ta Nit st OR NAY ies Halt antl ie ii ; HNN yy i} sy Coy am 3 NS halt yA 4 {2} rathie? AU), eee aN as Q eee Se, eet in ay eat ‘ ins \ ain " fit se BA ye Nya) Oh Dh Bi } > ui aun Mp i t) ! i \ oe ; tent SS SS ere as Se \ Ge iN my (i a TORS Is aah t noe easy hy ivan tga th ny y + We ea) x it 4 ntti i sats ash thtenlty oC WANA ITER NID OPIN NvE as Aur Aide } { het ay (analy wa at tis Al ane ‘ ‘ y M Y WY aati) ian WAH oh # fa Vals —~ loan VEN . \ as \ AA alle a r 7 VQ e al, / ( NAl Wie a eee | fe fe Nia Wg fy! LAS i) iP wa i \ r aa) Vaan Al) A Pp Uv \PR AP ae f’ ‘Bi am IN i- Tow Tae \ C CS eg cS we « a wf Cas { ae ; ve p : r CCC CE OO OE EE EOE COLE CO OC IO TE EER GIL OES ONE OE GE RO GE OC OE OS SENIOR FC OE EGC LE EOE EEE EOC CER EC OC LEI OE GUESS OE MTGE CLOSE NEON EC OEE” Oe CE OC EC ES SS Se Oe Crea ee CT CRS CO, a a eS ea agen © Ce IC GE GE Gh GE CTC OEE EGC & eG GA Ca OG ted ( i, & Sd CEO AN eG GE \ ak © aa G COE a SS MR Cage AM Gy TCI EE OEE (CROCE € OE CE COR OO CW EOE ENS EEOC NE (EI VDE EC CT LG i LO OEE MOOG OE OOO, OUMK OG CF EOS MR SOOO ALE EE COO OE ee CE MN we EO ES EC er CO CFV. @ UC OE OC OO TE M@ a EC Ce aac 7 oe Ae (&@ CEC (WCCO OCEOO HN Qe aC eee or 0 FO LE OME CORO OEE OEE EO 0 we EO CCE GET FOOT OEE OC OLS RE En EMG OEE COCOONS EC OT OES BC Serauc © OMA LO CG EE OQ EOE ONE EE COLE ON Ct OO EC en OE ee < Fie | OOO LE OO MEE OC CUO EE OC OO CO OCE ae GEO © Gi GLO i a I OOO ELE OC eg CCRC CE te Fe MEK OOO MO LO AOE ETL OOM COLE OE EER EOE CO BOS @ CE OE EEO EE CQ gg UE OLE O CON EEO OME (FLEE WU tg ODO Se da ee Ce, ROR, GE Ae CL EES cess. ae ( CMCC CO 1 OCMC LR OTE OG EOC OGER I EMO OREO EO GME OL EKA ON CCE Ca ge CC @ (LOCC OOO CME EO EK COOL EC O00 ECE OCC @ CEO COC CL ME Bild OAC CO UC oS BG Co CEOC 0 (CCE ~ eC aE eG, FY Bias 2 AYA a ) A ¢ € ¢ € ¢ ( C C WOE XE @ a ( & | (A \~A (aia f~ A\ . o > VN mae A a A AA Ce COC CE SE a RAE Oe CME c aan ES LORE OE EE eq GC 4 —& , > a Par 4 ? ° 7 CrC are a Ta = io ——— Me hi ac. ane S COM CGE EOE LOCI OOOO OC LEE OL dat Me er aCe, EG GC GAGT & Ci LdGKa” Cie a eee Se Cage, GA RCS GES eee SAR Ee SSCS Cal ¢ haa . eat & OOM LO (© OO] Ef COOOUOMOE CCOE O IE OEE EO ME IOC ORE STE LO Qe (le NG @ Tate Va Ka & tie (or ware Sar rir a COCad GEL ce CCQrs™ (C & ICE AEE « «Cl Ce Ee EG @ ce ie COC COCR CE CO CG (Qa a LE — et « i CEO cae ce RG " OU COE € CCG C a Ce ie oc UC OCE Lee “eietks Sonee ws Ce a GC COE C4 mt Ce Ce C&G ace < CECE CEE CL CLE ce & : Ke Ce aa« Car. ae ee (< Cae COT EA (CNG OC EE CCE « « & ( .G (C“ area CCC CO MOL KO CDC Cl Gk Cae CCE CH CE (CMS CMC CEG iS . . . « Bi scorpioides (Mont:)). i Thoracie somites produced dorsally into ean crests. . . JB. gibba (Sars). b. Lateral ridge faintly marked on carapace, not developed on GNORNAS GMS 4 6 5 6 6 he 86 6 6 ao ek 6 dab GOMORS, oe te b. Two distinct lateral ridges on carapace. . . . . =. . =. . ~ 3B. pulchella (Sars). B. Endopod of uropods unsegmented. a. Lateral ridge well-marked, extending on to thoracic somites. a. Basis of first leg (in female) 14 times as longas distal segments. B. arenosa Goodsir. b. Basis of first leg (in female) shorter than distal segments. . . JB. similis, sp. n. b. Lateral ridge faintly marked on carapace, not developed on thoracic somites. a. Antennal tooth of carapace acute. . . . . . . .- =. . « 3B, siamensis, sp. n. (reAntennaletoothvor carapace roumoded mi ee) eens enue) tl parva, sp. 0. B. pulew Zimmer has been omitted from the key as I have seen no specimens. It has the endopod of the uropods unsegmented, and the lateral ridge faintly marked on the carapace but well-developed on the free thoracic somites. In the collection of the Rev. Canon Norman I have examined four specimens labelled “ Cuma folinii Fischer” from the Bay of Biscay. I have been unable to discover whether any description of this species has been published. It is closely allied to B. arenosa, of which, perhaps, it may be only a variety. BopotTRIA SUBLEVIS, sp.n. (Plate I. figs. 1-3.) Description of adult Female.—Total length 2°6 mm. The carapace is a little more than two-sevenths of the total length. The dorsal edge, seen from the side, is nearly straight, with a slight notch at about one-third of the length from the hind margin. ‘The pseudorostrum is nearly horizontal. The dorsal keel is distinct but not prominent on the carapace and thoracic somites. On B2 4 DR. W. T. CALMAN ON NEW OR RARE the carapace the lateral keel is distinct only in its middle portion, and its posterior end curves slightly upwards; it is not developed on the first free somite, and only a slight lateral elevation indicates its place on the posterior thoracic somites; below it on the carapace is a slightly marked elevation, not forming a distinct keel, and behind the above-mentioned notch on the dorsal surface is a slight transverse elevation, not sharply defined. Antennal tooth triangular. Basis of first legs a little shorter than the remaining segments together. Uropods longer than the last two somites together, the rami two-thirds as long as the peduncle. Endopod of two segments, the proximal nearly four times as long as the distal, with five spines on its inner edge ; distal segment with two spines. Immature Male.—Total length 2°6 mm. Resembling the female, except that the carapace is still smoother, the lateral keel being the only one distinct. Occurrence. —Gulf of Siam, ‘ Koh Kam, 5 fathoms, 6/2/00.” Th. Mortensen Coll., Copenhagen Museum. Co-types in British Museum. BoporRia SIMILIS, sp. n. (Plate I. figs. 4-9.) Description of adult Female.—Total length 2-15 mm. Closely resembling in general characters B. arenosa Goodsir. Carapace less than two-sevenths of total length. Median dorsal keel rather prominent on the thoracic somites. Lateral keel well-marked and straight on carapace and on all except the last of the free thoracic somites. No distinct ridge below lateral keel on side of carapace. Antennal tooth triangular. Basis of first legs a little shorter than the remaining segments together. Uropods longer than the last two somites together, the exopod about three-fourths as long as the peduncle and a little longer than the endopod. Endopod unsegmented, with six small spines on the distal part of its inner edge, the last close to the slender terminal spine. Exopod with two unequal slender terminal spines and a series of setee on the inner edge. Adult Male.—Total length 2°8 mm. There is a distinct ridge on the side of the carapace below the lateral keel. The latter is distinct on all the free thoracic somites. Basis of first legs a little longer than the remaining segments together. Rami of uropods subequal and two-thirds as long as the peduncle. Endopod with twelve spines on the inner edge. Remarks.—Yhis species resembles very closely the 6. arenosa of British and Norwegian Seas. It is distinguished by its very much smaller size, by the shorter basis of the first legs, and by the longer rami of the uropods. Occurrence.—Gulf of Siam, “‘ Between Koh Mesan and Cape Liant” and “Tung Kaben.” Th. Mortensen Coll., Copenhagen Museum. Co-types in British Museum. CUMACEA FROM THE COPENHAGEN MUSEUM. 3) BopoTRiA SIAMENSIS, sp. n. (Plate I. figs. 10-15.) Description of adult Female.—Votal length 2:1 mm. Carapace about one-fourth of total length. Median dorsal kee! not very prominent on carapace or on thoracic somites. Lateral keel rather faintly marked on anterior part of carapace, becoming obsolete posteriorly; not continued on to the thoracic somites. Antennal tooth narrow and acute, almost spiniform. Basis of first legs about four-fifths as long as the remaining segments together. Uropods about one- fifth longer than the last two somites together, the rami sub-equal and five-sixths as long as the peduncle. Kndopod unsegmented, with nine spines on distal part of its inner edge, increasing in size distally ; terminal spine slender. Exopod with three unequal terminal spines and a series of sete on the inner edge. A transverse band of dark pigment occupies the region of the first leg-bearing somite. Adult Male.—Total length 2°75 mm. Carapace less than one-fourth of total length. Rami of uropods about three-fourths as long as the peduncle. EKndopod with ten spines along the whole length of its inner edge, increasing in size distally. Remarks.—This species resembles in many ways B. similis, in company with which it was found; but it is distinguished from that species by the lateral keel being inconspicuous on the carapace, and absent altogether from the thoracic somites. ? Occurrence.—Gulf of Siam, ‘‘ Between Koh Mesan and Cape Liant, 5-8 fathoms” ; “ Koh Kam, 5-10 fathoms.” Th. Mortensen Coll., Copenhagen Museum. Co-types in British Museum. Boporria PARVA, sp. n. (Plate I. figs. 16-18.) Description of adult Female.—Total length 1-5 mm, Carapace more than two-sevenths of total length. Dorsal edge with a depression bounded behind by a slight crescentic transverse ridge near its hinder end. ‘Two faintly marked longitudinal ridges on sides of carapace, not reaching to posterior margin nor joining the above-mentioned transverse ridge. Antennal tooth bluntly rounded. No ridges on the free thoracic somites. Basis of first legs a little shorter than the remaining segments together. Uropods slightly longer than the last two segments together, the rami subequal and rather less than two-thirds the length of the peduncle. Endopod with one spine on its inner edge about the middle of its length, aud another close to the apical spine, which is long and stout; the inner edge between the two spines is coarsely serrated. The exopod has two unequal terminal spines and no sete on its inner edge. Remarks.—In having the antennal tooth of the carapace bluntly rounded and in the armature of the uropods this minute species differs from all the other species of the i) DR. W. T. CALMAN ON NEW OR RARE genus. It appears to resemble B. pulchella in the disposition of the ridges on the carapace, though these are very obscure. Occurrence.—Gulf of Siam, ‘Koh Kam, 5 fathoms.” Th. Mortensen Coll., Copen- hagen Museum. Co-types in British Museum. Genus CYCcLASPIS. The new species described below form, with those already referred to this genus, a somewhat varied assemblage, within which, however, I am unable to define any groups which appear worthy of generic rank. The species of Cyclaspis form a very important part of the Cumacean fauna of tropical and southern seas. Key to the Species of Cyclaspis (excluding C. argus and C. bistriata Zimmer, of which I have seen no specimens *), A) Noveye oroculardobe) 55 5 09 5) ee eC Mlongacaudatas Gs On Sars: B. Eye and ocular lobe present. a. Carapace quite smooth. a. Peduncle of uropods more than twice as long as the rami.‘ C, picta Calman. 6. Peduncle of uropods not much longer than the rami. A!. Basis of first legs with a spiniform tooth at its distal end. a’. First legs long, propodus twice as long as carpus. . . C. hornelli Calman. b’. First legs short, propodus equal to carpus . . . . . C. herdmani Calman. B’. Basis of first legs without a spiniform tooth at its distal end. a’. Carapace one-third of totallength . . . . . . . C. pusilla G. O. Sars. b’. Carapace less than one-third of total length. 1. First legs very long, basis about half as long as distal Seem WoW 3 5 5 6 ob 6 ee oo a (GS Ung HOGS; Ss i 2. First legs short, basis nearly equal to distal seg- TSTMS WOON 6 5 bo 6 oo ee 8 tk ee eg Ge ML Whomegom, b. Carapace more or less strongly ridged or sculptured, or armed with a tooth on the dorsal crest. a. Ridges enclosing a quadrilateral depressed area on each side of the carapace. A!, Two strong transverse ridges crossing dorsal surface of carapace. apy Dransverse mid sessunickened emilee) a eine amen Cexsculpia G..OnSars. b’. Transverse ridges not thickened, the posterior one with two hookediteethimnmy | lien umr nn n-ne em en Cumensculpran@almant * lwo additional species of the genus have been described by Dr. Zimmer since this paper was read (Zool. Anz. xxxi. p. 367, 1907). =] CUMACEA FROM THE COPENHAGEN MUSEUM. B’. No transverse ridges on dorsal surface of carapace. a’. A prominent tooth at the posterior lower corner of the lateral enclosed area. . . C. elegans, sp. n. b!. No tooth at the posterior een corner sat the iefionell enclosed¥arcay cane i am hue okay sa . C. similis, sp. n. 6. Ridges not enclosing a quadrilater al area on the side of ‘he carapace. A'. Dorsal crest of carapace with a sharp tooth near anterior end, - Carapace with an oblique lateral ridge . . . . . . C. umiplicata, sp. n. . No ridge on carapace. . . 50 6 6 oo 6 5 Oh CBORD, JDo We BL Dorsal crest of carapace nea 1. Carapace longitudinally ribbed . . . . . . . . . GC. costata Calman. 2. Carapace with a very prominent encircling ridge cingulata, sp. D. thomsont, sp. Ni. . biplicata, sp. n. . australis G. O. Sars. anteriorly : eta ite : : 3. Carapace with a single aioe atime ridge ie ails : 4. Carapace with two subparallel oblique ridges laterally . 5. Carapace with two widely divergent oblique ridges 6. Carapace with three transverse ridges ie the dorsal surface and uniting below. . . . . . . . OC, siboge Calman. 7. Carapace with three ridges laterally, she two posterior uniting or dying out before reaching the mid-dorsal NTS Vee 7 eter: PRL OMe hts ki pets suet tates asin rik SS ers) < triplicata, sp. 0. CYCLASPIS LONGIPES, sp. n. (Plate V. figs. 1-5.) Description of adult Female.——Total length 3-1 mm. ‘The carapace is about two-sevenths of the total length, slightly compressed, its vertical height less than two-thirds of its length. The dorsal edge is slightly keeled, most distinctly so in front, and is hardly at all arched as seen from the side. Pseudo- rostrum very short, the ocular lobe reaching quite to the tip. Antennal notch shallow, widely open. Antennal tooth very sharp, almost spiniform. The ocular lobe is very broad and prominent on the dorsal surface. First leg-bearing somite well exposed. ‘The second has a slight dorsal keel. Abdomen a little longer than the cephalothoracic region ; the somites subcylindrical, with lateral articular processes. Antennules rather long, the last segment of the peduncle longer than the preceding. The first legs are very long and slender, extending beyond the tip of the pseuro- rostrum by nearly two-thirds of the length of the carapace. The basis is not much more than half of the length of the distal segments together. ‘The dactylus is shorter than the propodus and longer than the carpus. The peduncle of the uropods is one and a half times as long as the last somite and is finely serrated on the inner edge. ‘The endopod is about two-thirds as long as 8 DR. W. T. CALMAN ON NEW OR RARE the peduncle, with a slender apical spine and five small spines on the distal half of its inner edge. The exopod is one-fourth longer than the endopod, and has a slender apical spine with two spinules at its base and a single slender spine a little way down the inner edge. Adult Male.—Yotal length 3°35 mm. Resembling the female in general form. Carapace a little less deep. The ocular lobe is more swollen and the corneal facets are larger. The antennal tooth is not so sharp. ‘The first leg-bearing somite just visible between the carapace and the second somite. ‘The first legs are long and slender as in the female. The peduncle of the uropods is of the same relative length as in the female and bears a series of plumose sete on the inner edge. The rami are less unequal than in the female, the endopod being about five-sixths of the length of the peduncle and the exopod only a little longer. The endopod has ten spinules on its inner edge, and the exopod bears three plumose sete internally. emarks.—The smoothness of the carapace and the presence of five free thoracic somites bring this species into the neighbourhood of C. picta, C. herdmani, and C. hornelli Calman. While resembling the last two in general shape, it differs from them in having the basal segment of the first legs not produced into a distal tooth. C. picta differs in the acute pseudorostrum, the more arched carapace, the shorter rami of the uropods, and in the much shorter first legs. Occurrence.—“ Cruz Bay, St. Jan [Danish West Indies]. Chr. Levinsen, 6.1.96.” Two specimens. Copenhagen Museum. Cycrasris Levis G. M. Thomson. (Plate V. figs. 6-8.) Cyclaspis levis G.M. Thomson, Journ. Linn. Soc., Zool. xxiv. p. 264, pls. xvi. & xvii. figs, 1-26 (1892). Description of adult Female.—Yotal length 7:16 mm. The carapace is about two-sevenths of the total length, hardly compressed, its vertical height little more than half its length. The dorsal edge distinctly keeled ; keel flattened or faintly doubled posteriorly, very little arched as seen from the side. Pseudorostrum very short, the ocular lobe reaching quite to the tip. Antennal notch not widely open and antennal tooth not acute. Ocular lobe slightly prominent on dorsal surface. The surface of the carapace is everywhere beset with shallow and inconspicuous pits. There is a faintly marked ridge running backwards for a short distance from the antennal tooth. First leg-bearing somite exposed at the sides only. Second with a distinct dorsal keel. Abdomen a little longer than the cephalothoracic region ; the somites subcylindrical, with lateral articular processes, with a median dorsal and slight lateral keels. Antennules short; the last two segments of the peduncle equal. CUMACEA FROM THE COPENHAGEN MUSEUM. 9 First legs of moderate length, extending beyond the tip of the pseudorostrum by less than one-third of the length of the carapace. The basis is hardly shorter than the distal segments together and is not produced into a tooth, but bears two densely plumose setz at the distal end. The dactylus is a little shorter than the propodus and about equal to the carpus. The peduncle of the uropods is equal to or very little longer than the last somite. The endopod is a little shorter than the peduncle, acutely pointed at the tip, without an apical spine, with four spinules on its inner edge. Exopod a little longer than endopod, with a small apical spine and plumose setz on its inner edge. Adult Male.—Vhe single adult male in the collection is crushed and imperfect. So far as can be seen, however, it differs little from the female except in the usual secondary sexual characters. ‘The corneal facets of the eye are larger, the antennal notch of the carapace is shallower, and the antennal tooth less prominent. The uropods have the peduncle and the endopod fringed with sete on the inner edge, and the rami are subequal in length. In immature specimens of both sexes, and apparently also in the adult male, the first leg-bearing somite is quite concealed. Remarks.—This species resembles the last in having a smooth carapace, the ocular lobe reaching the tip of the pseudorostrum, and the basal segment of the first legs not produced into a distal tooth. It differs in having the first leg-bearing somite only partially exposed, the first legs much shorter, and the peduncle of the uropods hardly longer than the last somite. The identification of the specimens described above with Mr. G. M. Thomson’s species is only possible on the assumption that he was in error in describing the basis of the first legs as ending in a long spiniform process. The two long plumose sete at the end of the basis in my specimens are often encrusted with mud and matted together so as to form what appears at first sight to be a solid process having very much the shape and proportions of the process figured by Mr. Thomson. Until some other species is discovered agreeing more closely in this respect with Mr. Thomson’s figures, his name may be applied to the form here described. Occurrence.—New Zealand, “ Lyttleton Harbour, 1-5 fathoms, 8/97, H. Suter Coll. ”; «¢ Akaroa Harbour, H. Suter Coll.” Copenhagen Museum. ‘“‘ Otago Harbour, surface- net, G. M. Thomson.” British Museum. CYCLASPIS ELEGANS, sp. n. (Plate II.) Description of adult Female.—Total length 6:3 mm. Resembling C. exsculpta Sars and C. persculpta Calman in having the carapace strongly sculptured. ‘Lhe carapace is about one-third of the total length, and its vertical height is two-thirds of its length. On each side is a quadrilateral area, answering to the “lateral depressed area” in Sars’s description of C. exseulpta, bounded VOL. XVII.—PART I. No. 2.—August, 1907. Cc ia? 10 DR. W. T. CALMAN ON NEW OR RARE by ridges which rise at the corners into four prominent tubercles. There are no transverse crests across the dorsal surface, but there is a well-marked, obscurely granulated, median dorsal keel which rises at its posterior end into a blunt tooth. ‘Lhe depressed area occupies the greater part of the lateral surface ; its upper margin is sinuous and its posterior margin nearly parallel to the hinder margin of the carapace. The lower horizontal ridge is produced in front and behind to the edge of the carapace, and the upper is continued forwards on to the side of the pseudorostrum. The greatest width of the carapace is measured between the postero-inferior tubercles. The ocular lobe is little longer than broad and reaches nearly to the tip of the pseudorostrum. The eye is without pigment; there are about nine corneal facets. The first leg-bearing somite is almost entirely concealed, only a small portion being visible laterally. The second is large, nearly equal in height to the carapace, with a dorsal crest produced backwards into a sharp tooth and with the lateral plates expanded. The posterior thoracic somites have a median dorsal keel, which is continued on all the abdominal somites except the last. There are less distinct dorso- lateral keels on the last two thoracic and the first two abdominal somites. The first five abdominal somites have lateral articular processes anteriorly. The appendages are very similar to those of C. australis as figured by Sars. The antennules are a little more slender and appear to lack all trace of the inner flagellum. ‘The antenne have a long and slender external process which is not segmented off from the basal part. The mandibles have about eighteen spines on the inner edge. The lower lip has the tip of each lobe sharply bent inwards and armed with a group of peculiar spatulate spines. ‘The palp of the maxillula is about one and a half times as long as the distance between its base and the tip of the distal lobe and bears two long sete. The first maxilliped differs from that of C. australis in the longer basis, which exceeds in length the other segments together. The terminal segment is very small, The branchial apparatus is well developed; the epipod is produced forwards nearly as far as the end of the basis, and carries about thirteen branchial lobules. The second maxilliped is slender, its basis exceeding by two-thirds the length of the remaining segments together. There are about twelve long sete on the basal lamina. The third maxilliped has the basis sharply bent outwards about the middle of its length. Otherwise it resembles the corresponding limb of C. australis. The first leg does not extend much beyond the tip of the pseudorostrum in the natural position, and its distal segments are not so slender as in C. australis. The remaining legs are similar to those of C. australis, but carry longer sete. ‘The terminal segment of the second pair has three terminal but no lateral spines. The uropods have the peduncle stout and but little shorter than the last somite. The rami are subequal and about equal in length to the peduncle. The endopod CUMACEA FROM THE COPENHAGEN MUSEUM. Bt is slightly curved upwards and outwards at the tip; the distal half of its inner edge is strongly serrated and bears a single small pectinate spine. The exopod has a short apical spine and a series of plumose sete on the inner margin. Adult Male.