210 JIR. F. E. TJEDBARD OX NEW EARTII\VOR>rS. [Mar. 19, by Dendrohates, as obsei'ved by A\'"yraan, Kappler, and H. S. Smith, is not the least remarkable'. This J3atraehian was found to carry its tadpoles on its back, fixed by tlieir buccal suckers, with the object, it is believed, of transporhng them from pool to pool. Precisely the same mode of parental care is shown by a frog of the family Itanidai, P/njllohatex trinitatis, Garm., a native of Trinidad and Venezuela. A specimen from Venezuela, recently received at the Xatural History Museum, is preserved in spirit with the tadpoles sticking to the back in the manner described in the case of Dendrohates. The sex of the parent which transports the larvae had not been ascertained in the case of Dmdrohahs. It is therefore of importance to state that in the present instance the feat is performed by the male (as figured on Plate X.), which is distingnished by an internal vocal sac. It was further desirable to ascertain whether any buccal peculiarities existed in the larvae in connection with their habits, and as the specimens, six in number, are perfectly pre- served, this examination offered no difficulty. But, as in the case investigated by Wymau, no peculiarities could be detected: the tadpoles are perfectly normal, of the Eanoid type. There are two series of labial teeth above and three beneath the black horny beak, the lower outer series rudimentary ; the inner upper series is widely interrupted mesially, the inner lower very narrowly ; the lip is bordered by a series of papillae which is widely interrupted in the middle anteriorly. The spiraculum is sinistral and the anus dextral. The tail is about twice as long as the body. EXPLANATION OF PLATE X. Figs. 1, . HtjUt goeldii, Blgr.. female carrying the eggs. '2. An egg froiu the above specimen, eularged. 3. Young, on leaving the mother. 4. PJiyllohates trinUatis, Garm., male carrying the lavTse. 5. Mouth of the larval Phi/Uohaic^ trinUatis, eularged. March 19, 1895. Sir W. H. Tlowee, K.C.B., LL.D., F.R.S., President, in the Chair. The following papers were read : — 1. Preliminary Account of new Species of Earthworms belonging to the Hamburg Museum. By Prank E. Beddard, F.R.S. [Eeceived February 11, 1895.] The collection of " Terricoloe" made by Dr. Michaelsen in South America is in some ways richer than that of the " Limicolse " ^. ' Cf. Boulenger, Ann. & Mag. N. H. ser. C-,, i. 1888, p. 454, & ii. 1888. p. 122. - See Ann. & Mag. N. H. ser. 6, \iii. p. 205: "Preliminary Notice of South- American Tuliijicid(e collected by Dr. Michaelsen, including the Description of a Branchiate Form. By Frank E. Beddard, M.A., F.B.S." 1895.] MB. F. E. BEDDABD ON XEW EARTHWORSCS. 211 The number of individuals is much greater ; bat the proportion of new species is not so great. The actual number of new species is, however, much greater. Considering that a good many papers have been written upon the terrestrial Oligochjefa of this part of the world bv Michaelsen, Eosa, and by myself, I was unprepared for the very large number of new species whicli Dr. Michaelsen has got together and kindly entrusted to me for description. The collection consisted exclusively of members of the four following families : Lumbricidae, PericHfetidae, Acauthodrilidae, and Crypto- drilidae. Tlie total absence of Eudrilidae is not of course remarkable, except perhaps as regards the genus Endrilas itself. The entire want of Geoscolicidse is much more siogular. Hitherto, however, the only Greoscolicid which has been found south of the Kio G-rande do Sul is the Titanus forr/uesi of Perrier. Xorth of this point the family is apparently one of the most abundant. The Perichstidse are represented only by a single individual, but this family is not one which is common anywhei-e in the South- American continent. Their headquarters, as regards the New World, are certain of the West-Indian Islands. Lumbricidsj are fairly abundant in Dr. Michaelsen's collection ; but then this cosmopolitan familj^ is abundant everywhere. I am convinced that here as elsewhere the Lumbricidae have been introduced. Dr. Michaelsen informs me in a letter that he observed the proportion of Lumbricidte in his gatherings to diminish with the increased distance from the coast ; in cultivated gardens near to the seaboard this family was the most abundant. This fact (which Prof. »Spencer has confirmed for Austi'alia) is an argument for regarding these worms as the result of intercourse between Europe and the countries in question. A fact which obviously points in the same direction is the invariable identity of the exotic species with European or Xorth-Araerican forms. The characteristic indigenous forms of the temperate regions of South America are of the families Acanthodrilidae and Cryptodrilidse. Both of these families occur in Central and Xorth America as well ; but they are not by any means relatively so abundant in the south temperate region of the continent. More- over, the genus Acanthodrilus (s. s.)is only found in this part of the world, being represented in the tropical and north temperate parts by Benhamia, Trigaster, and Dlplocardia. An examination of this collection, in fact, seems to coufirui what previous researches upon the earthworm fauna of America appeared to indicate — namely, that it is possible to divide the Neotropical region into a tropical and a temperate section. The former is characterized by Greoscoli- cidae, the latter by the genus Acanthodrilus. But in this last genus we have a correspondence between temperate South America and New Zealand. This correspondence is also emphasized by the great prevalence of Microscolex in South America and its fairly common occurrence in New Zealand. The Cryptodrilidae collected by Dr. Michaelsen belong exclusively to this genus ; and they are quite as abundant, though perhaps there are not so many species as the AeanthodrilidEB. Microifcole.v, however, is not so restricted to 14* 212 MR. F. E. BEDDARD OX XEW EARTHWORMS, [Mar. 19, the temperate part of South America as is the genus Acanthoclrihis. 1 cannot see any way out of referring Eisen's recentl}' constituted genus Deltania to MicroseoleA', a point which I enter into later. It seems, however, judging from the information at hand, that the headquarters of Microscole.v are the more southerly parts of South America, and that it gi-adually dies out as we get north, finally disappearing in Xorth America. The very small number of genera coupled with the large number of species is a remarkable feature of the earthworm fauna of the temperate part of South America. Up to the present we are only acquainted with four genera, viz. Acanthodrilus, Kerria, PericJuvta, and Microscohx (leaving aside the Lumbricidse as a foreign importation), from this part of the world. Among temperate countries, 2sew Zealand forms a contrast ; it possesses certainly six, if not more, distinct genera. I shall now proceed to treat of the different families seriatim. Fam. ACAXTHODRILIB-E. This family, as already remarked, is represented in the collection by two genera only — Acanthoclrihts and Kerria. The latter genus is represented by three species, of which I regard two as new. This genus, recently constituted by myself for a small aquatic- species from the Pilcomayo and for E-osa's Acantliodrilusspegazziaii, has been increased by the addition of two new species from Cali- fornia. It therefore ranges through the South-American and part of the !Xorth-American continent. It appears to be rather a tropical form; Buenos Ayres and Valparaiso seem, so far as our present knowledge goes, to mark its southern limit. It is one of those genera that are both aquatic and terrestrial in habit. Acanthoclrihis is represented by a large number of species in South America. Altogether we are acquainted with the following, of which the names of those collected by Dr. Michaelsen are printed in italics : — 12. Acantliodrihis purpui-etis. n. sp. 13. Acanthodrilus ma^ellanicus, n. sp. 14. Acanthodrilus bicinctics. n. sp. 15. Acanthodrilus minutus, n. sp. 16. Acanthodrilus chilen-sis, n. sp. 17. Acanthodrilus cingiilatus, n. sp. 18. Acanthodrilus putablensis, n. sp. 1. Acanthodrilus littoralis. Kinb. ti. Acanthodrilus bilgeri, Mich. 3. Acanthodrilus picius, Mich. 4. Acanthodrilus dalei, F. E. B. 5. Acanthodrilus platyurus, Mich. 6. Acantbodrilus georgianus, ]\Iich. 7. Acanthodrilus falclandicus, F.E.B. 8. Acanthodrilus aquarum-dulciiun, ' 19. Acanthodrilus cameus, n. sp. F. E. B. \ 20. Acanthodrilus corralensis, n. sp. 9. Acanthodrilus bovei, Rosa. 10. Acanthodrilus decipiens, n. sp. 11. Acanthodrilus occidentalis, n. sp. 21. Acanthodrilus simulans, n. sp. 22. Acanthodrilus albus, n. sp. This part of the world must be undoubtedly regarded as the headquarters of this genus. ]N^ew Zealand comes next in number of species : but there are only seven referable to the genus as strictly defined. The South-American Acanthodrili do not form a definable section of the genus. They all agree, however, in having a clitellum 1895.] ME. r. E. BEDDARD ON NEW EAKTHWOBMS. iil3 which is rather limited in extei)t ; in only one species does it occupy more than segments xiii.