior segments pairs of lateral trunks (one pair to each segment) 
with the dorsal and ventral trunks. In segments vii, viii, and ix, these vessels are shorter 
and stouter, and form veritable hearts, comparable to those of the higher Oligochaeta. 
The organization of the vascular system of this species places it on a higher level than 
the other members of the genus to which it may be (in my opinion only temporarily) 
refeiTed. A more complete knowledge of its structure may very possibly necessitate 
the creation of a new genus for its reception. 

(9) Nais h.eterocliaeta> Benham* 

N. heterochaeta, Benham, Q. J. M. S. vol. xxxiv, 1893, p. 383. 

Definition. Length, 6 mm.; number of segments, 41. Dorsal bundles consisting of one 
capilliform and one uncinate seta, liyes present. Vascular trunks uniting dorsal with the 
ventral vessels of anterior segments connected successively, Hab. — England. 

This species has a remarkable vascular system, of which the following is a more 
detailed description. The dorsal vessel divides, as usual, in the prostomium into two 
lateral trunks, which unite in the fifth segment. The branches running from the 
dorsal to the lateral trunks in the second segment divide, each of them, into two with 
a separate union with the laterals. The corresponding circum oesophageal vessels of the 
next segment give off a branch which joins the perioesophageal of the fourth segment. 


The latter give off a similar branch which, before uniting with the perioesophageal 
of the fifth segment, is connected twice with the lateral vessel of its side. The 
perioesophageal vessel of the fifth segment gives off a branch which joins that 
of the sixth ; the latter bifurcates, and has two separate unions with the ventral vessel. 
This is an approach to the irregular network of N. Josinae. There are slight variations 
in the setae, sometimes two uncinate or no capilliform in the dorsal bundles, but 
the above is the normal arrangement. The pigmented covering of the intestine begins 
in the sixth segment. 

(lo) Nais reckei, Floebicke. 
Ophidonais reckei, Floebicke, Zool. Anz. 1892, p. 469. 

Definition. JJorsal setae sluyrt. straight, and pointed. Other characters as in N. serpentina. 
Hah. — Germany. 

The setae of this species are unlike those of any other Naid. They ai-e straight and 
cylindrical, .and suddenly end in a fine point. 

Genus Pbistina, Ehbenbeeg. 

Syn. Pristina, Ehrenbebg. 
Pristinais, Geevais. 
Wais, d'Udekem (in part.). 
Stylaria, Taubee (in part.). 
Naidium, Schmidt. 

DEFrisriTioiT. Dorsal setae commence in the second segment, capillary only or 
capillary and hastiform, the latter bifurcate or not bifurcate at extremity. 
Eyes absent. Glandular ventricle present. Septal glands well developed. 

I venture to differ from all recent authorities, in combining the two genera 
Pristina and Naidiwni. This step appears to me to be rendered necessary by 
Boubne's (5) description of two species which nearly entirely bridge over the gap 
which formerly separated the two genera. Previous to the publication of that paper 
the two genera might be thus distinguished. 

Peistina. Naidium. 

(i) Dorsal setae of segment iii very long. (i) These setae not longer than the others, 
(a) No bifid hastiform setae- in dorsal (a) Such setae present in dorsal bundles. 

(3) Prostomium elongate. (3) Prostomium short. 



The two new species of Bourne — Pristina eqimeta and P. breviseta — do away 
with the first and second diiFerential characters. In view of the identity in the 
cephalization of the two genera, the remaining points of difference are, perhaps, 
hardly of sufficient importance to warrant generic separation, particularly as only 
one species of Naidiwtn is known, and its range of specific variation, therefore, not 
ascertained. In both Naidium and Pristina the restriction of the pharynx to two 
segments, and the great development of the septal glands are especially remarked 
upon by Vejdovsky, and constitute a bond of union, which may be considered to 
counterbalance the difierences mentioned above. 

The well-marked species belonging to this genus are only five, viz.: — 
(i) Pristina longiseta, Eheenb., Europe and N. America. 
(3) „ equiseta, Boukne, England. 

(3) „ breviseta, Bourne, Madras. 

(4) „ lutea, 0. ScHM., Europe. 

(5) „ proboscidea, F. E. B., S. America. 

One or two doubtful forms have also been described too incompletely for satisfactory 
recognition ^ 

Pristina inaegualis of Eheenbeeg lias been tlius defined by bim : — 

'Setis quaternis inaequalibus, una longissima, reliquis brevissimis, pari secundo non diverso, 
uncinis, subquinis subulatis.' 

Pristina hreviceps {= Naidium breviceps, 0. Schmidt) is tbougbt by Vejdovsky to be an 

Nais temaria of Schmaeda is doubtfully transferred by Vaillant (6, p. 357) to Naidium. Tbe 
cbief reasons for believing it to be a member of tbe genus Pristina are tbe absence of eyes, and 
tbe commencement of botb dorsal and ventral setae upon tbe same segment. Tbe dorsal setae 
are capillary only. The intestine is said to have a spiral arrangement. Tbe woi-m occurs in 
Louisiana, Central America, Cuba, and Jamaica. 

(i) Pristina longiseta, Ehrenberg. 

p. longiseta, Ehrenberg, Symb. Phys. 1828. 

Pristinais longiseta, Gervais, Bull. Ac. Eoy. Belg. 1838, p. 17. 

Nais longiseta, d'Udekem, Bull. Ac. Roy. Belg. t. xxii, pt. ii. 1855, p. 553. 

Stylaria longiseta, Tauber, Ann. Dan. 1879, p. 'j^. 

Definition. Lenytk, 8 mm. ; number of segments, 20. Dorsal setae ■ of segment III much 
longer than the other. Prostomium long. Glandular ventricle in segment VIII. 
Five vascular arches connecting dorsal and ventral trunks. Nephridia commence in X. 
Hah. — Europe. 

' It is possible that 'Lumlricus Mrsutus' of Dalyell (p. 140) is referable to this genus. 


This is a small species, not measuring more than 8 mm., or a mm. according to 
Leidt. Vaillant has thrown some doubt upon the identity of the American and 
European forms on account of this difference in size. 

It may also be pointed out that Vejdovsky's figure of this species (24, Taf. ii, fig. 13) indicates 
no abrupt demarcation between the buccal segment and the comparatively short prostomium ; in 
the text (24, p. 31) it is remarked 'Der Eopflappen verjtingt sich allmalig zu einem konischen, 
fadenformigen Rtissel.' This ' gradual ' narrowing is not shown in Leidy's figure, nor in d'Udekem's ; 
and in a single example (English, but ? exact locality) examined by myself, the prostomium was 
as illustrated by Leidy and d'Udekem, not Vejdovsky. The specimen examined by me being 
about 4-5 mm. in length connects the two extremes of 2 mm. and 8 mm. As Vejdovsky leaves out 
Leidy's Pristina longiseta in the list of synonyms given under the species, and does not mention 
N. America as a locality for the species, but is acquainted with Leidy's paper ', I am disposed to 
infer that Vejdovsky is not certain about the identity of his and Leidy's varieties. The question 
must, I think, be left open for the present. 

The prostomium appeared to me to be grooved upon the lower surface, this gi-oove 
opening into the mouth. The number of segments is 17-20. The dorsal setae are 
entirely capiUiform, two or three are found in each bundle^; those of the third segment 
are, according to Vejdovsky's figure, about three times the length of those of other 
segments. The ventral setae are sigmoid and cleft at the extremity ; there are seven 
or eight to a bundle. The peculiarities of the alimentary canal have been referred to 
under the genus. The dorsal vessel is connected with the ventral by four pairs of 
vascular arches in segments iv-vii (Vejdovsky, 21, p. 1x3), of which the last pair is 
stouter than the rest. These vessels are entirely unbranched and arise from the dorsal 
vessel just behind the dissepiment. The nephridia commence in the tenth segment. 
The perivisceral corpuscles have black, brown, or green contents. 

VoELTZKOW has recorded from Madagascar a species very like P. longiseta, but no 
special description is given. 

(3) Pristina equiseta, Bourne. 
p. equiseta. Bourne, Q. J. M. S., vol. xxxii (1891), p. 352. 

Definition. Length, 8 mm.; number of segments, 21. Setae of segment III not elongated; 

prostomium long. Glandular ventricle in segment VIII. One vascular arch. Nephridia 

commence in IX. Hal. — England [?). 

This species was discovered by Bourne in the ' Victoria regia ' tank, in the Gardens 

of the Koyal Botanical Society, in Kegent's Park, London, whence I had myself 

I It is not only quoted in list of literature, but the species is referred to in the history of the group, 

P p 2 

P- 23- 

' Three to six are figured by Vejdovsky, three by Leidy and d Udekem, 


obtained specimens. It may not, therefore, be a native of this country. It measures 
8 mm. in length and consists of 18-21 segments. The most prominent difference from 
the last species is the absence of specially elongated setae from the dorsal bundles 
of segment iii. This is shown in an illustration (Zincograph) ' facing p. 337 of the 
memoir. The ventral setae of the fourth segment were often larger and stouter than 
the rest ; it is suggested that these may be genital setae, though no other indications 
of sexual maturity were met with. The segment in which the single pair of commis- 
sural vessels lie is not mentioned. The coelomic corpuscles are large and greenish ; 
sometimes quite black owing to quantity of secreted particles. On account apparently 
of these the worms look of a chalk-white colour when seen with the naked eye or 
a lens. 

(3) Pristina breviseta, Bouene. 
p. breviseta, Bourne, loc. cit. p. ^^^. 

Definition. Dorsal setae of two hinds, capilliform and hastiform, the latter cleft at extremity. 
Capilliform setae of segments II and III shorter than following. Nephridia commence 
in IX. Hah. — Madras. 

This Pristina is larger than the other species, and consists of about forty-six 
segments. It differs from all the species of the genus except P. lutea in having 
bifurcate setae in the dorsal bundles. The shape of these, however, is a little different 
in the two species. The coelomic corpuscles ' are black and very noticeable.' It has 
•been mentioned above that this genus is characterized by having only one cephalized 
segment, as in the vast majority of Oligochaeta. This condition, however, appears, 
so far as the present genus is concerned, to be secondary. Both in this species and 
in the last Eoukne has found that the newly budded head consists of seven segments. 

(4) Pristina lutea (0. Schmidt). 

Waidium luteum, 0. Schmidt, Froriep's Notizen, iii. 1847, p. 332. 
? Nais caeeilia, Mayee, SB. Niederrh. Ges., 1859, p. 45. 

Definition. Length, i^mm.; number of segments, 30. Dorsal setae, capillary and hastiform. 
Hearts in V-VII. Glandular ventricle in VIII. Hob. — Mbe. 

I have already indicated (p. 289) the reasons which appear to me to render necessary 
the inclusion of this species in the genus Pristina. Our knowledge of its structure 
is chiefly due to Vejdov^ky (24), who has illustrated his description. 


(5) Pristina proboscidea, new species. 

Definition. Dorsal setae capillary. Prostomiwm very large as in N. lacustris. Hab. — 

S. America. 
On a casual glance this would be undoubtedly referred to the species N. lacustris. 
As BouENE has found in N. lacustris dorsal setae occasionally present on some of the 
segments normally devoid of them, the present worm may be an extreme variation. 

Genus Ripistes, Dujaedin. 

Pterostylarides, Czeeniavsky. 
Stylaria, O. Schmidt (in part.). 
Nais, Geebnitzky (in part.). 

Definitioit. Prostomium elongated. Dorsal setae commence in the sixth segment, 
the first three pairs of bundles containing setae of great length ; ventral setae 
commence in the second segment, but are absent from IV and V. Eyes present. 

Though neither Vejdovskt (24), nor Vaillant (6) accept this genus, I think that 
BouENE (5) is quite right in emphasizing its distinctness from Stylaria. But in using 
Czeeniavskt's name Pterostylarides, the name Ripistes of Dujaedin, which clearly 
has the priority, was overlooked (as Vaillant, 6, p. 366, has pointed out). 
Czeeniavskt's definition does not include a reference to what is really the most 
important character of the genus, viz. the absence of both dorsal and ventral setae 
from segments iv and v. But the description of the three pairs of extremely long 
dorsal setae leaves no room for doubt as to the identity of Pterostylarides and 
Ripistes. The characters of the cerebral ganglia strengthen the arguments for 
regarding ' Stylaria parasitica ' of 0. Schmidt as the type of a genus distinct from 
Stylaria. It is, according to Vejdovsky's figure (24, PI. ii, fig. xo), very short — not 
prolonged backwards, as is the case with Nais. There are two species, viz. R. para- 
sitica and R. macrochaeta. 

(i) Ripistes parasitica (0. Schmidt). 

Ripistes sp., Dujardin, Proc.-Verb. Soc. Philom. Paris, 1842, p. 93. 
Stylaria parasitica, 0. Schmidt, Froriep's Notizen, iii, 1847, p. 331. 
Nais parasitica, Geebnitzky, Zapiski novoross. Obstch. Est., 1873, p. a68. 
Pterostylarides parasitica, Czeeniavsky, Bull. Soc. Nat. Mosc, 1880, p. 310. 
Definition. Length, 6 mm. ; number of segments, 20 ; prostomium about as long as peristomial 


segment. Ventral setae of segments II and III about one-third longer than setae of 
other segments. Five or six setae in the anterior bundles, seven or eight in the others. 
Twelve to fifteen setae in three anterior dorsal bundles, all very long. Hob. — Europe. 

This species has been chiefly studied by Vejdovsky (24, PL ii, figs- S-xa). 

(a) Ripistes maeroehaeta (Bourne). 
Pterostylarides maeroehaeta, BoxJENB, Q. J. M. S., vol. xxxii, p. 349. 
Definition. Prostomium mmch longer than peristomial segment. Ventral setae bundles of 
segments II and III consist of two or three setae, those of other segments of two to five 
setae. Three anterior dorsal seta bundles contain each two to five very long setae, the rest 
being shorter than the other dorsal setae. Hob. — -Vlngland. 
This species was described, with an illustration, by Bourne, in 1891. The descrip- 
tion is limited to the external characters which alone are . indicated in the figure. 
A note appended to the description of the species by ' E. R. L.' states that ' the long 
setae are frequently found thrown forward so as to partly encase and protect the head 
when the worm forms for itself a temporary tube. They are also used to strike the 
water in swimming.' 

Genus Unoinais, Levinsen. 

8yn. Paranais, Czerniavsky. 
Nais, AuCT. 

Clitellio, Vaillant (in part.). 
Ophidonais, Vejdovsky (in part.). * 

Enehytraeus, Minor (in pai-t.). 
Defiktitioit. Setae entirely uncinate ; dorsal setae commence in segment V. On 
segment V of sexually mature individuals genital setae replace the ordinary 
ventral setae. Glandular ventricle present. ISa nephridia (?). Testes (one 
pair) in VIII and IX. Ovaries in X. Spermathecae in V. Eyes absent. 

The only well-known species of this genus is TJ. littoralis ; under this heading the 
synonymy of the species and genus will be discussed. 

The above generic definition is entirely compiled from Bourne's (5) notes upon 
U. littoralis, which establish its generic rank. Bourne figures a budding worm and 
also the reproductive organs and setae of the sexually-mature worm. The most 
remarkable external characters of the genus are the existence of uncinate setae only, 
in both dorsal and ventral setae, and the absence of the dorsal setae from all segments 


anterior to the fifth. These points distinguish it from all other genera of Naidomoi-pha. 
The vascular system consists of a dorsal and ventral vessel which run to the end of 
the body. In segments ii-iv they are joined by three pairs of circular vessels, which 
are branched, and, according to Bourne's figure (5, fig. a), form a network comparable 
to that of N. Josinae, or perhaps more like that of N. heterochaeta ; but there is no 
irregularity in the origin of the circular vessels, and there are fewer pairs of them. 
In segments v-vii the dorsal and ventral trunks communicate by means of unbranched 
circular vessels. Bourne records the interesting fact that in the young buds com- 
missural vessels occur in all the segments. 

The absence of nephridia is, if confirmed, a very remarkable character. Behind 
the pharynx occur a pair of glands which probably represent the pharyngeal glands 
of Ifais and other genera. 

The arrangement of the sexual organs differs from that found in I^ais and in 
Clitellio, to which genus the worm has been referred by Vaillant. They indicate, 
however, a transition between the more typical Naids and the Tubificidae. 

As in the Naids the paired spermathecae lie in segment v, opening in front of the 
genital setae which replace the ventral setae of this segment ; the pouches are oval 
in form and restricted to the fifth segment. The genital setae are ' very stout, and 
longer than the ordinary setae, and they possess a mere rudiment of the crotchet at 
the free extremity.' The testes and ovaries lie further back than in other Naids, 
but a segment in front of the position which they occupy in the Tubificidae. The 
testes^, of which there are only a single pair, occupy nearly the whole of segments 
viii and ix; each testis is continuous through the septum which separates these 
segments — a condition which seems to be paralleled only in the case of the genus 
Phreodrilus. The ovaries lie in the following segment — the tenth. Neither oviducts 
nor sperm-ducts are mentioned or figured. 

(i) Uneinais littoralis (0. F. Mullee). 

Nais littoralis, O. F. Mullee, Zool. Dan. 1788, ii, p. 54. 
^'Paranais littoralis, Czerniavsky, Bull. Soc. Nat. Mosc, x88o, p. 311. 
Uneinais littoralis, Levinsen, Vid. Med. 1883, p. 218. 
Clitellio (Clitellio) arenarius, Vaillant, Annelds, p. 415. 

Definition. Frostomium Munt and rounded. Setae of segments II, III, IV longer and 
thinner than those which follow. Ventral setae of most segments thicker and shorter 
than dorsal setae. Hal. — Ewojae. 

• Possibly the structures called 'testes' and 'ovaries' by Boukne are sperm-sacs and egg-sacs. 


The synonymy of this species is a matter of great difficulty. It has been confounded with , 
CUtelUo arenarius by Vaillant. Having myself investigated CUtellio arenarius of Clapae4;de (3), 
I can state positively that it is not identical with Faranais Uttoralis of Boubne. Zengee's Peloryctes 
inquilina being probably synonymous with Hemitubifex benedii, must also be removed from Vaillant's 
list. As d'Udekem would hardly have overlooked a cephalization in his ' Tubifex hyalinus,' if it existed, 
it is probably safe to remove this species also from the list; it is indeed queried by Vaillant. 
Another synonym queried by Vaillant is Monopylephorus rubroniveus of Levinsen ; I am disposed 
to regard this as congeneric with Bothrioneuron of Stolc. As there is nothing in Czeeniavsky's 
definition of the genus Faranais, or of the species P. Uttoralis to warrant its identity with the species 
described by Bourne under the name of Faranais Uttoralis, I venture to use the generic name Uncinais 
of Levinsen, the definition of which does fit in. The following are the definitions referred to. 


'Setarum fasciculi utrimque bi-seriati, omnes et superiores et inferiores setis uncinatis formati. 
Corpus lineare teres, postice subtruncatum. Ocelli duo vel nuUi.' 

' Alle B0rsterne Krogbi^rster . . . Rygb(^rstEr mangle i de f(^rste 3 b^rstebserende Ringe ; kun 
fi^rste Rings Burster tydelig laengere end de 0vrige 5-6 i hvert Bundt, senere 3-4; c^verste 
B^rstegren noget smallere end nederste.' 

(3) Uncinais uncinata (Oehsted), 

Nais uncinata, Oersted, Nat. Tidsskr. 1842, p. 136. 

Enchytraeus triventralo-pectinatus, Minor, Am. Journ. Sci. Arts, xxxv, J 863, 

P- 35- 
Faranais uncinata, CzEENiAVSKT, Bull. Soc. Nat. Mosc, 1880, p. 311. 
Ophidonais uncinata, Vejdovsky, Syst. u. Morph., 1884, p. 24. 
Uncinais uncinata, Levinsen, Vid. Med., 1884, p. a 18. 

Definition. Length, 1 1 mm. ; number of segments, 2(^-25. Setae four in each bundle, always of 
the same size. Hab. — Europe ; America. 

Genus Bohemilla, Vejdovsky. 
Syn. Nais, Timm. 

DEFiBTiTiOKr. Dorsal setae begin on segment V; the longer dorsal setae serrated. 
Eyes present. 

This genus, instituted by Vejdovsky, seems to be a perfectly good one. It only 
contains one species, viz. 

Bohemilla comata, Vejdovsky. 

B. comata, Vejdovsky, SB. Bohm. Ges. 1883, p. 218. 
Nais hamata, Timm, Arb. Zool. Zoot. WUxzb., 1883, p. 152. 


Definition Length 6mm. ; number of segments 38. First pair of ventral setae bundles larger 
than those in two following segments; those of IF very small, sometimes absent. Hab. 
— Europe. 

This species has been chiefly investigated by Vejdovsky. 

Genus Dero, Oken. 
Syn. Proto, Oersted. 

XJronais, Gervais. 

Xantho, Ddtrochet. 

Aulophorus, Schmarda. 

Nais, 0. F. MiJLLER (in part). 

DEFlwiTioiir. Dorsal setae eapilliform and hastiform, commencing upon the sixth 
segment. Branchial processes present at hinder end of body. Eyes absent. 
Inhabit tubes 

The principal character of this genus is, of course, afforded by the branchial 
processes at the end of the body, which are dorsal in position. These are ciliated 
processes of the integument containing blood-vessels; they really spring from the 
expanded and funnel-shaped termination of the anus, and there are rarely more 
than four of them (only in D. multibranchiata). Sometimes the occurrence of six 
or even eight branchial processes has been asserted. In such cases it is the margin 
of the funnel-shaped expansion, which is drawn out into processes ; a pair of these 
are extraordinarily long in Dero furcata. The members of this genus usually (1 always) 
inhabit tubes, which are formed by a viscid secretion from the worm's body to 
which small extraneous particles adhere. A species investigated, by myself (25)j of 
whose identity I am not certain, was invariably found in the interior of the stems 
of dead and decayed aquatic plants. B. vaga walks about with its tube like a Caddis 
worm. The genus is also to be characterized by the complete absence of floating peri- 
visceral corpuscles (except in D. vaga), which are so distinctive a feature of many 
Naids. Other points in its structure have been already Sealt with. The genus Dero is 
widely distributed like many, if not all, of the aquatic Oligochaeta. It occurs in Europe, 
North America, the Philippines, Ceylon (if Aulophorus oxycephala of Schmarda be really 
a Dero), and finally, I have received specimens from Mombasa in East tropical Africa. 

The distinction of species is a matter of some difficulty. Vaillant allows as 
many as nine ; Bodsfield only seven, but this number must be increased by two, 
for BousFiELD was not acquainted with Reighard's observations upon D. vaga, 
and D. 'multibranchiata has since been described. Of Bousfield's seven species, 



four: viz. D. latissima, D. perrieri, D. mulleri, and D. acuta, are described for the 
first time by himself and not noticed by Vaillant. The remaining three are D. 
obtusa, D. limosa (=^D.philippinensis), D.furcata ( = D. palpigera, and D. rodriguezii); 
B. digitata [Nais digitata, O. F. Mullek) is mentioned as 'doubtful.' 