—Total length 6-2 mm. The carapace is about two-sevenths of the total length, its vertical height not much more than one-half of its length. The dorsal outline is less strongly arched than in the female. ‘The ridges defining the depressed area on the lateral surface are less prominent. The tubercles at the two lower corners and that at the anterior upper corner are well-marked, but there is no posterior upper tubercle, and the upper horizontal ridge does not meet the posterior vertical ridge. Seen from above, the carapace is not wider behind than in front. The ocular lobe and the corneal facets which it bears are considerably larger than in the female. The antennal notch is somewhat shallower, and the antennal tooth is slightly different in form. The first leg-bearing segment is entirely concealed. ‘The second is as high as the posterior part of the carapace. Its dorsal crest is rounded, not produced, and the lateral plates are not expanded. The dorso-lateral keels of the last two thoracic and the first two abdominal somites are more strongly developed than in the female. As usual, the abdominal somites are much stouter than in the female. The fifth somite is laterally constricted in its posterior half. The antennules are similar to those of the female, with a single additional sensory filament springing from the end of the third segment. The antenne resemble those of C. longicaudata as figured by Sars, except that the distal segment of the peduncle is shorter and stouter and the sete clothing it are shorter. The branchial apparatus is more fully developed than in the female, the lobules being about seventeen in number. The pleopods appear to differ from those of C. longicaudata in lacking the process from the outer margin of the endopod. The uropods are longer than in the female. The peduncle is about as long as the last somite and is beset along the whole of its inner edge with plumose sete. ‘The rami are about equal in length to the peduncle, the inner slightly shorter, and resemble those of the female except that the endopod, as well as the exopod, has its inner edge beset with plumose sete. Young Stages—In a specimen 1:75 mm. in length (PI. II. fig. 5) there isno trace of the last pair of legs. ‘The integument is well-calcified and brittle. Nocorneal facets are visible on the ocular lobe, and the lateral articular processes of the abdominal somites are not yet developed. The first leg-bearing somite is entirely concealed. On the side of the carapace the ridges bounding the depressed area are well-marked, but of the four tubercles only the anterior upper one is prominent. Specimens about 5 mm. in length (PI. Il. fig. 6) have assumed nearly the form of the adult, but the first leg-bearing somite is still concealed. It is noteworthy that at this stage the male resembles the adult female more closely than it does when full-grown, € C2 12 DR. W. T. CALMAN ON NEW OR RARE Remarks.—As already stated, this very beautiful species, which, from the abundance of the material, I have been able to describe with some detail, resembles C. exrsculpta Sars and ©. persculpta Calman, but it is at once distinguished from them by the different sculpture of the carapace, especially by the absence of transverse crests on the dorsal surface. In the structure of the appendages it shows great similarity to C. australis Sars, while at the same time it seems to present no differences of generic value from the type of the genus, C. longicaudata Sars. Occurrence.—* Lyttleton Harbour, New Zealand, 1-5 fathoms, H. Suter Coll.” Many specimens. Copenhagen Museum. Co-types in British Museum. CYCLASPIS SIMILIS, sp. n. (Plate III. figs. 1-3.) Description of adult Female.—Total length 5°75 mm. Resembling C. elegans in shape and general proportion of the body. The lateral depressed area on the carapace is less distinctly excavated, and the only prominent tubercle is that representing the anterior upper tubercle of the species named. ‘The lower horizontal ridge is produced forwards to the edge of the carapace, but posteriorly it is continued with an even curve into the posterior vertical ridge, and does not extend to the posterior edge of the carapace. ‘The posterior vertical ridge at its upper end forks into two branches diverging at an obtuse angle. The anterior one does not reach the ridge which runs backwards from the anterior upper tubercle, so that the upper margin of the depressed area is not completely enclosed. The sides of the carapace posteriorly are more or less rugose. The first leg-bearing somite is exposed only at the side. The dorsal crest of the second is produced backwards into a large rounded lobe. The remaining somites are similar to those of C. elegans, but the dorso-lateral crests are stronger and are continued as far as the penultimate somite. The thoracic appendages, so far as they can be seen in the undissected specimen, present no conspicuous differences from those of C. elegans. The peduncle of the uropods is shorter than the last somite and is finely serrated internally. The rami are shorter than the peduncle. ‘The endopod is not curved at the tip as in C. elegans, and is less strongly serrated internally. ‘The exopod has an apical spine, and some plumose setze on its inner edge. An immature male specimen 5:25 mm. in length differs little in general characters from the female described above. The first leg-bearing somite is hidden, and the second is not produced above into a cristiform lobe. Remarks.—This species resembles C. elegans, in company with which it was found. It is distinguished, however, by the different sculpture of the carapace. Occurrence.—* Lyttleton Harbour, New Zealand, 1-5 fathoms, H. Suter Coll.” Copenhagen Museum. Co-type (young) in British Museum. CUMACEA FROM THE COPENHAGEN MUSEUM. 13 CYCLASPIS UNIPLICATA, sp. n. (Plate LV. figs. 11-20.) Description of immature Female.—Total length 4-9 mm. The carapace is distinctly less than one-third of the total length, somewhat com- pressed, its vertical height not much more than one-half of its length. ‘The dorsal edge is sharply keeled, only slightly curved as seen from the side, with a prominent forwardly directed tooth overhanging the base of the ocular lobe. On each side of the carapace is a low but sharply marked ridge meeting its fellow about the middle of the dorsal edge and running downwards and forwards in the direction of the antero- lateral corner, which, however, it does not reach. Pseudorostrum short and truncated, the ocular lobe reaching quite to the tip. Antennal notch shallow, widely open. Antennal tooth double, the true antero-lateral angle having external to and below it a sharp spiniform tooth. The ocular lobe not longer than broad ; eye pigmented, with about nine corneal lenses. The first leg-bearing somite is exposed dorsally. ‘The second somite has a slight dorsal keel. The abdomen is distinctly longer than the cephalothoracic region, subcylindrical, and rather slender. The antennule has the third segment of the peduncle longer than the second, the first distinctly longer than the other two together. The inner flagellum is distinct. The first maxilliped has the basis shorter than the distal segments together. ‘The second maxilliped is rather stout, and has also the basis shorter than the distal segments. In the third maxilliped the basis is longer than the distal segments, and its distal process is very long, reaching to the end of the carpus. The first legs are very long, extending beyond the tip of the pseudorostrum in the natural position by about two-thirds the length of the carapace. The basis is less than two-thirds of the length of the distal segments, and is produced on the lower side of the following segment into a sharp tooth. The second legs are stout and have the basis shorter than the distal segments together. In the remaining legs the basis successively diminishes in length, being longer than the distal segments in the third pair and about half their length in the fifth pair. The peduncle of the uropods is longer by one-fourth than the last somite, equal to the endopod, and shorter than the exopod. The endopod tapers to a sharp point, and has about nine short spines on its inner edge. ‘The exopod has a slender apical spine and three short spines on its inner edge. Remarks.—In the general shape of the body and in having a tooth at the distal end of the basis of the first legs this species resembles C. herdmani and C. hornelli Calman, approaching the latter species especially in the length of the first legs and in the armature of the uropods. From both it is distinguished by the oblique ridge and the dorsal tooth of the carapace. F. Miller (‘Fir Darwin, p. 54 (English ed. 14 DR. W. T. CALMAN ON NEW OR RARE p. 81), fig. 52) figures the male of a “ Bodotria” having a single tooth on the dorsal crest. It may possibly have been a species allied to the present. Occurrence —Gulf of Siam, ** Koh Kam, 5-10 fathoms. Th. Mortensen Coll.” Several specimens. Copenhagen Museum. Co-types in British Museum. ‘Two young and imperfect specimens of this species were obtained by Prof. Herdman in Ceylon, and were recorded as ‘‘ Cyclaspis sp.” in my Report on his collections (Rep. Ceylon Pearl Fisheries, Royal Society, pt. 11. 1904, p. 160). CYCLASPIS UNICORNIS, sp. n. (Plate V. figs. 9-11.) Description of immature Female.—Total length 3:2 mm. The carapace is about one-third of the total length. The dorsal edge keeled, slightly arched as seen from the side, and armed about a third of its length from the anterior end with a sharp forwardly curved tooth. Pseudorostrum slightly prominent, the ocular lobe reaching quite to the tip. Antennal notch wide; antennal tooth acute, not reaching quite as far forward as the tip of the pseudorostrum. ‘The side of the carapace is rough with small granules. The first leg-bearing somite is hidden, the second has a slight dorsal crest. The abdomen is a little longer than the cephalothoracic region. The somites are subcylindrical, with lateral articular processes. The first legs extend beyond the tip of the pseudorostrum in the natural position for less than one-third of the length of the carapace. The basis is about three-fourths of the length of the distal segments together. ‘The ischium and merus are very stout, the former with the inner edge serrated. The three distal segments are very slender ; the carpus and propodus of equal length, and each about one-half longer than the dactylus. The peduncle of the uropods is longer by about one-quarter than the last somite and has its inner edge finely serrated. ‘The endopod is about two-thirds the length of the peduncle, serrated on the inner edge, with a slender apical spine and four spinules on the inner edge. The exopod is a very little longer than the endopod, has a slender apical spine with a small spinule at its base and another a little way down on the inner side. Remarks.—In the possession of a single dorsal tooth on the carapace this species resembles the last, but it differs in the absence of the lateral ridge on the carapace and inmany other characters. There is also a certain general resemblance to Stephanomma goestt Sars. It is stated, indeed, by Sars that in that species the lateral lobes of the pseudorostrum are absent and there is no “frontal fissure”; but I am inclined to suspect that Sars has been misled by the ocular lobe reaching quite to the tip of the pseudorostrum, and that Stephanomma will be found not to differ generically from some of the species at present referred to Cyclaspis. Occurrence.—* Cruz Bay, St. Jan” (Danish West Indies). One specimen. Copen- hagen Museum. CUMACEA FROM THE COPENHAGEN MUSEUM. 15 CYCLASPIS CINGULATA, sp. n. (Plate IV. figs. 1-10.) Description of immature Female.—Total length 4-2 mm. The carapace is about two-fifths of the total length, somewhat compressed, its dorsal surface as seen from the side strongly arched, and its posterior margin sloping back- wards so as to conceal more or less the first two thoracic somites when viewed from above. It is encircled by a very prominent ridge or collar, which crosses the dorsal surface a little in front of the middle of its length and slopes a little forwards as it passes down each side. In the mid-dorsal line this collar is interrupted by a deep notch. On the dorsal surface of the carapace in the posterior third of its length is a pair of ridges, slightly diverging and becoming more strongly marked posteriorly and bearing a few scattered sete. External to and below the frontal fissure on each side is a slight prominence. ‘The pseudorostrum is short, horizontal, and the long, narrow, ocular lobe extends quite to the tip, projecting in front of the lateral lobes. The corneal lenses are 1] in number and are grouped on the distal end of the ocular lobe. The antennal notch is small, and the antennal tooth is well behind the tip of the pseudorostrum. The first leg-bearing somite is well exposed, the fourth and fifth have well-marked dorso-lateral keels, and the fifth has also a slight median keel. The abdomen is about equal in length to the cephalothoracic region and rather stout. All the abdominal somites except the last have strong dorso-lateral keels. The first four have a well- marked dorsal keel which becomes faint on the fifth. The antennule has the third segment of the peduncle longer than the second, the first about equal to the other two together. The vestigial inner flagellum is very distinct. ‘The antenna appears to lack the external process. The third maxilliped has the basis little longer than the distal segments together. Its distal process is very long, reaching nearly to the end of the carpus. The first leg is long, extending beyond the end of the pseudorostrum in the natural position by nearly the length of the last two segments. The basis is about four-fifths of the length of the distal segments together, and is a little produced on the lower side of the following segment. The remaining legs are rather short and stout. In the second pair the basis is less than two-thirds the length of the distal segments together. The peduncle of the uropods is distinctly shorter than the last somite and has a single small seta on its inner edge. The endopod is a little longer than the peduncle and distinctly longer than the exopod, tapering to a sharp point, and with three small spines about the middle of its inner edge. The exopod has an apical spine and a few very small sete on its outer and inner margins. Male.—A single adult male, which may possibly belong to this species, is unfortu- nately so much damaged that it cannot be described m detail. It agrees with the 16 DR. W. T. CALMAN ON NEW OR RARE female in the very long and narrow ocular lobe, but the carapace is smooth, with only a faint trace of the encircling ridge so conspicuous in the female and without the paired dorsal ridges posteriorly. On the lower part of the side of the carapace is a horizontal ridge, also very faint, meeting the encircling ridge anteriorly. ‘The dorso- lateral keels are well-marked on the last two thoracic somites, but the abdominal somites, which are very stout, have no distinct keels. The terminal segments of the first pair of legs are longer and more slender than in the female. Remarks.—In the sculpture of the carapace this species is quite distinct from any other, and it is not easy to see in what direction its immediate allies are to be sought. In the long and narrow ocular lobe it resembles C. exsculpta Sars and C. persculpta Calman. In the distinctness of the first leg-bearing somite it approaches especially C. costata and C. picta Calman. Occurrence.—Several localities in the Gulf of Siam. “ Koh Kam, 5-10 fathoms.” “Between Koh Mesan and Cape Liant, 5-8 fathoms.” ‘‘Tung Kaben, 6 fathoms.” Th. Mortensen Coll., Copenhagen Museum. Co-types in British Museum. CYCLASPIS THOMSON], sp. n. (Plate V. figs. 12-16.) Description of adult Female.—Total length 6 mm. The carapace is little more than one-fourth of the total length, slightly compressed, its vertical height little more than half its length. ‘The dorsal edge with a distinct keel, which is flattened or faintly doubled posteriorly, slightly arched as seen from the side, with a slight concavity at the base of the pseudorostrum. Pseudorostrum short ; the ocular lobe reaching to or a little beyond the tip of the lateral lobes. Antennal notch rather widely open; antennal tooth triangular. The surface of the carapace is everywhere beset with shallow pits, which in full-grown specimens give it a rugose appearance. On each side, a little behind the middle of its length, is a short groove, limited behind by a faintly marked ridge running obliquely downwards and backwards. A slight ridge runs backwards for a short distance from the antennal tooth. First leg-bearing somite exposed at the sides only. Second with a distinct dorsal keel. Last three with lateral keels. Abdomen a little longer than the cephalothoracic region ; the somites subcylindrical, with distinct median dorsal and lateral keels and the usual lateral articular processes. First legs of moderate length, extending beyond the tip of the pseudorostrum by little more than one-fourth of the length of the carapace. ‘The basis is a little shorter than the distal segments together and is not produced into a tooth distally ; it bears two densely plumose hairs at its distal end. The dactylus is about two-thirds as long as the propodus and three-fourths as long as the carpus. The peduncle of the uropods is longer than the last somite by nearly one-third. The endopod is a little shorter than the peduncle, acutely pointed at the tip, without an apical spine, with four spinules on its inner edge. Exopod a little longer than CUMACEA FROM THE COPENHAGEN MUSEUM. MY the endopod, with a slender apical spine and a series of plumose set on its inner edge. Adult Male.—Total length 6°75 mm, Proportions and shape of carapace much as in female, but with pseudorostrum rather less prominent and ocular lobe distinctly projecting beyond the lateral lobes. Surface of carapace smoother than in the female, the pitting less distinct and the oblique groove faintly marked. First leg-bearing somite quite concealed, fourth with a slight lateral ridge, fifth with a double lateral ridge. Abdominal somites with median dorsal but no lateral keels. Proportions of uropods much as in female, but the rami rather less unequal. Endopod with about nine spinules and a series of plumose sete on its inner edge. Peduncle with its inner edge beset with long plumose sete, Remarks.—This species is closely allied to C. levis G. M. Thomson, but it differs from the specimens which I refer to that species in the much rougher surface of the carapace and the presence of an oblique groove on each side; in aslightly different out- line of the dorsal edge of the carapace, which is distinctly concave at the base of the pseudorostrum ; and in certain trifling differences of proportion in the first legs and uropods, In the male, where the oblique grooves of the carapace are inconspicuous, there is a certain similarity of outline to Mr. Thomson’s figure of the male C. levis, suggesting that his figure may have been drawn from a specimen of the present form. Since he expressly says, however, that the surface of the carapace in C. levis is ‘ smooth, destitute of ridges or sculpturing,” I have retained his name for the closely allied species to which this description more nearly applies. Occurrence.—New Zealand, “‘ Bay of Islands, 8 fathoms.” A number of specimens, mostly immature, taken at the same time as the type specimens of C. levis, were sent to me by Mr. G. M. Thomson. CYCLASPIS BIPLICATA, sp. n. (Plate III. figs. 4-15.) Description of adult Female.—Total length 4:2 mm. The carapace is about two-sevenths of the total length, somewhat compressed, its vertical height less than two-thirds of its length. The dorsal edge, seen from the side, is only slightly arched. There is a sharp median dorsal keel which rises at the posterior end into a broad truncated tooth. On each side of the carapace are two parallel ridges, running obliquely forwards and downwards and dying out before reaching the lower edge of the carapace. At their upper ends they converge and meet each other close to the median keel a little behind the middle of the carapace. ‘These ridges, though sharply defined, are not prominent, and hardly interrupt the lateral outlines of the carapace as seen from above. ‘The ocular lobe is a little longer than VOL. Xvii.—Ppart 1. No, 3.