-xvii. ; in the ^'e\v -Zealand species the cJitellum often extends back to the xixth segment. The American species, too, are never of large size ; a great many of them are very darkly pigmented, a circumstance which is only met with in Acanthodriliis smithii anions; Xew-Zealand Acanthodrili. o It is interesting that many of these species live equally well in fresh water and on land. Amnthodrilus dcilei, for instance, and Acanthodrilus pictus occurred in Dr. Michaelsen's collection from freshwater gatherings as well as from those in forests. This peculiarity of the genus is not, however, confined to the South- American representatives. The New-Zealand Acanthodrilus pidudosus and the Australian Acanthodrilus schmardie also were collected from freshwater sources. I shall now commence the enumeration of the species collected by Dr. Michaelsen. In the description it will be understood that in all species the nephridia are paired, the dorsal vessel single, the spermathecse in viii., ix., and, of course, the ovaries in xiii. (1) Acanthodrilus decipiens, n. sp. To a single specimen of Acanthodrilus from near Estancilla, Pi'ovince of Valdivia, I give the name of " decipiens " on account of its strong superficial likeness to a Microscole.r. I had at first passed it over as an example of that genus, which occurs in the same region. This species is one of the smallest of Acanthodrilus ; the measurements were : length ^.5 ram., diameter 2-3 mm. The number of segments of the worm were 110. The colour (in alcohol) is of a rich brown above, the setae being implanted within white circles. The prostomium is continued over the buccal segment by parallel furrows. The setse are paired, but not so strictly as in Acccntho- drilus dalei, for instance. The implantation of the setae is such as to give to the posterior part of the body a quadrangular aspect. The clitellum occupies segments xiii.-xvii. Dorsal pores are present. Thei'e appear to be small median genital papillae upon segments xvii. and xix., with a larger also median papilla upon xx. As there was only a single specimen of the worm, I am not able to give a complete account of the internal structure, which, liowever, shows no specially interesting peculiarities. The gizzard is very large and stout, relatively to the size of the worm . The spermatheca) have each a single good-sized diverticulum. The spermiducal glands extend through two or three segments only ; the penial setge, of which I found seven mature and immature in a single bundle, are of a fair length, but are not ornamented. Hah. St. 41, Estancilla, Yaldivia. A larger specimen from St. 47 measured 4() mm. 214 MB. r. E. BEUJDARD 0>' NEW EAETHWOKMS. [Mai". 19, (2) Acanthodrilus occidentalis, n. sp. In point of size this is the most conspicuous o£ all Soutli- Aiuerican species of Acanthodrilus; about a dozen specimens in all were collected by Dr. Michaelsen, and the largest of these slightly exceeds in bulk the largest example of Acantlwdrilus jnetus, Avhich is its only rival in size. The present species, however, differs from AcanfJiodrilus pictus in being entirely without integumental pigment, a fact which gives to it in the preserved condition a pale dirty brown colour. The worms are A^ery soft to the touch and the segments are very much annulated. The general appearance, indeed, is like that of the Xew- Zealand species Odocfuptus multiporus ; it is very different from that of any South-American species of the genus which I have had the opportunity of examining. The structure, however, does not in any way resemble that of OctocJuetus ; Acanthodrilus occidentalis is a perfectly tyjjical Acanthodrilus, though differing in detail from any other species known. The largest individual at my disposal measured 192 mm.; its diameter was at the widest point 9 mm. The number of segments of this specimen was 365. The colour of the species after death has been already referred to ; during life the colour was (according to Michaelsen) " blau gran ; Kopfende zart rosa." The clitellum during life a^ as hardly visible : I could not detect its limits in the preserved specimens. The obscurity of the clitellum gave the worm a certain resemblance to many (xeosco- licids. The soft feel of the body is due to the minuteness of the strictly paired setae ; the setae are implanted upon the ventral side of the body ; the ventral area within the setae is to the dorsal area as 7 : 11. On the anterior segments of the body I could not detect the setae at all. The prostomium is continued by grooves over the buccal segment. The dorsal pores commence at the eleventh or twelfth segment. The internal anatomj' shows no characters of very great import- ance Some of the anterior septa are thickened ; this is the case with the six which immediately follow the gizzard. The gizzard, in spite of its large size, is entirely contained between the septa which bound the sixrh segment. There are no distinct calcil'erous glands ; but the oesophagus is red and vascular pos- teriorly. The intestine begins in segment xvii. ; it has a very rudimentary typhlosole. The last pair of hearts is in segment xiii. The reproductive organs are hke those of other species of the genus. There are two pairs of testes in x., xi. ; these gonads are very much frayed out and extend right across their segments, looking as if they were attached to both w alls. The sperm-sacs are of fair size ; they are racemose in form and are attached to the front walls of segments xi., xii. 1895.] mi. F. E. liEDDABD ON NEW EARTHWOUMS. 2lo The spermiducal glands are small and narrow; they are confined to their respective segments (the xviith and xixth), and lie trans- versaly to the long axis of the body. The penial sette with which they are provided are slender and unoj-namented. The spermathecae are like those of many species of Pcriclui'ta m the narrow tubular appendix, wtiich is of tlie same length as the pouch. The diverticulum ends in a small dilatation. Bah. St. 7, Valparaiso, Salto ; St. 12, Valparaiso, Garten. (3) Acanthodrilus magellanicus, n. sp. There were several specimens of this species, of which the one selected for measurement was 66 mra. long a;jd3-5 mm. in diameter; it consisted of nearly 100 segments. The colour after preservation was a yellowish grey, owing to the absence of integumental pigment. The setse are not strictly paired ; the ventral setae are nearer together than the lateral sette ; but in the posterior i)art of the body there was less difference in the distance between the setae. The clitellnm occupies segments xiii.-xvii. On segment xi. are a pair of large genital papillae. On the middle of segments xiv. and xv. is a narrow band of glandular appearance ; finally, the pores of the spemiducal glands are situated upon large oval papillae united nearly or completely in the middle line ; they also bulge over the intervening segment so as to nearly obliterate it. Corresponding to the position of the ventral setae on segments xvii. and xix. are the apertures of the spermiducal glands. When this part of the body is mounted entire in glvcerine and examined, the summit of each of the four papillae already referred to is seen to be perforated by three orifices. One is the actual pore of the gla)id ; the two others each correspond to one of the two penial setae which accompany the glands ; it is not common for the setae to open thiLS independently of the glands. The orifices are pre- cisely similar in appearance and are of equal size. This species has a well-marked gizzard in segment vi. The septa separating segments vi./xii. are stout ; the first septum lies between segments v./vi. The last heart is in segment xii. The funnels of the sperm-ducts lie in segments x., xi. ; the racemose sperm-sacs are in xi., xii. The two pairs of spermathecae lie as usual in viii., ix. ; the pouch itself is oval, and it communicates with the exterior by a long cylindrical duct, which is not very much narrower but is rather longer. Just before the external orifice of the latter are two caeca, w-hich really open by a common pore ; they lie one above the other, and not laterally and symmetrically as in Acanthodrilus bovei. The two caeca are enveloped in a common sheath, but they are divided by a constriction superficially, which marks a real muscular septum which separates the two diverticula. 