The species not referred to by BoDSFiELD, but included in this genus by Vaillant 
are Pristina fiagellum of Leidy, Aulophorus oxycephala of Sohmabda, Xantho decapoda 
of DuTEOCHET ; D. palpigera and D. philippinensis are considered to be distinct species. 
P. fiagellum does not seem like to be referable to the present genus by reason of 
the fact that the dorsal setae commence from the very first— that there is no 
cephalization like that of Dero. It is just possible that P. fiagellum belongs to my 
genus Branchiura. B. decapoda is too imperfectly known to permit of its being regarded 
as a distinct species. It is said to have ten appendages from the branchial area. 

B. palpigera is so imperfectly described by Grebnitzky that I have no wish to 
criticize the view advanced by Sempee that it is the same as his B. rodriguezii, 
and by Bousfield that both are identical with B. furcata. 

It is very difficult to identify Xantho hexapoda of Ddteochet, Nais auricularis 
of Boso, which Bousfield regards, and with great probability as a Bero, and Aulo- 
phorus discocephalus of Schmaeda. No doubt Bousfield is also right in stating that 
it is impossible to identify B. digitata of Mullee. I do not admit for the present 
B. stuhlmanni of Stieben. The only distinctive character appears to be its very 
small size (2 mm.), though size is sometimes, I believe, a reason for specific distinction. 

B. intermedia of Ceagin is doubtful. Nais caudata. of Schmaeda, with lengthened 
• posterior segment, may also be a Bero. 

(i) Dero mulleri, Bousfield. 

D. miilleri, Bousfield, J Linn. Soc. 1886, p. 104. 
Definition. Length, z^ mm. ; number of segments g^. Branchial processes, oblong, quadrangular, 
lower longest. Hearts usually seven pairs. Hob. — Great Britain. 

This species is easily to be distinguished on account of the peculiar form of the 
branchial processes. 

(3) Dero limosa, Leidy. 

D. limosa, Leidy, Am. Nat. 1880, p. 431. 

D. philippinensis, Sempee, Arb. Zool. Zoot Wiirzb. 1877/78, p. 107. 
D. acuta, Bousfield, Eep. Brit. Ass., 1885, p. 1098. 
Beflnition. Length, 6 mm. ; mimber of segments 60. Branchial area with dorsal lip which 


is prolofiffed into two moderately/ long processes. Five pairs of contractile hearts. Hah.— 
Europe; N. America; Philippines. 

BoTJSFiELD (3) insists upon the identity of Leidy's and Sempee's species with a 
Dero studied and figured by himself. I am a little uncertain whether a species, 
concerning which I gave a few anatomical details a year or two since, and which 
I have already referred to as living inside the dead stems of water plants, is also 
identical. If so, there are six pairs of hearts and not five, and the first of these 
give off a branch on either side running forward. The hearts are in segments vi-xi, 
and those of vii, viii, ix are stouter than the rest. I regard D. acvia of Bousfield 
as merely a variety of the above; the branchiae are said to be longer. In this 
SxiEEEN agrees. 

(3) Dero furcata, Oken. 

D. furcata, Oken, Lehrb. d. Naturg. iii, Pt. i, 1815, p. 363. 
? D. palpigera, Grbbnitzky, Zapiski novoross. Obstch. Est. 1873, P* *^8. 
? D. rodriguezii, Semper, Arb. Zool. Zoot. Wiirzb. 1877/78, p. 106. 

Definition. Number of segments 35. Dorsal setae bundles commence in the fifth segment. 
Body ends in two long processes, and there are also a pair of supplementary branchiae 
at sides. Five pairs of contractile hearts. Hob. — Europe ; Trinidad 

This species was obtained by Bousfield from the inside of dead stems in water 
from the Royal Botanic Gardens, Regent's Park, by Sempeb from Minorca; I have 
received apparently the same species from Eastern tropical Africa. Bousfield, after 
remarking that 'D. rodriguezii is undoubtedly the same as the A. vagus of Leidy, 
D. furcata of Oken, and D. palpigera of Geebnitzky,' omits the second name in 
his list of synonyms. Sempee's figure (1, Taf. iv. fig. J5) represents the supple- 
mentary branchiae as springing from the anal area like the other branchiae. 
Stieben received this species from Trinidad. 

(4) Dero perrieri, Bousfield. 

D. perrieri, Bousfield, Rep. Brit. Ass. 1885, p. 1098. 
D. obtusa, Pebeiee, Arch. Zool. Exp. 1873, p. 6^. 

Definition. Number of segments, 2i5- Branchial area with entire margin. Hearts three to 
five pairs. Hob.— Europe. 

Bousfield is, no doubt, right in distinguishing between B. obtusa of d'Udekem 
and B. obtusa of Peeeieb. The entire margin of the gill-area in Peeeier's species 

Qq 2 


is not found in the true D. obtusa. It is probably this species whose anatomy I 
myself described (22). 

(5) Dero obtusa, d'Udekem. 
D. obtusa, d'Udekem, Bull. Acad. Eoy. Belg. xxii, pt. ii, 1855, p. 549. 

Definition. Number of segments, 50. Branchial area with distinct dorsal Up marked off by 
dee]} grooves on each side; branchiae rather short. Hearts usually four in number. 
Hab. — Europe. 
BousEiELD (1) remarks that 'this species shows the first trace of the modification 

which leads to the formation of supplementary branchial processes in the distinct 

demarcation of the dorsal lip, at the angles of which they are borne by those 

species which possess them.' 

(6) Dero vaga (Lbidy). 

Aulophorus vagus, J. Lbidy, Am. Nat. i880j p. 423. 
D. vaga, L. Vaillant, Annel^s, p. 383. 

Definition. Length, about 8 mm. ; number of segments, 25. Hody ending in two long 'processes ; 

branchiae rudimentary, only two slight processes. Dorsal setae bundles consisting of one 

capilliform, and two pectinate setae. Perivisceral corpuscles present. Contractile hearts 

in VIII, IX, X. Hab. — N. Am.erica ; Trinidad. 

This species, imperfectly described by Leidy (10), has been more fully studied 

and illustrated by Keighakd. The most marked characteristic of the species is the 

very rudimentary condition of the branchia, and the existence of two long processes, 

as in D. furcata. The vascular loops in the "branchial region are correspondingly 

simple. The dorsal vessel bifurcates a little in front of the tail, and joins a circular 

vessel, which is again connected with the ventral vessel. On either side of the 

point, where the ventral vessel arises, a slender trunk springs and forms a low loop 

in the two rudimentary branchial processes, and there joins the dorsal vessel, just 

after its bifurcation. The 'spade-shaped' setae of the dorsal bundles are peculiar to 

this species. Their outer ends are flattened, the expansion being supported by two 

or three ribs. Stieeen received examples from Trinidad. 

(7) Dero latissima, Bouspield. 

D. latissima, Bousfield, Kep. Brit. Ass. 1885, p. 1098. 

Defijiition. Number of segments, 40. Four pairs of hearts. Branchial area with entire 
margin, wider than long; branchiae long. Hab. — Great Britain. 


This species has been described by Bouspield in two papers (2, 3), in the later 
of which (3, PI. iv. fig. 8) the branchial area is illustrated. The definition does not 
appear to me to be enough to distinguish it from D. perrierl, but, as Bousfibld 
has examined both species, I leave it. 

(8) Dero nmltibranchiata, Stieren. 

D. multibranohiata, Stieben, SB. Dorpat. Nat. Ges„ 1892, p. 103. 

Definition. Length, imm. ; numher of segments, 6^. Branchiae fourteen in number, elongated. 
Hah. — Trinidad. 

This species, which has been somewhat fully described by Stieren (with a few 
illustrations), is, of course, to be distinguished by the gills. The shorter dorsal setae 
do not appear to be bifid. 

Genus Chaetobbawchus, Bourne. 

Deputitiom'. Fore end of body with a series of pairs of hollow processes 
of the body-wall, which enclose the dorsal setae, gradually diminishing in size 
posteriorly. Dorsal setae, which commence in second segment, capilliform, 
increasing in length up to about the tenth segment ; thence diminishing in 
posterior segments ; sickle-shaped setae also present ; ventral setae entirely 
uncinate. No glandular stomach. 

This is quite the most remarkable genus of the family at present known. Its 
distinguishing feature is the presence of a double series of long hollow processes 
of the integument, which enclose the dorsal setae, at least in the anterior seg- 
ments ; posteriorly — from the forty-second segment onwards, according to Bourne's 
figure (1, PI. xii. fig. 1), some of the dorsal setae are not so enclosed. These 
branchial processes, as Bourne assumes them to be, diminish in size after the first 
dozen pairs until 'they become mere warts on the surface of the worm, and in the 
posterior segments are entirely absent.' There are from sixty to seventy paii-s of 
them. Each process is hollow, its interior communicating with the coelom ; the 
walls are formed of epidermis alone, the muscular layers of the body-wall being 
apparently not prolonged into it ; the epidermis has a cuticle, through which project 
very fine cilia ; at the extremity are a few stiflE" hairs, doubtless of sensory function, and 
similar to hairs generally distributed over the body in the lower Annelids. Each 
branchial process contains a vascular loop, which is derived from the cii'cular vessel 


uniting the dorsal and ventral trunks of its segment. At present there are only 
three other Oligochaeta in which branchial processes exist, viz. Dero, Branchiura, 
Hesperodrilus branchiatus, and AIttm nilotiea ( = Digitibranchus niloticus); it is not, 
however, yet certain whether the latter worm is really an Oligochaet. 

The setae of Ghaetobranchus are arranged as in Pristina and Jfaidium, that is 
to say, there is no cephalization ; both the dorsal and ventral bundles commence in 
the second segment. The dorsal setae are of two kinds ; there are long capiUiform 
setae, which vary in length, being longest in some of the anterior bundles ; in the 
posterior bundles there are in addition sickle-shaped setae which have a straight 
shaft, and a curved, sickle-shaped extremity ; these two extremes are connected by 
intermediate forms. The ventral setae are all uncinate ; there are four to £ix of 
them in each bundle. 

The viscera do not present any features of special interest. The alimentary 
canal differs from that of most Naids in having no glandular ventricle. Nephridia 
are present, but there are no data as to the segment in which they commence. 
iJouRNE says that they resemble the nephridia of Naids generally. The coelomic 
corpuscles are round, and contain numerous oUve-green granules ; as they can pass 
from segment to segment, it is evident that the septa must be imperfect. 

The vascular system consists of the usual dorsal and ventral trunks, which are 
connected in each segment by a pair of lateral vessels. The dorsal vessel is much 

There are no eye-spots. 

Of the generative organs, unfortunately, nothing is known. The genus only 
contains one recognizable species at present, viz.j— 

Ghaetobranchus semperi, Bourne. 

C. semperi, BouKNE, Q. J. M. S. vol. xxxi, p. 88. 

Definition. Length, i^-^o mm. Branchial proeesses, 60-'] o "pairs. Ilab. — Madras. 

This species, of which a general account is given above, was discovered by Bodene 
in a ' tank ' in Madras town ^. It lives in the weed, moving about freely, and not 
constructing any tube, though forming a burrow. All the individuals found were 
in an active state of multiplication by simple fission, and were in no case sexually 
mature. Bouene considers that this Naid is either identical with, or closely allied 
to, the worm described by Sempee (1) as occurring along with Dero philippinensis. 

' I foiind specimens in the Vidoria regia tanlc at the Botanical Society's Gardens, Regent's Park. 


Genus Amphichaeta, Tauber. 

Dbfinitioit. Prostomium elongate. Setae in four rows ; several segments without 
setae after second setigerous segment No perioesophageal vessels. Sperm-sacs 
and egg-sacs present ; no genital setae. 

This genus was first described by Taubee (1) ; the description, however, is 
limited to the following brief definition : — 

'Praestomium dilatatum. Os inferum. Fasciculi uncinorum tam dorsalium quam 

This definition is nevertheless sufiicient to show that the worm in question is difierent 
from Chaetogaster. The genus has been fully established by the investigations of 
Kallstenius. Previously to the paper of the last-named, Levinsen (2) had sought 
to identify Amphichaeta with Uncinais. Since Levinsen wrote upon this subject 
Uncinais has been re-investigated by Bourne (5) ; and there is no longer any doubt 
of the complete distinctness of these two Oligochaeta. 

In the most distinctive feature of the organization of the Chaetogastridae, 
Amphichaeta agrees with Chaetogaster; there are, that is to say, a considerable 
number of segments intercalated between the second and third seta-bearing segment ; 
in Chaetogaster there are six intervening between the first segment and the second 
segment, provided with ganglia. The oesophageal segment has no setae ; in Amphi- 
chaeta there are the same number; but as there are two ventral bundles of setae 
immediately following the mouth, and as the oesophageal segment is provided with 
setae, the same number of ganglia separate the anterior and the posterior seta-bundles. 

This genus also differs from Chaetogaster in the absence of a perivisceral vessel 
in the oesophageal segment. A peculiar fact about the nephridia is their asymmetrical 
condition ; they commence in the segment following that which contains the ovaries ; 
occasionally they are symmetrical (paired), but more often one or other only of the 
two nephridia of a segment are present; they are closely adherent to the ventral 
blood-vessel, and are entirely without a pre-septal portion, that is, there is no 

The reproductive organs have been fully described by Kallstenius; the two 
testes lie in the segment immediately following the oesophageal; they arise from 
the peritoneum covering the nervous cord; the unpaired ovary arises in a similar 
fashion in the next segment; the efferent apparatus of the testes consists of the 
usual funnel-duct and terminal gland ; it opens on to the ovarian segment ; an 


important difference between Afiiphichaeta and Chaetogaster is the presence in the 
former genus of sperm-sacs and egg-sacs ; these are situated dorsally to the intestine, 
and consist of outgrowths of the anterior septa of the testicular and of the ovarian 
segments ; they extend, when fully developed, into the third segment after the ovarian 
segment; the sperm-sac comes to lie within the egg-sac. These sacs are not found 
in Chaetogaster. A further difference from Chaetogaster is in the absence of genital 
setae. The ventral setae of the ovarian segment are quite absent in fully mature 
worms. This segment, as well as the one in front, is occupied by the clitellum. 
The testicular segment contains, in addition to the testes, the spermathecae ; these 
are placed behind the anterior septum of that segment. 

The genus Am/phichaeta, although it should clearly be placed in the formerly-used 
family Chaetogastridae, on account of the peculiar • cephalization ' and the structure of 
the alimentary canal, differs from that family, and approaches the Naidomorpha in 
the presence of sperm-sacs and of egg-sacs ; the absence of genital setae removes 
it from other Chaetogastridae, and brings the genus near to such a genus as Dero, 
where, as I have shown, there are no genital setae; as a rule, these setae are 
present in the Naidomorpha. 

(i) Amphichaeta leydigi, Taubee. 
A. leydigi, Taubee, Ann. Dan. 1879, p. 76. 
Definition. 07ie seta in first Imndle, 2-4 in the rest. Hab. — Denmark ; Germany. 

(a) Amphichaeta saunio, Kallstenius. 

A. sannio, Kallstenius, Biol. Foren. Forh. 1892, p. 54. 

Definition. Length, 1-5 m. ; four setae in each of three anterior seta-bundles, aftenvards three. 
Hab. — Sweden ; Baltic. 

Genus Chaetogaster, von Baee. 

Snjn. Nais, Lamaeck, &c. (in part.). 
Mutzia, VoGT. 
Derostoma, DuGi;s. 

DEFiTsriTiON. Ventral seta-bundles only present, composed of uncinate setae. 
The first in segment II. Segments III-V without setae at all (except in 
Ch. fUiformis). One pair only of commissural blood-vessels. 


I follow Lankester (note appended to Bourne's paper 5) in relegating the genus 
Chaetogaster to the family Naidomorpha ; so long as the genus Chaetogaster alone 
was known, the entire absence of the dorsal seta-bundles might possibly be held 
to be of sufficient importance to keep it as the type of a distinct family, though 
doubtfully, in view of analogous differences in the family Enchytraeidae ; now, however, 
that another genus, closely allied to Chaetogaster, viz. Amphichaeta, is known, there 
is not this excuse for retaining the family; no other structural differences of first- 
rate importance support such a separation of the Chaetogastridae as a distinct 
family. The absence of the ventral setae in a number of the anterior segments 
(not from the first) was a peculiarity which might possibly merit emphasis of this 
kind, were it not for an analogous absence from two segments of the ventral setae 
in Ripistes Tnacrochaeta. The reproductive organs, whose modifications are frequently 
of family value, are constructed upon precisely the same lines as in other Naidomorpha ; 
and the spermathecae, testes, &c., are in the same segments in both of the ' families ' 
Naidomorpha and Chaetogastridae. I have mentioned as a characteristic of the 
genus the fact that there is only a single pair of commissural vessels uniting the 
dorsal and ventral vessels ; this is not absolutely distinctive, since Bourne found, 
and I have been able to confirm his discovery, that in Pristina equiseta there 
is but a single pair of such vessels. A marked characteristic of the present genus 
is the series of dilatations of the oesophagus ; in Ch. crystallinus, for example, there 
are two of the dilatations following the pharynx ; each is covered with a network 
of blood-vessels ; they, no doubt, correspond to the ' magen-ahnliche ' dilatation in 
many Naids. But they are absent in Ch. filiforonis, which is very Naid-like. 

The absence of an intimate correspondence between the nerve ganglia, the inter- 
segmental septa, and the setae is remarkable. The first septum occurs behind the 
pharynx and there are three ganglia in front of it, to which there are only one 
pair of setae bundles. 

The remai'kable double character of the ventral nerve cord in the anterior segments 
has already been commented upon. 

The nephridia of the genus, like those of Amphichaeta, are without a funnel ; the 
place of this is taken by a rosette-shaped bunch of cells attached to the anterior 
septum of the segment by a few muscle fibres. The development of the nephridium 
has been traced by Vejdovsky, who found no trace of a funnel at any time ; the 
earliest stage which he figures consists of three cells, all of them lying behind the septum. 
The mature organ consists of tubes running parallel to each other for the greater 
part of their course ; they are connected by numerous small tubules, which also branch 
within the secreting cells. 

R r 


As to the species of the genus, Vaillant allows seven ; but two of these are, 
as he himself admits, doubtful; the two doubtful species are Gh.filiformis of Schmaeda 
and the Derostoma laticeps of DuGis. "With regard to the former, which is figured, 
though very insufficiently, it is certainly a member of the genus. Schmaeda figures 
setae upon every segment of the body. It comes from Curaca, in South America ; 
I have examined a specimen from Valdivia. Derostoma laticeps, at first referred by 
DuGES to the Planarians, was afterwards (3, p. 30) placed in the genus iV"a«s ; but no 
particulars are given save that there is only a single set of setae on each side of the 
body, and that the ' lip ' is large (suggestive, as Vaillant has pointed out, of an 
Aeolosoma). Whether Ch. gulosus of Letdy (3, p. 124) is a 'good' species or not 
seems doubtful. There is really nothing in the description which is at all decisive 
in the matter. Its length is one line, and there are five or six setae in each bundle. 
The oesophagus is said to be short. 

I allow the five following species, and I use Vejdovsky's names. This because 
it does not seem to me to be possible to identify the Nais vermiciilaris of Mcjllee. 
The only distinguishing feature of it is its length (2 lines). That possibly brings it 
nearer to Ch. diastrophus, but the identification does not appear to me to be certain. 

(i) Chaetogaster limnaei, v. Baee. 

Ch. limnaei, V. Baee, Nov. Act. Nat. Curios., 1827, p. 6ii. 

Ch. furcatus, Ehkenbeeg, Symb. Phys., 1828. 

Ch. diaphanus, Oeested, Nat. Tidsskr., 1842, p. 138 (in part.). 

Ch. vermicularis, Geube, Arch. f. Nat, 1851, p. 353 (in part.). 

Mutzia heterodaetyla, VoGT, Arch. f. Anat. Phys., 1841, p. 36. 
Definition. Length, 2 mm. ; oesophagus much reduced. Hob. — Europe. 

This species lives parasitically upon fresh-water MoUusca and sometimes within 
their bodies. It is the smallest species of the genus. The anterior pair of setae 
are commonly somewhat stronger than the rest. The first pair of nephridia open 
in front of the second setae bundles. 

(2) Chaetogaster diaphanus, Gedithuisen. 
Ch. niveus, Eheenbeeg, Symb. Phys., 1828. 
, , Ch. diaphanus, Oeested, Nat. Tidsskr., 1842, p. 138 (in part.). 
Ch. vermicularis, Geube, Arch. f. Nat., 1851, p. 353 (in part.). 
Nais diaphana, Gbuithdisen, Nov. Act. Nat. Curios., 1828, p. 409. 
Nais lacustris, Dalyell, Powers of Creator, vol. ii, 1853, p. 130. 
Kais scotica, Johnston, Cat. Worms B. M., 1865, pp. 71 and 336. 


Definition. 1 5 mm. in length ; oesophagus very distinct but shorter than pharynx. Hah. — 

This species is the largest of the genus and completely transparent. 

Vejdovsky considei-s that the Ch. diaphanus of d'TJdekem (4), who described the 
genitalia, is not the same species, but is Ch. cristallinus. Vaillant, by omitting this 
particular reference from his list, seems to concur. I would point out that the ' yellow 
globules ' described and figured by Dalyell are probably the ripe ova. The genital 
organs of Gli. cristallinus are, however, unknown, and d'Udekem's Ch. diaphanus is 
only 5 mm. long, which better suits Ch. cristallinus. I include N. scotica as 
a synonym on the authority of Johnston, who regarded it as being, the same as 
Daltell's N. lacustris. 

(3) Chaetogaster cristallinus, Vejdovsky. 

Ch. cristallinus, Vejdovsky, SB. Bohm. Ges., 1883, p. aao. 
Definition. Length, 2—3 mm. Dorsal and ventral blood-vessels not continued into pharyngeal 
region. Oesophagus as long as pharynx. Hab. — Europe. 

This species was first described, and was later (24) more fully described and figured, 
by Vejdovsky. Vaillant regards, I think with some reason, Lankestek's Ch. niveus 
as probably identical with this species. The principal reason for the identification is 
the long oesophagus. The dorsal vessel of this species ends in a bunch of cells 
abruptly; just before its termination it passes through a sling attached to the first 

(4) Chaetogaster flliformis, Sghmarda. 

Ch. fiUformis, Schmaeda, Neue wirbell. Thier., I. ii, 1861, p. 11. 

Definition. Length, % mm. Prostomium well developied ; no missing ventral setae. Hah. — 
S. America. 

This species is curiously intermediate between Nais and the typical Chaetogaster. 
It has no dorsal setae as in Chaetogaster ; but it has a well- developed prostomium, 
and no nan-ow oesophagus, in which points, of course, it departs from Chaetogaster. 

(5) Chaetogaster diastrophus, Gruithuisen. 

Ch. vermicularis, Geube, Arch. f. Nat., 1851, p. 353 (in part.}. 
? Ch. miilleri, d'Udekem, Bull. Ac. Roy. Belg., xxii., pt. ii, 1855, p. 554. 
Nais diastropha, Gruithuisen, Nov. Act. Nat. Curios., 1828, p. 417. 