—August, 1907. D 18 DR. W. T. CALMAN ON NEW OR RARE broad, and reaches nearly to the tip of the pseudorostrum. ‘There are about nine corneal facets, but the eye is without pigment. The first leg-bearing somite is only visible at the sides. ‘The second has the dorsal crest produced upwards into an acute tooth curving slightly forwards at the tip. The posterior thoracic and the anterior abdominal somites have a median dorsal keel becoming fainter posteriorly. The antennules have the third segment of the peduncle longer than the preceding, and the inner flagellum represented by a minute vestige. The antenne have the outer process short and distinctly segmented off from the basal part. ‘The mouth-parts resemble very closely those of C. elegans. The first legs extend well beyond the tip of the pseudorostrum in the natural position. The basis exceeds only by about one- fifth the combined length of the distal segments. The terminal segment is equal in length to the preceding. ‘The remaining Jegs are similar to those of C. elegans. The peduncle of the uropods is a little longer than the last somite and bears five plumose setee on its inner edge. The rami are subequal and are about equal in length to the peduncle. The endopod has the inner edge serrated and carrying a single small seta near the tip. The exopod has a slender apical spine and several plumose sete on its inner edge. Adult Male.—Total length 4:16 mm. The carapace is less deep than in the female and more compressed. The tooth at the posterior end of the dorsal crest is low and rounded. ‘The oblique ridges on the sides of the carapace are similar in position to those of the female, but less strongly marked and do not meet above. ‘The antennal notch is more widely open and the tooth defining it is more obtuse than in the female. The first leg-bearing somite is hidden. ‘The second is produced dorsally into an acute forwardly curved tooth, more slender than that of the female. The first legs have the basis longer by one-half than the distal segments together. The uropods do not differ in their proportions from those of the female, but the seta on the inner edge of the peduncle are much more numerous and the inner edge of the endopod is also fringed with sete. Remarks.—This species resembles C. bistriata Zimmer (Zool. Jahrb., Abth. Syst. xvii. p-. 447, 1902), but appears to differ in the outline of the carapace, which has the dorsal edge less curved, in the position of the lateral ridges, and in the longer peduncle of the uropods, which in the species named is shorter than the rami. Except for the absence of the lateral ridges of the carapace, Zimmer's figure of the male C. argus (¢. c. p. 449, fig. A) resembles very closely the male of the present species. Without a re-examination of the type specimens it seems impossible to decide whether one or both of these species may not be identical with that here described. CUMACEA FROM THE COPENHAGEN MUSEUM. 19 Occurrence.—‘“ Lyttleton Harbour, New Zealand, 1-5 fathoms, H. Suter Coll.” Numerous specimens. Copenhagen Museum. Co-types in British Museum. CYCLASPIS TRIPLICATA, sp.n. (Plate III. figs. 16-24.) Description of adult Female.—Total length 4:0 mm. — Very similar in general form to C. biplicata. The carapace is more inflated poste- riorly. ‘The two oblique ridges on the side of the carapace have much the same course as in that species, but are much more prominent, while in front of these is a prominent tubercle occupying about the position of the anterior upper tubercle in C. elegans, from which a third nearly vertical ridge runs downwards and joins a prolongation of the anterior oblique ridge which turns horizontally forwards to the anterior lower edge of the carapace. ‘The tooth at the posterior end of the dorsal keel is slightly different in shape from that of C. biplicata and has the hinder edge finely serrated. The dorsal crest of the second leg-bearing somite differs from that of C. biplicata in having the upper angle blunt, the oblique dorsal edge sinuous, and the posterior angle prominent and overhanging the third somite. The median dorsal keel is very slightly marked on the posterior thoracic and anterior abdominal somites, but a pair of dorso- lateral keels, not seen in C. biplicata, are developed on the last two thoracic and the first three abdominal somites. The antennules have the third segment of the peduncle a little shorter and stouter than in C. biplicata. The first legs are shorter than in C. biplicata. The basis exceeds by one-half the combined length of the distal segments. The posterior legs have longer and more numerous setee. The uropods are similar to those of C. biplicata, but are a little stouter. Adult Male.—Total length 4:0 mm. The male closely resembles that of C. biplicata, from which, however, it is at once distinguished by the presence of the additional anterior ridge on the carapace. As in the female, this ridge rises above into a blunt tubercle, which is very conspicuous when the carapace is viewed from above. In this species, moreover, the last two thoracic and the first abdominal somites bear strongly marked dorso-lateral ridges, the dorsal tooth of the second leg-bearing somite is shorter and more curved, and the first legs are rather shorter than in C. biplicata. Remarks.—Both in general aspect and in the details of its appendages this species resembles very closely that last described, in company with which it occurred. Occurrence.—< Lyttleton Harbour, New Zealand, 1-5 fathoms, H. Suter Coll.” Many specimens. Copenhagen Museum. Co-types in British Museum. D2 20 DR. W. T. CALMAN ON NEW OR RARE Genus EocuMa. I have discussed the characters and the limits of this genus in a former paper (Herdman’s Rep. Ceylon Pearl Fisheries, Royal Society, pt. ii. (1904), Suppl. Rep. xii. Cumacea, p. 160). The material now examined reveals the existence of a number of closely allied species. Unfortunately they are represented, for the most part, by immature and sometimes damaged specimens, and it is only after some hesitation that I describe the four new species which follow, leaving aside for the present solitary and more or less imperfect specimens of three species, which may be distinct, from Aden, Trincomalee, and Penang respectively. ‘The new species have the lateral margins of the carapace carinated and the pseudo- rostrum projecting in front of the lateral cornua, and they therefore come under the division A.1 in the key to the genus which I gave in my previous paper (¢.¢. p. 161). This part of the key may be extended as follows to include them :— a. Antero-lateral teeth of carapace low and rounded (female) or absent (male) ; no paired ridges on dorsal surface. a’. (Male.) Cornua of carapace small, directed forwards; antero-lateral margin strongly convex ; peduncle of antennule not elongated. 4. taprobanica Calman. b’. (Male.) Cornua of carapace prominent, directed obliquely out- wards ; antero-lateral margin only slightly convex ; peduncle Of antennule) muchyelongated: 5 2 5) 5 i) Sy aE ongiconntsssp- sr 6. Antero-lateral teeth more or less prominent, acute. a’, Carapace nearly as broad as long, with paired dorsal ridges; autero-lateral teeth reaching nearly or quite as far forward as the pseudorostrum, a. Frontal margin with a small tooth on each side between the pseudorostral lobe and the antero-lateral tooth . . . . LE, stellifera, sp. n. 6’. Frontal margin without a tooth between the pseudorostral pte and the antero-lateral tooth. . Antero-lateral teeth acutely produced, directed ae OUMVEWES o 6 5 0 : . . . EH. hilgendorfi Marcusen. 6'’, Antero-lateral teeth eee onenlan aah acute spini- ; form points directed forwards. . . . . 0 E. lata, sp. n b’. Carapace hardly more than half as br oad as long, Pane all dorsal ridges ; antero-lateral teeth not reaching nearly as far fonwardias) pseudorostnumay jig.) er appellee Medan cuneate eM i ROCucLO mS em. ~ KocUMA LONGICORNIS, sp.n. (Plate VI. figs. 1-6.) Description of adult Male.—Total length 7°65 mm. Carapace little more than one-quarter of the total length, its greatest width, CUMACEA FROM THE COPENHAGEN MUSEUM. 21 excluding the lateral cornua, a little over half its length. Lateral keels well marked. Cornua prominent, directed obliquely outwards, situated behind the anterior third of the length of the carapace. The postero-lateral margins are nearly straight. The width of the posterior margin is about half of that measured between the bases of the cornua. ‘The pseudorostrum is prominent and formed of two rounded lobes. Antero- lateral margins between pseudorostrum and cornua very slightly convex. The trans- verse ridge on the lower surface on each side is faintly marked and does not reach the lateral margin, Dorsal surface with a faintly marked median ridge posteriorly. The ocular lobe is twice as broad as long, and the pseudorostral plates meet in front of it for a distance much greater than its length. ‘The eye is not pigmented, but there are three large indistinctly defined corneal areas. ‘The surface everywhere shows a regularly reticulated texture, which is not interrupted by a faint and inconspicuous pitting. Abdomen less than one-third longer than cephalothoracic region, scarcely narrowed posteriorly. Antennules with very long and slender peduncle ; third segment one and a half times as long as the second and a little longer than the first, about six times as long as broad. External flagellum of three segments, about two-fifths as long as last segment of peduncle, First legs long and slender. Basis little more than two-thirds of length of distal segments together. Carpus about equal to the propodus and nearly half as long again as the dactylus. The basis of the second legs has two or three rather coarse serrations on its inner edge near the distal end. The uropods are more than one and a half times the length of the last somite. The peduncle is about two-fifths of the length of the subequal rami. ‘The inner edges of the peduncle and endopod and both edges of the exopod are fringed with plumose sete and the endopod also bears three small spinules. Remarks.—The only species of Hocuma of which the male has been described is E. taprobanica. From that species the present form is distinguished by the more prominent frontal lobes, the less convex antero-lateral margins, the laterally projecting cornua, the much longer and more slender antennular peduncle, and other characters. It is unlikely that it will prove to be the male of any of the other known species, since, even if the absence of antero-lateral teeth should prove to be only a sexual character, the very narrow carapace and projecting lateral cornua, together with the long first legs, seem sufficient to distinguish it. Occurrence.—* Suez, 31/1/98, H. Mortensen.” Two specimens. Copenhagen Museum. LO iS) DR. W. T. CALMAN ON NEW OR RARE Eocuma tata, sp.n. (Plate VI. figs. 7-12.) Description of immature Female.—Total length 3-9 mm. The carapace is less than two-sevenths of the total length, very broad and flattened, with well-marked lateral carine. ‘The greatest width, across the lateral cornua a little behind the anterior third, is very little less than the length. The lateral cornua are short, with acute tips directed forwards. ‘The postero-lateral edges are nearly straight. The width of the posterior margin is less than half that measured across the cornua, The pseudorostrum is prominent, and formed, as seen from above, of two rounded lobes. External to these on each side the antero-lateral margin forms an obtusely triangular tooth with an acute spiniform point, turned forwards, not reaching so far as the level of the pseudorostral lobes. Between this antero-lateral tooth and the lateral cornu the margin is nearly straight. ‘The dorsal surface is slightly arched and has no distinct median keel, but a pair of well-marked dorso-lateral keels running from about the posterior ends of the frontal suture to the hind margin. On the under- side a transverse ridge runs from a little behind the lateral cornu to the free margin of the carapace. ‘The ocular lobe is not broader than long and the pseudorostral plates meet in front of it for a distance greater than its length. The eye is not pigmented and no corneal facets were observed. The surface of the carapace is beset with shallow pits which interrupt as clear spots the minute reticulate texture of the exoskeleton. The centre of each pit is occupied by a minute granule (or perhaps a pore). The slender abdomen is longer by more than one-fourth than the cephalothoracic region. The fifth somite is about three and a half times as long as broad. The last somite is depressed and expanded laterally, being, at its posterior end, one and a half times as broad as the preceding somite. The antennules have the first segment of the peduncle longer than the other two together, the third not quite half as long again as the second and less than three times as long as broad. The distal segments of the first legs are together about one and three-quarters as long as the basis. The carpus is equal to the propodus and nearly one-third longer than the dactylus. The uropods are about two-thirds longer than the last somite. The peduncle is a little more than one-third of the length of the subequal rami. There is a plumose seta on the inner edge of the peduncle and another, along with three spinules, on the inner edge of the endopod. ftemarks.—The specimens here described are very immature as compared with the sub-adult specimens of E. taprobanica formerly described. This is shown especially by the characters of the branchial apparatus, which has only about eleven lobules, becoming very small posteriorly, and probably also by the fact that there is no spine on the ischium of the second legs. Further, the outline of the carapace has a distinct resemblance to CUMACEA FROM THE COPENHAGEN MUSEUM. 23 that of the young specimens of E. taprobanica (t. c. pl. i. fig. 5). I believe, however, that the present species is sufficiently distinguished by the greater breadth of the carapace and by the presence of two dorso-lateral keels, since these characters are not likely to be greatly modified in the adult. LH. hilgendorfi, which also possesses dorso- lateral keels, can hardly be the adult stage of the present form, since it has the antero- lateral teeth greatly produced, while, from the analogy of E. taprobanica, we should expect these teeth to become less prominent as development proceeds. Occurrence.—Gulf of Siam, “ Koh Kam, 5-10 fathoms,” “ Between Koh Mesan and Cape Liant, 5-8 fathoms.” Th. Mortensen Coll., Copenhagen Museum. Co-types in British Museum. EOCUMA STELLIFERA, sp.n. (Plate VI. figs. 13-17.) Description of immature Female.—Total length 5°6 mm. Carapace little more than one-quarter of the total length, very broad and flattened, with well-marked lateral carine. ‘The greatest width, across the lateral cornua at about the anterior third of the carapace, is about eight-ninths of the length. The lateral cornua are rather stout and incurved at the tips. The postero-lateral edges are slightly concave. ‘The width of the posterior margin is about two-fifths of that measured across the cornua. The pseudorostral lobes are not very prominent; each is rounded at its inner end, then concave, and what may be regarded as its outer end forms a low rounded tooth over the base of the antennule. External to this, and separated from it by a concavity of the margin, is the antero-lateral tooth, which is acute and produced for- wards as far as the level of the frontal lobes. Between the antero-lateral tooth and the lateral cornu the margin is nearly straight. The dorsal surface is slightly arched and has a slight median keel posteriorly. Paired dorso-lateral keels are also present, but are not so marked as in H. data. On the underside the transverse ridge is much as in H. lata. ‘The ocular lobe is broader than long and the pseudorostral plates meet in front of it for a distance greater than its length. ‘The eye is not pigmented and no corneal facets were observed. The surface of the carapace is closely beset with shallow pits, and the meshes of the primary reticulated texture of the exoskeleton are arranged in a more or less regular radial manner around each pit. Under a moderate magnifi- cation the carapace appears covered with stellate clear spots. The proportions of the abdomen are much as in F. ata. The antennules have the first segment of the peduncle Jonger than the other two together and the third longer than the second. The first legs are very long and slender. ‘The distal segments are about two and a half times as long as the basis. The latter has its distal process very long and acute and directed obliquely inwards. The carpus is one-quarter longer than the propodus and more than twice as long as the dactylus. The uropods are little more than one-third longer than the last somite. The peduncle 24 DR. W. T. CALMAN ON NEW OR RARE is a little more than one-third of the length of the subequal rami. There are three plumose sete and three spinules on the inner edge of the endopod and a minute seta on the inner edge of the peduncle. A younger specimen, 3°7 mm. in length, agrees with that described above in the form of the carapace. Remarks.—The specimens described above, though immature, differ so much from the species hitherto described in the shape of the frontal region of the carapace, in the great length of the first pair of legs, and in the stellate texture of the integument, that there can be no doubt of their specific distinctness. Occurrence. —Gulf of Siam, “ Between Koh Mesan and Koh Chuen, 15 fathoms,” and “ Between Koh Mesan and Cape Liant, 8 fathoms.” Th. Mortensen Coll., Copen- bagen Museum. Eocuma propucta, sp. n. (Plate VI. figs. 18-20.) Description of immature Female.—TVotal length 4:6 mm. Carapace a little over one-quarter of the total length, flattened and with well-marked lateral carine. The greatest width, across the lateral cornua at a little behind the anterior third of the carapace, is a little more than half the length. ‘The lateral cornua are not very prominent, with the tips directed forwards, ‘The postero-lateral edges are slightly convex. The width of the posterior margin is nearly two-thirds of that measured across the cornua. The pseudorostral lobes are rounded and very prominent, projecting far in front of the antero-lateral teeth, which are small and blunt. Between the antero-lateral tooth and the lateral cornu the margin is straight. The dorsal surface is slightly arched and is without distinct median or dorso-lateral keels. The transverse ridge on the underside is faintly marked. The ocular lobe is longer than broad and the pseudorostral plates meet in front of it for a distance greater than its length. The eye is not pigmented. The integument is thin and transparent and minutely reticulated. The antennules were not dissected out, but they have the peduncle rather elongate, the last segment being about four times as long as broad. The first legs are rather short, extending but little beyond the pseudorostrum., The basis is four-fifths of the length of the distal segments together, The carpusis a little shorter than the propodus and a little longer than the dactylus. Tbe basis and ischium of the second legs have several teeth on the inner edge. The uropods are longer by nearly two-thirds than the last somite. The peduncle is less than one-third of the length of the subequalrami. ‘The inner edges of the peduncle and of the proximal half of the endopod are beset with plumose setz and the distal half of the endopod carries two spinules. Remarks.---Yhe specimen described above is distinguished from all the species of which the females are known by the narrow carapace and the very prominent pseudo- rostrum. In general shape it shows some resemblance to the much larger male CUMACHA FROM THE COPENHAGEN MUSEUM. 25 specimen described above as LH. longicornis, and it also approaches that species in the elongate peduncle of the antennule. It is distinguished, however, not only by the marked antero-lateral teeth, but by the shorter first legs. Occurrence.—“ Penang, Didrichsen.” Copenhagen Museum. Genus ZYGOSIPHON, gen. nov. Carapace with double pseudorostral projection and two widely separated and very long branchial siphons. Five leg-bearing somites distinct. Basis of third maxillipeds produced distally. Only the first pair of pereeopods with exopods. Endopod of uropods of two segments, distal segment short. Type species Z. mortensen?, sp. n. This new genus is distinguished chiefly by the remarkable development of the branchial siphons and their separation from one another, and by the correlated peculiarities in the shape of the anterior part of the carapace. In other respects it does not differ very strikingly from some of the existing genera of Bodotriidee. ZYGOSIPHON MORTENSENI, sp. n. (Plate VIL. figs. 1-19.) Description of adult Female.—Total length 2:67 mm. The carapace is about two-sevenths of the total length, somewhat compressed, its vertical height two-thirds of its length. The dorsal edge is slightly arched and rises in a rounded transverse ridge posteriorly. Viewed from above, the carapace is rather broader in front than behind, with nearly straight or, in ovigerous specimens, concave sides. In front it is squarely truncate, having at the corners, which are somewhat produced, the two widely separated branchial orifices from which are protruded the very long branchial siphons directed obliquely upwards and outwards. ‘There are thus two short pseudorostral processes instead of one. Seen from the side they are obliquely truncated, with the lower corner projecting in front of the upper. The lateral pseudo- rostral plates meet for a short distance in front of the ocular lobe. The antennal notch is rounded and widely open, and the antennal tooth is triangular and rather prominent. The ocular lobe is very broad, with its anterior margin notched so as to indicate a division into two parts. It contains two separate masses of ocular pigment, but there are no distinct corneal lenses. On each side of the carapace is a lozenge- shaped depression bounded by indistinctly marked ridges. From its anterior corner a ridge runs forward on to the side of the pseudorostrum, and above this the surface is slightly excavated on each side of the middle line. The first leg-bearing somite is completely exposed, and, like the second, is nearly equal in height to the posterior part of the carapace. ‘The three posterior somites are much lower and diminish rapidly in width posteriorly. In ovigerous specimens the greatest width of the body is reached at the second free somite. VoL. XviiI.