'J he lower one is globular in form ; the upper, which is larger, is more elongate. 216 MR. F. E. BEDDARU OIV NEW EAllTUWORMS. [Mai'. 19, The spermiducal glands are tubulnr in form and of fair thick- ness ; there is a narrow terminal duct. The glands were bent into an S-shape, and did not extend beyond their segment. The ])enial seta? are not particularlv long. They are regularly curved like a bow. There were four of them in each bundle that I examined — two fully mature and two immature. The end of the seta is abruptly truncated, looking very much as if it had been broken off short. There can, however, be no question of such an accident, as all the setae presented the same appearance. The free end of the seta, perhaps one-third of its entire length, is orna- mented by transverse ridges, which are apparent at the edges as notches. Hah. St, 99, Magellan Straits, Elizabeth Isl., under cow-dung. (4) Acanthodrilus minutus, n. sp. Of this species, again, I have only discovered a single specimen in the collection. I have been obliged, therefore, to exercise some care in my dissection of it. It is of a pale brown colour in alcohol, being apparently without anv integumental pigment. On a superficial inspection it might easily be mistaken for a small example of Acanthodrilus bovei. The worm measures 29 mm., its diameter is 3 mm.; the number of segments 70. It is therefore, with the exception of the Australian Acantliodrilvs madeayi (27 mm.), the smallest Acantho- drilus known. The prostomium is continued by furrows over the buccal seg- ment. The setse are strictly paired. In smaller immature specimens, but with more segments, the setae were seen to diverge from each other posteriorly as in A. platj/unis &c. The clitellum occupies segments xiii.-xvii. The gizzard is large for so small an earthworm, being 2-5 mm. long by 2 mm. broad. The intestine begins in segment xvii. Tsone of the septa appear to be specially thickened ; but those in the neighbourhood of the generative organs are, as is the rule, somewhat more obvious than rhe others. The testes and sperm-ducts are, as is often the case with the Acanthodrilus of America, single — -that is, there are only one pair of each. The sperm-sacs, racemose and of large size, lie in segment xi. The spermiducal glands are thick but not very long ; they are confined to their own segment. It is noteworthy that the pos- terior pair are distinctly smaller than the anterior pair, and less coiled. I have observed the same comuiencing disappearance (?) of one of the two pairs of glands in Acanthodrihis sch)nard(r. The penial setae are long and recurved at the end in a crozier-like fashion, which is so common in the genital setae of these earth- worms. The extremity is expanded and thin, but not ornamented — two in each bundle. 1895.] ME. r. B. BBDDARD ON NEW EABTHWOEMS, 217 The spermathecae have a single rather large tubular to oval diverticulum. Hah. St. 30, Putabla, Valdivia. (5) Acanthodrilus bicinctus, n. sp. Of this species there were two examples from Pictou Island and two from Juan Island. The two latter were rather larger than the former and showed a slight difference in the penial setae, which will be referred to in due course. The length is 42 mm., diameter 3 mm. ; number of segments 80. The larger worm from Juan was 55 mm. in length. The colour of the species is a dark purple, passing into an orange-brown on the ventral surface. The colour was darker than in most Amnthodnli. The prostomium is broad, narrower posteriorly where it is continued over the buccal segment for about two-thirds of its length. Posteriorly the prostomium bears a dorsal groove which is continued to the end of the buccal seg- ment, and is apt to produce the impression of the prostomium entirely dividing the buccal segment. The ditellum occupies segments xiii.-xvi.: in one specimen it was divided by a median furrow into two halves, each consisting of two segments ; this circumstance suggested the specific name. There are median unpaired genital papillae, extending between the ventral setae on segments xx. and xxi. in one individual, and in another on to the two following segments also. The gizzard is well developed. The reproductive organs are not very different from those of some allied species ; I fancy that the gonads are only one pair of testes, and of course one pair of ovaries, but am not quite certain. In any case there is but a single pair of large racemose sperm-sacs in segment xi., as is so general. The spermiducal glands are large and flattened, being coiled several times in one plane ; the t\\o following glands of each side of the body form an almost continuous mass occupying segments xvi.-xx. The narrow muscular duct of the gland arises from about the middle of the coil. It is accompanied by a sac contain- ing two penial setae which are fully mature, and two immature ones. The former are of a particularly bright yellow colour. One is bent like a bow in a regular curve and ends in a sharp point; its distal third is beset with sharp downwardly directed spinelets. The other seta is of quite a different form. To begin with, it is at least quite half as long again as the first and is more flexible ; this appears to be shown by the whip-like curvature of the distal extremity, which is disposed almost in coils. It termi- nates in an obtusely pointed extremity. Less than the distal third of this seta is beset with excessively fine spinelets, much finer than in the other seta. There can be no question that both setae are fully mature and they show a dimorphism. A similar dimorphism has been indicated by myself to exist in Acanthodrilus georgianus. It has also been met with elsewhere. In the specimen from Juan 218 MR. F. E. UEDDAllD ON XEW EARTHWORMS. [^Mar. 19, Island which I examined the dimorphism was not apparent. The mature sette were of the longer type with a flexible extremity ; the end was slightly dilated and beset with very fine spinelets. If I had had more specimens at my disposal it is possible that it might have been found necessary to divide these examples from Juan Island into a distinct variety. The spermathecae (as usual in viii., ix.) are oval sacs with a moderately short duct ; with the latter communicates the diverti- culum, which is longer than the pouch and has a ci'enate outline above. Hah. St. 174, Pictou Island; 58, Smyth Channel, Wide Baj', Juau Island. (6) Acanthodrilus purpureus, n. sp. Of this species there was only a single specimen in the col- lection. It was somewhat softened and therefore the measure- ments are perhaps a little higher than one would be inclined to allow as typical. The length is 95 mm., the diameter 5 mm. ; the number of segments 98. The colour (in alcohol) is a reddish purple above, passing into an orange -brown below : the clitellum is distinctly without pig- ment, being yellowish brown . The prostomium does not extend at all over the buccal segment. The setae are strictly paired. The chtellum occupies segments xiii.