R r 2 


Definition. Length, 2-5 mm. Oesophagus an long as pharynx. Blood-vessels quite normal. 
Hah. — Europe. 
Vbjdovsky thinks that d'Udekem's Gh. mulleri (which the latter ought to have 
called ' vermicularis ' — as he puts this name down as the only synonym) is identical 
with Ch. diastropha. It is to be distinguished, according to d'Udekem, from 
Oh. limnaei only hy fewer setae and absence of 'spicules ^pidermiques.' These 
latter processes ( = tactile hairs) are figured by Vejdovsky in Ch. diastrophus, so the 
identification is less certain. Vaillant believes (p. 450) that d'Udekem's Ck. mulleri 
is Gh. diaphanus j the difierence in size may be, he thinks, a matter of age ; but 
this does not seem likely, the difference being too great. Vejdovsky figures (24, PI. vi, 
fig. I a, chp) a small chitinous plate upon hinder median aspect of brain, the nature 
of which is mysterious. 



Defiwitioh'. Setae (absent in Anaehaeta) short, straight, or curved, not bifid at 
extremity. A single pair of ealciferous glands sometimes present. Dorsal 
blood-vessel only present anteriorly, sometimes with, cardiac body. Testes 
in XI, male pores on XII ; a reduced spermiducal gland present ; oviducts 
represented by pores. Spermathecae, one pair ^ in V generally opening into 
gut, with or without diverticula. Dorsal pores occasionally present. 

This family of Oligochaeta is very numerous in species, which are all of small 
size, ranging from a length of 3 mm. to 40 mm. The structure of the Enchytraeidae 
has been mainly elucidated by Vejdovsky (3, 20), ErsEN (13), and Michaelsen 
(1-5, 16) ; to a less degree by d'Udekem (3), Clapaebde (3), and Ude (1, 2). 
Michaelsen has recently published a detailed ' Synopsis ' of the family, the conclu- 
sions set forth in which are, in the main, adopted here. The family is a very 
natural one ; there appear to be no forms transitional between the group and other 
Oligochaeta. This is satisfactory to the systematist, but it renders the labours of 
the naturalist who desires to study the inter-relationships of the different groups 
of Oligochaeta extremely difficult. 

All the Enchytraeidae have a prostomium ; in most there is a single pore upon 
the prostomium; in a few forms (e.g. in Fridericia galha) there are also dorsal 

' Henlea puteana with two pairs is the only exception. 


pores; this family is the only one of the aquatic Oligochaeta which have dorsal 
pores. The setae are entirely absent in Anachaetaj there are only two bundles in 
Bi&tichopus (if this genus be really an Enchytraeid), but six in Ghirodrilus (con- 
cerning which genus also more information is wanted); all the others have the 
usual four bundles of setae to a segment. The setae are either straight or bent in 
the prevalent /-shape ; their extremities are never cleft. The setae in this family 
are always short and generally fairly numerous in each bundle. One genus only is 
entirely achaetous — the genus Anachaeta. The setae are here represented by large 
cells which depend into the body-cavity (see p. 5). The setae when present are of 
two kinds; there are curved sigmoid setae and straight setae; the former are found 
in certain genera, such as Fachydrilus j and the latter in other genera, such as 
Enchytraeus. The setae of a bundle are disposed in a fan-like fashion, the setae on 
one side of a bundle diverging from those upon the other side. The number of 
setae in a bundle varies from one [Ennchytraeus Ttionochaetus) or two [Fridericia 
bisetosa) to fourteen, which is the greatest number that has been hitherto recorded 
(in Pachydrilus minutus). Very commonly there is an inequality in number 
between the setae of the lateral and ventral bundles respectively ; in this case, 
the lateral bundles have usually the fewest setae ; thus in Fachydrilus pagen- 
stechen there are 7-10 setae in the ventral bundles and ^-$ in the lateral. The 
setae of a given bundle are usually of a size ; but this is not always the case ; 
the most conspicuous exception is seen in the genus Fridericia; here the outer 
setae of a bundle are always larger than the inner, which are at the same 
time the younger. There are sometimes differences in size between the setae of 
the four bundles of a given segment ; for example in Fridericia^ bullosa the setae 
of the lateral bundles are smaller than those of the ventral. In Mesenchytraeus 
setosus there is something like a formation of genital setae ; the lateral bundles of 
segments v-viii (inclusive) are made up of 1—4 setae, which are very much larger 
than those of adjacent segments. In Enchytraeus tnonochaetus the anterior setae are 
more slender than those of the posterior segments. Enchytraeus monochaetus is 
unique in the group by reason of the fact that the setae are entirely wanting upon 
the first few segments of the body. The nephridia are peculiar in their form ; they 
are stout and solid-looking organs, often lobed, with a lumen which (according to 
BoLsius) forms a plexus in their interior. They frequently begin as far forward as 
the second segment, though they are wanting in those segments which contain the 
generative organs. Some Enchytraeidae are characterized by the possession of 
a single gland, or a pair of glands, which seem to be the equivalents of the 
calciferous glands of other worms (see p. 61). The family is also characterized by 



the position of the reproductive organs; the testes are in the eleventh segment'; 
the ovaries in the following segment. The sperm-ducts open by a very reduced 
spermiducal gland on to the twelfth segment ; the funnel of these ducts is peculiar 
in form, being generally extremely long and of a glandular appearance ; the oviducts 
are represented merely by pores ; or rather by a short prolongation of the septum 
which meets the pore ; they seem to be degenerate. The spermathecae are far 
forward opening on to the intersegmental furrow iv/v ; as a rule they open into 
the gut, a fact which is paralleled in the genus Sutroa, and was first discovered in 
the present family by Michablsen (14). Egg-sacs and sperm-sacs are only present 
in Mesenchytraeus. The above characters distinguish the Enchytraeidae from all 
other Oligochaeta. 

Michablsen allows twelve genera, of which two (not described by himself) must 
be considered doubtful ; these genera are Ghirodrilus and Bistichopus ; I shall recur 
to them presently. To these 'Bryodrilus of Ude (2), and Parenchytraeus of Hesse 
must be added. 

The genera which appear to be valid may be thus distinguished : — 





Setae ./-shaped ; no dorsal pores, only a head-pore ; origin of 

dorsal vessel postclitellian, contains a glandular body. 
Setae y-shaped ; origin of dorsal vessel anteclitellian, contains 

a glandular bodj'. 
Setae ./-shaped ; no dorsal pores ; blood coloured ; dorsal vessel 

originates behind clitellum, contains no glandular body. 
Setae y-shaped ; dorsal f)ores absent ; dorsal vessel originates 

from the tip of the single dorsal diverticulum of the gut ; 

salivary glands present. 
Setae straight or /-shaped ; origin of dorsal vessel anteclitellian ; 

no dorsal pores ; oesophagus sharply marked off froin 

Setae straight; dorsal vessel postclitellian in origin. 
Setae straight ; dorsal pores present ; salivary glands present ; 

dorsal vessel postclitellian in origin. 
No setae ; no dorsal pores ; dorsal vessel anteclitellian in 

origin ; a single salivary gland ; spermathecae not opening 

into gut. 

' In a few forms the testes, &c., are further forward. 



Maeiobtia. Setae /-shaped; no dorsal pores and other characters as in 

Pachydrilus, except that testes are massive, not subdivided. 
BBYODBILTJS. Setae /-shaped ; no dorsal pores ; dorsal vessel arising behind 

oesophageal glands in xii ; blood colourless. 
PAEEWCHYTKABTJs. Setae straight ; no dorsal pores ; ventral vessel bifurcates only 

in first segment. 
The arrangement of the genera of this family naturally depends upon the affinities of 
the family as a whole to other Oligochaeta. Regarding, as I believe it to be 
necessary to do, the more simple forms as standing higher or lower in the scale 
(according to the sense attached to the words) than the more complexly-organized 
genera, in fact, as more specialized through degeneration, it appears to me that 
the genus Mesenchytraeus represents, on the whole, the most primitive Enchytraeid. 
I base this view of its position on the following characters : — 
(i) The setae are sigmoid in shape. 
(2) There are sperm-sacs and egg-sacs. 
On the other hand, the colourless blood, the absence of dorsal pores, and the 
rudimentary condition of the oesophageal glands (represented only by the cardiac 
body), furnish evidence against the placing of the genus. 

The genus Buchholzia exhibits two of these last characters with the addition of 
the sigmoid setae ; it has, however, colourless blood like Mesenchytraeus. Pachydrilus 
has coloured blood and sigmoid setae. How are we to decide between the conflicting 
claims of these three forms? 

A fourth, indeed, might be added, viz. Fridericiaj this genus has dorsal pores, 
a character distinctive of the higher Oligochaeta. 

It seems to me that the importance of these several characters is indicated by 
their position in the following list : — 
(i) Sperm-sacs and egg-sacs, 
(a) Oesophageal glands > ,^^^y 
Sigmoid setae i 

(3) Red blood. 

(4) Dorsal pores. 

(i) Sperm-sacs are, with the exception of certain Enchytraeidae (the majority), 
present in all Oligochaeta, even in the lowly form Aeolosoma. They must evidently, 
therefore, be looked upon as being very distinctive characters of these Annelids. 

(2) Oesophageal glands are met with in none of the lower aquatic Oligochaeta, 
excepting ceitain Enchytraeidae. They are very common in the higher forms, 
occurring in nearly all the ' earthworms.' Their presence or absence, therefore, 


seems to be correlated with the aquatic or terrestrial life. There are, it will be. 
observed, no reasons for inferring that such is the case with the sperm-sacs. 
I have bracketed together, as being of equal importance, the presence of these glands 
and the sigmoid character of the setae. It is perfectly true that sigmoid setae 
are found in all genera of Oligochaeta, excepting only certain Enchytraeids ; but 
the weight of this fact is partly lessened by the fact that among the aquatic forms 
there are very generally also capilliform setae, and the sigmoid setae are variously 
mjodified in shape (e.g. pectinate setae). 

(3) Red blood is only less universally present in the Oligochaeta than sigmoid 
setae. Apart from the Enchytraeidae, it stops short at the lower Naids, which 
have very faintly coloured blood; Aeolosoma has colourless blood. One can 
hardly help inferring from the facts that size is correlated with the colour of the 
blood ; the minute Oligochaeta have ' white ' blood, the larger and large forms red 
blood. Now the genus Pachydrilus contains some of the largest Enchytraeids 
(also, it must be admitted, some of the smallest). 

(4) There remains only the question of the dorsal pores ; I am of opinion that 
these are distinctly related to the habit of the worm ; thej' are to be found in no 
aquatic Oligochaeta. The genus Fridericia, in which alone they exist (among the 
Enchytraeidae), is terrestrial, and found in the driest localities ; so also it is true 
of many Pachydrilus, &c. But, on the . whole, the group of the Enchytraeidae 
is, in the matter of its mode of life, in an undecided state ; they are not purely 
terrestrial nor purely aquatic ; and, if aquatic, neither definitely marine nor fresh 

The presence, then, of egg and sperm-sacs, coupled with the sigmoid setae, leads 
me to place the genus Mesenchytraeus in the position of the nearest approach to the 
original Enchytraeid. 

I do not, however, think that any other genus can be derived directly from 
Mesenchytraeus; it is itself too degenerate in the matter of oesophageal glands and 
the colour of the blood. Nearest to it I should place Buchholzia, Pachydrilus, 
and Henlea, which have oesophageal glands, and Pachydrilus-like setae ; and the 
two last will stand nearer to Mesenchytraeus than Pachydrilus, in which all trace 
of the oesophageal glands has disappeared. Fridericia and Enchytraeus will be still 
further remote from the primitive stock, and Anachaeta furthest of all. I do not feel 
able to make any suggestions concerning JDistichopus and Chirodrilus, of whose 
anatomy we have at present insufficient knowledge. 

The following worms are either certainly or probably Enchytraeids. 

Halodrilm littoralis of Veekill, thought by Vejdovsky (24) to be a Tubificid. 


seems to have oesophageal glands as in Eenlea. The setae, disposed in a fan-like 
manner, suggest an Enchytraeid, as also their small size. 

Lumbricus multispinus of Geube (7) (called Echinodrilus multispinus by Vaillant) 
seems to be an Enchytraeid, as Vaillant (6, p. 89), has suggested. 

Lumbricus glacialis of Leidt (9), regarded by Michaelsen as belonging to this 
group, is said to possess generative organs extending from the fourth to the eighth 
segment. This does not read like an Enchytraeid. 

Michaelsen is, in my opinion, undoubtedly right in referring Enchytraeus 
moniliformis (d'Udekem), Nais albida (Caktee), Lumbricus Jordani (Williams, 1), 
and Enchytraeus juliformis (Kesslee), to this family. He is less certainly right in 
doing the same with Saenuris abyssicola and S. limicola of Vebeill, Tubifex pallidus 
(DuGEs), and Saenuris vagans (Johnston, 2). 

Enchytraeus sepultus of Menge, a fossil species from amber, is an undefined species. 

Lumbricus putridinis, synonymous, according to Johnston (2), with Enchytraeus 
verm/iculus, only needs to be referred to for the sake of completeness. 

The number of species in this family is considerable ; the names that have been given to supposed 
species is greater still. One hundred and three names are quoted in Michaelsen's Synopsis as 
applicable, or which have been applied, to worms belonging to this family. Vaillant allows no 
less than seventy-two of these in a way ; they are at any rate numbered and described in his work, 
though he doubtless casts some question upon the reality of certain among them. Michaelsen 
allows only sixty-one species (several of which are not included in Vaillant's work, as they have 
been described since 1886) ; in addition to these sixty-one, he mentions twenty-nine which are 
partly 'incertae sedis,' partly 'species inquirendae,' and partly ' species spuriae.' To the first category 
are referred such species as can be defined specifically, but whose generic position is uncertain. To 
the second category are referred those species which are insufficiently characterised, but which may 
be subjected to renewed examination since the original types are extant, or since the description 
given is enough with the locality to ensure recognition should they be again met with. To the last 
category are referred a few species which are quite hopeless in these possibilities. Some of these 
species have been dealt with in the pages which follow ; others may be suitably referred to here. 

The following five species are described by Eisen (13) under the generic name of ArcJii- 
enchytraeus : — 

(i) Archienchytraeus tenellus. 

(2) „ Jevinsenii. 

(3) ,, lampas. 

(4) „ gemmatus. 

(5) „ ochraceus. 

Michaelsen says of these species that they are definable as species, but that it is uncertain 
whether they belong to the genus Henlea or Enchytraeus. I am not able to elucidate the matter 
any further, and therefore leave these species in the position in which Michaelsen placed them. 

About the following genera we require further information. 

s s 


Genus Distichopus, Veeeill. 

DEFUfiTioir. 'Form and colour as in Enchytraeus, with a -well-produced girdle. 
Setapods in a single row on each side ventrally, in divergent fascicles of four in 
advance of the girdle and of three behind it.' 

There is but one species in this genus which is regarded by Michaelsen as a ' species inquirenda.' 
This species, called D. sylrestris, is 20-30 mm. in length and composed of sixty-eight segments. 

Genus Chirodbilus, Veeeill. 

DEriH'iTiOM'. 'Allied to Saenuris, but with six fan-shaped fascicles of setae upon each 
segment, two of which are ventral, two lateral, and two sub-dorsal ; setae in the 
ventral and lateral fascicles four to nine, simple, acute, slender, curved like 
an italic y"; those of the dorsal fascicles stouter and less curved, three to six in 
each fascicle. Intestine wide, somewhat moniliform. Anus terminal, large.' 

I follow Michaelsen in assigning this genus to the family Enchytraeidae. Both Vejdovsky 
and Vaillant place it among the Tubifioidae. As Michaelsen has pointed out (5), the shape 
of the setae and the colourless blood together are inconsistent with any other view upon the 
systematic position of the two species which comprise the genus. The clitellum, too, is developed 
(in C. larviformis) upon the eleventh seta-bearing segment and upon a small part of the following 
segment, which is precisely what occurs in other Enchytraeids. The two species of the genus 
described by Veerill are C. larviformis and C. ahyssorum ; they were both dredged in Lake Superior. 
In C. ahyssorum the dorsal-setae are described as being shorter than those of both lateral and ventral 
bundles. Both species are small, 6 or 8 mm. in length, with thirty-eight or forty-two segments — 
-another argument, not noticed by Michaelsen— for their being referred to the Enchytraeidae. 

Genus Mesenchytbaeus, Eisen. 
8yn. Analycus, Levinsen. 

Enchytraeus, Vejdovsky (in part.). 
Neoenehytraeus, Eisen (in part.). 
Paehydrilus, Vaillant (in part.). 

DEFiNiTiomr. Setae y^-shaped, usually more numerous in the ventral than in dorsal 
bundles of anterior segments. Head -pore usually near to anterior end of 
prostomium, no dorsal pores. Dorsal vessel arises behind clitellum, contains 
a cardiac gland, blood colourless. No salivary glands. Brain truncated or 
concave posteriorly, generally broader than long. Nephridia with short ante- 
septal and large irregularly lobate post-septal portion. Egg-sacs and sperm-sacs 
present ; sperm-ducts short, at most eight times as long as funnel. 


The term MesenchytraeiLS was first applied by Eisen to a few species, which were 
grouped together as a subgenus of Enchytraeus. 

It was thus defined:— 'The spermatozoa, as long as they remain in the perivisceral cavity of 
the body, or in the vesicle of the efferent duct, are not free, hut encysted or congregated into small 
globules surrounded by a membrane. The supra-oesophageal ganglion is deeply divided in front, 
but straight behind or nearly so, the emargination being very inconsiderable. The tube of the efferent 
duct is unusually short and broad, never more than six or eight times longer than the vesicle 
of the said organs.' 

It will be observed that the two last of these characters are retained in Michaelsen's definition, 
of which the above is substantially a reprint. As to the spermatozoa, Michaelsen does not use 
the point in his re-definition of the genus or in his revision of the Enchytraeidae (5). In an earlier 
paper, howeyer, (upon Archienchytraeus mSbii) he remarks :— ' Bedeutende Abweiohungen kann ieh 
von P. beumeri constatiren. Bei diesem Wurm zerfallen die Hoden in einzelne Zellgruppen. 
Diese Zellgruppen sammeln sich in zwei Sacken, die von Dissepiment XI/XII gebildet werden und 
rechts und links vom Darm in das XII. Segment hineinragen. In diesen Sacken maohen dann 
die Zellgruppen ihre Entwicklung zu Spermatozoen durch. Soweit ich ihre Entwicklung verfolgen 
konnte, ungefahr bis zu dem Stadium, welches von E. mSbii in III, 6. V gezeiohnet ist, bleiben 
die einzelnen Gruppen von einem feinem Hautchen umschlossen, wie es auch Eisen fur die Mesenchy- 
traeen, zu denen P. heumeri nach dem System dieses Autors geh6rt, feststellt.' 

The matter evidently needs further inquiry. Michaelsen added two other characters of 
importance to the definition of the genus, viz. the presence of sperm- and egg-sacs, and the peculiar 
form of the nephridia. Besides describing several new forms referable to this genus, Michaelsen 
rescued a species (JV; fenestratus) from Eisen's genus Neoenchytraeus, and assigned it to Mesenchytraeus. 
Eisen himself distinguished Mesenchytraeus from Neoenchytraeus partly on account of the form of 
the brain , in the former, the posterior margin is straight or slightly concave ; in the latter, it is 
convex ; now, in N. fenestratus, the hinder margin of the brain is, as Vaillant has pointed out 
(6, p. 249), so little convex that the worm ought to be assigned to the genus Mesenchytraeus. 
Michaelsen also drew attention to this similarity between two species, supposed by Eisen 
to belong to different sub-genera ; he furthermore asserted from an examination of the species that 
the nephridia have the form so characteristic of the genus Mesenchytraeus, a fact which is not apparent 
in the figure given by Eisen (13, PI. xiv, fig. 35). The sperm-ducts, too, are short, as they are in 
Mesenchytraeus. The identity of Levinsen's genus Analycus with Eisen's Mesenchytraeus seems 
to me to have been rightly established by Michaelsen. This identification rests chiefly upon 
the apparently similar form of the nephridia in the two genera. In addition there are other points 
of similarity which are dealt with in the following pages. 

The main characters which mark this genus are given in the diagnosis ; this 
diagnosis may now be expanded a little ; the brain has two pairs of muscles attached 
to its posterior margin, one pair above, the other below. The nephridia are remarkable 
for the fact that the voluminous part of the organ, lying behind the septum, is lobate ; 
this is brought out in Eisen's figures, which are stated by Michaelsen to be, in 
some particulars, inaccurate ; Eisen represents the tube as lost in a mass of 
suiTounding tissue ; Michaelsen says that the tube is so much and so closely 
coiled that there is but little o f the cellular sheath visible, ' dass hier die umhiillende 

s s a 


Zellsubstanz fast auf das Minimum reducirt ist.' The part of the nephridium which 
lies in front of the septum is little more than the funnel. As regards the lobate 
character of the post-septal part of the nephridium, attention may be here directed 
to Eisen's figure (13, PI. xiv. fig. 38) of the nephridium of Neoenchytraeus hyalinus, 
a species put by Michaelsen into the genus Enchytraeus, as restricted by him. 
It would appear from this figure that the species should be referred to the genus 
Mesenchytraeus. Another peculiarity of the genus, though restricted to two species, 
M. arviatus, M. setosus, is the existence of specially enlarged setae in certain segments 
of the body; these are referred to under the description of the species where they 
occur. The reproductive organs are remarkable in that there are, as in most 
Oligoehaeta, though not in other Enchytraeidae, sperm-sacs and egg-sacs ; the 
sperm-sacs are paired, the egg-sac is usually single ; the varying length of these 
sacs may be useful as a specific character, but it seems more probable that the 
length varies according to age and degree of maturity. The spermathecae open at 
one end into the gut ; they may or may not be provided with diverticula ; when 
these are present they vary in number. 

(i) Mesenchytraeus primaevus, Eisen. 

Mesenchytraeus primaevus, Eisen, Ofv. Svensk. Akad. 1878, No. 3, p. 68. 
Eneh3rtraeus primaevus, Vejdovskt, Syst. u. Morph. 1884, p. 41. 
Enchytraeus (Mesenchytraeus) primaevus, Vaillant, Annel^s, p. 269. 
Definition. Length, 10 mm. ; breadth, i\mm.; setae, 5-8 per bundle; number of segments, 
52. Spermathecae, very small, trilobed at free end; sperm^duct hardly longer than 
funnel. Hob. — Nova Zembla ; Siberia. 

Eisen distinguishes this species principally by the excessive minuteness of the 
spermathecae, which are said to be ' at least ten times smaller ' than those of other 
species. Eisen considers that this is not due to the immaturity of the organs though 
he confesses that no spermatozoa were found in them. The cardiac body is thick in 
cross section with irregular swellings, and composed of numerous cells. 

(3) Mesenchytraeus mirabilis, Eisen. 

Mesench3d;raeus mirabilis, Eisen, Ofv. Svensk. Akad. 1878, No. 3, p. 68. 
Enchytraeus mirabilis, Vejdovsky, Syst. u. Morph. 1884, p. 41. 
Enchytraeus (Mesenchytraeus) mirabilis, Vaillant, Anneles, p. 268. 