—ParT I. No. 4.—August, 1907. E 26 DR. W. T. CALMAN ON NEW OR RARE The abdomen is longer than the cephalothoracic region, the somites subcylindrical and having well-developed lateral articular processes. Antennule having the first segment of the peduncle longer than the other two together, the third twice as long as thesecond. ‘The accessory flagellum is represented by a minute nodule; the outer flagellum has two segments. Antenne simple, un- segmented and without a terminal process, bearing two plumose sete. Mandible of the usual form, bearing about twelve spines. Lower lip not seen. Maxillular palp longer than the distance from its base to tip of distal lobe, with two terminal sete. The first maxilliped has the basis much shorter than the remaining segments together. The posterior portion of the branchial apparatus proved very difficult to isolate, but its lobules are very slightly developed, only two small papille being observed. The anterior portion, however, is remarkably developed, its distal part forming a long siphon capable of extension and retraction. ‘The proximal part or stalk is, as usual, strengthened by a chitinous rod, with which is connected distally a spoon- shaped plate acting as a valve to close the branchial aperture. Beyond this the tubular portion is formed by a broad and very long strip of transparent membrane rolled up into a spiral with numerous ccils “telescoped” into each other. This membrane is marked along its length by parallel striz or plications, which form a complex spiral pattern when it is rolled up. When fully extended the siphon projects beyond the branchial aperture for a distance equal to at least twice the width of the anterior part of the carapace ; when the coils are closed together it forms a short cone protruding from the opening. Very often in the preserved specimens the siphons are unequally extended on the two sides. The second maxillipeds have the basis shorter than the remaining segments together. The basal plate bears four long sete on its distal edge and one short seta externally. The third maxilliped has the basis rather sharply bent, the proximal part being about half as long again as the distal measured along the inner edge. It is produced externally beyond the articulation of the ischium into a pointed lobe, which is nearly two-thirds as long as the segment itself and which bears a row of plumose sete on its inner edge and another on its lower surface. The merus is about half as long as the ischium along its inner edge and is expanded externally into a very large curved lobe, bearing a long feathered seta at its tip. ‘The terminal segment is very small, little more than half the length of the preceding. The first legs have the penultimate segment extending a little beyond the pseudo- rostrum. The basis is about two-thirds as long as the remaining segments together, with a small tooth at the distal end of its outer edge and a long plumose seta internal to it. The terminal segment is about half the length of the preceding. CUMACEA FROM THE COPENHAGEN MUSEUM. 27 The second legs are stout and not shorter than the succeeding pair. They have all seven segments distinct. The basis is a little shorter than the remaining segments together. The carpus and propodus are short and subequal, and together equal the ischium in length. The terminal segment is longer than the ischium, about three times as long as broad, and armed distally with three long and some shorter spines. The remaining pairs of legs diminish in length posteriorly, the basis of the last pair being not much more than half as long as that of the third pair. The uropods are slender, the peduncle slightly curved and longer by one-third than the last somite. The exopod is nearly three-fifths of the length of the peduncle and terminates in two unequal spines; its inner edge bears one or two long feathered sete. The endopod is three-fourths as long as the exopod, composed of two segments, the distal little more than one-third of the length of the proximal ; it terminates in a long spine and has three short spines on its inner edge. Adult Male.—Total length 2°7 mm. The carapace is about one-fourth of the total length, its vertical height less than two-thirds of its length. Seen from the side it differs from that of the female in having the dorsal outline less strongly arched and not elevated at the posterior end. The ridges on the side of the carapace are only faintly indicated and the surface is smoother than in the female. ‘The ocular lobe is much larger and more prominent. The eye is deeply pigmented and not completely separated into two portions ; nine large corneal lenses are visible from above. The first and second leg-bearing somites are subequal in height and much lower than the posterior part of the carapace. The abdominal somites are much stouter than in the female, with well-marked pleural plates fringed with rather long hairs posteriorly. The antenna is of the usual structure and its flagellum is equal in length to the body. The branchial apparatus differs greatly from that of the female, having about nine very broad lamellar lobules, increasing in size posteriorly and each truncate at the tip. The branchial siphon is similar to that of the female. The first leg has the basis longer than in the female, about equal in length to the remaining segments together. The pleopods are of the usual structure. The peduncle of the uropods is a little shorter than in the female and the exopod is a little more than three-fifths of its length. The inner edge of the peduncle has a fringe of feathered setee. The endopod has six spines on the inner edge. Remarks.—The two long branchial siphons projecting from the anterior corners of the head give this species a very remarkable appearance and distinguish it at once from any Cumacean hitherto described. Apart from the branchial system, however, its structure does not seem to be in any way unusual. The separation of the two branchial siphons, though nowhere else so marked, will, I E2 28 DR. W. T. CALMAN ON NEW OR RARE believe, be found to be not uncommon among the Cumacea. Sars describes (Crust. Norway, ili. p. 6) the anterior part (exopod) of the branchial apparatus as uniting with its fellow of the opposite side to form “a funnel-shaped tube,” and in his description of Nannastacus longirostris (Arch. Math. Naturvid. iv. p. 122) he states that this species differs from all other Cumacea in having two distinct siphons. This is a point which is very difficult to determine in preserved specimens, especially if the dissection has to be directed to the examination of other characters as well, but I have observed it in Cumacea belonging to widely distinct families and am inclined to suspect that it may even prove to be the rule. I hope to return to this point on a future occasion. Occurrence.—Gulf of Siam, “‘ Between Koh Mesan and Cape Liant, 5-9 fathoms,” “ Koh Kam, 5-10 fathoms.” ‘Th. Mortensen Coll., Copenhagen Museum. Co-types in British Museum. A solitary young specimen was obtained by Prof. Herdman in the Gulf of Manaar and is referred to in my Report on his collection (Rep. Ceylon Pearl Fisheries, Royal Society, pt. ii. (1904), Suppl. Rep. xii. Cumacea, p. 160) as “ Bodotriide un. g. and sp.” IPHINOE sp. Two immature specimens of a species of this genus, the larger only 4:4 mm. in length, are in the collection from the Gulf of Siam. In having the carapace more than twice as long as deep they differ from all the species hitherto described except J. serrata Norman (non Sars) and J. brevipes Hansen. The latter has the carapace three times as long as deep, and differs widely in many other characters. With J. serrata the present specimens agree in having a series of teeth on the outer edge of the basis of the first pair of legs. In co-typical specimens of Norman’s species the carapace is two and a half times as long as deep, but in a specimen from the Mediterranean which I refer to this species the proportion approaches that of the Siamese specimens, where it is about two and a fifth. Other characters, such as the longer last segment of the antennular peduncle and the shorter and stouter uropods, may perhaps be due to the immaturity of these very small specimens, to which, for the present, I prefer not to assign a specific name. Occurrence.—Gulf of Siam, “ Koh Kam, 10 fathoms, 4/2/00,” “‘ Between Koh Mesan and Cape Liant, 5-8 fathoms, 7/2/00.” Th. Mortensen Coll., Copenhagen Museum. Family VAUNTOMPSONIID. Dr. H. J. Hansen has suggested (Isopoden, Cumaceen, &c. der Plankton-Expedition, p. 57) that this family ought probably to be united with the preceding. I have also discussed some facts pointing in the same direction (Cumacea of Siboga Exp. p. 8). For convenience, however, the name is here retained pending a general reconsideration of the classification of the Cumacea. CUMACEA FROM THE COPENHAGEN MUSEUM. 29 VAUNTOMPSONIA CRISTATA Spence Bate. Vauntompsonia cristata G. O. Sars, Arch. Math. Naturvid. iv. p. 13, pls. xxiii—xxvi. (1879); Calman, Fisheries, Ireland, Sci. Invest. 1904, i. (1905), p. 17, pl. i. fig. 1. To this species I refer, not without some hesitation, a number of male specimens from the West Indies. They differ from British and Mediterranean specimens in their much smaller size (not exceeding 3, as against 5 mm.), but no differences of structure can be detected by a careful examination except that the peduncle of the uropods is relatively shorter and thicker, being shorter than the exopod and less than four times as long as thick. Occurrence.—“ Cruz Bay, St. Jan” (Danish West Indies), “ Chr, Levinsen, 6/1/96.” About twenty specimens, all males. Copenhagen Museum. VAUNTOMPSONIA ARABICA, sp. n. (Plate VII. figs. 20-24.) Description of adult Male.—Total length 3:1 mm. The carapace is distinctly more than one-fourth of the total length, its vertical height about two-thirds of its length. Seen from the side the dorsal edge is distinctly arched. Pseudorostrum very short and truncated. Anterior edge without teeth. Antero-lateral angle obtuse, with a single inconspicuous denticle. The anterior lobe of the fifth leg-bearing somite is broader, and the notch defining it above is less distinct than in V. cristata. The abdomen is about equal in length to the cephalothoracic region and is stouter than in V. cristata. The antennules have the peduncle shorter and stouter than in V. cristata, the second segment is without the stout sete present in the male of that species, and the third segment is longer than the second. The third maxilliped has the basis without serrations on its inner edge. The first legs are considerably shorter than in /. cristata, extending beyond the pseudorostrum by little more than the length of the terminal segment. The basis is two-thirds of the length of the distal segments together. ‘The dactylus is two-thirds — as long as the carpus and little more than half the length of the propodus. The second legs have the basis about equal to the distal segments; the last segment is less than four times as long as broad and not longer than the two preceding segments. ‘The posterior legs, and especially their basal segments, are relatively shorter than in V. cristata. The uropods have the peduncle equal in length to the terminal somite and rather less than four times as long as thick, with eight subequal spinules and a longer distal one on its inner edge. The exopod is equal to the peduncle, with three minute spines on its outer edge, one long and two short spines terminally and one small spine on the inner edge. ‘The endopod extends beyond the exopod by half the length of its distal 30 DR. W. T. CALMAN ON NEW OR RARE segment. The proximal segment has about ten subequal spines and a stronger distal one on its inner edge, and a minute distal spinule externally. The distal segment has one long and two short terminal spines and a small spinule on the inner edge. The pigmentation is variable; patches are generally present on the back and sides of the carapace, and some specimens have dendritic spots on the sides of the abdominal somites. Remarks.—This species is very closely allied to V. cristata, but it appears to differ in a number of small characters, of which the convex dorsal edge of the carapace and the shorter first legs are the most conspicuous. Occurrence.—* Suez Canal, 31/1/98, H. Mortensen”; “ Aden, 2/98, H. Mortensen.” Copenhagen Museum. Co-types in British Museum. LEPTOCUMA KINBERGI G. O. Sars. Leptocuma kinbergii G. O. Sars, Kongl. Svenska Vet.-Akad. Handl. ii. no. 5, p. 24, pl. vi. (1873). The specimen which I refer to this species is a female with developing but empty brood-pouch. It measures 22°5 mm. in length of body, and is thus more than half as long again as Sars’s immature specimen. It differs from the latter in the following particulars :— The carapace is less deep, its vertical depth being less than two-thirds of its length, its lower edge is less convex, and the antero-lateral margin as seen from the side is proportionately deeper. The ocular lobe reaches quite to the tip of the pseudorostrum, and the dorsal edge of each lateral plate, where it lies against the ocular lobe, is convex instead of concave. ‘The antennal notch is much shallower than in Sars’s figure. The pleural plates of the third free thoracic somite are produced backwards on each side as large rounded lobes, and those of the fourth somite are similarly but less strongly produced, while the last thoracic and the first four abdominal somites have the lower hinder corner slightly produced. The appendages, so far as they are preserved and can be seen without dissection, agree in general with those of Sars’s specimen. The second leg, however, appears to be longer, reaching as far as the anterior margin of the carapace in the natural position. The posterior legs are somewhat shorter and more robust. The uropods differ from the figure given by Sars in having sete on the outer edge of both segments of the endopod, and in having the distal segment little shorter than the proximal, while in Sars’s figure it is about two- thirds of that length. In view of the much smaller size and the immaturity of the type-specimen, the characters in which it differs from the present specimen cannot be regarded as of specific value. Occurrence.— Magelhaés Streedet, Schythe.” Copenhagen Museum. 1 specimen. CUMACEA FROM THE COPENHAGEN MUSEUM. oi Family LEUCONID. The new genera which are described below render necessary some important modifications in the definition of this family. While all the forms hitherto described agree in having exopods on the first three pairs of legs in the female and on the first four in the male, and in having two pairs of pleopods in the latter sex, the species now described from New Zealand show that the thoracic exopods may be reduced to two pairs in both sexes, and the male pleopods may be present asa single pair or altogether absent. Further, the division of the endopod of the uropods into two segments may be indistinct or suppressed. The fact that, in some of the species, the ischium of the second pair of legs is distinctly developed (as in Zimmer’s Pseudolewcon) cannot be made use of as a generic character, since Heteroleucon akaroénsis and Leucon (?) heterostylis present an inter- mediate condition in which the segment in question is developed only as a very narrow and incomplete ring of chitin between the basis and merus. I find the same structure in Leucon nasica Kréyer, L. assimilis Sars, and L. longirostris Sars, and possibly this vestigial segment has been overlooked in other members of the genus. I cannot identify it, however, in the species of Hudorella and Hudorellopsis which I have examined. It is noteworthy that the disappearance of the segment is thus shown to be due to “ excalation,” not, as Sars says (Crust. Norway, ili. p. 29), to fusion with the merus. In other respects, especially in the structure of the mouth-parts, the new forms agree with the Leuconide already known. ‘The inconvenience of genera distinguished only by the characters of one sex is sufficiently obvious, and indeed, in the case of one of the species described below, I have been unable to decide as to its proper generic position owing to the fact that no male specimens were found. Apart, however, from the frequent use of such distinctions in other groups of animals, a precedent is afforded among the Cumacea by Sars’s genus Hemitlamprops, which is distinguished trom Lamprops only by the characters of the male. The following key will serve to indicate the position which the new genera occupy relatively to the other genera of the family :— A. First three pairs of legs in female and first four in male carry exopods. a. Two pairs of pleopods in male. a. Autennules not geniculate . . . : Leucon Kroyer. }. Antennules geniculate between en aa third ne of josstm@le 5 5 6 «© a). fone ede . . . +» Hudorella Norman. c. Antennules geniculate betveen first and eect Saenee of peduncle. a’. Pseudorostrum obsolete or short, vertical, and truncated . . Hudorellopsis Sars. b’. Pseudorostrum long, oblique, and acute. . . . . . . . Pseudolewcon Zimmer. 32 DR. W. T. CALMAN ON NEW OR RARE b. One pair of pleopodsin male. . . . . . . +... « « Paraleucon, g. n. c. No pleopodsinmale. . . . Ce eemelemiencon rex B. First two pairs of legs only carry erent: in ott sexes. No pleopods SrallriP NIC.) RMON CSO lo care CVO eh dic savior eh (0). 0» \ 4 KANATROUGOICGNDS, (xo Ie Lrvucon (2) HETBROSTYLIS, sp. n. (Plate VIII. figs. 1-5.) Description of adult Female.—Total length 3°42 mm. Carapace a little over one-fifth of total length, its vertical height less than two-thirds of its length. ‘The dorsal edge is nearly straight, coarsely and somewhat irregularly serrated in the anterior half, with a small denticle near the posterior end and in front of it a shallow depression. Pseudorostrum straight, horizontal, acute, measuring along its upper edge about one-fourth of the total length of the carapace. The antennal notch forms a deep rounded sinus in the antero-lateral margin, which, above the notch, is cut into about four coarse teeth. On the antero-lateral corner begins a series of strong acute teeth, diminishing to faint serrations on the lower margin. The side of the carapace is smooth. The abdomen is longer than the cephalothoracic region and stout, the first two somites being not longer than broad. The antennules have the first segment of the peduncle longer than the second, which is again a little longer and stouter than the third. The outer flagellum is a little longer than the last segment of the peduncle and is composed of three segments, the first and second subequal and the third minute. ‘The inner flagellum is unsegmented and is nearly equal to the first two segments of the outer flagellum together. The antennee are composed of three segments, the basal one bearing three plumose sete. The branchial apparatus has two small papilliform lobules on its posterior part. The first legs are imperfect in all the specimens examined. The second legs have the ischium represented by two very narrow semiannular sclerites interposed between the basis and merus, forming a ring which is interrupted on the inner and outer sides. The uropods have the peduncle longer than the last somite and about three times as long as broad. The rami are very unequal, the exopod being shorter than the peduncle and little more than half the length of the endopod. ~The latter is composed of two segments, the proximal two and a half times as long as the distal. There are about ten slender spines on the inner edge and two unequal spines at the tip, while the outer edge bears a series of stout sete. The exopod bears a series of long sete on the inner edge and at the tip and some short sete on the outer edge. Remarks.—In the absence of the male it is impossible to decide whether this species ought to be referred to some of the new genera defined below. Its resemblance, however, in such characters as the shape of the antero-lateral edge of the carapace, to L. longirostris Sars leads me to believe that it will be found to belong to the genus CUMACEA FROM THE COPENHAGEN MUSEUM. a3 Leucon. In that case it will fall into the group of species which have a large inner flagellum on the antennule. From all the species of this group it is distinguished by having the outer rami of the uropods shorter than the first segment of the inner. Occurrence.—“ Akaroa Harbour, 8/97, 6 fathoms, H. Suter.” Copenhagen Museum, EUDORELLA TRUNCATULA (Spence Bate). Eudorella truncatula G. O. Sars, Crust. Norway, ii. p. 37, pk xxix. (1900). I am unable to find any differences of importance by which to distinguish three New Zealand specimens from North Atlantic specimens of this species. They are of small size, an ovigerous female measuring only 2°75 mm. in total length. The armature of the antero-lateral margin of the carapace resembles that found in HZ. truncatula and in E. pusilla (the distinctness of these two species appears to me doubtful), but the tooth of the antero-lateral angle is a little more prominent, though much less so than in E. nana or E. hispida. ‘The second legs have the basis not much shorter than the remaining segments together, and the merus and carpus relatively shorter than in northern specimens of H#. truncatula. The terminal spine of the endopod of the uropods is also somewhat stronger. Occurrence.—“ Akaroa Harbour, H. Suter Coll., 8/97.” 3 specimens. Copenhagen Museum. EUDORELLOPSIS RESIMUS, sp. n. (Plate VIII. figs. 6-10.) Description of adult Female.