-xvi. and is quite complete. There seem to be no papillae of any kind. The gizzard lies in segment vii. The last heart is in xii. The spermathecae have each a longish tubular diverticulum dilated at end. The sperraiducal glands are loosely and irregularly coiled, extending through several segments ; the penial setaj with which they are provided are curve-like and of moderate length only. The extremity is pointed and the distal end of the seta is f lu-nished with numerous spinelets, which have a broad base of attachment and the points of which are directed forwards. Hah. St. 81, Magellan Straits, Punta Arenas. This species comes perhaps nearest to AcantJwdrilus hicinctus. It indeed only differs by its much greater size, by the shorter penial setae, and by the form of the prostomium. (7) Acaathodrilus chilensis, n. sp. This species, again, is unfortunately I'epresented by a single specimen only, which had therefore to be discreetly dealt with. It is 80 mm. long by 5 mm. in breadth ; the number of seg- ments 150. The colour (in alcohol) is a dark purphsh above, passing into an orange colour below. The prostomium is completely joined to the second segment by furrows. 1805.] ilR. F. E. BEBDARD 0>" XEAV EARTHWOltMS. 219 The setae are strictly paired aateriorly ; posteriorly the_y diverge slightly from each other, as is the case with AcanthodrUus platy- urus ; but the divergence is by uo means so marked as iu that species. The clitellum occupies segmenis xiii.-xvi. The dorsal pores are quite obvious, commencing at any rate on xii./xiii. The gizzard occupies two segments, vi. and vii., the septum being present. The septa dividing segments viii./\ii. are thickened, but not much. The last heart is in segment xii. The sperm-sacs, as in so very many species, are in xi., and race- mose. The spermathecfB are longish and of a reddish colour ; the diverticulum is considerably shorter than the pouch. The spermiducal glands are not very long or much coiled ; they are also reddish in colour. The peuial setfe are brown, a colour which is, according to my experience, rare in AcanthodrUus. Their form hardly differs from that of Aca)Uhodrilns plati/urus. Hah. yt. 39, Teja Island, Valdivia. This species is clearly a close ally of AcanthodrUus ijlatjiurKs. The most ready way of distinguishing them is by the form of the sperniathecse. If it were not for the different position of the last heart I should have been inclined to regard the present species as merely a variety of AcanthodrUus platytirus. Three specimens from St. 48, Corral, appear to belong to the same species, but they are very much paler iu colour. The spermathecte haAe the same red colour and its diverticulum is similar. Possibly a larger series of specimens would allow of the separation of the two forms. (8) AcanthodrUus cingulatus, n, sp. Of this species five or six individuals were collected by Dr. Michaelsen. The largest of them — the only one that was fully mature — measured 58 mm. in length by 4 mm. in diameter; the number of segments 106. The worm is darkly pigmented, but the colour is more brown than purple. The setse are strictly paired and do not show any signs of divergence from each other at the tail end of the body. The prostomium is continued by furrows over the buccal seg- ment. The clitellum is unusually far back for the American members of this genus ; it extends from segment xiv.-xviii. inclusive. Dorsal pores are obvious. There is a large gizzard in segments vi., vii. The spermatheca; are oval pouches \\ith a smalHsh diverticulum not one-half of the length of the pouch. 220 MR. F. E. UEDUARU OX NEW E.VRTIIWORMS. [Mar. 19, The testes and the funnels are in segment x. The spermiducal glands are coiled a good deal ; the first pair are distinctly larger than the second pair ; as this was found to be the case in two specimens not selected in any way, it may, i imagine, be safely regarded as normal for the genus. I have pointed out in the present paper that Acanthodrilus minutus shows the same peculiarity, also seen in the x'\.ustralian Acantho- clrilas svhmardce. The penial setae are expanded and recurved at the extremity, and are very similar to those of Acanthodrilus 2>latyurus. Hah. St. 39, Island of Teja, Valdivia. (9) Acanthodrilus putablensis, n. sp. Of this species there were two examples, one of which was larger than the other. I had at first passed by the worms on the assumption that they were examples of Acanthodrilus platyurus. The coloration, however (of the preserved worms), is a little different from that of any of the examples of A. platyurus in the collection ; and a nearer inspection showed other external differences which rendered their separation from A. platyurus even more obvious. The main external difference is in the arrangement of the setae ; but in order to appreciate it properly, for it is, after all, slight, it is requisite to compare individuals of both species somewhat closely ; it can then be mnde out that, while there is the same divergence of the setse of each pair from each other posteriorly, the distance which ultimately separates the setse of each pair is distinctly greater in the species now under consideration than in the allied A. platyurus. The worms were of stout build, and evidently rather contracted by the preservative reagent. The length is 82 mm. by 8 mm.; number of segments 150. The skin of one individual was invaded by numerous encysted Gregarines. These formed a series of white warts upon the skin, a situation where I have never before observed Gregarines; the interior of the body was also full of the parasites. The prostomium is quite completely developed, dividing the buccal segment. The clitellum occupies segments xiv.-xvi. The gizzard is stout and lies . in segment vi . ; after the gizzard are two thin septa ; those separating segments viii./xiv. are strengthened. The last hearts are in segment xiii. It is the reproductive organs which show the greatest differences from A. platyurus. The testes, howeA'er, are, as in that species, one pair in x. ; the funnels in the same segment are highly vascular, as are also the oviducts, which occupy the usual position. The sperm-sacs are not like those of A. platyurus, in that one pair exist in the xiiith segment ; there is another pair in ix. Although the sperm-sacs are in xiii., I presume that the ovaries are there also, for the oviducal funnel undoubtedly projects into that segment. J 895.] 5IB. F. E. BEDDARD ON NEW EARTHWORMS. 221 The sperraiducal glands are much like those of A. platyurus. They are somewhat slender and confined to their segments ; the posterior pair was decidedly smaller than the anterior. The penial setae are a facsimile of those of A. platyurus. The spermathecse have two symmetrically arranged short diverticula. Hah. St. 30, Putabla, Yaldivia. (10) Acanthodrilus carneus, n. sp. A number of specimens of this species were collected at Quilipue. They show when in alcohol a reddish- violet colour, which, however, is due to their tissues having taken up the colouring-matter from a Myriopod placed in the same bottle. The real colour of the worm during life is, according to Michaelseu, " schmutzig fleischfarbe : Giirtel weiss." The length of one of the largest specimens was 52 mm. by 3"5 ram. in diameter. The number of segments of this specimen was about 100. The prostomium is large and completely divides the buccal seg- ment. The setfB ai'e paired, but not very strictly ; the ventral setae of each pair are rather closer together than are the lateral setae. The clitellum occupies segments xiii.-xvi. and is complete all round. As is so often the case, this species is provided with a certain number of genital seta papillae ; as is also frequently the case with the worms of this geuus, these papillae are subject to some variation. In one specimen there was a single eye-like papilla on the boundary liue of segments x./xi. In another there were two median papillae, one upon x. and the other upon xi. ; in two others, and this appears to be the more typical arrangement, there were the two papillae already mentioned and in addition to them paired papillae upon ix. The dorsal pores are quite conspicuous and begin in front of the clitellum. The internal characters are not very different from those of other species. The gizzard is well developed and lies apparently in segment vi. The last heart is in the xiith segment. There appears to be only one pair of testes and of funnels in segment x. The sperm-sacs are in segments ix., xi. ; those of the latter segment are very coarsely racemose. The penial setae are unusually delicate and slender; they taper towards the free end, which is not ornamented ; perhaps on account of their slenderness they are not of the yellow colour so usual. The spermathecse in viii., ix. are oval pouches with a wavy tubular diverticulum nearly as long. Hah. >St. 3, Quilipue. 