Definition. Length, 10-15 w«.; breadth, \\-%mm.; number of segments, 64; setae, 5-7 


per bundle. Spermathecae conical, with 4-5 globular diverticula; sperm-duct generally 
much longer than funnel. Hab. — Siberia. 

This species, like the last, has been investigated by Eisen (13), and Michaelsen 
(2, 5). The last-mentioned author deals with a few points not referred to by Eisen. 
He figures and describes the cardiac organ, stating that it agrees with that of the 
last species in being larger than that of M. heumen, and in being composed of 
a considerable number of cells. There is (according to Michaelsen) a single sperm- 
sac, which extends as far back as the twenty-sixth segment ; the median egg-sac 
reaches further back still, to the twenty-ninth segment. Both sacs are constricted 
where they pass through the septa; this constriction is shown in one of Eisbn's 
figures (13, PI. xv, fig. 44); these figures indicate gaps between the anterior and 
posterior parts of the sperm-sacs. 

(3) Mesenchytraeus falciformis, Eisen. 

Mesenchytraeus falciformis, EiSEN, Ofv. Svensk. Akad. 1878, No. 3, p. 68. 
Enehytraeus falciformis, Vejdovsky, Syst. u. Morph. 1884, p. 41. 
Enohytraeus (Mesenchytraeus) falciformis, Vaillant, Annel^s, p. 272. 

Definition. Length, 4-5 mm. ; breadth, 3 mm. ; number of segments, 50 ; setae, ^-6 per 
bundle. Spermathecae, long 'and narrow, without diverticula; sperm- duct six times as 
long as funnel. Hab. — Nova Zembla. 

The cardiac body is, according to Michaelsen (2), smooth and thin and com- 
posed of but few cells. The median egg-sac extends back as far as the nineteenth 
segment in the specimens examined by Michaelsen ; Eisen figures the egg-sacs as 
paired structures reaching back to the sixteenth segment only. 

(4) Mesenchytraeus fenestratus, Eisen. 

Neoenchytraeus fenestratus, Eisen, Ofv. Svensk. Akad. 1878, No. 3, p. 74. 
Enohytraeus fenestratus, Vejdoysky, Syst. u. Morph. 1884, p. 41. 
Analycus glandulosus, Levinsen, Vid. Med. 1883, p. 23a ^ 
Mesenchytraeus fenestratus, Michaelsen, Abh. Nat. Ver. Hamb. 1889, p. 17. 
Enohytraeus (ifeoenchytraeus) fenestratus, Vaillant, Annel^s, p. 266. 

Definition. Length, 15-26 mm.. ; breadth, i mrn.^; number of segments, 60 ; setae, 5-7 jier 

bundle. Spermathecae elongated without any diverticula ; sperm ducts very short, not longer 

than funnel. Hab. — Siberia ; Denmark. 

' Levinsen considers that his species = Enehytraeus albidus of Taueek in part, the latter also including his 
(Levinseh's) 'Analycus armatus.' 



The principal reasons which lead Michaelsen to identify this species with Levinsen's Analycus 
glandulosus are (i) the fusion of the prostomium with the buccal segment, and (2) (apparently) the fact 
that the spermathecae are long sacs without diverticula ; there is, moreover, nothing in the diagnosis of 
Levinsen which goes against the identification. 

Michaelsen found that the head pore was (exceptionally for the genus) placed 
between the prostomium and the buccal segment. Sperm-sacs and egg-sacs appear 
to be paired. 

(5) Mesenchytraeus flavus (Levinsen). 

Analycus flavus, Levinsen, Vid. Med. 1883, p. 232. 

Paehydrilus flavus, Vaillant, AnneMs, p. 245. 

Mesenchytraeus flavus, MiCHAELSEN, Abh. Nat. Ver. Hamb. 1889, p. 18. 
Definition. Length, 15 mm.; iimally three setae in lateral and jive in ventral bundles. 
Spermathecae with a single pear-shaped diverticulum ; funnel comparatively small ; sperm- 
duct very short. Hob. — Nova Zembla ; Denmark. 

Michaelsen has identified an Enchytraeid found by him among three specimens 
of ' Neoenchytraeus vejdovskii' as ' Analycus fiavus' ; he naturally adds something to 
Levinsen's very short account of the species. The egg-sac is paired and extends 
through at least segments xiii, xiv, xv. Vaillant, in copying Levinsen's diagnosis, 
has fallen into an error, by omitting to mention the appendix of the spermatheca ; 
he says that the spermathecae are ' formees d'un conduit long et etroit sans appendices 
glandulaires et d'un sac pyriforme ' ; the latter is evidently the diverticulum and not, 
as Vaillant's description would lead one perhaps to infer, the swollen end of the 
pouch. The other characters of the species are to be found in the above definition. 

(6) Mesenchytraeus beumeri, Michaelsen. 

Paehydrilus Beumeri, Michaelsen, Arch. Mikr. Anat. 1885, p. 294. 
Paehydrilus (Mesenchytraeus, Eisen) Beumeri, Michaelsen, Ench. Mob., 1886, 

p. 44 et passim. 
Mesenchytraeus Beumeri, Michaelsen, Arch. Mikr. Anat. 1887, p. 372. 
Definition. Length, 30 mm,. ; Setae, 3-5 in lateral, 5-8 in ventral bundles ^. Spermathecae 

' The actual numbers of setae are thus stated by Michaelsen : — 

















Lateral . 
Ventral . 


























long with two oval diverticula ; sperm-duct eight times the length of the comparativeli/ 
small funnel. Hah. — Germany. 

This species has been described by Michaelsen in two papers quoted above. 
The species has two short sperm-sacs confined to the twelfth segment ; the egg-sac 
is single, but in one instance it was found to be divided into two at its extremity ; 
the egg-sac extends as far back as the nineteenth segment. The egg-sac in 
this species shows no constrictions at the septa, as is so often the case. The 
•perivisceral corpuscles are ellipsoidal. The septal glands are in iv and v ; from 
these a band passes back on each side as far as vii, from which arise irregular lumps 
of gland-cells. The dorsal vessel springs in xviii ; it has swellings in this and the 
three segments lying in front. The clitellum occupies segments xi-xiii. 

(7) Mesenchytraeus flavidus, Michaelsen. 

Mesenchytraeus flavidus, Michaelsen, Arch. Miki-. Anat. 1887, p. 37a. 

Definition. Length, i a mm. ; setae, up to Jive per hmdle. Sperm-duct at most five times us 
long as funnel ; spermathecae without diverticula, dilated at distal extremity. Hob. — 

The dorsal vessel arises in the thirteenth segment. The sperm-duct is only about 
five times as long as the funnel ; it has small prostates. The clitellum occupies 
segments xi-xiii. 

(8) Mesenchytraeus setosus, Michaelsen. 
Mesenchytraeus setosus, Michaelsen, Ai-ch. Mikr. Anat. 1 888, p. 494. 

Definition. Length, i^mm.; 1-12 setae per bundle, those of the lateral bundles of segments, 
V — VIII, much larger^ than others, and fewer in a bundle. Spermatheca with a single 
shm-t diverticulum, close to its oesophageal opening. Hab. — Germany. 
This species shares with the doubtful M. armatus the peculiarity in the setae 

' The exact numbers in the bundles of anterior segments are, according to Michaelsen. as follows 
(large setae indicated in thick type) : — 


















Right lateral . 

















Eight ventral . 

















Left ventral . 

















left lateral . 


















referred to in the definition. It is also remarkable in having more setae than any 
other Enchytraeid. Michaelsen does not mention the comparative lengths of the 
sperm-duct and the funnel which is often, apparently, a useful specific character. The 
clitellum is longer in extent than in any other Enchytraeid ; it reaches from the 
eleventh to the fourteenth segment, occupying, however, but half of the first and the 
last of these segments. The egg-sac is single and extends below the gut to the 
eighteenth segment ; the sperm-sacs are paired and do not extend beyond the fourteenth 
segment. The lymph corpuscles are flat and N'avicella-hk.e in form. 

(9) Mesenchytraeus armatus (Levinsen). 

Analycus armatus, Levinsen, Vid. Med. 1883, p. 232 ^ 

Mesenchytraeus armatus, Michaelsen, Abh. Nat. Ver. Hamb. 1889, p. 18. 

Definition. Setae, 1-6 in a lundle, those of segments IF — FI (one or two to each segment) 
double the length of the rest. Hah. — Denmark. 

This species agrees with the last in the possession of large setae ; it is to be 
distinguished from it by its smaller size, and also in the smaller number of setae 
in some of the segments. 

Genus STEECUTITS, Michaelsen. 

DEFiisriTiow. Setae as in Pachydrilus. Head-pores present. Dorsal vessel ante- 
clitellian in origin, with cardiac body; blood colourless. No salivary glands. 
Spermathecae not opening into gut. • 

This genus, of which our knowledge is entirely due to Michaelsen, is in many 
respects remarkable. It appears to connect the genera Mesenchytraeus and Buchholzia ; 
it has the cardiac body of the former genus, and the ante-clitellian origin of the 
dorsal vessel found in the latter. The other characters of the worm I shall regard 
as specific, and describe under the species — 

Stercutus niveus, Michaelsen. 
Stercutus niveus, Michaelsen, Arch. Mikr. Anat. 1888, p. 483. 
Definition. Length 6 mm. ; number of segments a8 ; setae 3-4 per bundle. Funnel of 
sperm-duct small; spermathecae oval, with no constriction between pouch and duet; 
glands at base of latter. Hob. — GerrAany. 

' See footnote, p. 317, 


The worm was called ' niveus ' on account of its very white appearance ; it occurs 
in earth which had been manured with fish debris, and was, on that account, and also 
because of its look, mistaken at first for a Dipterous larva. No head-pore was found, 
and, if present, it must be but small. The brain is deeply cleft behind, and the 
two lobes formed by this furrow are slightly divergent, and have the appearance 
of distinct lobes, owing to this fact and to the presence of a shallow lateral depression 
at the side of the brain, just where they join the main mass ; there are two pairs 
of muscles attached to the brain. The nephridia all have a short anteseptal part, 
and the duct arises near to the septum. The dorsal vessel springs from the intestinal 
sinus in the ninth segment; it contains a cardiac body. The intestine is mainly 
remarkable for the huge size of the peritoneal cells which clothe it. In the original 
description of the worm, Michaelsen. mentioned as a character the absence of a lumen 
in the whole of the alimentary tract lying behind the pharynx; in the rectum 
the lumen was wholly obliterated, while in the intestine the lumen was blocked 
by stellate cells ; that this is not characteristic either of the genus or even of the 
species, was subsequently shown by Michaelsen, who met with individuals living 
in ordinary garden mould, whose alimentary tract was quite normal ; it is, therefore, 
doubtless with justice, that Michaelsen attributed the peculiar condition of the 
gut in the first studied specimens to the nature of their diet. It should be mentioned 
that the specimens found in ordinary earth had been placed there by Michaelsen ; 
after two years they showed the normal gut-structure. The reproductive organs 
occupy the usual position. The funnels of the sperm-ducts are remarkable on 
account of their small size. 

Genus Pachydbilus, CLAPAniDB. 

Syn. Lumbricus, O. F. Mullee (in part.). 
Clitellio, LtJTKEN et alii (in part.). 
Nais, O. F. Mullee (in part.). 
Saenuris, Hoeemeistek (in part.). 
Enchytraeus, Ratzel et alii (in part.). 
Arehienehytraeus, EiSEN (in part.). 
Epitelphusa, Deago. 
Lumbrieulus, Oeested. 

Depinitioit. Setae /-shaped. Head-pore between prostomium and buccal segment, 
no dorsal pores. Brain incised behind; nerve-chord often with outgrowths in 



neighbourhood of genital segments. Blood yellow or red, dorsal vessel arises 

behind clitellum and is without cardiac body. No salivary glands. Testes 

multiple ; sperm-duct long. Copulatory glands often present. 

The genus Pachydrilus was instituted by Clapaeede, but not defined in such 

a way as to distinguish it from Snchytraeus j in a subsequently published memoir 

he indicated the differences between Pachydrilus and Enchytraeus, whose near 

relationship he had not formerly appreciated. These differences are in the red 

blood of Pachydrilus, its living upon the sea-shore, and finally the absence of 

dorsal pores. 

Vaillant adopted this definition of the genus, with the exception of the marine 
habit ; in his recent work upon the Oligochaeta he still adheres to it, modifying 
the definitions so as not to be universally applicable. With the exception of Ana- 
chaeta and Distichopus all the Enchytraeidae are referred to one of the two genera, 
Pachydrilus and Enchytraeus. 

CLAPAEi;DE's distinctions were criticised by Eisen (13), who remarked that Pachydrilus lacteus 
had colourless blood, and that all the red-blooded species do not live in water. Eisen therefore 
re-united the two genera, and divided the genus thus formed into a number of subgenera on the 
strength of the characters afforded by the shape of the brain. Michaelsen has criticised 
Eisen's position, pointing out that the features used by CLAPARiiDE were not the sole points of 
distinction between the genus Pachydrilus and its allies. Eisen's criticisms were to some extent 
justified by reason of the fact that Clapak^ide placed in his genus Pachydrilus the species P. lacteus, 
which has white blood. This species has also, as Michaelsen reminded us, setae unlike those of 
other Pachydrilus in being straight, and therefore like those of the genus Enchytraeus; the fact 
that it has, like other species of Pachydrilus, no dorsal pores is not a reason for including it in 

that genus, for in many species of Enchytraeus these pores are wanting. 


The genus Pachydrilus was for the first time satisfactorily defined in Vejdovsky's 
' Monograph of the Enchytraeidae ' (p. 50) ; his definition is as follows : — ' Borsten 
stark hakenformig gebogen. Blut ockergelb oder roth. Porus cephalicus zwisehen 
dem Kopf- und Mundlappen in der Mittellinie des Riickens. Die Segmentalorgane 
in alien Segmenten vorhanden, vom dritten anfangend. Hoden in biischelformigen 
Gruppen, gestielt.' This definition, it wiU be seen, includes all the important points 
mentioned ia Michaelsen's definition, but, as a matter of fact, Vejdovsky applied 
it inaccurately when he included ^Pachydrilus' sphagnetoru/m ; this species, as was 
afterwards shown by Michaelsen, has not lobed testes (Vejdovsky's specimens were 
immature), and has, therefore, been relegated by Michaelsen to another, but very 
nearly-i'elated genus Marionia. 

Michaelsen has argued that Deago's genus Epitdphusa is no more than 
a Pachydrilus; it has red blood, and the testes are '^ bouquet,' two facts which 


seem to be only explicable on the theory that the worm is a Pachydrilus. Deago 
certainly speaks of the setae as 'quasi diritte e corte'; but later says: 'Setole 
alquanto ricurve alle estremitk.' Construing the word ' quasi ' rather freely will, 
without any great effort, bring the genus Epitelphusa within the genus Pachydrilus. 

The chief characters of the genus are given in the definition; a few of these 
characters, and some others, may now be stated more in detail. 

One of the most characteristic structural features of the genus, though confined 
to a few species, is the outgrowths of the ventral nerve-chord in certain segments ; 
these occur in P. lineatus, P. inaximus, P. nervosus, and P. pagenstecheri ; they 
were first described by Eisen (13). 

EisEN does not do much more than briefly refer to these structures and figure 
them in P. nervosus; he remarks (p. ^^), 'In Archienchytraeus nervosus the ganglionic 
swellings of the fourth and the eleventh and twelfth segments attain an enormous 
development, and surpass the supra-oesophageal ganglion several times in size. It 
is remarkable that the said swellings are found only in the segments containing 
the organs of generation. In A. profugus and nasutus I have sometimes found 
certain irregular nervous enlargements in some of the segments, but not to be 
compared in size with those above.' 

The structures ai-e figured (not in section) on Plate viii, fig. i6 c and d. Miohaelsen 
recorded some years after a similar structure in P- *■ germanicus'( = P. lineatus); in the 
segments following the clitellum are lateral wing-shaped outgrowths of the nerve- 
chord, consisting of ganglion-cells ; they are compared to somewhat similar structures 
found by Timm in Phreoryctes; in his paper upon the Oligochaeta of South Georgia 
(15, p. SS)' MiCHAELSEN states that similar structures exist in P. maximus. 
In this paper these outgrowths are described in all of the three species in which 
they were then known to occur; they show constant, though small, difi'erences in 
these three species, agreeing in their main characters in all three. They are formed 
as a proliferation of the ventral mass of cells of the nerve-chord ; in P. nervosus 
this mass is lobate, and projects freely into the peritoneal cavity on either side 
of the nerve-chord, and independently of it. In the two other species they are 
in contact with the nerve-chord everywhere; in P. lineatus they approach each 
other dorsally. 

In the region where these structm-es exist, the nerve- chord sends a single 
median nerve to the body-wall, which is accompanied by a branch from each 
of the lateral masses; arrived at the epidermis, the mass of nerve-fibres spreads 
out right and left, and the epidermis is here modified, the cells being long and 
spindle-shaped, without any admixture of gland-cells. This region of the epidermis 

T t 3 


is also marked externally by a slight swelling^; Michaelsen thinks that these 
tracts of cells serve as sense-organs related to the generative function ; the swellings 
are thus distributed in the three species in which they occur: in P. maximus in 
xiv, XV, xvi ; in P. lineatus in xiii, xiv ; in P, nervosus in xiii, xiv, xv (according to 
EiSEN in iv, v, xii as well). 

Since the investigations of Michaelsen, Ude (1) has recorded the discovery of 
the ventral organs in P. pagenstecheri ; but no details are given, except that it is 
said that the organs most resemble those of P. lineatus. I am disposed to agree 
however with Hesse and to look upon these organs as copulatory glands. 

Michaelsen allows eleven species to the genus. They are these : — 

(i) Pachydrilus lineatus, O. F. Mulleb. 
{%) Pachydrilus nervosus, Eisbn. 

(3) Pachydrilus profugus, Eisen. 

(4) Pachydnlus verrucosus, Clapaeede. 

(5) Pachydrilus pagenstecheri, Ratzel. 

(6) Pachydrilus krohnii, Clapaeede. 

(7) Pachydrilus suhterraneus, Vejdovsky. 

(8) Pachydrilus catanensis, Deago. 

(9) Pachydrilus maxiTnus, Michaelsen. 

(10) Pachydrilus minutus, O. F. MCllee. 

(11) Pachydrilus fossarum, Taubee. 

Of these eleven, Michaelsen separates P. krohnii and P. catanensis as ' species inquirendae.' 
He also remarks, ■with reference to P. fossarum, that in the absence of further details than are 
given in Taubeb's and Levinsen's descriptions of tMs species, it cannot be regarded as certainly 
a ' good ' species. Vejdovsky, however, accepts it, but without giving any reasons for so doing. 
Tauber's definition is this : — ' Coi-pus, 20-40 mm. longum, postice attenuatum, ex segmentis 40-80 
constans. Setae minutae, apice leviter curvatae, 4-9 in quoque fascicule. Sanguis respiratorius 
purpureus. Color antice albidus, postice fuscus. Vitellus ruber. Ova mense Aprili-Junio in capsulis e 
stratis superioribus deciduisque clitelli formatis, multa in quaque capsula deponuntur. Variat rarior 

Levtnsen (2, p. 231) adds to this definition that the worm has a spermatheca with a duct distinct 
from the main pouch, and that the duct of the nephridia comes off from the middle of the organ, and 
not, as is usually the case, from the extremity. That the species is really a Pachydrilus is evident from 
the facts given by Taubek ; but there is not, in my opinion, any security that Levinsen's P. fossarum 
is the same species as that so named by Taubee. If there were, the species might be regarded as 
distinct on account of its large size and the origin of the nephridial duct ; I prefer to leave the matter 
as I find it. It should be mentioned that Levinsen considers that Taubek's P. verrucosus is identical 
with his P. fossarum, and that, according to Michaelsen, P. crassus also (of Taubee) is the same 

^ Of. Fridericia novae-selandiae, and Parenchytraeus. 


CLAPAEiiDE's P. krohnii is incidentally described in his memoir upon the anatomy of Lumbn'cus 
(1, p. 571, footnote) ; this worm was found in the soil near Kreuznach. It is 5-9 mm. long, and consists 
of about fifty segments. But CLAPABtoE gives no other characters that are of any use in differentiating 
the species ; it is hardly necessary to point out that the above are insufficient for this purpose. 

CzERNiATSKY has described the following species of 'Pachydrilus' : — P. gracilis, P. proximus, 
P. affinis, P. similis, P. lacustris, P. charkoviensis, and P. opacus. These are allowed by Vaillant, who, 
however, does not regard the descriptions as sufficient — a conclusion which most will share with him. 
All these species, together with P. cavicola (Joseph), are placed by Michaelsen at the end of his 
Revision of the Enchytraeidae as ' species inquirendae.' 

The well-characterized forms can be divided into two sets according to whether 

the spermatheca has or has not a duct distinct from the pouch. 

The following table will serve to discriminate the species : — 

A. Spermatheca without a duct distinct from the pouch. 

(i) Copulatory glands present. 

{a) nearly covering nerve-chord P. lineatus. 

(6) not covering nerve-chord P. litoreus. 

(2) No such structures. 

{a) Setae, 3-5 per bundle P. verrucosus. 

(6) Setae, 5-8 per bundle P. subterraneus. 

B. Spermathecae with a distinct duct marked off from pouch. 

(i) Copulatory glands present. 

i. Setae more numerous ia ventral bundles . . P. pagenstecheri. 
ii. Setae equal in both bundles. 

(a) Copulatory glands lobed P. nervosus. 

(6) Copulatory glands, not lobed P. maximus. 

(2) No copulatory glands. 

i. Setae not more than 9 in a bundle . . .P. profagus. 
ii. Setae as many as 14 in a bundle P. minutus. 

(i) Pachydrilus nervosus, Eisen. 

Arohienohytraeus nervosus, Eisen, Ofv. Svensk. Akad. 1878, No. 3, p. 73. 
Bnchytraeus nervosus, Vejdovsey, Syst. u. Morph. 1884, p. 41. 
Enehytraeus (Archienehsrtraeus) nervosus, Vaillant, Annel^s, p. 386. 
Pachydrilus nervosus, Michaelsen, JB. Wiss. Anst. Hamb. 1888, p, 58. 
Definition. Length, i^mm.j setae, 4-7 per bundle. Brain concave in front; outgrowths of 
nerve-chord large and lobed. Spermathecae with distinct demarcation between pouch and 
dzict. Hab. — Nova Zembla. 


This species, originally described and figured in many of its details by Eisen (13), 
has been more recently studied by Michaelsen (15), who has added to Eisen's 
account. The species is to be distinguished by the form of the outgrowths of the 
nerve-chord which are very large and lobed and do not cover the fibrous part of the 
chord. There seem to be no perivisceral corpuscles. 

(a) Pachydrilus profugus, Eisen. 

Enehytraeus pagenstecheri, Eisen, Ofv. Svensk. Akad. 1872, No. i, p. 133. 
Archienchytraeus profugus, EiSEN. Ofv. Svensk. Akad. 1878, No. 3, p. 73. 
Pachydrilus profugus, Levinsen, Vid. Med. 1883, p. 231. 