—Total length 1:75 mm. General form much as in HL. deformis (Kroyer). Pseudorostrum well-marked, directed upwards with a slight inclination forwards, distal end truncated, posterior corner not produced, length along posterior edge about one-fourth of total length of carapace. Posterior part of dorsal edge of carapace carrying a prominent tooth directed obliquely forwards. Antero-lateral edge coarsely serrated, the teeth becoming lower and more irregular on the upper part, lower part curving backwards and ending in a small tooth defining the antennal notch. The side of the carapace bears, above the middle of its height, a longitudinal ridge which curves upwards to the anterior edge of the pseudorostrum. The antennules are more slender than in £. deforms and bear less numerous spines. The outer flagellum, composed of three segments, is shorter than the last segment of the peduncle. The unjointed inner flagellum is equal in length to the first segment of the outer. First legs short, reaching beyond the anterior end of the carapace by not more than the length of the last segment, less richly setose than in LH. deformis. Second legs also short, with carpus half as long as merus and equal to propodus. Uropods short and stout, peduncle about two-thirds as long as the last somite. VOL. XVIII.—PART 1. No. 5.— August, 1907. F 54 DR. W. T. CALMAN ON NEW OR RARE Exopod nearly twice as long as the peduncle, roughened on its outer surface with irregular tubercles or blunt teeth, with two unequal spines at the tip and a single seta on the inner edge. Endopod only a little shorter than the exopod ; first segment twice as long as the second, the latter with a long stout terminal spine and a smaller one internal to it, immer edge unarmed. Remarks.—This species differs from the two species of the genus already known in having the pseudorostrum produced vertically and in possessing a lateral longitudinal ridge on the carapace. In the large size of the pseudorostrum it approaches the genus Pseudoleucon of Zimmer, and suggests that that genus should not be separated from Eudorellopsis. Occurrence.—* Lyttleton Harbour.” 1 specimen. Copenhagen Museum. Genus HETEROLEUCON, gen. nov. Only the first and second pairs of legs have exopods in either sex. The endopod of the uropods is unsegmented. The male has no pleopods. Type species H. akaroénsis, sp. 0. The general form of the body is that of the genus Leucon, the carapace having a prominent pseudorostrum and a serrated dorsal crest (in the female). The very oblique pseudorostrum recalls the genus Psewdolewcon of Zimmer, from which, how- ever, the present form is distinguished by the fact that the peduncle of the antennules is not sharply geniculate between the first and second segments. HU&TEROLEUCON AKAROENSIS, sp.n. (Plate VIII. figs. 11-23.) Description of adult Female.—Total length 2:75 mm. Carapace a little less than one-fourth of total length, compressed, its vertical height about two-thirds of its length. ‘The dorsal edge is slightly arched as seen from the side, keeled, serrated in its anterior half with eight to ten teeth, and with a single larger tooth just in front of the posterior margin. In front of the posterior tooth there is generally a shallow rounded excavation of the dorsal edge, but in some cases this is less marked than in the specimen figured. The pseudorostrum is straight, directed obliquely upwards, and sharply pointed. The length of its upper edge is a little less than one-third of the length of the carapace. The antennal notch forms a, rather shallow rounded sinus in the antero-lateral margin, which has one or two teeth above the notch and below it is coarsely serrated and curves backwards into the lower margin without any distinct antero-lateral angle. The abdomen is rather stout and is a little shorter than the cephalothoracic region. The antennules have the first two segments of the peduncle very stout and subequal: the third only half the diameter of the second and two-thirds of its length. The CUMACEA FROM THE COPENHAGEN MUSEUM. 30 outer flagellum is about equal to the second segment of the peduncle and consists of two segments. ‘The unjointed inner flagellum is equal in length to the first segment of the outer. The antenne are composed of three segments. The proximal segment bears two plumose setze. The mouth-parts are of the usual Leuconid type. The posterior division of the branchial apparatus bears only two small papilliform lobules. : The first legs extend beyond the pseudorostrum by nearly the length of their two distal segments. The basis is about two-thirds as long as the distal segments together. The second legs have the basis shorter than the remaining segments together ; the ischium is represented by a narrow chitinous ring between the basis and merus. The uropods have the peduncle longer than the terminal somite and abont three times as long as thick. The rami are subequal and a little longer than the peduncle. The exopod is obliquely truncated at the tip, which bears about five unequal seti. The endopod has a strong, dorsally curved, apical spine and about ten spines on its inner edge. Description of adult (?) Male.—Total length 1:65 mm. Carapace one-fourth of total length, its vertical height about four-fifths of its length. Dorsal edge slightly arched, smooth, or with one or two small serrations anteriorly. Pseudorostrum very short, horizontal, broadly rounded at the tip. No antennal notch. Antero-lateral angle broadly rounded and serrated. Antennules with three segments in outer flagellum. Antenne with flagellum very short, not longer than the peduncle, composed of nine segments. It is thickened at the base, where the segments are indistinctly separated, but distally they are slender and well-formed and provided with sensory sete. The remaining appendages do not differ greatly from those of the female. The peduncle of the uropods is little more than half the length of the rami. emarks.—Though there can be little doubt that the males described above belong to the same species as the females, their much smaller size is very suggestive of immaturity, and to this cause may be due the remarkable shortness of the antenne. The antenne, however, do not show the usual characters of immaturity, for the peduncle is beset with sensory filaments, and the flagellum, though short, is distinctly segmented and carries sete. Occurrence.— Akaroa Harbour, 6 fathoms, H. Suter Coll., 8/97”: many females and three males. “ Lyttleton Harbour, 1—-d fathoms, H. Suter Coll., 8/97”: many females. Copenhagen Museum. Co-types in British Museum. 36 DR. W. T. CALMAN ON NEW OR RARE Genus PARALEUCON, gen. nov. Differing from Leucon in having only one pair of pleopods in the male sex. Type species P. sutevz, sp. n. The species described below resembles Heteroleucon akaroénsis in the arrangement of the teeth on the dorsal crest of the carapace. The segmentation of the endopod of the uropods is much less distinct than in most other Leuconide, and may be regarded as forming a transition to the unsegmented condition found in Heteroleucon. The number of thoracic exopods, however, and the presence of a pair of pleopods in the male seem to require that this species should be distinguished generically from the last. PARALEUCON SUTERI, sp.n. (Plate LX. figs. 1-20.) Description of adult Female.—Total length 2°9 mm. Carapace more than one-fifth of total length, compressed ; its vertical height a little more than two-thirds of its length. The dorsal edge is nearly straight as seen from the side, keeled, with a large tooth near the hind margin overhanging a rounded excavation, in front of which the dorsal keel bears a varying number of irregular serrations. The pseudorostrum is straight, directed obliquely upwards, and sharply pointed. The length of the upper edge is less than one-fourth of the total length of the carapace. The antenna] notch is well-marked and angular, defined below by a triangular tooth, below which the anterior part of the lower margin is obscurely serrated. On the side of the carapace are three subparallel ridges curving obliquely downwards and forwards. The abdomen is longer than the cephalothoracic region. The antennules have the first segment of the peduncle little stouter than the second and shorter than the second and third together. The outer flagellum is composed of three segments and is shorter than the third segment of the peduncle. The inner flagellum is unsegmented and about equal to the first segment of the outer. The antenne are composed of three segments. ‘The proximal segment bears two plumose sete. The mouth-parts are of the usual Leuconid type. The maxille have two or three small sete on the distal part of the inner edge. The branchial system is reduced, the lobules being represented only by two small papillee. The first legs extend beyond the pseudorostrum by little more than the length of their last segment. The basis is little shorter than the remaining segments together, and the dactylus is nearly as long as the propodus. The exopod is shorter than the basis. The second legs have the ischium distinct and the dactylus longer than the carpus. CUMACEA FROM THE COPENHAGEN MUSEUM. 37 The uropods have the peduncle a little longer than the last somite and about 34 times as long as thick, with one or two minute sete on its inner edge. The endopod is about 14 times as long as the peduncle, somewhat indistinctly divided into two subequal segments, with about 12 spines on its inner edge and two unequal terminal spines. The exopod is longer than the endopod, with a slender terminal spine and a series of setz on its inner edge. Description of adult Male.—Total length 2°4 mm. Teeth of dorsal crest of carapace in some specimens only a little less marked than in the female, in others almost obsolete. Pseudorostrum horizontal, truncated, about one-sixth of total length.of carapace. No antennal notch or tooth, the antero-lateral corner rounded off, with a few obscure serrations. Only two oblique ridges on the side of the carapace, corresponding to the two posterior ridges of the female. The antennules have the second and third segments of the peduncle shorter than in the female, and together shorter than the first segment. The flagella are similar to those of the female. The antennz have the penultimate segment of the peduncle more than one-half as long as the succeeding segment. The flagellum is very short, not more than twice as long as the peduncle, and not extending back beyond the first free thoracic somite in the natural position. Exopods are present on all except the last pair of legs. Only the first abdominal segment has appendages and these are of very small size. The peduncle is twice as long as broad and bears two minute unjointed rami. Hach ramus has a single long plumose seta, and the outer bears in addition one or two short simple sete. The uropods hardly differ from those of the female, having only a few additional small sete on the inner edge of the peduncle. Occurrence.—“ Lyttleton Harbour, 5/97, 1-5 fathoms, H. Suter”; “ Akaroa Harbour, 8/97, 6 fathoms, H. Suter.” Copenhagen Museum. Co-types in British Museum. HEMILEUCON, gen. nov. Differing from Leucon and Paraleucon in having no pleopods in the male sex. Type species H. wniplicatus, sp. n. HEMILEUCON UNIPLICATUS, sp. n. (Plate IX. figs. 21-25.) Description of adult Female.——Total length 2:6 mm. Carapace about one-fourth of total length, its vertical height a little more than two- thirds of the length. ‘The dorsal crest is slightly arched, its anterior half serrate. Pseudorostrum straight, slightly upturned, obliquely truncate; the length of its 38 DR. W. T. CALMAN ON NEW OR RARE upper edge is less than one-fourth of the total length of the carapace. The antennal notch is well marked, defined below by a triangular tooth. The antero-lateral margin above the notch bears three or four serrations, and the front part of the lower margin is serrated. The side of the carapace is marked by a single horizontal ridge above the middle of its height, curving upwards posteriorly to join the hind margin, and terminating anteriorly below the end of the frontal fissure. Antennules with the outer flagellum of three segments, the inner of one, equalling the first segment of the outer. The legs, so far as they are visible without dissection, present no marked differences from those of Paraleucon. The uropods have the peduncle a little longer than the terminal somite and three times as long as thick. ‘The rami are subequal and 1} times as long as the peduncle. The endopod is distinctly segmented, the proximal 1} times as long as the distal segment. The terminal spine is not much longer than the distal spine of the inner edge. ‘The exopod has several unequal sete at and near the tip. Description of adult Male.—Total length 2°35 mm. Dorsal edge of carapace slightly arched, with only some faint traces of serration anteriorly. Pseudorostrum horizontal, truncated, less than one-fifth of total length of carapace. No antennal notch or tooth, the antero-lateral corner rounded off, the antero-lateral edge serrated from the base of the pseudorostrum to the lower edge. The side of the carapace bears a horizontal ridge similar to that of the female. The antenne have a very short flagellum and hardly reach back to the hind margin of the carapace in the natural position. Exopods are present on all except the last pair of legs. The uropods resemble those of the female. Occurrence.—“ Lyttleton Harbour, 5/97, 1-5 fathoms, H. Suter”; “ Akaroa Harbour, 8/97, 6 fathoms, H. Suter.” Copenhagen Museum. HEMILEUCON COMES, sp. n. (Plate IX. figs. 26-32.) Description of adult Female.—Total length 2°8 mm. Carapace less than one-fourth of total length, its vertical height less than two-thirds of its length. Dorsal crest very slightly arched, irregularly serrate anteriorly, with a depression near the posterior end. Pseudorostrum straight, directed obliquely upwards, sharply pointed. The length of its upper edge is about one-third of the total length of the carapace. Antennal notch rather widely open, defined by a triangular tooth, the lower edge of which is obscurely serrated. The side of the carapace bears two oblique ridges similar in position to, though less strongly marked than the two anterior ridges of Paraleucon. CUMACEA FROM THE COPENHAGEN MUSEUM. 39 Abdomen hardly longer than the cephalothoracic region. Antennules as in Hf. wniplicatus. The first legs are imperfect in the specimens examined. ‘The second legs have the ischium distinct. ‘The dactylus is equal to the merus and carpus together. The uropods have the peduncle longer than the last somite and about 34 times as long as thick. ‘The endepod is longer than the peduncle and shorter than the exopod, and is distinctly divided into two segments, the proximal less than 1} times as long as the distal. ‘There are about nine spines on thé inner edge, increasing in length towards the slender terminal spine. The exopod has two unequal terminal sete and series of sete on the inner and outer edges. Description of adult Male.—Total length 2-7 mm. Dorsal edge of carapace slightly arched, smooth. Pseudorostrum horizontal, truncated, about one-sixth of total length of carapace. No antennal tooth or notch, antero-lateral corner rounded, no serrations on antero-lateral or lower edges. The side of the carapace bears two oblique ridges similar to, but fainter than, those on the carapace of the female. The antenne have the flagellum short, reaching to about the second free thoracic somite in the natural position. Exopods are present on all except the last pair of legs. The uropods resemble those of the female. Occurrence.—* Lyttleton Harbour, 5/97, 1-5 fathoms, H. Suter.” Copenhagen Museum. NG te ‘ BY Sith Pe aticd at # PLATE I. vor. Xvii.—Ppart 1. No. 6.—August, 1907. NEW OR RARE CUMACEA FROM THE COPENHAGEN MUSEUM. 9 IS Cr R oF WS eH 39 39 PLATE I. . Bodotria sublevis (p. 3). Female. Anterior part of body. 3p First leg. Hs Last somite and uropod. similis (p. 4). Female. Male. Female. Carapace. i First leg. , Last somite and uropod. Male. Last somite and uropod. stamensis (p. 5). Female. P) parva (p. 39 39 5). Male. Female. Carapace. . First leg. i. Last somite and uropod. Male. last somite and uropod. Female. Carapace. % First leg. 5 Last somite and uropod. Trams,lock. Soc. Yel AVM #01 ees W.T. Calman, del, Hult} sever amp? I~ 3, BODOTRIA SUBLEVIS, 4-9, B.SIMILIS, 10-15, B.SIAMENSIS. 16-18, B.PARVA. wi) tt il i Vis aie i i i) PLATE IL 44 NEW OR RARE CUMACEA FROM THE COPENHAGEN MUSEUM. PLATE II. Fig. 1. Cyclaspis elegans (p. 9). Female. From the side. 2 My - Male. From the side. oe a Female. From above. 4, i a Male. From above. 5. 5p ss Young, before development of last pair of legs. 6. He Young male with developing pleopods. We x rs Female. Anterior part of head, from above. 8. < a ee Antennule and antenna. 9, a o e Autenna, another view. 10. . ne Male. Antennule. dak i FF a Antenna. 12. i Female. Lower lip.—1l2a. Tip of one of the lobes, further enlarged. 126. One of the spathulate spines. 1B, ae Be a Mavxillula. 14, % a5 i First maxilliped with branchial appa- ratus.—14qa. Endopod of same, further enlarged. 15. Si s 5 Second maxilliped. NG, si 5 AB Third maxilliped. Wo . Me us First leg. 18. Hi 38 oe Second leg. 19. Bs is i Third leg. 20. 3 ip 2 Fourth leg. Zale 55 a i Fifth leg. We Fe a Male. Pleopod. 23. 5) 5 Female. Last somite and uropods. 24, a Male. Last somite and uropods. Trans Loot. Foo. Vol. VIL FCM. Huth, sc, et imp. W.T. Calman, del. > 24 OxeGin NESS bd CUAING S- PLATE III. 46 NEW OR RARE CUMACEA FROM THE COPENHAGEN MUSEUM. 1. Cyclaspis similis (p. 12). ~~ a Seow fat feed bo -) OR OR SOM SO SOM Se one SCODH AT oO wo MN SV Ss On 9” b)) biplicata (p. 17). bP) triplicata (p. 19). PLATE IIL. Female. From the side. = From above. 54 Last somite and uropods. Female. From the side. Male. From the side. Female. Anterior part of body, from above. Male. Anterior part of body, from above. Female. Dorsal outline of first free somite. Male. Dorsal outline of first free somite. Female. Antennule.—10q. Distal portion further enlarged. 5 Antenna. . First leg. a Second leg. a Last somite and uropod, Male. Last somite and uropod. Female. From the side. Male. From the side. Female. Anterior part of body, from above. Male. Anterior part of body, from above. Female. Dorsal outline of first free somite. Male. Dorsal outline of first free somite. Female. First leg. is Last somite and uropod. Male. Last somite and uropod. Trums.coct. Soc. Vol. VUE FC ML #7 WT. Calnan, del. Huth) se. et 10xp. 78}, (CACCIWASID INS) SONU), LS, 16-24, ©. TRIPLICATA. hay aN iter} RAGA HAN heats PLATE IV. 48 NEW OR RARE CUMACEA FROM THE COPENHAGEN MUSEUM. PLATE IY. Fig. 1. Cyclaspis cingulata (p. 15). Female. From the side. 2. ne a us From above. 3 3 i ss Outline of carapace, seen from in front. 4, a .; pe Anterior part of head, from above. 5. 3 es i Anterior part of head, from the side. 6. ‘ He or Antennule. le me a . Third maxilliped. 8. 5 a i First leg. 9. 35 a ae Second leg. 10. a i = Last somite and uropod. 11. us uniplicata (p. 18). Female. From the side. Ue _ As < From above. 13. Me ie rs Anterior part of head, from the side. 14, “ Ss a Antennule. 15. fe Fs; Pe Third maxilliped. 16. i a se First leg. ie Be 3 Second leg. 18 a 5 a Third leg. 19. i 4 ee Fifth leg. 20. is rs un Last somite and uropod, Trans. Loot, Soc ‘ Vol. AVI OAs WT. Calman, del. Huth, sc.et imp. RS 10, CYCLASPIS CINGULATA. TLS ZO), (Sy IWINMMEIiL eva a he ey uN PLATE V. VoL, xvi.—ParT I, No. 7.—Augqust, 1907. 50 NEW OR RARE CUMACEA FROM THE COPENHAGEN MUSEUM. PLATE V. Fig. 1. Cyclaspis longipes (p. 7). Female. From the side. A. BS “ Male. From the side. SP a Female. First leg. 4. 5 ‘6 ae Last somite and uropod. 5, eS 8 Male. Last somite and uropod. 6. », levis (p. 8). Female. From the side. We ie , re First leg. 8. a fe i Last somite and uropod. 9. » Uunicornis (p. 14). Female. From the side. 10. 3 ie i First leg. ve . NS Last somite and uropod 12. » thomson (p. 16). Female. From the side. IL}, i e Male. From the side. 14. 7 Bs Female. First leg. NS ae & ‘ Last somite and uropod. tay 3 fs Male. Last somite and uropod. Trans Loot Sec. Yor. VHT ICN. oe | : ye W.T.Calman,del. Huth,sc.et imp. = Gy ©. Liewigsss 9-11, C. UNICORNIS. 12-16, C. THOMSONTL. 1-5, CYCLASPIS LONGIPES. 4 i if Pal ES eval. Bigs ° ° 10. Uk 4 5 6. Ue 8 $) NEW OR RARE CUMACEA FROM THE COPENHAGEN MUSEUM. 1. Hocuma longicornis (p. 20). 2. 3 lata (p. 22). Female. producta (p. 24). 39 PLATE VI. Male. From the side. From above. Antennule. First leg. Second leg. Last somite and uropod. From the side. m From above. e Minute texture of carapace, much enlarged. Fi Antennule and antenne. 2B First leg. ap Last somite and uropod. Female. From the side. Female. 39 22 From above. Minute texture of carapace, much enlarged. First leg. Last somite and uropod. From the side. From above. Last somite and uropod W.T. Calman, del. 1-6, EBOCUMA LONGICORNIS. . Z. oo K Wr VINE = LORMAN Gr = Ay Frans, Loot. Soc. Vit. XVI FEV MWS, OWN) wae Sea Huth, sc eb amp. P12, WA, WSN, Jee (Sedalia save 18-20, E.PRODUCTA. at al i H i} ny eg aie _— 2 NEW OR RARE CUMACEA FROM THE COPENHAGEN MUSEUM. H 99 BD be on ° SU Se Soe. PLATE VII. . Zygosiphon mortensent (p. 25). Female. From the side. os a i From above. ie a, a: Anterior part of head, from above. mn : Male. From the side. Ee 5s i Anterior part of head, from above. ss a Female. Antennule. BA ah is Antenna. e a Male. Antenna. 4 : Female. Mandible. Me a i First maxilliped, with branchial apparatus. Male. Posterior part of branchial apparatus. , Female. Second maxilliped. a 55 Third maxilliped. First leg. Second leg. Third leg. Fifth leg. 33 ¥ rs Last somite and uropod. Ae Fe Male. Last somite and uropod. . Vauntompsonia arabica (p. 29). Male. From the side. Antennule. First leg. Second leg. Last somite and uropod. Trams doot. oc. Ve XVI FOIL = SE ialins i Kits ~ PAI CUPP RSSS Se A an ON Tee i 1 EN) Oe) N ‘ \ Y \ {| { W.1 Calman, del. Huth, sc ot amp. 19) ZMGO SiR EON MORTENSENI. 20-24, VAUNTOMPSONIA ARABICA. PLATE VEL. = 56 NEW OR RARE CUMACEA FROM THE COPENHAGEN MUSEUM. PLATE VIII. Fig. 1. Leucon(?) heterostylis (p. 32). Female. From the side. 2. 