222 Mn. F. E. BEDDAKD OX KEW EATlTHWOHirS. [MuP. 19, (11) Acanthodrilus corralensis, n. sp. Of this species there was ouly a single individual. It is tolerably' large, measuring 103 mm. in length by 5 mm. iu diameter ; number of segments 145. The colour is very much that of Acanthodrilus picbis ; and it also resembles that species in the fact that the prostomiuni only extends over about one halt of the buccal segment. On the other hand, Acanthodrilus corralensis differs from A. pictv.s in that the sette are strictly paired from end to end of the body, being only to a very slight extent separated at the posterior end of the body. In this region the body has a roughly quadrangular section, the setfe being implanted at the angles. The symmetry of the iigure, however, is broken, ouing to the two ventral pairs being closer together than the two dorsal. The clitellum occupies segments xiJi.-xvii. and is complete. The dorsal pores commence at any rate in the clitellar region. Xone of the intersegmental septa are especially thick. Those separating segments x./xiii. are rather stouter than the others. The gizzard lies anteriorly to the first pair of spermathecje, but I am not able to fix its segment with precision. In the xivth and xvth segQients, parricularly in the latter, the oesophagus is enlarged and highly vascular, its lining membrane being folded. The last pair of hearts are in segment xii. The testes and sperm-duct funnels are one pair only in seg- ment X. The racemose sperm-sacs are in ix. and xi. The spermiducal glands are like those of a few other species of the genus in that the anterior pair are distinctly larger than the posterior. The penial setse, of which there were two in the bundle that I examined (the posterior on the left side), are crooked at the end like a crosier. The ends of the set* have numerous denticulate ridges. The setae resemble, in fact, very closely the penial setse of A. pictvs. The spermatheciP are not very large ; each is an oval sac, to the duct of which is fixed the very slender duct of the appendix, widening out above into an oval sac. The diverticuluru of the Spermatheca is about half the size of the pouch. Hah. St. •17, Corral, " "Wald, unter Steinen."' (12) Acanthodrilus simulans, n. sp. This species was found in a large gathering of worms from Corral in company with the last species and with a considerable number of examples of Microscolex spatidifer. The external characters of A. simulans are so like those of the last-named species that they were at first passed over accidentally. It is the colour which is so strilcingly like. In the specimens of M. sj^atu- lifer from this locality the colour of the dorsal surface -was a reddish brown instead of the more usual violet. Precisely the same colour was found in the Acanthodrilus. As this species is also 1895.] MR. F. E. BEUDARD OS NEW E.VRTHWOEMS. 223 characterized bv a complete prostomium and strictly paired setae the likeness is increased. The length of the largest specimen is 82 mm., diameter 4 mm. ; number of segments 150. The clitellum was not developed. There are dorsal pores. The gizzard is well developed and lies in vi., vii. ; the oesophagus in segments xiv. and xy. is extremely vascular, being a bright red colour, which entirely disappears in the narrow section of oesophagus lying in xvi. The intestine begins in xvii. The last heart is in segment xii. Xone of the septa are greatly thickened, nor are any of thein expended backwards in the cup-like fashion so prevalent among earthworms. Septa ix. xii. are somewhat stouter. There appears to be only a single pair of testes and sperm-duct funnels in x. The sperm-sacs in ix. and xi are very large and coarsely racemose. The spermiducal glands show the same inequality of size between the anterior and posterior pairs that has been noted in other species. They are both stout, but short and only slightly coiled. Thepenial setae are not particularly long ; they are spatulated at the free extremity, and, when seen in profile, doubly curved in the usual way. They are not, however, ornamented. The spermathecas occupy the usual segments. The single diverticulum is nearly as large as the pouch. It is chalky white (owing to the presence of spermatozoa) and has a crenate outline. The posterior pair of spermathecae are larger than the anterior pair, which is doubtless correlated with the difference in size of the spermiducal glands. ffab. St. 47, Corral, " Wald, unter Steinen." (13) Acanthodrilus albus, n. sp. This species again is represented by a single specimen, which I at first passed over under the impression that it was A. hovi-i. It has the same long and slender form and absence of integumental pigment. A. bovei, however, has not been met with so far north on the eastern side of the continent as the present species, which may perhaps be looked upon as its representative there. Length 85 mm., diameter 2 mm. ; number of segments 145. The prostomium seems to be not continued across the buccal lobe, but I am not certain as to this point. The setae are paired ; the ventral more strictly than the lateral. The clitelkun occupies segments xiii.-xvii. ; there are dorsal pores. The gizzard is well developed, contrary to what we find in A. bovei ; it occupies the vith segment. The intestine commences as usual in segment xvii. Xone of the septa are very much thickened ; those of Lx./xiii. are most so. The segments occupied by the sperm-sacs are somewhat unusual. The sacs, which are racemose in character, occupy segments Lx., xi., xii. As there were no sacs in segment x., I am inclined to believe that the testes and funnels lie there, but have not seen them. The spermiducal glands are slender and moderately coiled. The 224 MR. F. E. BEDDABD 0>' NEW EAETHWOBMS. [Mar. 19, penial setae are long for the size of the worm and also slender. They are curved in the usual way and terminate in a very pointed extremity. The setae are not ornamented. The spermathecae are tubular in shape and have a tubular diverticulum of equal length. Rah. St. 47, Corral, " Wald, unter Steinen." Genus Kerria, Beddard. The collection contained examples of at least three species of this genus, of which I regard two as new. The species which is not new is Kosa's Acanthodrilus spegazzinii. I have before expressed the opinion that this species is really a member of the genus Kerria. I am now in a position to confirm that supposition. The genus is at present chiefly known from an excellent paper by Eisen upon new species which he discovered in California. The existence of this paper renders it necessary for me to enter into the minute anatomy of the species, to which I refer, or which I describe, for the first time in the present paper. The genus is, so far as we know at the present time, exclusively American. It also appears to be a tropical genus, or at least to be a native of the warmer parts of the American continent. The original species of the genus, Rosa's '■^Acanthodrilus" sjiegazzinii, was met with in the neighbourhood of Buenos Ayres ; my species Kerria halophila was brought back by Mr. Graham Kerr from the upper reaches of the Pilcomayo River. The new species which I describe in the present paper were found at Buenos Ayres and in Valdivia. Eisen's species were found by him in Lower California. The two northern species of the genus diiFer from the southern species in having no gizzard and in having diverticula of the spermathecae. (1) Kerria rosae, n. sp. Eisen has used as a character for distinguishing the different species of this genus the number of setae present upon the seg- ments which bear the pores of the sperm-ducts and of the spermiducal glands. This character is probably valuable, but it must be used with circumspection. In two examples of the present species the setae upon the segments in question differed. In one individual they were all present ; in another the inner of the two ventral setae was alone present on segments xvii. -xix., excepting on one side of one segment, where the outer seta alone was present. There is no doubt that this difference has some relation to the condition of the maturity of the worm, but the same remark may possibly apply to other cases. The species to which these observations apply is a longish, thin worm, rather more than an inch in length and about 1 mm. in breadth. A large number of specimens were collected under stones on the banks of the river Baraccas do Sul near Buenos Ayres. The species is described by Michaelsen as having been " flesh-red " during life ; none had a clitellum. 