Paehydrilus pagenstecheri, Vejdovsky, Syst. u. Morph. 1884, p. 41 (in part.). 
Pachydrilus profugus, Michaelsen, Abh. Nat. Ver. Hamb. 1889, p. 34. 

Definition. Length, 1 8 mm. ; number of segments, about 50 ; setae, 8-9 per bundle. Brain 
concave in front. Spermalhecae with a distinctly marked duct rather longer than pouch 
and leset for its whole length with oval glands. Hah. — Greenland. 

This species, formerly confounded with P. -pagenstecheri by Eisen, was subsequently 
recognized by him as distinct ; the two species differ in — among other points — the 
characters of the setae, which in the present form are more or less uniform in size. 
In the definition of the species I therefore follow Eisen's more recent description of 
the species; in his earlier account he gives the length as 13 mm., the number of setae 
as 3-9, which number suggests that, after all, some specimens at any rate of the true 
P. pagenstecheri were included. 

(3) Pachydrilus verrucosus, Clapaeede. 

Pachydrilus verrucosus, Clapabede, M^m. Soc. Phys- Gen. 1863, p. 83. 

Definition. Length, i2,mm.; number of segments, about 40 ; setae, ^-^ per bundle ; integument 
covered with minute papillae. Spermatheca without distinct duct. Hah. — Hebrides. 

This species has been investigated by CLAPARiiDE, who has described and figured 
a good many points in its anatomy; he has not, however, described the brain, or 
mentioned whether the spei-matheca is furnished with glands as it is in other species. 
I include Taubee's Pachydrilus verrucosus among the synonyms, and do not relegate 
it, as does Michaelsen, to his own species Pachydrilus fossarum ; as Taubee merely 
gives the name and the locality where he met with the worm, it does not appear to 
me that there is any justification for doubting his identification. 


(4) Pachydrilus pagenstecheri, Ratzel. 
Enehytraeus pagenstecheri, Ratzel, Z. Wiss. Zool. 1868, p. 587. 
Pachydrilus pagenstecheri, Vejdovsky, Ench. 1877, p. 53. 
P Enehytraeus pagenstecheri, Taubek, Ann. Dan. p. 7 a. 
Pachydrilus limosus, Dieffenbach, An. Syst.-Stud. 1885, p. 106. 
Definition. Length, 1^-2,0 mm.; number of segments, 55-60; setae, 3-5 in dorsal, 7-10 in 
ventral bundles. Brain concave in front. Spermathecae with distinct duct twice the length 
of pouch, beset with glands. Hah. — Germany ; Bohemia. 
This species is easily to be distinguished on account of the numbers of the setae 
differing in the dorsal and in the ventral bundles. Ude never found more than 
six setae in the ventral bundles, which appears to upset the above distinction. Pachy- 
drilus limosus of Dieffenbach, which Ude identifies vrith P. pagenstecheri, has not 
more than five setae in a bundle. I am not at all sure that P. limosus is not a good 

(5) Pachydrilus subterraneus, Vejdovskt. 
Pachydrilus subterraneus, Vejdovsky, Rev. Biol. Nord, Vol. I. No. 4, p. i. 
Definition. Length, 20 mm. ; number of segments, 50-55 ; setae, 5-8 per bundle ; prostomium, 
buccal segment, and half of the following covered with papillae. Spermathecae without 
distinction between pouch and duct, funnel of sperm-duct unusually long. Hah. — Prague ; 
Lille; underground waters. 
This species has been investigated by Vejdovsky and is well characterized. It has 
the outward appearance of a Tubifex or a Phreatothrix, and the blood-vessels of the 
anterior part of the body, not specially described by Vejdovsky, are very long and 
coiled, contributing largely to this resemblance. A curious local peculiarity dis- 
tinguishes the specimens from Lille ; they have a circle of glandular cells surrounding 
the aperture of the spermatheca which are wanting in the individuals found in Prague ; 
the sperm-duct funnel is extremely long reminding one of the same organ in Marionia 
ebudensis ; it is represented by Vejdovsky as slightly coiled. 

(6) Pachydrilus maximus, Miohaelsen. 
Pachydrilus maximus, Miohaelsen, JB. Wiss. Anst. Hamb. 1888, p. 56. 
Definition. Length, 40 mm. ; number of setae per bundle tip to 7. Brain convex in front ; out- 
growths of nerve-chord completely cover chord laterally. Spermatheca with a short but 
distinct duct, surrounded at its extremity with a mass of glands. Hab. — South Georgia. 
This species is the largest, not only in the genus Pachydrilus, but among the 


Enchytraeidae in general. The great length of the funnel (eight times as long as 
broad) also distinguishes this species from any of its immediate allies. It resembles, 
however, in this particular, F. subterraneus, with which it cannot be confounded, 
owing to other differences. The anteseptal part of nephridium is short and broad ; the 
postseptal part is oval, with a longer duct arising from its hinder extremity. The 
glands of the epidermis are distinguished from those of P. nervosus and P. lineatus 
owing to their being deeply stained with Picrocarmine. Michaelsen found no 
giant-fibres in the nerve-chord, of which there are three in P. lineatus, but only one in 
P. nervosus. The dorsal vessel arises at the end of the fourteenth or fifteenth 

(7) Pachydrilus minutus, 0. F. Mullee. 

Lumbricus minutus, 0. F. MtJLLER, Zool. Dan. Prodr., 1776, p. 216. 
PClitellio minutus, Lutken, Kevised Cat. Annel., 1884. 
Pachydrilus minutus, Lbvinsen, Vid. Med., 1883, p. 231. 
Clitellio minutus, Vaillant, Annelds, p. 420. 
Definition. Length, 14 mm.; number of segments about 24;' setae 12-14 per bundle. 
Spermathecae with distinct duct beset with glands. Hab. — Greenland; Denmark. 
This species would have had to be placed among the ' species inquirendae/ had 
it not been for Leyinsen's description ; this rendered the distinctness of the species 
beyond question. Its chief character is the very large number of setae in a bundle ; 
this, coupled with the separation between the pouch and the duct in the spermatheca 
enables it to be easily defined. The above synonymy is given on the authority of 
Michaelsen. Vaillant retained the species of 0. F. Mullee in the genus OliteUio in 
spite of the segments occupied by the elitellum. Levinsen does not query Clitellio 
minutus of Lutken as a synonym. 

(8) Pachydrilus lineatus, O. F. Mullee. 

Lumbricus lineatus, O. F. MiJLLER, Verm, terrestr., 1. ii, 1774, p. 29. 

Nais littoralis var. mutica, O. F. Mtjllee, Zool. Dan., ii, 1788, p. 56. 

Saenuris lineata, Hofpmeistee, Arch. f. Nat., 1843, p. 195. 

Pachydrilus rivaUs, Levinsen, Vid. Med., 1883, p. 231. 

P. germanicus, Michaelsen, Ench. Mob., 1886, p. 43 et passim. 

P. lineatus, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 23. 
Definition, length, io mm.; number of segments ^o ; setae, ^-% per bundle. Copulatory glands 
nearly meet above nerve-chord. Spermathecae without distinction betioeen pouch and duct ; 
glands at base. Hab. — Germany ; Denmark 


MiCHAELSEN gives a much longer list of synonyms than that given above ; I have, 
however, included all of his (excepting 'Gordius pallidus linea longitvdinali rufa] 
0. F. MtJLLEB, excluded because not binomial) that are not queried ; the principal 
reason which led Michaelsen to identify Muller's species with that named by himself 
P. geTTnanicus is the exact correspondence of locality. 

(9) Pachydrilus litoreus, Hesse. 
p. litoreus, Hesse, Z. wiss. Zool. 1893, p. 3. 
Definition. Length, 1 7 mm. ; numher of segments 40. Setae 5-6 per bundle. Copulatory glands 
do not meet above nerve-chord [in XITI-XF). Hab, — Naples. 

Genus Marionia, Michaelsen. 

Syn. Pachydrilus, ClapaeIide (in pai't.). 
Euchsrtraeoides, ROULE. 

Depiwitioit. Setae y-shaped. Head-pore between prostomium and buccal segment ; 
no dorsal pores. Blood coloured; dorsal vessel arises behind clitellum. No 
salivary glands. Testes massive. 

The only difference that there is between this genus and Pachydnlus lies in 
the form of the testes, which in Marionia are compact organs, not divided at the 
free extremity into several lobes, as in Pachydrilus. Whether this character is 
sufficient upon which to found a generic division is not quite certain ; in the 
meantime, however, I adopt Miohaelsen's opinion. The five species allowed by 
Michaelsen may be distinguished as in the table; Marionia enchytraeoides (=Enchy- 
traeoides Marionii of Eoule) is considered by Michaelsen to be a 'species 

I. Gonads three or four segments in front of normal position . M. sphagnetorum. 
II. Gonads normal in position. 

(]) Funnel of sperm-duct very long M. ebudensis^. 

(3) Funnel not specially long. 

i. Spermathecae with a thick covering of glands at base. 

(a) Spermiducal glands very large M. semifusea. 

(b) Spermiducal glands not so large M. crassa. 

ii. Glands at base of spermathecae small and few . . M. georgiana. 

' This is only doubtfully a Marionia. 

u u 


(i) Marionia sphagnetorum (Vejdovsky). 

Pachydrilus sphagnetorum, Vejdovsky, SB. Bohm. Ges., 1877, p. 304. 

P. (Archienchytraeus, Eisen) sphagnetomm, Michaelsen, Ench. Mob., J 886, p. 43. 

M. sphagnetorum, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 29. 

Definition. Length, 15 mm.; number of segments 50; setae, 3-4 per bundle. Duct of 
nephridiiim long, and arising just behind septum. Gonads moved a few segments in front 
of usual position, llab. — Germany. 

The examples of this species studied by Vejdovsky were sexually immature ; 
but this deficiency in our knowledge of the species was filled up later by Michaelsen. 
Michaelsen's observations were conducted upon what he considered to be a variety 
of the species to which the name 'glandidosa' was applied, but he was able, in 
the same paper in which his results were described, to state that the same 
characters were to be found in the typical form '■sphagnetorum.' The variety 
"■ glandulosa' is a stronger worm than the type-form ; it measures 20 mm. as against 
] 5 ; the number of setae in a bundle are two or three instead of three or four. The 
principal difierence, however, between the two forms is in the septal glands; in the 
type there are five or six pairs of these glands, a pair to each segment ; in the 
varietal form about nine pairs, owing to the fact that the duct connecting the 
several glands of one side of the body have given rise to additional glands. The chief 
character of the present species is in the abnormal position of the sexual organs, 
which have, as in Buchholzia appendiculata, been moved a few segments in advance 
of the normal position. They are not, howe^^r, constantly found in one particular 
segment ; in some individuals they are three, in others four segments, in front of 
those which contain them in other Enchytraeids. As in the Buchholzia the spermatheeae 
have preserved their normal position. The latter are composed of a very long and 
narrow tube, which swells out at the blind extremity into an oval pouch which 
does not communicate with the gut ; a little way in front of the external pore 
(which has a mass of glands on one side) the tube has a spherical dilatation. These 
spermatheeae are strikingly like the spermatheca of Anachueta bohemica minus 
the terminal pouch through which the tube opens on to the exterior^. The funnels 
of the sperm-ducts are about three times as long as broad. The duct of the 

' In a longitudinal section of the spermatheca (optical) figured by Michaelsen (3, Taf. xxiii, fig. 2 c) 
the lumen of the spherical swelling is reduced by a projection into it of a plug containing distal part of duct. 
This arrangement would seem to facilitate entrance of sperm into spermatheeae, but to hinder its exit, at 
least until there is sufficient to fill the crescentic lumen of the dilatation. 


spermatheca is so long that the organ often reaches back as far as the seventh 
segment. The lymph-corpuscles are flat, -with a pear-shaped outline, subject to slight 
variations. This species is sexually mature in latter half of August. 

(2) Marionia ebudensis (Clapaeede). 

Paehydrilus ebudensis, Clapaeede, M^m. Soc. Phys. Gen., 1862, p. 85. 
M. ebudensis, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 29. 

Definition. Length, 12, mm.; number of segments 47. Funnel of sperm-duct very long. 
Sperrnatheca hardly distinct from pouch. Eah. — Island of Shye. 

This species has been briefly characterized by Glapar^de, the only naturalist 
who has investigated the worm. I am not quite sure that Michaelsen is fully 
justified in placing it in the present genus, rather than in Paehydrilus. Nothing is 
said by Claparede as to the gpnads, whether they are 'uniques' or 'multiples.' 
Indeed, one would rather infer that the testes were multiple ; for, in describing the 
preceding species (P. verrucosus) Claparede remarks that the gonads are multiple and 
not simple, ' comme dans les e'speces prdcedentes ' {M. cratsa and M. semifusca) ; it 
would rather seem to follow, therefore, that, in the species following, the testes 
were constructed on the plan of those of P verrucosus ; or, at the very least, the 
reverse could not be inferred. 

(3) Marionia semifusca (Clapabjide). 

Paehydrilus semifuscus, CLAPAEtDE, Mem. Soc. Phys. Gen., 186'a, p. 76. 
M. semifusca, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 29. 

Definition. Length, 10 mm, Spermathecae with a duct twice as long as pouch, and with 
a thick coating of gland-cells at pore; spermiducal gland very large. Hob. — Island 
of Shye. 

This species has been imperfectly described by Claparede ; it is stated in this 
description that the testes are in the tenth (eleventh), and the ovaries in the 
twelfth (thirteenth) segment; there must be an error here, one would suppose. 
The funnel of the sperm-duct is hardly twice as long as broad. The spermiducal 
glands are so large that they occupy the whole of the coelom of their segment, and 
the body is even bulged out in this region. The nephridia (which are figured, 

u u 2 


but not described), have a very short anteseptal part ; the duct arises some little 
way behind the septum. 

(4) Marionia crassa (CLAPAEtDE). 

Pachydrilus crassus, CLAPAEfeDE, M^m. Soc. Phys. Gen., 1862, p. 79. 
M. crassa, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 29. 

Definition. Length, i^mm.; number of segments, 48; setae, 2-5 per bundle. Spermathecae 

furnished at base with minute glands. Lymph corpuscles of two different hinds. Kab. — 

Island of Shye. 

It seems to me to be a little doubtful whether this species is really referable to 

the genus Marionia or to Pachydrilus. The sharp demarcation of two kinds of lymph 

corpuscles is not met with elsewhere in the genus, and is used by Michaelsen to 

distinguish the species from other Marionia. Some of the cells are roundish with 

numerous granules, the others have the form more usual in the genus and are fusiform 

with no granules. Another point in which the species differs from other members 

of this genus or the last is in the form of the ' testes.' Clapabede describes them 

as extending from the eighth (ninth) to the eleventh (twelfth) segment, an extent 

which is suggestive rather of sperm-sacs than testes. The fact that the organs in 

question are unpaired ^ is in favour of this suggestion. I do not pretend to speak 

decisively upon the matter, which requires looking into. 

(5) Marionia georgiaiia, Michaelsen. 

Pachydrilus georgianus, Michaelsen, JB. Hamb. Wiss. Anst., 1888, p. 65. 
M. georgiana, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 29. 

Definition. Length, 8 mm.; setae, 5-7 j»e?- bundle. Spermathecae with few glands at pore. 
Dorsal vessel springs from a cup-like depression upon intestine. Hob. — South Georgia. 

The last character in the definition distinguishes the species. The duct of the 
spermatheca is about half the length of the pouch. Septal glands are developed in 
segments iv, v, vi. The dorsal vessel originates at end of thirteenth segment. 

' Clapae^de uses the singular in writing of them. But so does he in the species M.fusca, so that this is 
perhaps not quite enough reason for inferring that the organs are unpaired, d'Udekem, however, describes 
as 'unique,' and figures as single, the ovary oi Fridericia gcdba. 


Genus Buchholzia, Michaelsbn. 
Syn. Enehytraeus, Buchholz (in part.). 

Definition. Setae, /-shaped. Head-pore between prostomium and buccal segment ; 
no dorsal pores. Dorsal vessel arises from the tip of a median dorsal 
diverticulum of the oesophagus in front of the clitellum ; blood colourless. 
Salivary glands present. 

This genus was founded by Michaelsen (1, p. 293) for the species termed by 
Buchholz Enehytraeus appendiculatus. It was defined by Michaelsen in the 
following words : — ' Enchytraeiden mit farblosem Blut und leicht hakenformig ge- 
krlimmten Borsten. Die Speicheldriisen sind breit gelappt, mit kurzem Ausfiihrungsgang, 
und miinden im iv. Segment seitlich in den Oesophagus. Das Gehirn ist im Hinterrande 
gerade abgestiitzt. Der Giirtel entwickelt sich am viii. Segment und an der vordern 
Halfte des ix. (schon von Buchholz der Hauptsache nach richtig angegeben). Die 
Samentrichter liegen vor dem Dissepiment vii/viii, und gehen, dieses Dissepiment 
durehbohrend, in sehr lange, feine Samenkanale liber. Die Samenkanale miinden 
im viii. Segment in der ventralen Borstenlinie nach aussen aus. Die Eileiter habe 
ich nicht erkennen konnen. Die Hoden bilden sich am Dissepiment vi/vii, die Ovarien 
am Dissepiment vii/viii. Die Samentaschen liegen im v. Segment und miinden in der 
Intersegmentalfurche iv/v nach aussen.' 

It will be noticed that in this definition the position of the reproductive organs 
is the main point made use of, while in the generic definition given above the 
position of these organs is not so much as mentioned. The discovery of a new 
species of the genus led Michaelsen to abandon the unusual position of the generative 
organs as a part of the necessary definition of the genus. The new species Buchholzia 
fallax agrees so closely with the type species B. appendiculata in all characters except 
the forward position of the reproductive organs that it was, as I think with reason, 
included by Michaelsen in the same genus. The main characteristic of the 
genus is in the single dorsal diverticulum of the oesophagus whence arises the 
dorsal vessel. This organ was figured and described by Buchholz (PI. iv, fig. 2 app.) 
and led to his selecting the name of ' appenndiculatus ' for the species. Later it 
was described and figured by Vejdovsky (3, PL ii, fig. 8). Both these observers noted 
the fact that the diverticulum was composed of a bundle of parallel tubes, of which 
Michaelsen subsequently gave a rather more detailed account in his paper upon the 
' Chylusgef ass-system' of the Enchytraeidae. Vejdovsky regarded the tubes as blood- 


vessels ; Michaelsbn pointed out that the gland is formed of a mass of glandular 
tubes in connection with the lumen of the gut, which are suiTOunded by a blood- 
sinus continuous on the one hand with the sinus surrounding the alimentary canal 
lying behind the divei-ticulum, and on the other hand with the dorsal vessel ; the 
tubes appear to be intracellular, hut do not seem to be ciliated ; the whole gland 
is covered by a layer of peritoneum. In transverse sections its paired character 
becomes apparent, but the two halves are enclosed within one peritoneal sheath. 
This genus Buchholzia is the only genus with ' Pachydrilus-setae ' which has salivary 
glands ; these glands are small and formed of a short duct which expands at the 
blind extremity into a group of oval sessile diverticula ; they open some way behind 
the pharynx. 

(i) Buehliolzia appendiculata (Buchholz). 

Enchytraeus appendiculatus, Buchholz, Schr. Phys.-Oek. Ges. Konigsb. i86a, 

P- 93- 
E. pellucidus, Vejdovsky, SB. Bohm. Ges. 1877, p. 301. 

E. (Mesenehytraeus) appendiculatus, Vejdovsky, Syst. u. Morph. 1884, p. 40. 
E. (Mesenehytraeus ? Eisbn) appendiculatus, Michaelsen, Ench. Mob. 1886, p. 47. 
B. appendiculata, Michaelsen, Arch. mikr. Anat. 1886, p. 393. 

Definition. Length, 10 mm.; numher of segments, about 2,S ; setae, ^ in lateral hundles, 4-6 
in ventral ; clitellum on Til, Till. Testes in VII; spermathecae in V with long duct 
to which are apperuled two large glands. Hai. — Germany; Denmark; Bohemia; Italy. 

As may be inferred from the above list of references, the present species has been 
the subject of a considerable amount of study. It is indeed one of the best known 
of the Enchytraeidae. The differences which it shows from Buchholzia fallax will 
be mentioned under the description of the latter species. For its anatomy see also 
Vejdovsky (3), Michaelsen (1). 

(2) Buchholzia fallax, Michaelsen. 

B. fallax, Michaelsen, Arch. mikr. Anat. 1887, p. 374. 

Definition. Length, 10 mm. ; setae, 4-6 per bundle of different lengths. Sexual organs 
occupying the usual situation ; spermathecae with diverticula near to distal swollen 
extremity, two large glands at external pore. Hob. — Germany. 

This species has been investigated by Michaelsen. Its chief difference from the 
last species is that the sexual organs occupy the position which these organs occupy 


in other Enchytraeidae, and are not moved forwards as in Buchholzia appendiculata. 
The setae of each bundle are of unequal sizes and are so aiTanged that the longest 
setae of a ventral bundle come nearest to the longest setae of a lateral bundle. The 
salivary glands are still more reduced than in the last species ; the median dorsal 
diverticulum is only different from that of B. appendiculata in a few small points ; 
it is not divided into two halves and the tubes are not so closely pressed together 
as in the former species. The spermathecae are very unusual in the peculiar character 
of the diverticulum ; this is a canal surrounding the end of the spermatheca a little 
way off the end which opens into the oesophagus ; this canal communicates here and 
there by pores with the main pouch. Michaelsen has also mentioned as existing 
in the sixth or the seventh and eighth or in the ninth segment, unpau-ed papilla like 
outgrowths of the epidermis which have the appearance of imperforate penes ; the 
describer of the species thinks that these are not on account of their unpaired nature 
to be looked upon as penes; otherwise, as he points out, we can more easily under- 
stand the abnormal position of the reproductive organs in B. appendiculata. They 
appear to me to be more comparable to the genital papillae so frequently met with 
in the higher Oligochaeta. 

Genus Enchytraeus, Hbnlk 

Syn. Ifeoenchytraeus, Eisen (in part.). 

Archienchytraeus, Michaelsen, Rosa (in part.). 
Paehydrilns, Taubee (in part.). 

DBPUsriTioii. Setae of each bundle of equal length, straight, only hooked at free 
extremity. Head-pore between prostomium and buccal segment. Oesophagus 
passing gradually into intestine. Dorsal vessel arises behind clitellum. 

The principal character of this genus is in the shape of the setae. The other 
characters are merely negative. It contains, according to Michaelsen, ten species, 
which are reduced to nine by Ude, who unites E. hwniicvltor with E. vejdovskii. 
The species may be thus distinguished: — 

I. Brain deeply cleft behind. 

i. Setae only one in each ' bundle ' B. monochaetus. 

ii. Setae more than one in each 'bundle': 

(i) Blood yellow B. arenarius. 

(3) Blood colourless E. spiculus. 


n. Brain convex behind or but slightly concave. 

i. Funnel of sperm-duct very long E. humicultor. 

ii. Funnel short : 

(i) Spermathecae almost sessile E. adriaticus. 

(2) Spermathecae with longish duct: 

(a) Preseptal part of nephridium with straight duct. 

Duct of spermatheca dilated E. hyalinus. 