55 3B ie Carapace, further enlarged. Bp s : Be Antennule. 4. 5 33 5 Second leg.—4 a. Middle part of same from inner side to show reduced ischium. 5. A es fe Last somite and uropod. 6. Hudorellopsis resimus (p. 33). Female. From the side. te Be i i. Carapace and antennule, further enlarged. 8. ss a 3 First leg. 9. Be a vy, Second leg. 10. a A em Last somite. 11. Heteroleucon akaroénsis (p. 34). Female. From the side. 12. eS s Me Carapace. 13. 5 a Antennule. 14. Hi i Antenna. 15. BG a 6 Mandible. 16. 2 a me Maxillula. lei 53 ss 5 First maxilliped with branchial apparatus. 18. Ms iS . Third maxilliped. 19. ue “ fe First leg. 20. Pa rel Hy Second leg. 21. a3 ss Last somite and uropod. DOF ie e Male. From the side. 2% - Ee ys Carapace, antennule, and antenna. Frans. loot. Soo. Vol MVM. LE VUL W,T.Calman, del. Huth, sc.et imp. 1-5, LEUCON (?) HETEROSTYLIS. 6-10, BUDORELLOPSIS RESIMUS. 11-23, HETBROLEUCON AKAROENSIS. i iN ay DAN! CG i eanin) Ribs hy G aa SA) \ ay) ny if t/ Fe oe al PDSTE 1X, 58 Fig. Site Oo tsa 20. NEW OR RARE CUMACEA FROM THE COPENHAGEN MUSEUM. PLATE IX. Paraleucon suteri (p. 36). Female. by) 39 9 be) + 1s 39 99 29 ie, 21. Hemileucon untplicatus (p. 37). 22. 29. 39 39 3 comes (p. 38). Female Male. 33 Led DO From the side. Carapace, further enlarged. Antennule. Antenna. Mandible of right side and part of that of left side. Maxillula. Maxilla. First maxilliped with branchial apparatus. First maxilliped, further enlarged. Third maxilliped. First leg. Second leg. Fourth leg. Last somite and uropod. From the side. Carapace, further enlarged. Antennule. Antenna. Pleopod. Last somite and uropod. Female. From the side. = Carapace with antennule, further enlarged, showing outline of mandible. 5 Last somite and uropod. Male. From the side. with antennule and antenna, further enlarged. 5 Carapace . From the side. Carapace, further enlarged. Antennule. Second leg. Last somite and uropod. From the side. Carapace, further enlarged. Frans Tol Se. We KUM GCM f\ Huth, sc. et mp PARALEUCON SUTERT. 21-25, HEMILEUCON UNIPLICATUS. 26-32, H. COMES. a mn) NA) ent} TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON (continued). To Fellows. BE se VOLUME XY. (1898-1901, containing 52 Plates) . Price5 15 6 Part 1. (1898, containmg 8 Plates) . . .. . pe Oop. D iv 2. (1899) contamine 95: Plates)? 5520) samme. 55 O39. O » o. (1899, containmg 4 Plates) .... . pone ab) 5, 4: (1899; containmg 4 Plates). 9. ;). . ss Oreo en O roe, (19005 containins 24 Plates). . 6 ose. ts eee ke 6 » 6. (1901, containmg 4 Plates) eee : aie Ollie » 7. (1901, containing 3 Plates and Title and fa op Ue eal) VOLUME XVI. (1901-1903, contaiing 88 Plates) . Price5 8 O Part 1. (1901, containmg 6 Plates) . . . . . i OPS 0 peed COO combiaimuine: DEPIates): soe. uae esos Peed Os 2 Bik: arora (1901. convaminges O-Plates) 5.20. acme Se Vealinn ee) » 4 (1902, containmg 5 Plates) . . . . . ec Ove DeaiQ js D (1902, containing 4 Plates)... :. doe Opallay5.(6) eee L902" contamimer Oo (Plates\aies tire iit o. ays IO Loe » 7. (1902, contaming 3 Plates) eae PR OA on » 8. (1908, containing 3 Plates andTitle and Index),, 0 13 6 VOLUME XVII. Parr 1. (1903, containing 5 Plates) . . . . . Pricel 2 6 » 2. (1903, containing 3 Plates)... . . poy, Walang (0) » 9. (1904, contaming 13 Plates) . . . . . Seema Oss ; 4. (1905, containing 8 Plates) . . .. . Pali te i) » 5. (1905, containing 50 Text-figures) . . Ay AO ee) » 6. (1906, containing 12 Plates and Title and fades Lens. G VOLUME XVIII. Parr]. (1907, containmg 9 Plates)... . . PriceO 15 O To the Public. £ SPN PES SSeS) SE ee d. SSS SSIS) sy MSY ares) (oy Vici sey) nay (enh easy (=) LOSS) CONTENTS. I. On New or Rare Crustacea of the Order Cumacea from the Collection of the Copenhagen Museum.—Part 1. The Families Bodotriide, Vauntompsoniide, and Leuconide. By W. T. Cauman, D.Se., .ZS., British Museum (Natural History): (Plates: LAV Xe) re ee 0 een cae teu ee cere eae THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tue scientific publications of the Zoological Society of London are of two kinds—“ Proceedings,” published in an octavo form, and “‘ Transactions,” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published in the “ Proceedings” by the Committee of Publication. A large number of coloured plates and engravings are issued in the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Among such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “Proceedings” for each year are issued in four parts, in the months of June, August, October, and April, the part published in April completing the volume for the last half of the preceding year. The pagination is now consecutive throughout the issue for the year, so that the year and page are the complete reference. The “Transactions” contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are issued at irregular intervals. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1881, if they exceed the value of five pounds. Fellows also have the privilege of subscribing to the Annual Volume of the ‘ Zoological Record” for asum‘of £1 10s. (which includes delivery), payable on the Ist July im each year; but this privilege is forfeited unless the subscription be paid before the 1st December in each year. Such of these publications as are in stock may be obtained at the Society’s Office (8 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. P. CHALMERS MITCHELL, August, 1907, Secretary. TRANSACTIONS THE ZOOLOGICAL SOCIETY. OF LONDON. Vou. XVIII.—Parr 2. | LONDON: PRINTED FOR THE SOCIBKTY, SOLD AT THEIR HOUSE IN HANOVER SQUARE; AND BY MESSRS, LONGMANS, GREEN, AND CO., PATERNOSTER ROW. May 1908. Price 40s. Taylor and Francis, Printers. | [Red Lion Court, Fleet Street. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. To Fellows. © ‘To the Public. ef ss aid. Sa i. : VOLUME I. (1883-1835, containing 59 Plates). . Price 313 6 . .. 418 O* VOLUME II. (1835-1841, containmg 71 Plates). . 5 4 0 0... 5 6 6* VOMUME Ill. (S40 1a40, contumins 6s Plate) = 5 3s an VOLUME IV. (18511862, contains 77 Plates). 6 80 ek ee VOLUME _ V. (1862-1866, containmg 67 Plates). . , 5 4 3 Ge ee OS ZO) VOLUME VI. (1866-1869, containing 92 Plates) . i 50 15 0 0 VOLUME VII. (1869-1872, containing 73 Plates). . ,, 10 4 0... 13°12:40 VOLUME VIII. (1872-1874, containing 82 Plates) . ni DS 8 oe ee oe tlionalal 0 VOLUME IX. (1875-1877, containing 99 Plates) . Bp ree eG eho ee fli Gn ou) VQLUME X. (1877-1879, containing 95 Plates} . 5p il Oe ONO Coruna. lo en GENERAL INDEX, Vols. I. to X. (1883-1879) ._,, OES ee a AO) ION) VOLUME XI. (1880-1885, containing 97 Plates). . ,, Ue sae Gee UG (0) VOLUME XII. (1886-1890, containing 65 Plates). . ,, Dey SO LO aay Oe aa el nO) VOLUME XIII. (1891-1895, containing 62 Plates). ., 6 8 8... 811 0 VOLUME XIV. (1896-1898, contaming 47 Plates). . , 5 5 0... 7 0 0 * No copies of these volumes remain in stock. Continued on page 3 of Wrapper. [59 | Il. On recently discovered Subfossit Primates from Madagascar. By WeERBeERt F. Sranpine, D.Sc. (Leeds), M.Sc. (Vict.)—With an Appendix: On the Form of the Brain in the Eatinct Lemurs of Madagascar, with some Remarks on the Affinities of the \ndrisine. By G. Kutiot Situ, W.A., UD., F.R.S., Professor of Anatomy, School of Medicine, Cairo. Received and read March 19, 1907. [Piates X.-X XVIII. and Text-figures 1-52.] INTRODUCTION. THE existence in Madagascar of a considerable subfossil fauna has long been known to zoologists, but it is only within quite recent years that the extent and scientific interest of these remains have been recognised. In his valuable thesis on fossil Lemuroidea, published in 1906, Dr. G. Grandidier has given a summary of the results of the work of various travellers and investigators in this field during the last few years. I need not here do more than refer to the memoirs of Forsyth Major, Andrews, Milne-Edwards, Filhol, Grandidier, Lorenz, Guldberg, Burckhardt, and others, who have done so much to throw light on the nature and affinities of the numerous mammalian and other vertebrate remains which have been brought to Europe during recent years. The various localities where these remains have been found are situated chiefly on the south-west and south coasts of the island, and include both lacustrine and cave- deposits. In the volcanic regions of the central plateau two districts have been discovered and partially explored which have already yielded very valuable additions to our knowledge of the subfossil fauna of Madagascar. The more recently discovered of these localities is situated at Ampasambazimbas about 100 kilometres to the west of the capital, in the district of Miarinarivo (text- fig. 1). The Académie Malgache, a scientific society the headquarters of which are in Tananarive, has undertaken the systematic study of the numerous remains exhumed at this place; and at the request of this Society I have had occasion during the last four years to publish in the Bulletin of the Académie several preliminary notices referring to the various subfamilies of Lemuroids represented among these remains. The richness and variety of the material already discovered and the interesting nature of the conclusions which it appears to me may legitimately be drawn from a detailed study of these recently extinct Lemuroids, seem to warrant a more systematic treatment than it was possible to give in the brief notices above referred to. VOL. XVIII.—pParT 11. No. 1.— May, 1908. K 60 DR. H. F. STANDING ON SUBFOSSIL In a preliminary section the nature and probable origin of the deposits at Ampasambazimba will be briefly discussed. These deposits are especially rich in mammalian remains. In the comparatively small area already explored some 60 skulls of Lemuroids, mostly in an excellent state of preservation, and representing Text-fig. 1. Maj ue Sketch-Map of Madagascar showing position of Ampasambazimba and other fossil deposits. several new species and one new genus, have been found. In addition to these Primate remains at least 200 skeletons of Hippopotami of all ages have been exhumed, besides numerous remains of Cryptoprocta, Centetes, and other Insectivora. Along with these vestiges of an extensive subfossil mammalian fauna large numbers of bones of pyornis, Chelonians, Crocodiles, Lizards, &c. have been found. It is my intention, however, at present to confine myself to a description and discussion of the fine series of Lemuroid remains which have been brought to light at Ampasambazimba. These not only form a most valuable contribution to our knowledge of the strange extinct fauna of Madagascar, but also throw light on the origin and affinities of the existing Primates of that interesting island. After discussing the character and probable origin of the fossiliferous beds at Ampasambazimba my aims in the present memoir will be :— 1. To give a detailed description of the various new species or genera of Lemuroids, drawing special attention to such characters as tend to explain the affinities of the different forms. PRIMATES FROM MADAGASCAR. 61 2. To discuss the origin and relationships of existing species in the light of these subfossil remains. 3. By a comparative survey to trace the bearing of recent discoveries on the position of the Lemuroids and their relations to the higher Primates, an attempt being made to throw some light on the course of development of the Primates in general, and incidentally to distinguish between such so-called lemuroid characters as are due to degeneration or specialisation on the one hand and such as may be considered as belonging to a common ancestral Primate stock on the other. I wish in this place to express my thanks to the various members of the staff of the British Museum at South Kensington who have generously given me every facility for the comparative study without which the preparation of this memoir would hare been impossible. I am especially indebted to Dr. Smith Woodward and Dr. Andrews, of the Geological Department, who have, with unfailing courtesy, given me every assistance in their power. In the course of this essay, I have, I trust, made ample acknowledgment of my indebtedness to Dr. Forsyth Major, whose views on the affinities of the Lemurs are in almost every instance confirmed by a study of the ampler material now at my command. I must not omit to record also my hearty thanks to my former professor, Dr. L. C. Miall, of the University of Leeds, for much valuable help during the preparation of this memoir, and to Dr. Elliot Smith of Cairo, whose special knowledge of the comparative morphology of the Primate brain has been brought to bear on the elucidation of the problems raised by the discovery of these interesting remains. It should, however, be understood that the responsibility for the particular views here advanced is entirely my own. In conclusion, I cannot refrain from alluding to the sad loss which the cause of scientific research in Madagascar has sustained in the death of my late lamented friend and colleague Mons. A. Jully, to whose initiative as President of the Académie Malgache the work of exploration, which has resulted in such valuable additions to our knowledge of the extinct fauna of Madagascar, was largely due. THE FOSSILIFEROUS DEPOSITS AT AMPASAMBAZIMBA. Ampasambazimba (“The Vazimba’s Tomb”’) takes its name from a reputed tomb of one of the aboriginal inhabitants of the country which is situated on the edge of the fossiliferous marsh. It lies near the western side of the area of volcanic disturbance which runs north and south for several days’ journey to the west of the province of Imerina. This area represents a region of more recent disturbance than the mass of the K2 Text-fig. 2. +Ambohitriaimerina LXtznandona ELA nparikitel 7 Yo ibatlampy. —-57 On evarangs, wees ener, Qe oharinkitroka 2 SS odmbuttoatotely ~ if aie <_ ae SS Reilitar§ ouenpe) \ \ Ss odmbofidroa Contour Map of part of Aarbitany between Lake Itasy_and Ifanja Marsh After a survey by M. Hugues Berthier Colonial Administrator Scale 1:70,000, ae FPrinttive Hock References : x Site of Fassel Denosit (Anmpasambazimba) Nee 2 /TASY Ik ee ON SUBFOSSIL PRIMATES FROM MADAGASCAR. 63 Ankaratra Mountains. Volcanic cones are thickly scattered over a large part of this district, and at many points extensive lava-streams of varying age are plainly visible. The fossiliferous marsh occurs in the course of the River Mazy, an affluent of the Sakay, the waters of which discharge themselves into the Mozambique Channel. The Mazy, after traversing a region of primitive rock in a westerly direction, encounters several streams of lava which have entirely deflected it from its ancient course. There is evidence of the recent drying up of extensive lakes in the neighbourhood, of which fact the traditions of the natives give confirmation. A reference to the accompanying contour map (text-fig. 2) showing the results of a careful survey of the region by M. Berthier, Administrator of the District, and which by the courtesy of the late M. Jully I am able here to reproduce, will show that above the fossiliferous marsh the Mazy has cut its course for several kilometres through an ancient lava-flow and that on encountering the primitive rock it has followed the line of division between this formation and the volcanic output. The fossiliferous marsh itself is roughly semicircular in form. It is bounded on its semicirenlar contour by the River Mazy, the bed of which now lies some five metres below the surface of the marsh. On its other sides it is limited by a stream of lava, the surface of which shows it to be of much more recent date than that traversed by the river in its higher reaches (text-fig. 3). A calcareous deposit several metres thick and of great hardness separates the marsh from the river, and it is no doubt to the presence of this broad bank of limestone that the preservation of the fossiliferous deposits is due. The marsh when trenched presents considerable variety in the sections exposed. Over large areas there are deposits of travertin in varying thickness. ‘This in many parts forms an unbroken sheet which greatly impedes the work of exploration; here and there this sheet is replaced by detached nodules, and over considerable areas the lime-deposits are absent. Mineral springs are abundant, but all are now cold. They are of varying mineral constituents, some being strongly impregnated with iron and many highly charged with carbonic acid. From these springs the deposit of lime is still forming. On making a section through a part of the marsh where the layers of limestone are discontinuous, various strata are successively encountered. To a depth of about 1 metre below the present surface the deposits consist of recent vegetable remains succeeded by a fine black humus which rapidly turns to mud on admixture with water. ‘This is superposed on a stratum of forest débris, consisting of leaves, twigs, seeds, and fruits of numerous species. Below this again a layer of larger branches and tree-trunks is frequently encountered. These vegetable remains are in a wonderful state of preservation, the leaves often retaining their green colour. The bones begin to appear in general with this layer of wood, though they are occasionally met with nearer the surface. 64 DR. H. F. STANDING ON SUBFOSSIL Text-fig. 3. fossil tferous te Z Depostt = z ~ Zz tab) LZ, N = _& Lava-Stream & ~ > , a 5 Pes SS be 8 ic oe SS SS ! SS as SS S 70 48 aS % aS S G y SS ~S Oy Y BY = ' WD “MS ~~ Q RN te g 160 4Q 4; 59368 S Sy S eg ty SS SS PEC 1150 44 Shaper es RS i g Q he GS RS . MO Cars} 1140 = < N39-01 pay 1130 - 128-54 8 1129-5} 1120 | 110 Life. Zawa Stream Seale 1 fa 10,000. Seale of lengths 1] in 10,000. Section from toll { Sale of heights lin 1,000. Plan of Fossil deposit at Ampasambazimba showing altitudes along a line crossing the River Mazy (1 to 4), the N.W. lava-stream (4 to 7), the northern edge of fossil marsh (7 and 8), the eastern tongue of lava (8 to 10), the tributary used for channel through diggings (10), and the edge of extensive ancient lava-flow (11).—From survey made for present memoir by Mons. Hugues Berthier, Administrator of the District. PRIMATES FROM MADAGASCAR. 65 Tn several places the bed of the marsh has been uncovered. It is found to consist of a compact layer of volcanic pebbles cemented together by mineral deposits and frequently presenting a metallic sheen. ‘These pebbles apparently form a nearly hori- zontal floor about 3 metres below the present surface. Primate remains are found at all depths; the bones of Hippopotami, Crocodiles, and Chelonians occur chiefly, though not exclusively, in the lower strata. Two theories have been proposed to account for the origin of this rich deposit of fossils at Ampasambazimba. The first is that put forward by Messrs. Baron & Mounéyres, who visited the site soon after its discovery. They say in their report :— “ A notre avis, les ossements ainsi rencontrés et qui sont relativement condensés dans un petit espace, sont les restes d’animaux qui se sont rendus en ce point pendant un long espace de temps et qui y sont morts, soit de mort naturelle, soit de suite de maladies ou de luttes, le cas de semblable concentration est fréquent aux points ot sourdent les eaux minérales.” It is quite possible that this explanation may account for the presence of some of the remains exhumed at Ampasambazimba. The thickness of the lime-deposits would show that the mineral springs are of ancient date, but there are strong objections to the belief that this theory supplies a sufficient explanation. In the first place, the very large numbers of aquatic and amphibious animals—Hippopotami, Crocodiles, and Chelonians—would seem to lead to the inference that a considerable sheet of water formerly existed at this point. A further objection arises from the consideration of the state of the fossils themselves. In the case of the Lemuroid remains it has not been found possible to reconstitute a single complete skeleton, the bones being scattered without any apparent order. ‘Thus, for example, a large number of humeri and several radii of Megaladapis have been discovered, but not a single femur. Considerable numbers of femurs and tibio-tarsi of various species of Apyornis have been found, but comparatively few vertebre, and only one imperfect skull. In describing the method found necessary in exhuming the bones in the alluvial deposit at Lambo- harana on the south-west coast Dr. Grandidier says:—‘‘ The bones are not at a great depth, never more than two metres; they are encased in a compact humus, which during the work is rapidly transformed into mud on account of the water which immediately invades the diggings as the operations are continued. ‘I’his is, indeed, the chief obstacle to the methodical carrying out of the search, and which, both in the western region and in the centre of the island, prevents the finding of complete skeletons, for one is obliged to feel for the submerged bones either with a spade or with the hands or feet, according to the depth at which they are situated under the water, and to draw them out one by one.” If the above description were applicable to the method employed at Ampasambazimba the argument as to the scattered and incomplete character of the remains exhumed would not have much weight; but, as I shall presently describe, it has been possible 66 DR. H. F. STANDING ON SUBFOSSIL here to employ a much more perfect method, by which the contents of the marsh over a considerable area have been thoroughly explored, and had any perfect skeletons existed they must of necessity have come to light. What, then, is the origin of this deposit which has yielded such valuable results in spite of the fragmentary nature of the remains? A careful study of the topography of the immediate neighbourhood has revealed the fact that the flow of lava which now bounds the marsh on the north and west sides must at the time of its output have completely blocked up the course of the River Mazy and formed a lake of considerable area. The lava-flow now rises above the present bed of the river to a height of 50 metres and forms abrupt cliffs overhanging the river and showing the usual broken columnar structure due to the erosive action of the river and prolonged weathering. But there are two interesting facts which show clearly that this lava-stream once impinged on the primitive rock on the left bank of the river. ‘The first is that con- siderable masses of lava remain in sétw embedded among the primitive rock high up above the present level of the stream (text-fig. 3, No. 2). The second fact is that there exists on the right bank of the river, at the point where it now first encounters the basaltic cliffs, a triangular section of what once has been a promontory of primitive rock continuous with the present cliffs on the Jef¢ bank; thus showing not only that the river has been deflected from its course by the lava-flow, but that its original bed has been entirely obliterated and filled in by the stream of lava. Under these circum- stances one is driven to the conclusion that a lake must have been temporarily formed and that the height of its surface must have been at least equal to that of the masses of lava mentioned above and probably considerably higher, since detached fragments of volcanic rock are found further up the hill-side. A tongue of lava at a lower level has also descended on the east side of the present area of the marsh. Reference to the sketch-map (text-fig. 5) will show that the present course of the river turns abruptly south at the eastern edge of the marsh. As already stated, the stream is now far below the level of the floor of the alluvial deposits, but a broad bank of water-worn pebbles which at this point overlies the tongue of lava forming the eastern boundary of the marsh shows that formerly the river owed at a much higher level, and must, in fact, have been some 25 feet above the then floor of the basin. It is thus extremely probable that at least the lower layers of fossiliferous alluvium were deposited in this lake. ‘The presence of large numbers of aquatic animals would thus be explained, and, conversely, some support is given to the presumption, arising from anatomical evidence to be adduced later, that some at least of the gigantic Lemuroids whose remains occur mingled with those of Hippopotami, Crocodiles, and Chelonians, were themselves aquatic. In the case of the pyornis and other land- animals, one may suppose that during the course of the lengthened period which must have elapsed ere the river wore down its exit sufficiently to drain the lake, numerous carcases of these animals were brought down by the annual floods, and finally settled PRIMATES FROM MADAGASCAR. 