1895,] MR. F. E. BBDDAED ON NEW EAETHWOE.MS. 225 The alimentary canal has, as in Kerria TidlopMla, a gizzard, hut it is rather more slightly developed. The muscular walls are only about twice the thickness of the epithelial lining; the chitinous cuticle secreted by the latter is not at all thick. As in all the other species of Kerria, there are a pair of calciferous glands in the ninth segment. The structure of these is, however, a little more complicated than in Kerria zonalis or Kerria macdonaldi, in which species it has been figured by Eisen. The walls of the pouch are of considerable thickness and project here and there as folds into its lumen. In the thickness of the walls run numerous blood-vessels ; but the tissue of which it is composed is permeated by channels which are intra-cellular. The tissue in fact looks like a closely welded mass of nephridial tubules. The minute structure recalls that of the dorsal diverti- culum of the alimentary canal in Bucholizia, and it is — so to speak — an exaggeration of the structure \\hich I have described in the distal part of the calciferous gland of Gonluodrilus. The oesophagus is narrow and nowhere sacculated ; its cihation commences at the orifice of the calciferous glands. The large intestine begins in segment xii. Septal glands are present as in other species. I did not find them further back than segment vii. The anterior septa are thicker than those which follow. The stoutest are those between segments vi./ix. ; but the three which come next are also moderately stout. The last heart is in segment xi. The spermathecse He in segments viii. and ix. ; they are sessile upon the body-wall, with no long stalk, nor have they any trace of a diverticulum. The extreme end of each pouch is enveloped in a thick muscular layer derived from the body-waU ; this is continuous over the pouch for but a short way from the pore. The muscular sheath is so perfectly continuous with the two muscular layers of the body-wall that it is really impossible to say where the body-wall ends and where the sheath of the spermatheca begins. The lining epithelium of the spermatheca is folded and forms numerous narrow ridges. The testes, sperm-sacs, ovaries, and oviducts are of the form, and occupy the positions characteristic, of the genus. The spermiducal glands are long and coiled ; the muscular part of the organ is also of considerable length. Hab. St. 204, Buenos Ayres, Baraccas do Sul, " Unter Steinen, Ufer des Husses." (2) Kerria saltensis, n. sp. The general aspect of this apparently new species recalls that of the species Kerria halojphila. But it evidently difiers from that species, though not in very important points ; indeed, all the American species of the genus come near together. I leave out in the following description the majority of the characters which are, so far as our present knowledge goes, of generic value and Peoc. Zool. Soo.— 1895, No. XV. 15 22(3 Mil. J?, B. BBDJ)Aai) ON KEW EABTHWOBMS. [Mar. 19, confine myself: to those characters which appear more or less to distinguish the species. It is a small species, about an inch in length ; there is no pigmentation at all discernible. The clitellum occupies segments xiv.-xx., and is incomplete ventrally on the genital segments, i. e. segments xvii.-xix. On those segments only the inner o£ the two ventral setae are present. There is a not very well developed gizzard in segment vii. The calciferous glands, as usual, are in ix. They are rather thick- walled, but are without any folds of the lining epithelium. The walls are vascular, but whether there is the mass of tubules which I have described in the last species I am unable to say. The intestine appears to begin in segment xiii. At any rate in this segment the lining epitheUum undergoes a sudden change in thickness, being from that segment onwards much thinner ; at the end of the twelfth segment the thick epithelium projects into the lumen of the gut of the next segment and forms a kind of trap which would allow of the passage of food backwards, but would not allow of its passage in the opposite direction. The first segment of the intestine, however, as I find it, is of less calibre than the section which commences in the fourteenth segment ; but it differs from the section which begins in that segment by being rather folded. The thick septa occupy segments v./ix., but the next two are thicker than those which follow. The last heart is in xi. The spermathecse, without any diverticula, are in viii., ix. ; they consist of a thick-walled duct and of a thin-walled portion which is stored with sperm. The length of the two regions of the spermathecae is about the same. The epithelium which lines the thick-walled section of the organ which may be regarded as the duct is thrown into folds ; there is no folding of the distal section. The testes are, as usual, in segment x. This segment also contains the sperm-duct's funnels. There seems to be only a single pair of sperm-sacs, which are in segment xi., and are not racemose in character. The cavity of the tenth segment contained a mass of developing sperm ; but this cannot be regarded as the equivalent of a sperm-sac, for it was not surrounded by any membrane. The sperm-ducts were not developed in the specimen which I examined by means of transverse sections ; on the other hand, the oviducts were fully developed, and their openings on to the exterior in segment xiv. were quite obvious. The spermiducal glands reached back as far as the twenty-fifth segment. Hab. Valparaiso, Salto. Fam. Cetmodeilid^. The family Cryptodrilidse is represented in the present collection by the genus Mia^oscolex only. Nor has any other genus belong- ing to this family been recorded from the southern parts of the American continent. Michaelsen's Crifjotodrilus spatulifer is the only Cryptodrilid that we at present know from the temperate 1895.] MB. F. E. BEDDARD ON NEW EABTHWOKMS. 227 regions of South America. Michaelsea does not pretend to definitely assign this species to the genus Oryptodrilus ; he only placed it there (with a query) pending the revision of the Crypto- drilidae. The classification of that family is one of the most difficult tasks for the systematist ; the differences are so slight between species and species that the boundaries of genera are extremely hard to draw. I believe, however, that it is possible to define the genus Microscolex by the following characters : — (1) Nephridia paired, beginning in the second to foui'th segment. (2) Male pores upon segment xvii. (3) Sphermiducal glands tubular, provided generally with penial setse. This definition is naturally somewhat wider than that adopted by Eosa, the founder ol' the genus. But a number of new forms which I shall describe immediately necessitate an expansion of the genus. They do not present a sufficient number of salient differ- ences to warrant the formation of a new genus. I have already proposed to include my genus Rhododrilus in Microscolex, and I do not see sufficient reasons for retaining the genus Deltania of Eisen. Gryptodrilus spatulifer of Michaelsen must, I think, espe- cially in view of the new species to be described here, be relegated to Microscolex. The size is really the only bar to its inclusion. I find a good many individuals of Microscolex duhius which are really quite as large as many individuals of Microscolex spatulifer. The distribution of this genus is interesting. It is the commonest earthworm in point of number, and nearly so in point of species, in the temperate parts of South America. I quite agree with Rosa that it has probably been introduced into Italy and possibly also into Australia ; I described a species from Teneriffe and one from Algeria which are very 1 lively to be regarded, in the same way. The remaining species occur in New Zealand {Microscolex miau- tus and M. novce-zealandice) and in California (the genus Deltania of Eisen). Dr. Michaelsen's collection contained examples of the following species : — 1. Microscolex spatulifer (Mich.). 2. Microscolex ditbius (Fletcher). 