No such dilatation E. buchholzii. 

(b) Preseptal part of nephridium with a convolute 

duct . . . . E. argenteus. 

It will be observed that I omit Levinsen's E. affinis, which is admitted by Michaelsen. This 
species (termed E. danicus by Vaillant, on the ground of the pre-occupation of the name 'affinis' by 
Eisen) is only mentioned by MlCHAEtSESf in his Revision to distinguish it on account of its coloured 
blood. It does not seem to me that Michaelsen is justified by the extremely slender description 
given by Levinsen in referring this species to the genus Enchytraeus (s. s.) ; no doubt, there is nothing 
in the description which positively forbids this identification ; the species, in all probability is referable 
either to this genus or to Fridericia ; but why not to the latter ? Levinsen does not mention the 
dorsal pores, but neither does he in the case of ^Enchytraeus' bisetosus, placed by Michaelsen in the 
genus Fridericia. The most remarkable fact about the species is the presence of 'chlorophyll 
corpuscles' in the integument. Concerning these further information is to be desired. 

'Enchytraeus' minut'us of Tauber may or may not belong to this genus. It is not referred to 
by Michaelsen. It is, if an Enchytraeid, the minutest species, measuring only o-i mm.; the 
setae are paired (cf. Fridericia iisetosa) : ' Semen vivum in segmento quai-to, ova in quinto sextoque.' 
This does not seem intelligible on the theory that the worm is of this family. 

(i) Enchytraeus humicultor, Vejdovsky. 

Enchytraeus humicultor, Vejdovsky, Ench., 1879, p. 57. 

E. spiculus, MoBius, JB. Comm. wiss. Unt. deutsch. Meer., 1873, P- i°7 (™ part.). 

Neoenehytraeus vejdovskii, Eisen, Ofv. Svensk. Akad., 1878, No. 3, p. 75. 

N. Stuxbergi, Eisen, ibid., p. 75. 

Pachydrilus lacteus, Taubee, Ann. Dan., 1879, P- 7i- 

E. fucorum, Levinsen, Vid. Med., 1883, p. 235. 

E. sordidus, Levinsen, ibid., p. 235. 

E. vejdovskii, Vejdovsky, Syst. u. Morph., 1884, p. 41. 

E. Stuxbergi, Vejdovsky, ibid., p. 41. 

Archienchsrtraeus mbbii, Michaelsen, Zool. Anz., 1885, p. 237. 

E. mobii, Michaelsen, Ench. Mob., 1886, p. 1 1. 

' To the above list of synonyms should possibly be added E. vermimlaris of Taueek (^in part.), which 
Levinsen identifies with his E. sordidus. 



Definition. Length, ^^ mm.; number of segments, ']^; setae, ^-^ {rarely 6) 2}er bundle. Brain 
slightly concave -behind. Nephridia with only funnel anteseptal Sperm-duct funnels very 
longs duct of spermatheca as long as, and sharply marked off from, pouch, and beset with 
glands along entire length. Hab. — Nova Zembla ; Siberia; Denmark; Germany. Sea-shore. 

The above list of synonyms is adopted on the authority of Michaelsen, and E. humicultor 
is added on that of Ude ; the name, therefore, of the species must be altered from E. vejdovskii 
to E. humicultor. Michaelsen's identification of six differently named species -with E. vejdovskii 
depends, not on a mere collating of the various descriptions given, but upon an examination of most of 
the types; it is clear, therefore, that there is no choice but to follow him. Eisen has stated that 
E. stuxhergi has red-coloured blood; but, as Michaelsen has pointed out, this apparept difference 
from E. mobii, &c., is not to be regarded, as the worm was not examined in the fresh condition. 
Moreover, half of the material labelled by Eisen E. stuxhergi is really made up of specimens of 
Pachydrilus nervosvs, which, of course, has red blood. As for the two species described by 
Levinsen, and referred to this species, I do not gather that Michaelsen examined type-specimens ; 
the synonym, P. lacteus of Taxjbee, I understand Michaelsen to have introduced on the authority of 
Levinsen. The chief reason for believing in the identity of E. sordidus and E. fucorum with 
E. vejdovskii, &c., appears to be in the similarity of the spermathecae with their continuous coating of 
gland-cells upon the duct. The sole difference between Vejdovsky's E. humicultor and E. vejdovskii 
is in the presence in the former of dorsal pores ; Ude, who considers these species to be the same, 
points out that the presence of these pores requires verification. They are, however, figured by 
Vejdovsky (3, PI. V, fig. 6). I may mention another small difference ; Michaelsen, in his account 
of the anatomy of 'Enchytraeus mobii' (the most complete account of the species), states that the 
dorsal vessel springs from the blood-sinus of the intestine in the fifteenth segment, and that it is 
dilated in this and the two segments lying in front into hearts. On the other hand, Vejdovsky 
described and figured the dorsal vessel as arising in the seventeenth segment, and the hearts as lying in 
segments xvi, xv, xiv. Ude has neglected this point, which, however, inay not be a serious matter of 
difference. In other respects the two species undoubtedly agree. 

E. humicultor has the same long sperm-duct funnel, and the duct of the sper- 
matheca is, in the same way as in ^E. mobii,' Sec, beset with glands along its whole 
course ; the nephridia in every case have an anteseptal portion, which consists 
only of the funnel. Between the male-pores is an area of the clitellum devoid of 
glandular modification. Four perivisceral trunks connect dorsal and ventral vessel, 
of which the two middle pairs belong to the fourth segment. 

(2) Encliytraeus spiculus, Fket and Leuckart. 

B. spiculus, Frey and Leuckakt, Beitr. Kennt. wirbell. Th., 1847, p. 150. 

E. ? spiculus, Taubee, Ann. Dan., 1879, p. 73. 
Definition. Length, to mm.; number of segments about 30; setae 4-6 (rarely 8) per bundle. 
Anteseptal portion of nephridium with straight duct. Brain deeply cleft behind. Hab. — 
Heligoland; Cuxhaven; Wilhelmshaven. 

X X 


This species has been principally investigated by Michaelsen (13)- 

It does not, however, seem to be by any means certain that the species termed by Michaelsen 
'spiculus' are the same as the E. spiculus of Leuckaet. Vejdovsky places this species among 
the 'species inquirendae,' and Vaillant (6, p. 247) remarks that 'il est assez difficile de pouvoir 
appreoier la valeur des E. spiculus, &c.' emphasizing this remark by relegating the species to the 
section ' iucertae sedis ' on p. 253 of his work. 

If Michaelsen's identification be right, the species can easily be distinguished 
from others; as a general rule there are not more than six setae in a bundle, but 
occasionally the number is as high as eight. The spermathecae consist of a pear-shaped 
pouch, communicating with the exterior by means of a rather short, straight, simple 
duct. The nephridia are straight, or bent sharply at right angles ; there is no 
constriction at the septum; there is no distinctly differentiated duct leading to 
the exterior, the calibre of the organ diminishing but slightly at the point where 
it comes into contact with body-wall ; the anteseptal part of the organ, besides the 
characters mentioned in the definition of the species, is to be distinguished by the 
absence of granules, and is, therefore, quite clear. The clitellum is remarkable for 
the fact that there are regularly alternating transverse lines of darkly-staining and 
hardly-staining cells. A curious point about the species appears to be that the 
cocoons contain only one egg, differing, therefore, from all the other marine 

(3) Enchytraeus buchliolzii, Vejdovsky. 

E. buchholzii, Vejdovsky, SB. Bohm. Ges, 1877, p. 302. 
ArcMencliytraeus Bucholzii, KosA, Boll. Mus. Zool. Torino, 1887, No. 29. 

Definition. Length, 10 mm.; number of segments, 28; setae, 2-3 per bundle. Brain slightly 
concave behind,. Nephridia with narrow anteseptal portion, with straight lumen, no 
differentiaied duet. Spermathecae with duct as long as pouch, a group of glands at pare. 
Hab. — Denmark; Gertnany ; Bohemia; Italy. 

This species has been chiefly studied by Vejdovsky and by Ude (1). The salivary 
glands are long and much coiled in the distal half; these organs are figured by 
Vejdovsky (3, PI. iii. fig. i) as gradually diminishing in calibre towards the free 
end, but Ude found that, in all the examples of the species examined by himself, 
the coiled part of the gland arose from a broad tract. The funnels of the sperm- 
ducts are not large; the duct is composed of comparatively few coils, and the 
spermiducal glands at its termination are large. 


(4) Enelijrtraeus adriaticus, Vejdoysky. 
E. adrlaticus, Vejdovsky, SB. Bohm. Ges., 1877, p. 303. 

E. adriaticus (forma jaltensis), CzEENiAVSKY, Bull. Soe. Nat. Mosc, 1880, p. 32a. 
Definition. Length, -\.^mm.; number of segments, 25; setae, 3 jser bundle. Brain convex 
posteriorly. Anteseptal part of nepJiridium liardly narrower than postseptal, with coiled 
lumen. Spermatheca with short duct covered by three rows of glawlular cells. Hah. — 
Austro-lllyrian coast-line ; Jalta. 

(5) Enchytraeus monochaetus, Michaelsen. 
E. monochaetus, Miohablsen, JB. wiss. Anst. Hamb. 1888, p. 66. 

Definition. Length, 7 mm. ; setae in four rows of a single seta each ; setae usually tvanting 

upon the first four or five segments, lateral setae wanting upon a few segments after. No 

salivary glands. Brain deeply cleft behind. Nephridia bent at right angles, anteseptal part 

not much more than funnel, duct not sharply marked off. Spermathecae with a long duct as 

long as, and sharply distinguished from, the pear-shaped pouch, a few glands at base. 

Hob. — South Georgia, sea-shore. 

This species is most remarkable for the ' cephalization,' met with in no other 

Enchytraeid. Michaelsen states that the cuticle is unusually thick. In the buccal 

cavity is a taste-papilla. Besides the other characters mentioned in the above 

definition, the species has one pair of small septal glands in the fourth segment, 

two pairs of larger glands in the fifth segment, and three pairs of very large ones 

in the sixth segment ; the sperm-duct funnel is short. 

(6) Enchytraeus arenarius, Michaelsen. 
E. arenarius, Michaelsen, ibid. 1889, p. 12. 

Definition. Length, i o mm. ; setae, 3 per bundle. Brain deeply cleft behind. Anteseptal part 
of nephridia little more than funnel, duct arises near to septum as in Mesenchytraeus. 
Spermathecae cylindrical with no distinction between pouch and duct, external pore toith 
glandular cells. Hub. — Germany {Elbe shore). 

Michaelsen describes as a remarkable character of this species the form of 
the perivisceral corpuscles; these are furnished along one side with numerous 
pointed processes. The blood, moreover, is yellow — the present species being the 
only one of the genus which shows this Pachydrilus-\ikQ character, if we ' except 

X X 2 



the doubtful species E. affinis. This resemblance to Pachydrilus is further increased 
by the fact that the ventral nerve-chord has outgrowths such as those which occur 
in P. nervosus (see p. 323), which are found in the first post-clitellian segments. The 
nephridia are peculiar in the points mentioned in the definition; besides these, 
MiCHAELSEN figures (fig. 5 c) a transparent sheath covering the dorsal aspect of the 
post-septal part of the organs. The funnels of the sperm-ducts are longer than in 
any other Enchytraeus ; the proportion of length to breadth is about la : i. 

(7) Enchytraeus argenteus, Miohaelsen. 
B. argenteus, Michaelsen, ibid. p. 15. 

Definition. Length, 5 mm. ; number of segments, 30 ; setae, a-3 jier bundle. Brain convex 

heUnd. Anteseptal part of nephridia of equal diameter with postseptal part, containing 

also a coiled lumen, duct comes of at right angles, and is long and distinct. Hob. — 

Germany {Elbe shore). 

This species appears to be the smallest of the Enchytraeidae. The name of the 

species was given to it on account of its silver colour, due to the dark pigmentation 

of the perivisceral corpuscles. 

(8) Enchytraeus hyalinus (Eisen). 

Neoenehytraeus hyalinus, Eisen, Ofv. Svensk. Akad., 1878, No. 3, p. 76. 
B. hyalinus, Vejdovsky, Syst. u. Morph., 1884, p. 41. 
E. (Weoenchytraeus) hyalinus, VaillaN'O^ Annel^s, p. 264. 

Definition. Length, 8 mm. ; number of segments 43 ; setae, 3 per bundle. Brain convex 

behind. Anteseptal part of nephridium long, with nearly straight duct. Spermathecae with 

a dilatation upon the d^ict, at opening of which are a few glands. Hob. — Nova 


Michaelsen (5, p. 40) distinguishes this species from E. adriaticus by the 

presence in the latter of numerous small pear-shaped glands ; I do not think that 

this difference will hold; Eisen figures (13, PI. x, fig. 201) quite similar glands in 

E. hyalinus. I do not perfectly understand the shape of the spermathecae from 

Eisen's figures ; he describes the spermatheca as consisting ' of two distinct parts, the 

lower one is funnel-shaped and wrinkled, and furnished at the base with small glands,' 

&c. I suppose that merely a dilatation of the duct is meant such as occurs in the 



Genus Pbidericia, Michaelsen, 

Syn. Enchytraeus, AucT. (in part.). 

Neoenehytraeiis, Eisen (in part.). 

DEFlBriTlOBr. Setae straight, either two to a bundle or more, in which case the 
middle setae are much smaller. Head-pore between prostomium and buccal 
segment, dorsal pores also present. Salivary glands always present. Dorsal 
vessel springs in nearly every ease behind the clitellum ; blood colourless. 
Spermathecae, as a rule, with appendices. Copulatory glands sometimes present. 

K it is a little difficult to define Enchytraeus (s. s.) there can be no difficulty in 
distinguishing the present genus. The most conspicuous character is undoubtedly 
the presence of dorsal pores, which are not found elsewhere among the Enchytraeidae, 
save only in the doubtful (according to Udb) instance of Enchytraeus humicultor. 
Another matter in which the genus is peculiar concerns the setae ; these are developed 
in each bundle two at a time ; the next youngest pair lie within the first-formed, the 
following within these, &c. ; the result is that, in a bundle, the two outermost setae 
are the oldest and longest, the innermost the youngest and shortest; sometimes the 
symmetry of the arrangement is spoiled by one seta falling out; in the species 
F. bisetosa it appears that the older pair falls out before the next is developed ; hence 
this worm has two setae only in each bundle. This remarkable and interesting mode 
of development of the setae in Enchytraeids was first made known by Vejdovskt in 
his work Upon the Enchytraeidae (3). 

MiOHAELSEN allows eleven species^ of Fridericia, to which may be added my 
species F. antarctica, which may be distinguished from each other by the following 
table : — 

I. Spermathecae without diverticula. 

i. Salivary glands branched P. striata. 

ii. Salivary glands not branched F. bulbosa. 

' Enchytraeus vermiiybUaris of Clapae^de (3, p. 55) seems to be undoubtedly a Fridericia, since it has dorsal 
pores (3, PI. ii. figs. 10, 11). The absence of diverticula to the spermathecae ('les receptacles de la semence 
. . . ressemblent beaucoup a ceux du Pachyckilus semifuscus ') places the species in the neighbourhood of F. striata 
and F. bulbosa, from both of which it differs in that the anteseptal part of the nephridium is long, but has 
a straight duct. 


11. Spermathecae with diverticula. 

i. Anteseptal part of nephridium no more than funnel. 

(i) Diverticula of spermatheca two I", leydigii. 

(a) Diverticula very numerous P. hegemon. 

ii. Anteseptal region large, with coiled duct, 
(i) Glands at base of spermathecae. 

(a) One large gland P. dura. 

(b) Several glands P. ratzelii. 

(2) No glands at base of spermathecae. 

{a) Diverticula in the form of solid glands ... P. lobifera. 
(6) Diverticula hardly developed P. callosa. 

(c) Two diverticula. 

(a) Setae paired ... . P. bisetosa, 

(/3) Setae 4-6 P. perrieri. 

(d) Four diverticula. P. galba. 

(e) Seven diverticula .... P. antarctiea. 

In two species of the genus the spermathecae are simple, without diverticula ; 
in all the others these diverticula are present, and sometimes (in F. hegemon, for 
example) extremely numerous. This recalls the peculiar ring, with occasional openings 
into the spermatheca, which surrounds the spermathecae of BucUiolzia fallax. The 
dorsal pores commence in the seventh ring; the perivisceral corpuscles are of two 
kinds ; the larger are circular or more or less elliptical, the smaller i\raviceZZa-shaped. 

• (i) Fridericia striata (Levinsen). 

Enehytraeus striatus, Levinsen, Vid. Med., 1 883, p. 236. 
P. striata, MiCHAELSEN, Abh. Nat. Ver. Hamb., 1889, p. 43. 

Definition. Length, lo-^ mm.; mimber of segments ^o ; setae, 6-8 jper bundle. Nejahridia with 
large anteseptal part, in which lumen is coiled. Spermathecae without diverticula. Hah. — 
Denmark ; Germany. 

This species has been described by Levinsen and by Ude (1). The former says 
that there are sometimes nine setae in a segment ; this must result from the dropping 
out of a tenth ; I give eight as the highest number in accordance with Ude's 
statements. Sometimes, however, there are only four setae to a bundle ; in this case 
the inner pair are about one-third shorter than, the outer. The colour of the 
species (a greenish-grey) is said by Levinsen to be due to the presence of chlorophyll 


bodies in the skin ; the same author has mentioned a similar presence of chlorophyll 
in Enchytraeus affi,nis ^. The brain is described by Ude as longer than broad, with 
a slight convex hinder margin; the dorsal vessel arises behind the clitellum in the 
sixteenth segment. The funnels of the sperm- ducts are a-3 times as long as broad; 
the prostates are large. The salivary glands are branched at the extremity. 

{2) Fridericia bulbosa (Rosa). 

Neoenchsrtraeus bulbosus, Rosa, Boll. Mus. Zool. Torino, 1887, No. 39. 
P. bulbosa, MiCHAELSEN, Abh. Nat. Ver. Hamb. 1889, p. 42. 

Definition. Length, Stum.; number of segments, 43; setae, 4. per lunclle in anterior segments, 2 
posteriorly. Lumen of anteseptal part of nephridia short, wide, and straight. Spermathecae 
without diverticulum, duct about twice the length of pouch, with glands at base. Hob. — 
Italy; Germany. 

This species has been described by both Rosa and Ude (1) ; their accounts 
differ in some small particulars, which possibly are indicative of geographical 
varieties. Thus Rosa speaks of the salivary glands as sHghtly branched, generally 
only bifurcate, while Ude says that they are simple and not branched. Rosa 
found no glands upon the spermathecal duct, which are stated by Ude to be 
pi'esent. The brain is said by Rosa to be rounded behind, by Ude to be slightly 
concave ; this difference is in all probability to be set down to the different degrees 
of contraction of the cerebral muscles. The funnels of the sperm-ducts are three 
times as long as broad, their ducts are furnished with large spermiducal glands (Ude). 

(3) Fridericia callosa (Eisen). 

Ifeoenehytraeus callosus, Eisen, Ofv. Svensk. Akad., 1878, No. 3, p. 76. 
Enchytraeus callosus, Vbjdovsky, Syst. u. Morph., 1884, p. 41- , 
F. callosa, Michaelsen, Abh. Nat. Ver, Hamb., 1889, p. 42. 

Definition. Length, 20 mm.; number of segments 64 {about); setae, 4 in a bundle. 
Anteseptal part of nephridium short, with straight duct ; duct leading to exterior comes 
of just behind septum. Spermathecae with three indistinct diverticula. Mob. — Nova 

Our knowledge of this species, originally due to Eisen, has been supplemented 
by Michaelsen. The anterior as well as the posterior margin of the brain is 
convex; the length of the brain is about a quarter greater than its breadth. The 

' See description of Anachaeta, where chlorophyll does occur. 


spermathecae ai'e intermediate in their characters between those of the simpler forms 
without any diverticulum and the more differentiated species ; the pouch swells out 
at the base into three compartments, which are not constricted off from the main 
chamber. , 

(4) Pridericia bisetosa (Levinsen). 

Enehytraeus bisetosus, Levinsen, Vid. Med., 1883, p. 233. 

E. (Mesenchytraeus) bisetosus, Vaillant, Annel^s, p. 268. 

E. (Neoenchytraeus, ErsEN) Leydigii, MiCHAELSBN, Ench. Mob., 1886, p. 47. 

E. tenuis, Michaelsen, Arch. Mikr. Anat., 1886, p. 294. 

Neoenchytraeus bisetosus, KosA, Boll. Mus. Zool. Torino, 1887, No. 29. 

P. bisetosa, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 43. 

Definition. Length, 20 mm.; number of segments, 60; setae, paired. Anteseptal region 
of nepkridium nearly equal to postseptal, with undulating duct. Spermathecae with two 
diverticula. Hah. — Denmark; Germany ; Italy. 

It does not appear to be quite certain whetlier Michaelsen's ' Enchytraeiis tenuis' is really 
identical with Levinsen's 'Enehytraeus setosus'; Michaelsen himself, in a paper subsequent to that 
in which the species was described, considered that the question of the identity of the species was not 
settled ; the principal difficulty in the way of uniting them was the fact that in the supposed species 
'tenuis' the anteseptal part of the nephridium was not equal in extent to the postseptal. Ude 
commented upon differences in the form of the diverticula of the spermathecae, and the form of the 
funnel of the efferent duct. In the meantime I retain here the view expressed in Michaelsen's 
paper, viz. the identity of the species. 

The species is characterized by the fact that there are, as a rule, but two setae to 
each bundle ; this is, however, not invariably the case ; four were found by Ude 
in immature individuals, one pair of which were, however, evidently just on the point 
of falling out. The brain is about double as long as broad, with a nearly straight 
or slightly concave hinder margin (convex in 'tenuis'). The dorsal vessel arises in 
the eighteenth segment ; the salivary glands are slightly branched ; the funnel of 
the male efferent duct is double as long as broad ; the ducts themselves are long and 
coiled; the spermiducal glands are large. 

(5) Pridericia leydigii (Vejdovskt). 

Enchsrtraens Leydigii, Vejdovsky, SB. Bohm. Ges., 1877, p. 303. 
E. (Ueoenchytraeus) Leydigii, Vaillant, Anneles, p. 255. 
Neoenchytraeus Leydigii, RosA, Boll. Mus. Zool. Torino, 1887, No. 29. 
P. Leydigii, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 43. 


Definition. Length, 12 mm.j number of segments, 45; setae, 3-4 per bundle. Anteseptal ])art 
of nephridia not large, but with a convoluted duct, and covered with brown pigment. 
Spermathecae with two large diverticula, without glands at base. Hab. — Bohemia; 

The brain of this species is nearly twice as long as broad, with a convex posterior 
margin; the salivary glands are sparely branched. The funnel of the male efferent 
duct is about three times as long as broad; the ducts themselves very long and 
thin; there are spermiducal glands, but Vejdovsky has given no details as to their 
size, &c. This species is not identical with Michaelsen's (4, p. 47) Enchytraeus 
leydigii (for which see above, p. 344). 

(6) Pridericia perrieri (Vejdovsky). 