67 down in the alluvial deposit forming on the floor of the lake. This view seems the more probable in that numerous remains of #pyornis have been found actually embedded among the water-worn pebbles where the bank of shingle shelves down at this point to form the floor of the basin. It will be seen on referring to the contour-map to what an extent volcanic action has taken place in the region drained by the Mazy in the higher reaches of its course. It is certain, not only that the country was formerly much more wooded than at present, but that large areas were covered by lakes which have since been drained. Judging by the recent appearance of some of the streams of lava and the very irregular character of the surface in this volcanic region, it is quite possible that considerable tracts of forest may have been actually devastated and the débris of the flora and fauna carried down and deposited in the lake at Ampasambazimba. A curious incidental confirmation of the belief that these animal remains were deposited in water is afforded by the fact that very many bones, notably the pelves of giant Chelonians and skulls of Paleopropithecus, show the imprints of Crocodiles’ teeth. The photograph of a skull of Paleopropithecus maximus shown in Plate XII. fig. 1 will illustrate this fact. It has been possible in the case of the fossils at Ampasambazimba to employ a method of excavation which has great advantages over that described above by Grandidier. It will be noticed on reference to the map that a small affluent joins the River Mazy at the point where the latter turns abruptly southwards before skirting the edge of the marsh. An ancient native watercourse indicated where a part of the waters of this streamlet had been led round the head of the marsh for the purpose of irrigating the rice-fields, traces of which may still be seen. By deepening and widening this watercourse and continuing it by a broad channel down to the edge of the marsh, it was found possible to conduct a considerable volume of water right through the fossil diggings. It was necessary at the same time to provide an outlet for this stream at the southern edge of the marsh. Here a formidable obstacle presented itself in the broad barrier of compact travertin which has already been referred to as separating the fossiliferous deposits from the present course of the river. By the help of native workmen and a considerable expenditure of dynamite, a channel some 40 metres long and about 2 metres deep was at length cut. It was now possible to expose a long vertical section of the marsh, and a line of workmen could be employed throughout the day working against the vertical wall thus formed; while each evening the earth thus carefully examined could be got rid of by sending it down the stream into the river below. By this method the smallest and most fragile articles can be obtained intact. In one instance, a small bird’s egg was found unbroken at a considerable depth in the compact black earth. The percentage of the breakages of valuable specimens has thus been greatly reduced, and one can feel certain, with a company of trained workers, that nothing of importance will escape their scrutiny. The accompanying photographs (text-figs. 4-8) will sufficiently supplement the description here given of this method of excavation. VOL. XVII.—PART I. No. 2.—May, 1908. L 68 , DR. H. F. STANDING ON SUBFOSSIL Text-fig. 4. General view of fossil marsh, looking South-west. Gang of natives in fore-ground digging channel across the eastern tongue of lava. The River Mazy lies between the marsh and the line of hills of metamorphic rock. Text-fig. 5. Nearer view of marsh, looking South, showing fossil diggings along its eastern edge. To the right is the untouched surface of the marsh, PRIMATES FROM MADAGASCAR, 69 Text-fig. 6. _ View of fossil diggings, looking North. Tn the distance is the main lava-stream ; to right, the lower tongue forming the eastern boundary of marsh. Text-fig. 7. Method of getting rid of earth by deflecting stream through diggings at end of day. Nearer view of lava-stream beyond the marsh. L2 70 DR. H. F. STANDING ON SUBFOSSIL Text-fig. 8. Channel through barrier of travertin separating fossil deposits from river. Turning, finally, to the question of the age of these deposits at Ampasambazimba, we find that the data for the solution of the problem, while fairly numerous, are not very precise. ‘These data may be briefly enumerated as follows :— (a) The age of the neighbouring lava-flow as judged by disintegration of surface ; (b) The depth of material worn through by the river in cutting its new channel ; (c) The depth of alluvial deposit in the marsh ; (d) The state of preservation of the fossils ; (e) The traces in the deposits of contemporary human occupation of the district ; (f) ‘Traditions of the natives as to the former existence of animals now extinct. The full discussion of these various points, while of much interest, would lead us beyond the limits of this paper. A brief reference to a few salient facts must suffice. PRIMATES FROM MADAGASCAR. 71 It is evident in the first place, from the facts already detailed, that the river, in order to reach its present level, must have cut through some 80 feet of material. Its course, however, now lies, as above explained, between the recent volcanic and the ancient primitive rock, and it is at the expense of both these formations that its new channel has been cut. ‘To anyone who knows the effect of the annual tropical rains on the disintegrated granites and gneisses of the central region of Madagascar, the denudation of a depth of 80 or 100 feet in a comparatively small number of years is quite conceivable. Again, since, as above shown, the lake which formerly occupied the site of the present marsh received year by year the alluvium brought down by the floods during the rainy season, a deposit of several feet of earth and organic débris need not necessarily have occupied a very lengthened period. The wonderful state of preser- vation of many of the remains, both of plants and animals, would seem to suggest a comparatively recent origin; thus one Lemur’s skull contained when found a white pulpy substance, evidently representing the brain of the animal. On the other hand, the well-known preservative action of the organic acids of the bog must be remembered. Fragments of pottery, beads, and other traces of the occupation by man of the surrounding district occur in the upper strata. One small round piece of earthenware of foreign origin is especially interesting. I have submitted this to Mr. C. H. Read, otf the British Museum, who reports that it is a coarse earthenware of Chinese manu- facture.~ Its presence at this remote site in the interior of Madagascar must probably be attributed to intercourse with Arab traders. Dr. Forsyth Major in his memoir on Megaladapis madagascariensis has referred to the legends current among the natives relating to the former existence in the island of extinct animals of large size. That memories of the quite recent existence of the Hippopotamus exist in many localities is unquestionable, indeed the native name (/alomena) finds a frequent place in local traditions and folklore. There is also a reference to a creature called the Tokan-dia (“the one with a single footprint”), which may well have been the Apyornis. Taking all these facts into consideration I should be inclined to fix the age of the uppermost of the Lemuroid remains (including a skull of Palwopropithecus and one of Mesopro- pithecus found about 90 cm. below the present surface of the marsh) as probably not more than five centuries. But evidently there is large scope for conjecture in the attempt to solve such a problem. One may at any rate from a biological point of view regard all these subfossil Malagasy Lemuroids as the contemporaries of extant species in other parts of the island. DR. H. F. STANDING ON SUBFOSSIL ~I Ww Classification of Extant and recently Extinct Lemurs. Famitigs. SuBFaMILIES. GuNERA. eye : { Archceolemur. Archocolemuninoe ) tortie Sebel elerehereuel suey necks TEieie oreo ( Mesopropithecus. | ) Palaopropithecus MOUPISIOMD cig 6.090 0.4.0.0. Iie asuctecoE esd oouopKoous Goad Indris. | Propithecus. } L Avahis, (| Girronmnypines codéaoscccepo0 cc 09000000 Chiromys. ( Mm CROLMIDOED sconedadco00000b005900 Megaladapis. ! ( Lemur. ' . | ITER oocv0g60 bt0000000009008000 < Lepidolemur. | ( Hapalemur. erm Unid Copa en ee < | ‘ Chirogaleus. Galavinte aren. ayerpetae erurrne townie Galago. | | ( Loris. ! \ilvorisincs aa jn eke ose cs dene ete, PnoeTs < Nycticebus. | Perodicticus. AUSENCEY Gae oconeud oot. Marsiin ery cyan. Resa ade ean Uae Tarsius. Noe.—The Genera &c. represented by subfossil forms are printed in Italics, Family INDRISIDZ. Malagasy Primates recent or subfossil showing evidence of common descent from ape-like ancestors, their affinity being, however, disguised by the varying degrees to which the different subfamilies have become specialised in adaptation to various modes of life. The dentition in every case shows reduction from the “primitive” or “generalised” Primate formula (retained by the Cebide and many of the Lemuride), such reduction being carried furthest in Chiromys, and least apparent in the Archeo- lemurine. Outer pair of upper incisors when present smaller than inner, Lower first premolar generally caniniform. Auditory bulle generally present. Lacrymal fossa variable in position—within the orbit, on the orbital margin, or on the face. 1 [Sh) PRIMATES FROM MADAGASCAR. There is frequent evidence of retrogressive change in brain-structure. Note.—The reasons which appear to justify the creation of this new family will be made apparent in the detailed descriptions which follow, and more especially in the final Comparative Survey. Subfamily INDRISIN «A, Subfossil and recent Malagasy Indriside, the pithecoid affinities of which are in most of the genera disguised by adaptive specialisation. Brain-case generally depressed; subfossil forms showing marked constriction of postorbital frontal region. Nasal fossa large, with tendency in recent genera to elongation of muzzle and depression of facial angle, the latter characters being less marked in Mesopropithecus. Position of lacrymal fossa variable. Dental formula of all genera: = os er = S third molar. Long inclined mandibular symphysis. Mandible of subfossil forms massive. Auditory bullie generally present and prominent. All show reduction of upper Genus PaL#opropituxcus (G. Grandidier). Definition.—Subfossil Indrisine much specialised and probably adapted to an aquatic life. Brain-case depressed with postorbital constriction. Occiput vertical, lambdoidal crest frequently present. Orbits high and upwardly inclined with raised margins. Postorbital bar triangular with sharp inner edge. Premaxille send up strongly developed projections which are fused to the anterior margin of nasals. Zygomata high and arched. Auditory bull absent. Series of molars and premolars parallel. Mandible very massive with inclined symphysis. Lower incisors sub-equal, but little inclined, and not pectinate in disposition. 1 PAS le jays PAS sin, IZEECHO ppm arenes Dental formula: PALZOPROPITHECUS MAXIMUS (Standing). (Plates X.-XV.) Until the discovery of the Lemuroid remains at Ampasambazimba nothing was known of the skull of Palwopropithecus. In this locality 13 almost perfect crania have been exhumed, besides numerous fragments. These, while showing considerable variations in size and proportions, must all be referred to one species, to which the name of Paleopropithecus maximus has been given. So far as it is possible to make a comparison with the species named by Dr. Grandidier, the animal found at Ampasambazimba appears to have been larger and much more massive. This will be evident on a comparison of the various numbers denoting the breadth of the jaw and length of the symphysis in the following table:— 74 DR. H. F. STANDING ON SUBFOSSIL Tasie 1.—Showing comparative dimensions of mandible of Palwopropithecus ingens and P. maximus. P. ingens. P. maximus. | mm. mm | From the condyle to the summit of the symphysis .......-.--- 130 166 | Length of series of molars and premolars ........--+++ ++ -+5+ 76:5 88 From condyle to posterior edge of third molar .............--. 55 65 | Length of symphysis .....+--0. eee eee eee eee eects 58 86 | ' Width between second premolars and first molars ............-- 26 36 | Width between anterior margins of first premolars ............ 14 25 Width of series of incisors at base ........00-+ eee eee eee 13 23 | Maximum length of first premolar ....-....--+ 2.0 eee sees 14 17°5 a width ,, Br Weg ecthitons on icici Pena ocean 6 6-5 a length of second premolar .......-- +2. +000 eres eeee 15 17-4 width. se ip pester Sas Braeye Aencter site pho 75 ae | 3 length of first molar... 6 2... .se ee etme eee eee 16:8 17°6 | Wee idehy ieee ais ee ec eres HAS are 9 115 a length of second molar .....-..--...---+-++++--5- 16 168 | os Waewicth: ag RE See Won. Us NHN, Aerie aN ta 8 10-7 * length of third molar ........-...-.--+-+22---s-00 13:5 14 e width oa Be ys chains ec ARU HA ute opane loys uel eee Rekonets 8 9-4 Description of the Skull of Palzopropithecus maximus. The skull of Palwopropithecus in its general features conforms to the type of the Indrisine, though curiously departing from that type in certain particulars. It is broad and flattened, with depressed brain-case and narrow postorbital frontal region. The facial portion is short, herein contrasting strongly with Megaladapis. ‘The occiput is vertical; the zygomatic arcade high and widely curved away from the brain-case, The orbits are small and extraordinarily elevated in position and upwardly directed. The nasals at their anterior margins are turned upwards and form a suture with a curious shell-like upward prolongation of the premaxilla. ‘The dentition resembles in general that of Propithecus, though in detail presenting many deviations from the type of the recent Indrisine. The entire absence of auditory bulle is the most marked peculiarity to be observed in the base of the skull (see Plate X.). On attempting to give detailed measurements of the skull one is impressed with the great variety both in absolute size and in the proportion of the various parts presented by different specimens. Among the earliest examples obtained at Ampasambazimba were two showing such differences in outline and proportions that I referred them to two distinct species. In view of the complete series of more than a dozen skulls PRIMATES FROM MADAGASCAR, 75 eventually obtained, this distinction appears to be no longer tenable, and all the specimens are referred to the one species P. maximus. The average maximum length of the skull of Palwopropithecus maximus, measured from the lambdoidal crest to the alveolar margin of the incisors, is 198 mm. (max. 216, min. 181 %*). The average bi-zygomatic breadth for six skulls where both arcades are complete measured 121 mm. (max. 135°5, min. 113). The ratio between the maximum breadth and maximum length expressed as a percentage index is 61. Similar measurements taken on a skull of Jndris give the number 59°4. ‘The maximum external breadth across the orbits in the case of Palcopropithecus is 98 mm., giving, when compared with the maximum length of the skull, an index of 49:5; similar measurements on /ndris and Propithecus give indices of 61°5 and 63:6 respectively. The form and proportions of the individual bones composing the brain-case conform in general to the condition observed in Jndris and Propithecus. Occipital Region.—The occiput is truncated and forms a planeat right angles to the long axis of the skull. The foramen magnum is, generally speaking, circular in outline, but varies much in size and shape, corresponding diameters showing a discrepancy of 5 or even 7 mm. in different skulls. ‘The occipital condyles are broad and prominent, frequently in adult skulls projecting backwards beyond the lambdoidal crest (Plate X.). The paroccipital processes are extraordinarily developed, their free internal margins attaining in some cases a length of 20 mm. and extending downwards and backwards nearly as far as the condyle itself. In its basal portion this process follows along the margin of the occipito-temporal suture as a prominent ridge, in some instances reaching the lambdoidal crest. On its outer edge it is fused to the strongly developed mastoid process of the squamosal and also to the styloid process. The prominent convexity for the lodgment of the central lobe of the cerebellum, which is a marked feature of the occiput of the extant Indrisine, is absent in Palwopropithecus, this region of the skull-wall being much thickened and of porous texture. The lambdoidal ridge is prominent in all but quite young skulls, and in old individuals forms a sharp crest which descends along the mastoid process, an upper branch passing above the external auditory meatus to join the upper edge of the high zygomatic process of the squamosal. Where it is still possible to trace the lambdoidal suture, the supraoccipital is seen to enter to but a small extent into the roof of the cranium. In one young skull the supraoccipital occupies 18 per cent. of the distance from the lambdoidal crest to the root of the nasals (text-fig. 9), while in Propithecus more than 26 per cent. of this distance is covered by the supraoccipital. Parietal Region —TVhe brain-case in the parietal region is narrow and depressed, * When dimensions are given in the following description three numbers will generally be quoted, the first being the average for a series of several skulls; and following this in parentheses are the maximum and minimum measurements noticed in any individual. VOL. XVIII.— PART I. No. 3.—May, 1908. M DR. H. F. STANDING ON SUBFOSSIL Text-fig. 9. VAN je == —= <= } es SSS ——. SS we aee SSS SSSA S S Ss Ww SAX Ss SV Wy UAT RAS OWS i| i} Wh ay Wf HT f { | | t Vertex view of skull of young Paleopropithecus maximus, showing sutures. Nat. size. PRIMATES FROM MADAGASCAR. 77 though approaching more nearly to the condition of recent Lemurs than is the case with Megaladapis (Plates XII. & XIII.). The upper lines of attachment of the temporal muscles form strongly marked ridges, occasionally uniting in old skulls to form a longer or shorter sagittal crest ; but in the majority of cases the fusion of these ridges has only taken place for a short distance above the occiput, the two lines diverging at various angles, leaving a flattened space, wedge-shaped or fusiform, according to the direction taken by the lateral ridges before they widely separate to join the postorbital bar (Plate XIII.). The parietal suture becomes obliterated in old specimens, though the line between the frontals can be traced in almost every case. The squamosal extends up the wall of the brain-case to about the same relative height as in Propithecus, and the spheno-parietal and coronal sutures closely conform to the conditions observed in the extant genera, though the coronal suture has an acute wedge- shape form recalling that of some of the Cebide (e. g. Pithecia rufiventer) (text-fig. 9). Frontal Region—tThe postorbital region of the frontal resembles the corresponding part of the skull of Megaladapis, being very much constricted compared with the modern representatives of the subfamily. The average breadth in this region for 13 skulls was 36 mm. (max. 40:5, min. 28). This is about 18 per cent. of the maximum length of the skull, being only about half a similarly obtained index in the case of Propithecus, viz. 35°6. The frontal sinuses are voluminous and have encroached on the cranial cavity in a manner analogous to that observed in Megaladapis. ‘There is indeed a similar though less pronounced atrophy of the temporal lobes of the brain with a partial obliteration of the lumen between the cerebral and olfactory fosse*. ‘The median line of the frontal behind and between the orbits is convex, though the degree of convexity varies greatly in different specimens. There is also a slight convexity in a transverse direction. The Orbital Region.