3. Microscolex griseiis, n. sp. 4. Microscolex michaelseni, n. sp. 5. Microscolex corralensis, ii. sp. 6. Microscolex diversicolor, n. sp. 7. Microscolex longiseta, n. sp. 8. Microscolex robustus, n. sp. 9. Microscolex modestus, Eosa. 10. Microscolex gracilis, n. sp. 11. Microscolex papillosus, n. sp. It will be seen, therefore, that this collection contains a large number of new species ; evidently South America is the head- quarters of this genus as it is of Acanthodrilus. 15* 228 SIK. F. E. BBDDABD ON NEW EABTHWOEMS. [Mar. 19, Eisen allows three genera o£ worms where I only allow the genus Microscolex. These genera are of course my Mhododrilus and Eisen's Deltania in addition to Microscolex. Deltania is dis- tinguished from both by the closer approximation of each pair of ventral setae in the segments near to the xviith. Rhododrilus is distinguished from both by the separate opening of sperm-ducts and spermiducal glands. This latter character will hardly suffice as a mark of distinction ; there are so many intermediate conditions in the genus Microscolex (s. s.). In M. gracilis the two unite well within the body-cavity ; in M. novce-zealandiw just at the common orifice ; in M. ^xipillosiis the external orifices are separate but still close together, in M, michaelseni at some little distance apart ; finally in M. modestus, as I have an opportunity for stating here, the pore of the sperm-duct is positively in the next segment to that which contains the orifice of the spermiducal gland. Nor do I admit that the approximation of the ventral setae in the genital segments is a character of sufficient importance to imply generic distinction. To begin with, M. diversicolor is somewhat inter- mediate, the approximation being less marked. Then the species of the genus are not all of them furnished %vith distant setae ; in a good many of the South-American species the setae are strictly paired ; this difEerence is itself of greater importance, I think ; and the close pairing of the ventral setae in the neighbourhood of the male pores is an intermediate condition between the paired setae and the distant setae. If it is thought necessary to divide what I here ca\l Microscolex into two genera, a more suitable line of division, as it appears to me, would be to sepai-ate off those species in which the testes and sperm-duct funnels are only one pair ; this division would include all the new species described in the present paper with the addition of Microscolex spatulifer, and would include all the South- American forms, excepting only M. duhius and M. modes- tus. I do not, however, propose even this division of the genus. Most of the new species differ from Microscolex modestus and M. dubius, the types of the genus, in having a prostomium which entirely divides the buccal segment, and in that the setae are strictly paired. As, however, these two characters are not always correlated, I do not think it possible to create any new genera. Another interesting feature about some of the new species of Microscolex is that, like M. spatulifer and many of the Acantho- drilvs, they are very brightly coloured. The internal structure does not present much of great interest. Indeed, the uniformity of the Cryptodrilidae as a whole is in striking contrast to the greatly varied structure of the Eudrilidae, with which I do not think that anybody now would venture to associate them. (1) Microscolex griseus, n. sp. This species, like Microscolex spatulifer, is one to which the generic name Microscolex is etymologically inapplicable. It is a large species, nearly if not quite as large as Microscolex spatidifer. 1895.] ME. F. E. BEDDAED Olf NEW EAETHWOEMS. 229 A considerable number of specimens were collected in a garden in Valparaiso. They vary greatly in size, some being mature and others not so. The biggest specimen is 84 mm. in length by 5 mm. in diameter. The number of segments of this individual was 117. The colour of the Uving worm is described by Michaelsen in the following terms : " schmutzig grau ; vorne schmutzig fleischfarbe ; Giirtel weiss bis braunlich." The preserved worms are of much the same colour. During life the worm was remarkable as " lebhaft schlangelnd bei Beriihrung." The setae are strictly paired. Dorsal pores are present. The clitellum occupies segments xiii.-xvii. There is a median genital papilla upon segment xvi. ; the male pores are on xvii. The prostomium, as in so many species of the genus, is completely carried over the buccal segment by grooves. The first setigerous segment has in most specimens a furrow on the dorsal surface which extends right across that surface, but is not visible ventrally. This gives a little the impression that the prostomium is an incomplete one. The gizzard is large ; it lies in segment vi. It is immediately followed by a thin and delicate septum, behind which are five septa, moderately thick. The last heart is in the twelfth segment. The sperm-sacs are in the ixth and xith segments ; those of the latter segment are so coarsely racemose that they appear to be formed of a number of small separate sperm-sacs ; this does not apply to the anterior pair. The spermiducal glands are different from those of any other Microscolex which I examined in their regular tubular form and in the complex coiling ; they look much more like the corresponding glands of certain Acanthodrilidse. Moreover, the duct of the gland is unusually conspicuous ; it is also long and is not more than one- third of the diameter of the gland. It has a distinctly nacreous glitter. In common with the spermiducal glands opens on each side of the body a sac of penial setse ; there are two of these on each side. They are very yellow in colour and have a bent bow- like form. The extrea)ity is only moderately pointed and a little flattened and expanded. Fine spinelets ornament the extremity. The spermathecse in segment ix. are particularly large. The racemose diverticulum is borne at the end of a long muscular stalk. Hah. St. 12, Valparaiso, in garden ; St. 6, Quilipue ; St. 22, Coronil ; and St. 25, Valdivia, San Jose.' (2) Microscolex longiseta, n. sp. This is a small, rather transparent species, with little or no trace of integumental pigment. * From each of the last two stations a single indiyidual frohably of this species. 230 MB. p. B. BBDDAH.D ON NEW BABTHWOKMS. [Mar. 19, The length of a specimen selected for measurement was 40 mm. by 3 mm. in diameter. Number of segments 95. The prostomium is complete. Dorsal pores are visible. The clitellum occupies segments xiii.-xvii., and is not developed all round the body. There are no papillae of any kind that I could discover. The gizzard is in segment vi. The last heart is in xii. The sperm- sacs in xi. The sacs of penial setae are, as in Microscolex papillosus, of enormous length. They measure 6 mm. and extend through fourteen segments. The sperraiducal glands, on the other hand, only extend through six segments. The penial setae have the same form as in Microscolex ptapillosus. The spermathecae are also exactly as in that species. Hah. St. 193, Tierradel Fuego, Puerto Pantalon ; St, 187, Tierra del Fuego ; St. 140, Uschuia ; St. 178, Navarin, Porto Toro. (3) Microscolex papillosus, n. sp. Of this rather remarkable new form three individuals were taken in forest at Uschuia. The largest specimen measures 87 mm. ; its diameter is only 3 ram. ; the number of segments 95. The worm is thus a slender species. Like so many of the American species of Microscolex, it is pig- mented ; but the pigmentation is not very great, the anterior end of the worm dorsally being alone much coloured. The colour is purplish. In one individual there was no pigment. The setas are quite strictly paired. The prostomium is not continued by furrows over the entire buccal segment ; it extends over about half of it. The clitellum occupies segments xiii.-xvi. ; it is complete. On the sixteenth segment in the middle line is a deep hole, looking as if caused by the contraction of the stout muscles connected with it and passing to the dorsal parietes. Tliis is very characteristic of the species, and was not confined to the largest individual which was selected for dissection. It lies in the clitellar region, but is lined by a strong tall epithelium, which is not glandular like that of the clitellum. There seem to be no dorsal pores. The genital papillae are very numerous, an unusual state of affairs in this genus. There are at least six of them round somewhat translucent areas, possibly rather sense-organs than glandular papillae, upon the middle of segments xii.