Enchytraeus Perrieri, Vejdovsky, SB. Bohm. Ges., 1877, p. 30a. 
E. (Neoenehytraeus) Perrieri, Vaillant, Annelds, p. 256. 
Ifeoenchytraeus Perrieri, RosA, Boll. Mus. Zool. Torino, 1887, No. 29. 
?E. (Neoenchytraeus, Eisen) Perrieri, Michaelsen, Ench. Mob., 1886, p. 47. 
Pridericia Perrieri, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 43. 

Definition. Length, 15 mm.; number of segments, 58; setae, 4.-6 per bundle, more numerous in 
ventral bundles. Anteseptal part of nephridium long and tian-ow, but with contorted duct. 
Spermathecae with two diverticula, dilated at their blind extremity. Salivary glands with 
two whorls of branches. Hab. — Denmark ; Germany ; Italy. Terrestrial. 

This is a very marked species, on account of the peculiar arrangement of the 
branches of the salivary glands ; these branches form two sets, which are inserted 
on to the duct of the organ at the same level, in segments iv, v ; another remarkable 
character of the species is in the presence of dilatations of the dorsal vessel in 
segments v, vi, and vii; this suggests, as Michaelsen has pointed out, that the 
dorsal vessel is ante-clitellian in origin, since such swellings seem to be invariably 
found at the commencement of the dorsal vessel. Vbjdovsky's figure of the male 
efferent organs (3 PI. viii. fig. 9) looks as if there were no spermiducal glands; 
he figures a number of strands which suggest retractor muscles attached to the 
integument round the external orifice of the duct. These important differences from 
other species of Fridericia almost seem to necessitate the creation of a new genus. 
The brain is convex behind ; it is nearly twice as long as broad. There are three 
perivisceral trunks uniting the dorsal and ventral vessels : the two anterior of these 
spring from a common trunk, which is very, short. 


(7) Fridericia galba (Hoffmeistee). 

Enehytraeus Galba, Hoffmeistee, Arch. f. Nat., 1843, p. 194 (in part.). 
E. galber,^ d'Udekem, Bull. Ac. Eoy. Belg., xxii. pt. ii. 1855, p. 547. 
E. (Neoenchytraeus, Eisen) galba, Michaelsen, Ench. Mob., 1886, p. 48. 
E. vermicularis, Ratzel, Z. wiss. Zool., 1868, p. 588 (in part.). 
Keoenehytraeus galba, Rosa, Boll. Mus. Zool. Torino, 1887, No. 29. 
Fridericia galba, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 43. 

Definition. Length, ao mm. ; nnmber of segments, 50 / setae, 4-6 per bundle, at posterior end 
all setae of a bundle equally long. Anteseptal part cf nephridia long and broad, with 
contorted duct. Spermathecae with four stalked diverticula. Hob. — Denmark; Germany; 
Belgium ; Bohemia ; Italy. 

This species has been much studied in comparison to many others ; see d'Udekem 
(4), Leydig (2), BacHHOLz, Vejdovsky (3). 

The identity of Ratzbl's E. vermiculans in part with Hoffmeistee's E. galba is stated by Ratzel 
in a later paper (1, p. 28, footnote). Miohablsen queries the identity of Taubee's E. galba with the 
species so called by Hoffmeistee and others ; as, however, Taubee contents himself with a mere 
mention of the occurrence of the species, it is diflBcult to see why this should be called in question 
unless from a general distrust of Taubee's identifications. 

The brain is double as long as broad, with a slightly convex hinder margin. The 
Salivary glands are branched posteriorly. The funnel of the sperm-duct is four times 
as long as broad ; the duct is long and coiled, and there are large spermiducal glands. 
The diverticula of the spermathecae appear to vary in number from four to six^. 
There appear to be no glands at the external ^ore. 

Vejdovsky was of opinion (3, p- 5) that Hoffmeister's 'E. galba' included all 
the larger Enchytraeids. It must be admitted that Hoffmeister's definition is totally 
insufficient to characterize the species as we now know it. 

(8) Fridericia lobifera (Vejdovsky). 

Enchytraeus lobifer, Vejdovsky, Ench., 1879, p. 57. 

E. (Archienclisrtraeus) lobifer, Vaillant, Annelds, p. 379. 

Fridericia lobifera, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 44. 

Definition. Length, 30 mm. ; number of segments, 60 ; setae in anterior region of body 4, in 

1 Of course a misprint, with which, it may be remarked, that particular paper teems. 
'' 4-8 according to Levinsen. d'Udekem also (4, PI. iii. fig. 8) figures eight. 


ventral 6, m dorsal bundles, in posterior region 3. Aanteseptal part of nephridium short, 
will brown pigment. SpermatJiecae with numerous diverticula, which are solid. Hab. — 

This species is to be easily known by the fact that the diverticula of the 
spermathecae, or, to speak more accurately, the appendages of that organ, which 
appear to correspond to the diverticula of other species, are solid, and probably 
(Michaelsen) function as glands. The brain is concave behind. The salivary glands 
are provided with numerous branches. The duct of the nephridium arises just 
behind the septum ; Vejdovsky does not state whether the anteseptal part of that 
organ has a straight or contorted lumen. 

(9) Pridericia ratzeli (Eisen). 

Enehytraeus Batzeli, Eisen, Ofv. Svensk. Akad., 1873, No. i, p. 123. 

Neoenchjrtraeus Ratzelii, Eisen, Ofv. Svensk. Akad., 1878, No. 3, p. 77. 

E. Ratzelii, Levinsen, Vid. Med., 1883, p. 237. 

E. (Neoenchytraeus) Ratzelii, Vaillant, Annelds, p. 262. 

Friderieia Ratzelii, MiCHAELSEN, Abh. Nat. Ver. Hamb., 1889, p. 44, 

Definition. Length, ^omm.; number of segments, 60; setae, 6-8 anteriorly, ^posteriorly. 
Anteseptal part of nephridia narrow, but with contorted duct ; its duct arises just behind 
septum. Spermathecae with 6-13 diverticula, with a few small glands at base of duct. 
Hab. — Norway; Germany; Italy. 

The brain has a convex hinder margin. The salivary glands are branched. Ude 
states that at the external pore of the spermathecae are two large stalked glands 
instead of the numerous small glands figured (PI. xii. fig. xxii c) by Eisen. 
Levinsen, on the other hand, says : ' Kecept. seminis ved Udmundigen uden 
stilket Kjertelmasse.' For structure see, in addition to above references Ude (1), 
Eisen (13), and Hesse, who describes ' copulatory glands ' like those of F. antarctica 
in the thirteenth segment. 

(10) Pridericia dura (Eisen). 

Ifeoenchytraeus durus, Eisen, Ofv. Svensk. Akad., 1878, No. 3, p. 77. 
Enehytraeus durus, Vejdovsky, Syst. u. Morph., 1884, p. 41. 
E. (Ifeoenchytraeus) durus, Vaillant, Annelds, p. 363. 
Pridericia dura, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 44. 

Definition. Length, 16 mm.; number of segments, 45 {about) ; setae, 6 per bundle. Anteseptal 

Yy 2 


part of nep/tridium long and stout with coiled duct, duct of nephridium arising just hehind 
septum. Spermathecae with many diverticula, at base of duct one large gland {or many 
snudl glands). Hai. — Norway. 

This species is very nearly akin to the last; it is, indeed, a matter of no little 
difficulty to distinguish them by the descriptions given. Michaelsbn distinguishes 
them by the ' absence ' of any glands at the external pore of the spermatheca in 
F. ratzdi, these structures being present in F. dura. Eisen, however, figures 
a few small glands in this position in F. ratzeli (PI. xii. fig. 22 c). This point 
is still further confused by Ude's statement that F. ratzeli has two stalked glands 
instead of the many figured by Eisen. Another point of difference urged by 
MiCHAELSEN is that the brain of F. dura is triangular, while that of F. ratzeli 
is oval, and longer than broad. The number of setae also is greater in F. ratzeli. 
If the supposed difference in the brains has no existence, it will, I think, be necessary 
to unite the two species. 

This species, like F. ratzeli, F. callosa, and E. hegemon, has a double layer of 
longitudinal muscles. The salivary glands are stated by Michaelsen to be like 
those of F. callosa. 

(11) Pridericia hegemon (Vejdovsky). 

Enchytraeus hegemon, Vejdovskt, SB. Bohm. Ges., 1877, p. 303. 
E. (Neoenchytraeus) hegemon, Vaillant, Annel^s, p. 357. 
Pridericia hegemon, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 44. 

Definition. Length, opmm.; setae, 4 per bundle. Anteseptal part of nephridium short, with 
straight duct. SpermMhecae very numerous with stalked diverticula, with a pair of large 
glands at external pore. Hob. — Bohemia ; Germany. Terrestrial. 

This species is very distinct from any other; its chief characteristic is the very 
numerous stalked diverticula. The species has been studied by Vejdovsky (3) 
and Michaelsen (4). The brain is oval in contour, with a marked projection 
anteriorly, and is convex posteriorly. The salivary glands are much branched — 
more so than in any other Fridericia. The dorsal blood-vessel springs from the 
peri-intestinal sinus in the eighteenth segment; in this segment, and in the 
seventeenth, sixteenth, fifteenth, fourteenth, are heart-like swellings which are also 
developed, though to a less extent, in the two segments lying in front of these 
again. There is an organ of taste in the buccal cavity similar to that of other 
species of Enchytraeids. 


(la) Fridericia antarctica, Beddaed. 

Priderieia antarctica, Bebdaeb, Proc. Roy. Phys. Soc, 1892-93, p. 41. 

Definition. Length, \% mm.; number of segments, 6^ ; setae, /\.-6 per lundle. Anteseptal part 
of nephridium as large as postseptal, with convoluted duct. Spermatheca with numerous 
diverticula. On segments XIII-XV papillae connected with peculiar cells surrounding 
nerve-chord. Hab. — New Zealand. 

This species, though agreeing in most of its characte;r8 with F. galha, is apparently 
to be distinguished by masses of cells surrounding the nerve-chord in certain segments 
which appear to be comparable to those of F. ratzeli and Pachydrilus (see p. 32^). 
The cells in question are large, pear-shaped, and granular ; they have, indeed, 
a glandular, rather than a nervous, look ; but, being enclosed in a common sheath 
with the nerve-chord, favours their nervous nature. 

The clitellum occupies segments xi-xiii, and is developed all round the body. The 
setae of xii are, as usual, absent. The salivary glands are branched. The septal 
glands occupy iv-vii, and the intestine commences in xiv. The sperm-duct funnel 
is three times as long as broad. 

Genus Henlea, Michaelsen. 

Syn. Enohytraeus, d'Udekem et alii (in part.). 
Archienchytraeus, ElSEN (in part.). 

DEPlWiTiOir. Setae diflferent in form and arrangement, according to the species. 
Head-pore between prostomium and buccal segment ; no dorsal pores. Oesophagus 
sharply marked off from the intestine. Dorsal vessel arises in front of the 
clitellum; blood colourless. Duct of nephridium springs from the postseptal 
portion near to the septum. 

This genus has been recently instituted by Michaelsen in his Synopsis of the 
family (5). It forms undoubtedly a somewhat heterogeneous assemblage of 
species, which are chiefly bound together by the fact that the oesophagus is 
distinctly separated from the intestine by a constriction, and by the ante-clitellian 
origin of the dorsal vessel ; as a rule, there are oesophageal glands at the commencement 
of the intestine; all these points ally the genus Henlea to the genus Buchholzia. 
There appear to be no other characters distinctive of the genus ; some have and 
some have not salivary glands ; H. puteana has two pairs of spermathecae, which 


is unique in the whole family; this species also has straight setae like those of the 
genus Enchytraeus, while in H. dicksonii the setae are ./-shaped, as in the genus 
Pachydrilus, &c. The following species, at least, appear to be well characterized : — 

(i) Henlea ventrieulosa. 

(a) Henlea leptodera. 

(3) Henlea puteana. 
Besides these three, Michaelsen has assigned three others to the genus: viz. 
H. socialis, H. iiasuta, and ff. dicksonii. The latter has heen investigated by Eisen 
and by MiCHAELSEN ; more recently, Ude has added something to our knowledge 
of this species, which may be now removed from the doubtful position it held at 
the time that Michaelsen wrote. 

'Enchytraeus socialis' of Leidy (6) is queried as a member of the genus Enchytraeus by Vaillant 
(6, p. 291) ; but it must, of course, be remembered that Enchytraeus, with that naturalist, is used in 
a very wide sense. 

The only salient point in the description of the species is that it possesses a dilatation upon the 
oesophagus in the eighth segment, a fact which led Michaelsen to place it in the present genus. 
No more than this can be said. 

The well-known species can be distinguished by the following table : — 

I. Two pairs of spermathecae H. puteana. 

11. One pair of spermathecae. 

(i) No oesophageal glands H. dicksonii. 

(3) Two oesophageal glands H. leptodera. 

(3) Four oesophageal glands ... . . H. ventrieulosa. 


(i) Henlea ventrieulosa (d'Udekem). 
Enchytraeus ventriculosus, d'Udekem, Bull. Ac. Roy. Belg., 1855, xxii. pt. ii. 

P- 547- 
E. latus, Letdig, Vom Bau des thierischen Korpers, 1864 (Fide Michaelsen). 
E. (Arehienchytraeus, Eisen) ventriculosus, MichAELSEN, Ench. Mob., 1886, p. 46. 
E. (Archiencliytraeus) albidus, Vaillant, Annelds, p. 281 (in part.). 
Arehienchytraeus ventriculosus, KosA, Boll. Mus. Zool. Torino, 1887, No. 39. 
Henlea ventrieulosa, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 31. 
Definition. Lengih, 15-30 mm.; number of segments, 60; setae straight or slightly bent to 
right or left, 4-9. Four oesophageal glands. Eab. — Germany; JDenmarh ; Bohemia; 
Italy ; Belgium; New Zealand; Asia. Terrestrial. 

The first describer of this species, d'Udekem, omitted a large number of details in both his 
description and his figures; some of these omissions, however, are probably due to the real absence 


of the structures in question ; thus he makes no mention of the salivary glands, which Vejdovsky 
said were inconspicuous ; the rudimentary character of these glands distinguishes the species from 
R. leptodera and H. dicksonii. From d'Udekem's account and figures the fourfold division of the 
oesophageal gland is not apparent ; the whole structure, indeed, is described as a kind of muscular 
stomach; Vejdovsky, while naturally improving upon the account given by d'Udekem, did not 
recognize that the supposed stomach was really formed out of four diverticula of the oesophagus, 
a fact which was later demonstrated by Michaelsen (1). These four glands lie in the eighth segment. 
They are simple diverticula of the gut, with a lining of ciliated cells ; the walls are a little folded, but 
there is in the figures of Michaelsen, at any rate, no trace of the intra-cellular lumina described by 
the same author in Buehholzia. 

The dorsal vessel springs from the perioesophageal plexus in the ninth segment; 
it has three heart-like swellings in segments vii, viii, ix ; the brain is longer than 
broadj and broader behind than in front; it is concave anteriorly, and is markedly- 
notched behind ; the funnel of the sperm-duct is rather small ; the sperm-duct is 
long and coiled. The spermathecae consist of a pear-shaped distal part and of a duct 
of about the same length, which is somewhat dilated at its opening on to the exterior. 
d'Udekem has remarked that these organs present the appearance of glands attached to 
the alimentary canal, thus foreshadowing the discovery of Michaelsen that they open 
into it. The nephridia are said by d'Udekem to commence in the fourth segment. 

(2) Henlea leptodera (Vejdovskt). 

Arcliienehsrtraeus nasutus, Eisen, Ofv. Svensk. Akad., J 878, No. 3, p. 72. 

Enchytraeus leptodera, Vejdovsky, Ench. 1879, p. ^^. 

E. (Archienehytraeus, Eisen) leptodera, MiCHAELSEN, Ench. Mob., 1886, p. 46 et 

Archienehytraeus leptodera, Rosa, Boll. Mus. Zool. Torino, 1 887, No. 29. 
Henlea leptodera, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 32. 
H. nasuta, Michaelsen, ibid., p. 32. 

Definition. Length, 20 mm.; number of segments, 60; setae straight, ^to J. Two oesophageal 
glands, with simple lumen. Hah. — Europe; Sibevia. Terrestrial. 

Michaelsen considers that the present species is the same as EisEn's Archienehytraeus affinis; 
this opinion is based upon an actual examination of both species by Michaelsen. Eisen did 
not mention the gut-diverticula, which are nevertheless, according to Michaelsen, present, and 
of exactly the same form as in H. leptodera. On the other hand, it jnay be pointed out that 
the brain appears from the figures to be a little different in form in the two 'species.' In 
H. leptodera the brain is more sharply notched behind than in the form called by ElSEN Archi- 
enchytramis affinis (cf. Vejdovsky, 3, PI. x. fig. i, and Eisen, 13, PI. vi. fig. 2 c). As has already 
been mentioned, however, the brain is susceptible of alterations in form, according to the state 
of contraction of the muscles attached to it. Eisen has remarked that the species 'affinis' is very 


near to 'nasuta'; upon this matter Michaelsen writes : 'Eisen's Angabe von der Verwandtschaft des 
A. affinis mit seinem A. nasutus ware zu bestatigen, falls die oben angefuhrte Synonymie angenommen 
werden musste.' 

Ude unites the two species on the following grounds : Miohaelsen admitted that the two species 
stood very near to each other ; but he distinguished them by the fact that in H. nasuta the appendices 
of the oesophagus were much more folded internally than in H. leptodera, and that in H. nasuta the 
setae, instead of being about the same length, are of different sizes. Ude found in the same individual 
of H. leptodera some bundles in which the setae were of the same length, and others in which the setae 
were unequal, like the typical H. nasuta ; in one preparation it was found that the left oesophageal gland 
had a much less folded lumen than the right gland, thus doing away with the second difference between 
the supposed two species. 

This species differs from the last in having well-developed salivary glands. The 
oesophageal glands are in the seventh segment, and open into the gut between 
this and the following segment. The brain is stated by Ude to be broader than 
long, but by Vejdovsky to be as long as broad — a difference, no doubt, to be accounted 
for on the supposition that the brain varies in its proportions according to the 
state of contraction of the muscles which are attached to it. There seems to be 
a little confusion as to the precise segment in which the dorsal vessel arises, and 
in which segments the dilatations are; Vejdovsky (3, p. ^z) says that the dorsal 
vessel can be followed back as far as the seventh segment, and that the hearts are 
in segments vii, viii — two statements which do not appear to be reconcileable ; in the 
systematic part of the same work (on p. 56) he places the heart-like swellings in 
segments vi, vii; the latter statement is followed by Vaillant (6, p. 266). Ude, 
on the other hand, notes the commencement of the dorsal vessel in the same segment 
as in the last species, i. e. the ninth. The spermatheca is much as in the last 
species, but the duct is rather shorter than thfe pouch, and the latter is not dilated 
before its external pore. 

(3) Henlea dicksonii (Eisen). 

Arehienchytraeus Dicksonii, Eisen, Ofv. Svensk. Akad., 1878, No. 3, p. 70. 
Enehytraeus Dicksonii, Vejdovsky, Syst. u. Morph., 1884, p. 41. 
Henlea Dicksonii, Michaelsen, Abh. Nat. Ver. Hamb., 1889, p. 33. 

Definition. Length, i^mm.j number of aegments, 52; ietae straight or a little bent to 
one side, of different sizes, 6-8 per bundle. No oesophageal glands. Hah. — Nova Zemlla ; 

Eisen's account of this species is not sufficient to place it accurately; it was subsequently 
investigated by Michaelsen (5), who, however, had only a specimen minus the anterior end ; hence 
he was unable to do more than state that the dorsal vessel, not being there visible, must originate 
in front of the clitellum. A complete description of all the important facts in the structure of this 



species was afterwards given by Ude (1). Ude states that the description of the setae given by Eisen 
was not correct ; Eisen figured (13, PL iv. fig. 7 c) the setae of the worm as like those of Pachydrilus. 
MiCHAELSEN, in commenting upon this diflFerence from other species of the genus, remarks that the 
arrangement is not like what is met with in the genus Pachydrilus. It is, however, more satisfactory 
not to have to explain away a distinct resemblance in this way. 

The brain is double as long as broad (1-1-3 according to Eisen); it has a slight 
concavity posteriorly. The oesophageal glands are completely absent; in transverse 
sections the end of the oesophagus is seen to project into the interior of the intestine, 
thus forming a kind of valve. The dorsal vessel arises in segment viii; there are 
dilatations of this vessel in segments vi, vii, viii; the salivary glands are present, 
but their form is not described. The spermathecae are furnished with glands at 
their external pore; they are dilated into a pouch at the other end. 

(4) Henlea puteana (Vejdovsky). 

Enchj1;raeus puteanus, Vejdovsky, SB. Bbhm. Ges., 1877, p. 301. 
Henlea puteana, MiCHAELSEN, Abh. Nat, Ver. Hamb., 1889, p. 34. 

Definition. Length, 15 mm.; number of segments, 2,0; setae of ventral bundles, 8-10, 
of dorsal, 5-7. Two pairs of spermathecae. Hob. — Bohemia. 

This species is at once to be distinguished by the presence of two pairs of 
spermathecae situated in segments iv, v. This is unique, not only in the genus, but 
in the family. The dorsal vessel arises in segment ix ; this and the two segments 
in front have each a dilatation of the dorsal vessel. Nothing appears to be known 
as to the presence of oesophageal glands, or as to whether there is that marked 
distinction between oesophagus and intestine which characterizes other species of the 
genus. Nor are salivary glands mentioned. Considering the inequality of the setae 
in the dorsal and ventral bundles, it may perhaps be doubted whether the species is 
really a member of the genus Henlea; I leave i^ there in deference to the opinion 


Genus Anachaeta, Vejdovsky. 

Syn. Achaeta, Vejdovsky. 

DBPiwiTiOBr. Setae absent ; only represented by large gland-cells. Head pore at the 
tip of the prostom.iuni ; no dorsal pores. Dorsal vessel arises in front of the 
elitellum; blood colourless. An unpaired salivary gland opens behind the 
pharynx. Spermathecae not opening into the gut. 

z z 


Our knowledge of this genus is almost entirely due to Vejdovskt; his account 
has been confirmed and extended in a few directions by Michaelsen (5) and Ude 
(1). The most salient character of the genus is the absence of setae. These appear 
to be represented by a series of large clear cells which occupy the precise position 
that the setae should, were they present; these cells are of large size— according to 
Vejdovsky's figure of A. bohemica (24, PI- vii. fig. i) one-third of the diameter of the 
body; they are pear-shaped in form, and have a single nucleus at the base. 
Michaelsen states that there are more than one nucleus to each of these cells, 
which, of course, heightens their resemblance to seta-sacs, these structures being 
usually formed of a group of cells. A. eisenii has two pairs of these bodies, while 
only one pair, the ventral, is to be found in the other species of the genus. The 
integument of AvMchaeta is further remarkable for the existence of 'chlorophyll- 
glands.' There is in A. eisenii one of these to each segment, lying in the mid-dorsal 
line on a level with the seta-sacs; the chlorophyll-gland consists of a group of 
gland-cells, which open on to the exterior by a common duct. The chlorophyll in 
the interior of these cells was proved to be such by Vejdovskt. In A. bohemica 
there is a row of these glands, surrounding the middle of the segments, and it would 
appear from Vejdovsky's figure, illustrating these glands, that they are unicellular 
spherical bodies. Vejdovskt has stated that he never met with similar structures in 
other Enchytraeidae ; it is possible, however, that Fridericia striata may have similar 
chlorophyll-glands (see above). 