-xii. ; I am inclined to think that others were to be distinguished upon the clitellum. But I did not, with a view of preserving the specimen intact, submit this part of the body to a microscopical examination. In another specimen there was an additional papilla upon xiii., a smaller median papilla upon xv., and a pair upon xvi., one on either side of the median depression. The male pores upon segment xvii. are very prominent ; each is 1895.] MB. F. E. BEDD/VBB ON NEW EAETHWORMS. 231 situated on a conspicuous papilla ; the end of the spermiducal gland appeared to be protruded, thus forming a second " papilla " of smaller size upon the first. The two orifices are near together. The internal characters of this Microscolex serve in great part to differentiate it from its allies, among which Microscolex loncjiseta is the nearest. The first septum divides segments iv.-v. ; septa ix.-xii. are thickened. There is a moderately sized gizzard in segment v. The intestine seems to begin in xvii., but in longi- tudinal sections I cannot draw a line between it and the oesophagus. The latter is straight until the end of xi. ; afterwards it is con- stricted by the septa. The sperm-sacs are in segment xi., as is so generally the case with the worms of this genus. The testes and funnels are in x. There is also a pair of sperm-sacs in ix., attached to the posterior wall of that segment and racemose, though con- taining little or no sperm. The single pair of spermathecae are in segment ix. Each is an oval pouch with a diverticulum as long as itself ; the diverti- culum is coiled in a spiral and ends in a swollen extremity. This is lined with a much-folded epithelium, so that the lumen appears to be divided by numerous radiatiug partitions. The most remarkable feature in the internal organization of the worm concerns the spermiducal glands. These are very long and of the usual tubular form. The entire efferent apparatus extends back to the xxviith segment, thus occupying ten segments. It measured 8 mm. in length. The spermiducal gland, however, did not reach so far back ; but the muscular sac containing the penial setae passed beyond the end of the spermiducal gland. The latter is coiled slightly upon itself once or twice; it opens on to the exterior by a moderately short and narrow muscular duct. The sac containing the penial setse contained two of them, which wei'e of such different lengths that ohe might speak of a dimorphism in the setae. One measured the full 8 mm., the other was not half its length. They were alike, however, in their form being without ornamentation at the free end. The sac contained an immature seta of a pale colour contrasting with the bright yellow of the mature seta, which was much longer than the smaller of the mature setae. The base of implantation of these penial setae was supplied with abundant blood-capillaries. The penial setae, the single sperm-duct, and the spermiducal gland open separately on to the exterior, though near together. (4) Microscolex michaelseni, n. sp. This is one of the most abundant species of Earthworms among the w"orms collected by Dr. Michaelsen. I have great pleasure in associating it with his name as a mark of my respect for the valuable work which he has done in this department of natural history. The present species is a very well-marked form, and it is a matter of surprise to me that it has never yet been described from 232 ME.. F. E. BEDDABD ON NEW EARTHWOEMS. [Mar. 19, this part of the world. It appears to be a very southern Patagonian form. It occurs in fresh water as well as upon dry- land, as is also the case with Acanihodrilus pictus and a few other worms. The species is long and slender, and is quite bleached by the alcohol. A typically sized specimen selected for measurement was 85 mm. long by 3 mm. broad, and consisted of 92 segments. The prostomium is continued for some way over the buccal segment by grooves, but these do not reach the end of that segment. The setae are more paired veutrally than laterally, the distance between any two lateral setse being about twice that between the ventral setae. The clitellum occupies segments xiii.-xvi. and is quite complete. The nephridiopores open in front of seta 3. The oviducal pores lie exactly in line with seta 1. The most characteristic external feature in the organization of this worm is afforded by the male pores. These pores in the present genus are, as a rule, very conspicuous, but are more or less rounded orifices confined to the xviith segment. In Microscolex michaelseni the male apertures are represented by crescentic slits, which commence upon about the middle of the xviith segment and extend back to at least the posterior extremity of that segment, and sometimes even encroach upon the xviiith segment. This groove is wider in front than behind. An investi- gation of it by means of transverse sections showed that the spermiducal glands and the penial setae opened in common at the anterior end of the groove where it is widest. The sperm-ducts open at the posterior end. In this wide separation between the external apertures of the sperm-ducts and the spermiducal glands, the present species resembles Microscolex {Wiododrilus) minutus. It will be observed also that the present species is almost an Acanthodrilid in the relations between these organs. The only actual difierence between the Acanthodrilidse and the Cryptodrilidae which is invariable, is that in the one (Acanthodrilidae) the sperm- ducts open a segment behind or in front of the spermiducal glands ; in the Cryptodrilidae, on the other hand, if the two apertures are separated, they are upon the same segment. In Microscolex michaelseni, however, in some specimens at any rate, the apertures must be at least only just not upon consecutive segments, and in other specimens the groove extended well into segment xviii. On segments xvii., xviii. are a pair of rounded papUlae which lie just behind the furrows, separating these segments from the one in front. Sometimes the anterior pair are closer together than the posterior. The alimentary canal is not provided with a well-developed gizzard ; there is a rudimentary one only in segments vi., vii. The oesophagus gradually passes into the intestine, which acquires its full calibre in the xvitb. The last pair of hearts are, as is usual 1895.] ME. F. E. BEDDABD ON XEW EAETHWOEMS. 233 with the genus, in the xiith segment. The testes iare large and bushy ; there are only a single pair of them lying in segment x. Opposite the testes are the conspicuous funnels of the sperm-ducts. I could find neither testes nor funnels in segment xi., -where, however, are a pair of large and racemose sperm-sacs. The spermiducal glands are plainly separable into a glan- dular and a non-glandular portion. The latter is rather less than one-half of the length of the former and very much more slender. The entire gland is straight and occupies about four segments. It opens in common with an opaque thick-walled sac containing penial setae. This sac is not so long as, but much stouter than, the muscular end of the spermiducal gland. The sac coritained, in a specimen dissected by me, 8 penial setae on one side of the body. Only two of these, however, appeared to be fully mature. These were to be distinguished from the rest by the fact that the implanted end was curved. In the others this end was broadened out. In the former setae the free end of the seta tapering gradually to the end was unornamented ; in another specimen, however, these setae were faintly ornamented. I am therefore disposed to think that the ornamentation upon the immature setae is a mark of their immaturity, the spines being as a rule worn off in the setae which happen to be in use. A curious difference in the constitution of the two kinds of setaj was shown by treating them with a strong solution of potash : when this reagent was applied, the presumably immature setas became almost invisible, while the others did not exhibit any changes. There is nothing worthy of comment concerning the ovaries and their ducts. The spermathecae are a single pair lym