Michaelsen states that the longitudinal layer of muscles consists of a single layer 
of hollow fibres, without the additional layer that exists in so many species, such 
as F. ratzelii. The clitellum (in A. bohemica^aX any rate) occupies only the lateral 
and ventral regions of the body-wall (cf. Aeolosoma). The dorsal vessel arises from 
the intestinal sinus in front of the clitellum ; it has three dilated heart-like swellings, the 
position of which varies according to the species. The salivary gland is unpaired, 
and opens behind the pharynx ; there is a little confusion about the unpaired character 
of these glands ; Vejdovskt says in the description of the species A. eisenii (3, p. 6o) : 
' Die Speicheldi-iisen stellen kurze,' &c., and again, in the definition of the genus 
(p. 50) : ' Die Segmentalorgane modificiren sich im ^—^. Segmente zur Speicheldriise.' 
On a previous page also the same facts are substantially stated. In a later work, 
however (24, p. 106), Vejdovskt remarks upon the unpaired condition of these glands 
in the genus Anachaeta (' bei Anachaeta dagegen unpaarig '), though the observations 
recorded in that work concern, so far as this genus is treated of, the species 
A. bohemica. Ude (1, p. 98) says that the salivary gland is unpaired in A. eisenii. 

The spermathecae in this genus do not communicate with the lumen of the gut. 



as is so very generally the case with the Enchytraeidae. Both species have taste-papillae, 
and in both the brain is convex posteriorly. 

(i) Anachaeta eisenii, Vejdovsky. 
Achaeta Eisenii, Vejdovsky, SB. Bohm. Ges., 1877, p. 300. 
Anachaeta Eisenii, Vejdovsky, Ench., 1879, p. 60. 
Definition. Length, 13 mm.; number of segments, 30; four seta-glands in each segment. 

Spermathecae oval, with a duet not sharply marked off from pouch. Hah. Bohemia ; 


The structure of this species is described and illustrated in Vejdovsky's Monograph 
of the Enchytraeidae. A good many of the characters of the species have been already 
dealt with in the account of the genus. A few remain, which are either of less 
importance or differ in the two species, to be noted here. The dorsal vessel arises, 
according to Vejdovsky (3, pp. '^'3, and 61), in the sixth segment, and has two 
dilated hearts, one in this segment and one in the preceding segment ; Ude, however, 
who has more recently examined the worm, says that this vessel arises in segment viii, 
and that there are three swellings, which, therefore, presumably lie in segments vi, vii, 
viii. The nephridia have a long anteseptal region, which is constricted off from the 
postseptal half of the organ by the septum; the lumen is very undulating in its course 
in the anteseptal part of the nephridium. In a longitudinal section through the 
body-waU (3, PI. i. fig- 2 gl) Vejdovsky represents a series of epidermic glands of large 
size which lie at the junction of every two segments, occupying the position, in fact, 
that is occupied in the genus Fridericia by the dorsal pores, elsewhere wanting. 
Vejdovsky says nothing of these glands in his Monograph (24) ; it is, therefore, 
to be supposed that there is no special peculiarity in the arrangement, but that they 
are simply some of the numerous glands which occur in the integument generally. 
The funnel of the sperm-duct is five or six times as long as broad; the duct 
is spirally twisted, forming a very compact coil ; the external pore is beset with 
prostates. The spermatheca is an oval pouch, passing gradually into the duct ; the 
external pore is beset with glands. 

(2) Anachaeta bohemica, Vejdovsky. 

Anachaeta bohemica, Vejdovsky, Zool. Anz., 1879, p. 183. 

Definition. Only two rows of seta-glands, corresponding to the dorsal series of setae. 
Spermatheca consisting of a small oval pouch, into which opens laterally a very long 
diverticulum, expanded into an oval pouch at extremity. Hab. — Bohemia. 

z z 2 


This species has been studied by Vejdovsky (24) and, later, by Miohaelsen (5). 
The shape of the spermatheca is not referred to, unless there is an error in the 
lettering of the Plate (24, PI. vii. fig. 0,2) ; I am inclined to think that this must be 
the case. For both Vejdovsky and Ude agree in their descriptions of the sperma- 
thecae of A. eisenii, which differs from the figure given of the spermathecae of that 
species in the great work of Vejdovsky. In that figure the spermatheca is represented 
as consisting of a short oval pouch opening on the exterior ; into the side of this, and 
at right angles to it, opens a long and narrow duct which is only a little dilated just 
where it opens into the pouch; at the end of this long duct is an oval swelling 
which lodges the sperm ; it seems to me that the long duct, widening out at its blind 
extremity may, perhaps, be compared to the diverticulum of the spermathecae of other 
Enchytraeidae. If this figure really refers to A. eisenii, then the differences 
between the two species will be the absence of the ventral pair of seta-glands, 
and the presence of three pairs of hearts in A. bohemiea in segments iv-vi. From the 
last heart, and from the dorsal vessel in front of it, arise three pairs of periviscerals ; 
the anteseptal part of the nephridium is shorter than in the last species, and the external 
orifice is beset with glands similar to those found round the aperture of the 
spermathecae of both species ; the nephridia are furnished with a distinct terminal 
vesicle, which appears to be wanting in the other species of the genus. 

Genus Bryodeilus, Ude. 

Definition. Setae y-shaped. Head-pore between prostomium and buccal segment. 
Four oesophageal glands in VII. Blood colourless. Dorsal vessel arises in XII, 
without cardiac body. 

This genus is, of course, easily to be distinguished on account of the four 
oesophageal glands, which are arranged dorsally, ventrally, and laterally ; the dorsal 
vessel, by an exception unique in the family, does not arise from these pouches, but 
behind them. There is only one species — 

Bryodrilus eMersi, Ude. 

B. Ehlersi, Ude, Zool. Anz., 189a, p. 344. 

Definition. Length, 13 mm.; setae 4-5 per bundle. Three pairs of septal glands. Brain 
two to three times longer than broad, concave anteriorly. Spermathecae, without diverticula, 
communicating with gut. Hab. — Germany. Terrestrial. 



At present we only know this species from a preliminary description of Ude. The 
funnel of the sperm-duct is described as two to three times as long as broad. The 
perivisceral corpuscles are large and disc-siaped. The nephridia have a small ante- 
septal part; the duct arises from the postseptal part, just behind the septum. 

Genus Pabenchytraeus, Hesse. 

DBPmiTiON-. Setae straight. Head-pore between prostomium and buccal segment. 
Blood colourless ; dorsal vessel arises in XII ; ventral vessel undivided after first 
segment. Wo oesophageal glands. Copulatory gland present. 

The genus is to be distinguished by the limited extent of the divided part of the 
ventral vessel. Only one pair (of the three pairs) of perioesophageal trunks join the 
divided ventral vessel. 

Parenchytraeus litteratus, Hesse. 

p. litteratus, Hesse, Z. wiss. Zool., 1893, p. 2. 

Definition. Length, n mm. ; number of segments, 45; prostomium and buccal segment with 
papillae. Three pairs of septal glands {in IV-VI). Brain ij times as long as broad, 
concave in front. SpermMtheca, without diverticula, communicating with gut by wide orifice. 
Hob. — Naples, seashore. 

The specific name is taken from the sculpturing (Kke Arabic letters) of the cuticle. 
The copulatory glands are in iv, v, xiii, xiv. 


I agree with EosA in the necessity for uniting together the families Perichaetidae, 
Acanthodrilidae, and Cryptodrilidae ; but I do not go so far with him as to include 
also the Eudrilidae. I consider that this family differs as much from either of the 
three enumerated as do any two families of earthworms. The three which I here unite 
to form the Megascolicidae are very difficult to separate from each other; the line 
which separates Dichogaster from the Acanthodrilidae is a very slight one ; Microdrilus, 
too, is not far removed from Benhamia. The detailed resemblances between these 


various types will be found treated of in the introduction to the description of the 
genera and species of the subfamily Cryptodrilidae. On the other hand, the Peri- 
chaetidae are obviously nearly connected, through the remarkable forms with eight 
setae in anterior segments, with Cryptodrilus and its immediate allies. 

Certain peculiarities of structure are found only in the family Megascolicidae, and 
some of these occur in all of the three subfamilies into which it is divided. The principal 
of these is the peculiar form of the spermiducal gland, which I term ' lobate ' ; this kind of 
spermiducal gland is found in all the Perichaetidae with a very few exceptions, and in 
a great many of the Cryptodi-ilidae ; at first sight it may appear not to exist in any 
Acanthodrilid ; but the genus Diplocardia has spermiducal glands, which are, at 
any rate, a step in this direction. In no other family of the Oligochaeta do glands 
of this particular description exist. In this family only are there nephridia which 
ramify within the coelom and open on to the exterior by numerous pores ; the 
integumental network of certain Eudrilids, which is derived from a series of paired 
nephridia being a different thing. 

Among the higher Oligochaeta the Megascolicidae are the only genera which have 
spermathecae furnished with diverticula of a different structure to the pouch ; and, 
furthei-more, there are only a few exceptions to the universality of these spermathecal 
appendages in the group. 

The very general position of the male pore upon the eighteenth segment of the 
body is a characteristic of the Megascolicidae; in no genus are the male pores far 
from this segment, if they are not upon it. 

These are the principal facts which lead me to associate together the three families 
Perichaetidae, Cryptodrilidae, and Acanthodrili^ae, within what I may term a super- 
family, Megascolicidae. 

It is noteworthy too that certain peculiarities of structure which occur in the 
group are confined to it, and are found in more than one of its three families. 

The continuous circle of setae, so universal in the Perichaetidae, occurs nowhere 
else among the Oligochaeta except in the Acanthodrilidae. Even in so rare a point 
of structure as the enclosure of the dorsal blood-vessel in a coelomic space we fi.nd 
that the only two examples, Beinodrilus henhami and Megascolides austraZis, are 
in this group. The peculiar intestinal glands of Megascolex coeruleus are repeated in 
Typhosus. The peculiar condition of the gizzard in Perichaeta, occupying, as it 
does, two segments, with a suppression of the intervening mesentery, is repeated in 
the genus Typhaeus. 

As to negative characters, these are numerous. The most striking, perhaps, though 
by it the family does not contrast with the Eudrilidae, is the absence of specially 



modified cUtellar setae ; to this rule there is but one exception, viz. Perichaeta houUetij 
and in this worm the modified genital setae are not at all like those of the Lumbricidae 
or the Geoscolicidae. 

The mutual relationships of the three subfamilies are more fuUy discussed in the 
following pages. 


DEPlBTiTiOBr. Earthworms with numerous setae per segment arranged in a con- 
tinuous circle, or with dorsal and ventral gaps; male pore nearly always on 
XVIII. Gizzard always present ; caleiferous glands absent or present ; intestine 
frequently with two (occasionally more) caeca, and a rudimentary typlilosole. 
Nephridia diffused or paired. Spermiducal glands generally lobate ; penial setae 
present or absent; spermathecae with one or two diverticula ^ 

This family of earthworms is one of the largest; it comprises over loo species, 
which will be in this work relegated to four genera. This family, like many others, 
contains some of the largest as well as some of the smallest earthworms. There are 
no species known which are of greater size than Perichaeta musica, while other 
species do not reach to a greater length than two inches. 

Anatomical Characters^. 

The continuous circle of setae does not absolutely distinguish the family as it was 
formerly thought to do. I am quite of Benham's opinion that his genus Plagiochaeta 
is an Acanthodrilid. The setae are in the species of the genus Megascolex interrupted 
dorsally and also ventrally by a gap where setae are absent. In the remarkable 
species which I formerly placed in the genus Anisochaeta there is an interesting 
transition to the worms with eight setae to the segment ; that is what is found in 
the anterior segments of those species, while posteriorly the numerous setae of the 
typical species exist. As a general rule the setae are less numerous in the most 
anterior segments — ^the reason for which may be no more recondite than that these 
segments are of less circumference ; the number of setae to the segment varies and 
often aflfords a valuable series of characters for the systematist; the largest number 
of setae in a single segment is shown in Pleionogaster horsti, where there may be 
as many as 151. 

' Very exceptionally absent — P. ijima. 

' For the affinities of the Pericliaetidae see under Cryptodrilidae and Acanthodrilidae. 


The setae are as a rule similar on all the segments of the body ; there are, however, 
exceptions to this rule. In many species, for example, the setae upon some of the 
anterior segments are stouter than those which occur elsewhere ; in P. sinensis 
this is the case with the setae of the segments vi, vii, viii, ix. In other species the 
ventral setae are larger than the rest ; in P. houUeti there are three or four setae 
on each side of the nerve-cord, which are distinctly longer as well as thicker than 
those upon the remaining part of the segment ; in this case there is a gradual increase 
in size, commencing a little way from the median ventral line and culminating 
at the last seta on each side. In the genus Perichaeta there are either very few 
or no setae upon the clitellum when this is mature. P. taprohanae is one of the 
few species of the genus which has setae forming continuous circles upon the 
cliteUar segments ; it is common, if not universal, among the species of Megascolex 
to have setae upon the clitellum. When setae are present upon the clitellum they 
are never, except in P houUeti, in any way different from those of other segments. 
In the above-mentioned species the cliteUar setae are shorter than those upon other 
segments, and have a truncated form and a bifid extremity. The clitellum is in the 
majority of species (in all belonging to the genus Perichaeta itself) complete, i. e. 
it completely encircles the body ; in this genus too, with very few exceptions, there 
are only three segments in the clitellum — only two apparently in P. upoluensis. 
Megascolex has the most extensive clitellum (nine segments in one species), and here 
the clitellum is only complete anteriorly ; it is deficient ventrally behind. 

The male pore is nearly invariably upon the eighteenth segment ; the only 
exceptions are M. attenuata, where it is upon the seventeenth, and M. cTiormis, where 
it is between the seventeenth and eighteenth. , 

The prostomiuTn in the Perichaetidae is rarely continued by grooves as far as the 
end of the buccal segment ; it does, however, as a rule, extend for a certain distance 
on to the buccal segment. 

Genital papillae are very general in this family ; they are found in nearly all the 
species. When present they may occur in the neighbourhood of the spermathecae 
as well as near to the male pores. With the papillae are always connected glands 
of a peculiar kind whose structure has been described above (p. 144) ; the papillae 
in fact are but the external pores of these gltads. 

The alimentary canal of the Perichaetidae is chiefly remarkable for the two caeca 
of the intestine, which arise in the twenty-sixth segment, and are directed forwards ; 
these were first discovered by Vaillant (3), and have since been found to be, with 
a few exceptions, characteristic of the genus Perichaeta itself, and they serve as a means 
of distinguishing that genus from Megascolex. It is a remarkable fact that similar 


caeca occur in the not nearly allied genus Urobenus (a Geoscolicid) ; with thia 
exception they are limited to the present family; in a few species, all natives of 
Japan, the two caeca are replaced by six or seven, as was first pointed out by 
HoBST (9) ; but in these cases there is always one pair which are very much larger 
than the others. A gizzard is, so far as we at present know, invariably found in 
this family; its position varies. In the genus Megascolex and in Biporochaeta it is 
placed far forwards — ^in the fifth, sixth, or in the seventh and eighth segments ; 
in the geiius Perichaeta the gizzard appears with rare exceptions to occupy two 
segments — the eighth and the ninth. It is often stated to occupy three segments ; 
and from one point of view it often does; for the septum which bounds the 
ninth and tenth segments is pushed so far backwards by the growth of the 
gizzard that the setae belonging to this (the tenth) segment lie to the forward side 
of the septum that intervenes between the ninth and tenth segments. In this 
genus (Perichaeta) it is nearly always the case that the septum dividing segments 
viii/ix is absent. 

It is a doubtful point whether true calciferous glands occur in the majority of 
Perichaetidae ; in the genus Perichaeta they certainly do not ; they are represented 
simply by an increased vascularity of the oesophagus in certain segments ; in Biporo- 
chaeta, however, there are two pairs of calciferous glands in x,' xi ; these are, however, 
not such distinct caeca of the gut as the same structures are in Pontoscolex ; they 
present the appearance of local thickenings of the walls of the oesophagus ^ ; the 
same state of affairs appears to characterise the genus Megascolex according to 
Fletcher. In Biporochaeta crystals occur in the calciferous glands similar to those 
which are found in the calciferous glands of other earthworms. Unique in the 
family is M. coeruleus for the series of intestinal glands, which, however, are again 
met with in Typhaeus. These are placed upon the dorsal surface of the intestine in 
segments xxi-xxxix (about). Their lumen is lined by a much-'folded cellular membrane ; 
their structure in fact is not unlike that of the calciferous glands. 

A typhlosole has been stated to be wanting in the Perichaetidae ; it is as a 
rule present, but never appears to attain to such dimensions as it does, for example, 
in the Acanthodrilidae. 

With regard to the vascular systeTn, a subnervian vessel, sometimes stated to be 
absent, is present at any rate in some forms. The details of the circulatory system, 
at present only known in M. coeruleus, will be gone into below. The last pair of 
hearts is usually ia segment xii or xiii. 

' Besidesj this genus is possibly not referable to this family, but rather to the Acanthodrilidae 
(see below). 




Genera of Perichaetidae. 

KiNBERG (pp. 101-103) distinguished no less than six genera of those species 
characterised by possessing more than eight setae to each segment; it is possible, 
however, that some of them might be AcanthodriUdae as defined in the present work. 

These ' genera ' are thus defined by Kinbeeo : — 

Amyntas, n. Lobus cephalicus, e parte superiore anteriore segment! buccaHs 
formatus, marginibus lateralibus sohs distinctis, segmento illo 
multo angustior et brevior ; segmenta anteriora posterioribus 
duplo longiora ; setae radiatim et seriatim collocatae, minutae, 
laeves, 50-60: nae posteriores magis numerosae. 

NiTOCKis, n. Lobus cephalicus transversus, latus, obtusus, superus, postice arcua- 
tus, integer; segmenta anteriora et posteriora reliquis longiora, 
medio carinata ; setae series transversas formantes, parvae, 1 8-53 : 
nae posteriores magis numerosae ; cingulum nullum. 

Pheketima, n. Lobus cephalicus terminalis, transversus, ad marginem anteriorem 
superiorem segment! buccalis affixus ; setae radiatim et seriatim 
positae, segmentorum posteriorum illis anteriorum magis numero- 
sae ; foramina dorsalia ; cingulum ; tubercula ventralia duo. 

Bhodopis, n. Lobus cephalicus baud distinctus, e margine anteriore superiore 
segment! buccalis formatus ; orificium oris terminale plicis papillae- 
formibus instructum; setae radiatim et seriatim positae, minutae, 
segmentorum posteriorum i41is anteriorum magis numerosae. Tuber- 
cula ventralia duo in sutura segmentorum obvia. 

Pekichaeta (Schmaeda), ex parte. Setae numerosae, anteriores et posteriores numero 

Lampito, n. Lobus cephalicus transversus, ovalis, integer ; segmentum buccale 

antice non incisum ; tubercula ventralia duo pone cingulum sita ; 
setae radiatim et seriatim positae, anteriores posterioribus numero- 
siores, laeves, fusiformes, apice parum curvato. 

Several of the species referred to these six genera have been subsequently examined 
by Pereier and by myself; they proved to be 'merely variations on the theme of 
Perichaeta.' It is hardly necessary at the present time to insist upon the worthlessness 
of the above characters as serving to discriminate genera. And yet Va-illant (6, p. 6^) 
admits them as subgenera, iacluding Peeeier's genus Perionyx and a new subgenus 


— Perriera (for P. biserialis). These subgenera are indicated in a table which is 
slightly altered from that of Kinberg. 

In 1 87 1 Peeeieb added a new genus Perionyx to Perichaeta, considering it possible, 
however, that this genus might later be regarded as not really distinct. The principal 
character in which it differs from Perichaeta is in the presence of paired nephridia 
('la nettet^ de ses organes segmentaires '), the inclusion of the male pores in a 
circumscribed ventral area, the position of the spermathecae, and the extent of 
the clitellum. 

The two first characters alone distinguish the genus as now defined. When Peeribe 
wrote of the position of the spermathecae as serving to distinguish the genus he must 
have meant the structure of these pouches, i. e. the absence of a diverticulum. There 
is nothing remarkable in their position. This genus has been accepted by all subse- 
quent writers except Vaillant, and is admitted here. In 1881 I added a new genus 
Pleurochaeta ; the identity of this form with Tbmpleton's Megascolex, suspected by 
Vejdovsky (24), was later proved by myself (5). 

Below I have something to say with reference to the confusion which has been 
made with the genera Megascolex and Perichaeta; here I am concerned with the 
genera of Perichaetidae alone. The distinction between Perichaeta and Megascolex 
was first pointed out by Schmaeda, who introduced the latter name ; his distinctions, 
however, were based upon a misunderstanding of Templeton's description of Mega- 
scolex; indeed he re-described M. coeruleus as P. leucocycla. Baied (1) denied that 
there was any distinction between any two such genera ; but his opinion is rendered 
of little value by the fact that he paid no attention to internal characters, and for 
the matter of that but little to external characters also. 

In 1883 I pointed out that two genera, Megascolex and Perichaeta, might be 
distinguished as follows : — 

(i) Ring of setae upon each segment discontinuous at one or more points, 
(a) Clitellum occupying more or fewer segments of the body than three. 


(i) Ring of setae continuous. 

(2) Clitellum occupying segments xiv-xvi. 

It will be observed that I have here defined as Perichaeta the worms now referred 
to Megasco%x and vice versa. This was on the strength of Baied's statement that 
Megascolex and ' Perichaeta were the same. I should, however, in that case have 
allowed the name Perichaeta to be dropped, and have introduced a new name. 

3 a a 


A year after the publication of the paper just referred to I re-defined the genera 
Perichaeta and Megascolex in consequence of the discovery that my Pleurochaeta was 
identical with M. coeruleus. The definitions were as follows : — 

Pebichaeta, Sghmabda. 

Syn. Megascolex, Auct. 

Depinitiobt. Setae generally arranged in a continuous row round tbe middle 
of each segment ; elitellum occupying two, three, or four segments (XIV-XVIl). 
Male generative apertures paired, and situated upon eighteenth segment of 
body, which is always behind the elitellum ; genital papillae occasionally 
developed in neighbouring segments. Female generative aperture single, and 
wiJhin the elitellum, upon the fourteenth segment. Two pairs of testes, more 
or less solid and compact, in segments XI and XII ; terminal portion of 
vas deferens on either side connected with the duct of a large prostate 
gland. Copulatory pouches varsdng in number from two to four pairs, and 
provided each with a variously shaped supplementary pouch or pouches. 
Intestine with a caecum on either side in twenty-sixth segment. 

During the last few yeai's a large number of species of ■' Perichaeta ' have been 
described by Fletchek and Spencer in Australia. The principal characters of these 
are shown in the table