—tThe size, position, and direction of the orbits have already been referred to as differentiating Palwopropithecus from living Indrisine. These characters must now be considered more in detail. In size the orbits are pro- portionately smaller than in any of the recent representatives of the subfamily, and indicate an animal of diurnal habits. A percentage-ratio, comparing the transverse diameter of the orbit with the bi-zygomatic breadth, gives the number 23. Similar indices in the case of Propithecus, Indris, and Avahis respectively give 34:2, 37-3, and 41, Except in quite young skulls the margin of the orbital process of the frontal is raised to form a strong bony rim, which is continued around the external and anterior margin by a similar growth of the malar. ‘The latter bone attains an extraordinary depth and breadth below the orbit. The distance measured from the external anterior * For further details the reader is referred to the report by Dr. Elliot Smith on the brain-casts of Paleopropithecus and other recently discovered Primates published as an appendix to the present memoir, M2 78 DR. H. F. STANDING ON SUBFOSSIL margin of the orbital rim to the lower end of the suture of the malar with the maxilla exceeds the diameter of the orbit itself (text-fig. 10). No extant Lemur exhibits this peculiarity, which is only paralleled by some of the Old World Apes (e. g. Papio). In the case of Palwopropithecus a broad plate of bone frequently stretches across between Text-fig. 10. Front view of skull and mandible of Paleopropithecus maximus. the lower posterior margin of the orbit to join the zygomatic arch, thus walling in part of the temporal fossa. The plane of the orbital rim is more upwardly and forwardly directed than in Indris or Propithecus. The angle made by this plane with a plane passing through the lower margins of the orbits and the upper margins of the external auditory meatus (the horizontal plane of the Frankfort Convention) may be compared with that occurring in the three extant genera thus :— EXQUCODRODILNCCU SHE are mat le teen nO LQTS’ iratiie ena niet Roe Segara Sa Sh.) MEAD RRR OG aah THe 7 PRODI NeCUS eae a near See Nh mea en oa Te nN Mae TAD ARGS) UMS Salen aren aes Pleas Wee ort tacos Magee tate, 4th ee hehe fic The intersection of the plane of the orbital rim with the median sagittal plane gives an angle which is a measure of the extent to which the orbits are directed forwards. For the same series of skulls these angles are approximately as follows :— Paleopropithecus sa. te tee er ie 1 Oe TETAS 0 UNG RR DE NS OOS Propithecus!.) sideways oir aa ae oer ie) act OOF Avahig 3) 2 OONIE Oise eas Meio nena Ment ese Hee) OS In form and arrangement the various elements constituting the inner wall of the PRIMATES FROM MADAGASCAR. 79 orbit agree in general with the corresponding bones in Propithecus. The following deviations from the type of the recent Indrisine may be noted. The lacrymal enters to a very limited extent into the face, but has a considerable downward extension within the orbit, resembling the condition seen in some of the Hapalide and Cebidee (see text-fig. 48,p.152). It has at the same time a very considerable extension ina direction in line with the orbital rim. The inner edge of the orbital rim shows an extraordinarily developed elevation, forming in some old specimens a roughened prominence 5 mm. high and 10 mm. or more in length. The lacrymal fossa is entirely surrounded by the lacrymal bone, and in most specimens the foramen is well within the orbital margin. In the youngest skulls which I have been able to examine the foramen is in line with the sharp inner edge of the orbital margin, and, strictly speaking, this condition generally still obtains in the adult, though the great forward and upward development of the orbital rim as adult age is reached leaves the whole fossa well within the orbit. The small size and elevated position of the orbits on the one hand and the constricted postorbital region on the other, naturally alter the contours of the maxillary and frontal bones in this part of the skull. In Jndris the posterior part of the maxilla is inflated by an extensive aérial sinus and forms a veritable floor for the eyeball. In Palwo- propithecus the maxilla rises up to form an extensive fronto-maxillary suture between the lacrymal and the ethmoid, but forms no horizontal floor as in the extant genera. The ethmoid has been mentioned as entering the orbit in Palewopropithecus. ‘There is apparently in young skulls in the position where this bone occurs in Propithecus (viz., at the point where the frontal, maxilla, orbito-sphenoid and palatine all approach each other) a small os planum which early fuses with the end of the palatine, and in older animals is drawn in away from the orbital wall by the sinking in of the margins of the spheno-palatine and postpalatine foramina. The great depth of the malar below the orbit has been referred to. Below and within the outer orbital margin this bone is triangular in section, presenting a distinct inner ridge which partially walls in the orbit behind. A downwardly extending ridge of the frontal serves a similar purpose in the upper part of the orbit. We shall have occasion in succeeding sections to notice a similar condition of the postorbital bar in several of the other subfossil Malagasy Lemuroids. The Nasal Region.—The nasals are broad and flattened as compared with those of the modern genera. Their median suture is early obliterated, as also are those they make with the frontal and maxilla, but their contour can be readily traced on a young skull. Fach nasal sends out a rounded wing to meet the anterior angle of the lacrymal. The nasalia are slightly constricted in the front of this point, the two edges running parallel for a short distance, when they again broaden out, resembling closely the corresponding part in Archwolemur. At their anterior end the nasals are curiously turned up, and along their outer anterior margin form a suture with an upwardly 80 DR. H. F, STANDING ON SUBFOSSIL projecting wing of the premaxilla (text-fig. 11). The two bones together form an extraordinary snout-like convexity, the smooth inner surface of which is continuous with the walls of the nasal fosse. ‘The outer surface of these nasal convexities is Text-fig. 11. Muzzle of Paleopropithecus maximus, showing upwardly extending wings of premaxillee fused to anterior borders of nasals. roughened as though for the attachment of some epidermal excrescence. ‘The anterior contour of the narial opening thus formed is heart-shaped as viewed from the front. In old animals the premaxille are fused along the median line of the palate almost to the alveolar margin, though frequently a small lacuna is left unjoined in front of the anterior palatine foramina. The premaxille are thus more strongly developed than the corresponding bones of the extant members of the subfamily. Correlated with this condition of the premaxille the median incisors have had a more vertical position and have been stouter and stronger teeth proportionately than in Indris. The Temporal Region.—The zygomatic process of the malar has a smaller backward extension than in Propithecus (text-fig 13). ‘The squamosal element, on the other hand, almost reaches the posterior margin of the flattened vertical wall below the orbit. The zygomata are proportionately stronger and higher and more outwardly curved than in the extant Indrisine. Hach arch is attached to the vertical plate of the squamosal by a broad horizontal shelf of bone which extends forward almost to the temporo- sphenoidal suture. The presence of this horizontal plate combined with the con- striction of the postorbital frontal region gives to the temporal fossa at this level an almost circular horizontal section (Plate XIIJ.). The massive lower jaw and widely extended mandibular ale were evidently associated with very powerful temporal muscles. The temporal region of Palwopropithecus has already been referred to as possessing one feature of considerable interest, namely, the absence of the inflated auditory bulle which are so prominent and characteristic a feature of the Lemurs (Plates X. & XIV.). This fact is the more interesting as it is associated in Palwopropithecus with the possession of an external auditory meatus of considerable length. ‘The tympanic bone PRIMATES FROM MADAGASCAR. 81 has the same contour as that observable in recent Lemurs, but instead of being inflated and globular in form it has actually a smooth concave surface. The disappearance of the bulle is associated with several minor changes, to which reference may be briefly made. The outer pterygoid lamine of the alisphenoids are widely everted and sharply truncated at their posterior border, the flat plate or bony bridge which in Propithecus or Indris connects this bone with the auditory bulla being, of course, entirely absent. Text-fig. 12. Inclination of lower incisors of (A) Paleopropithecus maaimus and (B) Pithecia rufiventer compared. In disposition the latter bear more resemblance to the usual Indrisine type. (See also text-figs. 18 and 22.) Text-fig. 13. Restoration of cranium and mandible of Paleopropithecus maximus. X 3. The flattening down of the suture between the alisphenoid and the tympanic leaves the foramen lacerum medius widely exposed, and the orifice of the Eustachian duct, foramen rotundum, and foramen ovale are all crowded together at the posterior end of this opening. On tracing the Eustachian duct outwards and backwards towards the auditory meatus it is found to open into a narrow chamber greatly flattened in a 82 DR. H. F. STANDING ON SUBFOSSIL vertical direction and occupying a position just internal to the postglenoid process. Its inferior wall is much thickened. This shallow chamber apparently represents the vestigial auditory bulla. There is a curious massing together and proliferation of the parts around the meatus, forming a tube of more than 1 cm. in length, An irregular and variable mass of bone is, in fact, always present in this region produced by the anchylosis or close apposition of the following elements :—the postglenoid process, the external margin of the tympanic, the mastoid process, and the paroccipital process. ‘This mass of bone is pierced by the stylo-mastoid foramen and the postglenoid foramen, The actual margin of the external auditory meatus is formed by the petrous portion of the squamosal and by the tympanic. The Dentition of Paleopropithecus. (Plate X.) As already stated, the dental formula of Palwopropithecus maximus corresponds with that of Indris, Propithecus, and Avahis. A detailed comparison with these three genera shows that in several particulars Propithecus exhibits the closest analogy with the fossil form. We may instance specially the shape of the premolars. The median pair of incisors are much larger than the lateral. Both pairs are cylindrical in section. The inner pair are curved towards each other at their lower extremity. Their worn surfaces show that all four have been apposed to the extremities of the lower incisors as in Arch@wolemur. The inclination of both upper and lower incisors bears a close resemblance to the condition seen in some of the Platyrrhine Monkeys (e. g. Pithecta) (text-fig. 12). The canines are triangular in section, presenting a strong anterior ridge which is separated by a deep groove from a broader internal ridge. A small posterior cusp is present, as in all the extant representatives of the subfamily. The anterior premolar does not differ materially from the corresponding tooth of Propithecus, except that it is narrower in a transverse direction. The whole series of molars and premolars is in fact decidedly narrower than in any of the recent forms. ‘The anterior premolar is constricted near the centre of the lingual margin, thus being divided into a narrower anterior and a broader posterior portion, the same feature being seen to a less degree in the posterior premolar. ‘The second premolar bears a very small cusp on its inner margin, the homologue of which is traceable in Propithecus. The anterior ridge of this tooth curves outwards to join the external cingulum, but without giving rise toa pointed cusp as in Propithecus. The strongly developed anterior and middle labial cusps of Propithecus are represented in Palwopropithecus by slight elevations of the external cingulum. In the first molar the internal cusps are not subequal as in the recent genera, but the anterior one is greatly in excess of the posterior both in height and length. The same is true of the second molar, though here the extant Indrisinz bear some resemblance to the fossil form. This flattening down and reduction of the postero-internal cusp of the molars is of special interest, as indicating a transition to a PRIMATES FROM MADAGASCAR. ro) ° tri-tubercular form; indeed the condition of the molars in some of the Lemurinz which have a strong postero-internal cingulum does not materially differ from that seen in Palwopropithecus. The transverse ridge joining the posterior labial cusp with the hinder edge of the anterior lingual cusp in the first molars of Propithecus is traceable in Palewopropithecus. The third molar of the fossil genus is greatly reduced, the two outer cusps having quite disappeared and being represented by a line of small nodules; the inner posterior cusp has also gone in most specimens. ‘The series of the premolars and first and second molars form nearly straight lines which converge but little towards the anterior part of the palate. The appended table (p. 86) gives the dimensions of the teeth of a series of skulls. It will be seen that considerable variations occur, as might be expected from the great diversity in the size and proportions of the skulls themselves. ‘The teeth of the lower jaw of Paleopropithecus ingens have been described by Dr. Grandidier, and these agree so closely with the larger species that it does not seem necessary to describe them minutely again. ‘The annexed tables (pp. 84-85), giving detailed measurements of a series of skulls and mandibles, will serve to supplement the foregoing description. Genus MEsopRoPITHEcUs (Standing). Definition —Subfossil Indrisine retaining many pithecoid features. Brain-case rounded, and showing constriction of postorbital frontal region. Sagittal and lambdoidal crests generally present. Interorbital frontal region elevated, with steep facial angle. Orbits small and forwardly directed. Postorbital bar triangular in section. Nasals short and broad. Zygomata high and arched. Palate broad. Series of molars and premolars parallel. Mandible massive, with inclined symphysis. Auditory bullz prominent. 12h cle pm. 2s) Mo i. 2): c. 0): pm. 2): m. 3° Dental formula : MESOPROPITHECUS PITHECOIDES (Standing). (Plates XXI.-X XIII.) Among the Lemuroid remains in the collection of the Académie Malgache are four skulls which, while agreeing in many respects with the existing genera of the subfamily of Indrisinz, differ from these latter in so many important particulars that they must be referred to a new genus. To this I have given the name of Mesopro- pithecus. All the four skulls belong to a single species, which I propose to call pithecordes. As this specific name implies, the new genus presents certain well-marked pithecoid characters. While agreeing in the main with the recent Indrisine, it shows also many striking resemblances to Archeolemur (Nesopithecus). Its main interest indeed lies in the fact that it definitely connects the Archeolemurs with the Indrisine, and hence throws a clear light on the origin of the modern genera of this subfamily. VOL. XVIII.—ParT I. No. 4.—May, 1908. N DR. H. F, STANDING ON SUBFOSSIL (90) [-et] G-LT [cg] [og] [¢-9¢] G-6S OF 68 GS-61 GG 9 9 LUT G96 86 SL 8-04 8-69 G-26 (trT] GCE [cet] S181 “OE [3-46] 81 9g [se] LLB e-GE 86 G-F8 6G | 9-06 29 £9 (ett) g¢ LG G16 92 FTL €L 96 TFL LET. 606 6-061 “Ul SPL S-81. GEE 0& 96 G-BE 86g 6 9% 9% G.gg L9 I6L [se] 83 G-1¢ 9-3L GL 8-89 [ss] ISL 9-81 (¢-g6T] 9-961 ‘OT 8-06 8-06 G16 €& 66 GEE P96 C-98 8IL 9-86 6 CFG 06 G.cg [ee] 6:66 [eg] e-FE G-86 ad &% G9 g-89 [eet] 9 9% 1g G94 g re | [6-06] S61 | $-8T 6-9¢ | [tg] aie 66 63 | 1-83 9-98 |3-FS OF |¢:8S Ze 18 cB 7 #6 [9g] | 49 [62]. | $69 Loet] | Loat] | 19 |6-9¢ as 5a G.Gg 63 9d 19 Oye, | Dekel 9-61, |¢-49 G-I0L | 68 ¢-1&L]] 2-861 (¢-6eT]] 921 G06 | IST | (e6r] | ¢-2or | 7 13 6G 06 1g Gz G13 8-78 re SL TZ 03 o¢ 19 [tet] LLG LG 63 G9 | L-@L 9-69 1-68 1-661 L961 [est] 6-691 6 OIL |? wesw oiyeumosXz Jo (41S1ey) APIA TUNWIXe PL waa “' +> gITRseU JO TApBerq WNUNLUT_L L¢.9g] ‘sess BITesBu Jo TApreiq WNUMIXxe yy BG Se a vee S9ULUV JO SULGIVUL [CRS AROR EG UOT ayered jo Yprosg fg ||P enecnen BooooDeGRCOUSaCOO DS Gaiman -UL) SIB[OT pay} wooajoq oyeied jo yypvorg AG OEE * BUIWMIVAOS [BUIATOVT ToaMjoq YApvetg FeLG | 71157 Sytqdo Jo Toaoy 4B [NYS Jo yypeorg [6L] snqvour Ar0j1pne Jo saoyito WoeAoq oouRyST(T F#G | °° UWNUSeI WoUIVIO} Jo Ia}OUIIP [VOT}IO A. eG |'* WhuseUl usMLI0] Jo JoJaMIVIp osIoAsURIY, | ge |'octtt+: sassaooad propAqs weemqeq yypRerg | F9 asBo-UTeIq JO YJPVaIq [VULO}Xo WINMIXLTL Cleese sues as ([eureyxe) Yypeerq onvmoshz-1g ig [°° °° osvo-urerq Jo 4US10q TeUIE}xXe TINUUIXE py IZ | occ c ct tt t+ sqtqro Jo Jejewerp os1easuesy, ag foc: 0060500 S}IqIO Jo TOJOMVIP [woTyZO A eg |i ccc c%* repour patyy Jo aspo ro1103s0d 04 |ssoooid prouezsysod jo eovyans Jo1ieyue WoL, c.oL |*'* saejoutard pue siv[ow Jo sotses Jo T4SU8T G.F9 eee toe et et ee ee oe eee ee epixemoid jo IoPIO ILTOIATL 0} FIQIO Jo UTSIVW IOMOT WOT PEG Bin a||| Meese eiche: "+++ ogeyed Jo ospe soreqsod 0} WNUSeU USMVAOJ JO ULSILU Jamo, MOLT G.QeT | °° °' 7% eull wetpour Suoje oyeyed Jo yASuT | OLN4NS [BSLU-OFUOIJ 0} 48010 [BPLOpquIV] wWor,;T ZL wee weet ee ee eee wmnuseut UOMLBLOF jo ursaem taddn 09 TREO YO ULSIVU IOMOT WOT 281 Gs * [[MYys Jo yySuey, wNUrxezy @O9OT |°° °° °° ¢* eyrxewoad jo rapioqg repooaTe | 0} WNUSYM WOUIVIOF JO UISIVOL LOMO, TOIT Tt > SS) co aed Hid qt ‘snuncnu snoaypdoidonjog JO s[[UYS Uso}A1Y] Jo (SoAjOUNT]]IUE UL) SMOIsSMAMIP SULIALQ—'Z WAV, i> ise) PRIMATES FROM MADAGASCAR. i[tg] 6P g9 “GT PF G-09 mas se (8-07 | er |\c-9F12-8F Bl |G WG Ise ex. Gage | Be || Se |) Srl [ze] | *- | [ee] | g6|¢-F0r SF |o-6r |¢-l7 |c-9F| 6F erL | [gz] | 28 | e8 |¢:¢6 1 | 98 | GD || of | Bi | Tee we Iw bw oF OT LT G6 8-67 GEE (¢-18] 99 &-76 [g9] G-L8 sees (qj00} Sutpnyour) aepour (qstg Jo yoaoy ye mel jo 4qS10}q i Dor Com On OG aa oC a qeyour qsay Jo Joao, ye Mel Jo ssouyorqy, ** suejomerd omg Aq potdnooo soda * *SIB[OUL PLT] 0} SLIOSIOUT ULIPOUT Wo. saties [eyuop Aq potdnodo sordy “+ ** sIBjoul 981} Jo TduaT Sueloul pItyy Puryad TMI WOaMyoq YIpreAg So[SUB IOLIOJUL WHIMJoq dOUYST(T so[Aptuoo WeoAyog soULyST(T sossoooaid prlowor109 MaaM oq OOULIST(T stsfydurds Jo yysu0T Renin oon es IR[OUL 4sey Jo aspe sorteysod 03 eyApuooa wo.arT reese: g7Suv 1oldejut-o103sod 04 IB[om 4seT Jo oSpa somojsod woay sxejou Q[SUL IOTLOFUL 07 eApuoo UOT; oOURYISIP WINUIEXeTT Ce opsuv IOWaJU 0} ssed0rd prlouot0d WOT T sopsue Lorzeyut Surutol uty e 03 siskyduds Jo pue aMOT WOT sishyd -uids Jo yimums 0} o[Apuoo wor “+++ grostout Jo sd1q 09 et{puoo wor, -ord pue sIefour Jo solies Jo qSus'T ‘snuncou, snosypdosdoajng Jo soyqipurut Jo (seajouII]]1UL UL) sUOIsUEWIp SUTALD—'e WAV I, o DR. H. F. STANDING ON SUBFOSSIL 86 Ps G9 | G9 Lh | 84 L GL L ssouyortyy @) | Sei g ©) ) Ge 9 GG YApeetq + sOSIOUL yeUAe}xXG | °g : a 6 6-6 ai 6 oe OL or “gq Srey a : ) |) (88) a 9 9 9:9 | G9 2 SSL PAE) = G g a G G ee G G “+ yypeeaq | aostour werpey | “2 o OL OL nee 8 . aoe OL 6 . sey 6 6 6 | &6 OL | 2 | Bs 6 or | 6 woes “ yypeerq ; FL FI GET al CI GFL al 1 seat _ || aaye|proovenane Dintete cietelelsialsiatelsisfefeloielatcleteretelerstalaietevaistse Wy8uey Tejou PALL T, 9 or 6 8:6 IL | G11 IL | @0L IL | 601 ah : 41 9 GG At At GALT GI | GAL OT cookie |p EUR BEES | ZL | ¢6 601 | ¢-41 | GI SIt | IL | GE | &t i. SURETY seqour yean | “y 6-81 | 3-91 41 81 8 41 LT | @-81 8T see TySU9y : oe e oh 8 6 6 6 or | £8 eel avjourerd puossg | “g | LT | G9L GOL | GALT LT 61 8I LT SI LT LT oe) YSU] | ee LT nes SI | 9-41 61 81 LI | 8:81 OTe 59:9 Nea einen ti cose qeourerd uoytutuee Jo yIpLerg | °G | on el - oT el FL TI el eT G.8T ral sere meee e seen setae wane essseee aefouerd WIIOFTUTUTA JO qqsIoy a | | oe SAG |) AE ALE BIS TG | 8 kh 9 Ag v s 6 ‘T | | | Z Z ie Jere el ae as ta. = ee i 5 = > Z | | | G.S& 0s O€ | GCE GE | 6S | aze! |c-08 Og | &-8E GE | GLE savpouedd Jo yysuey [BIO], | “11 | | on0 ¢ 8.5 Boo boo BB @ Bot 260 ** ss0U Wold} | +e G : : QF ee bee QF CF : ee see oo" Uypvadq - AOSTOUL [CAO}VT | “OT | $ 0 3.8 000 060 Cup CF 000 000 sss g Brat | eee wae aoe GC. . eee G.G Bod wisie ssouyorqy a ono n00 000 1.8 0 bl 000 Boo ‘ Ygpvatq > LOstoul UBIpe}l | “6 . GL see Goo 86 aee wee 50 500 eee eee eee e eee teens been ceeeee sereereees QUSTOq, eg |e6 |cor|se |s6 |¢6 |ee |e6 |68 (86 |68 fo ee aejout pany | °g 01 |36 |S80OL | G8 |80l | 86 | SOL | G01 | 66 | 96 | 86 |. * ays] 3 Pal || Hee |) HSL SIE |) TES | ASE | 8 @.- | Dt BCE EGE cee axejou puooeg | *y | gt | GSE | L8t | SL | G8t | Get | 41 | SST | SL | Sst | st yy8u9y | €1 | G3l | S&L | 82L €l | GSl | Let | 48 FL | SEL | GST el aepour yeti | “9 91 | 81 | 8-81 | 6L | SSL | GST | G6L | G6L | ST | 66T | GST ea acer é IL | GZL | 820 | 401 | LIT | IL | SIE | GL | Ger | ef | SIL * (gstoaster9) TIPE | seromeard pucoag | “¢ FL | ZOT |Get | oGT | 69. | Srl | GL | 2ST | 91 | GAN | GPT fo sSeeeage (qsod-que) 43809] | ae lee leo lao | 2a leq ee WG) eG 173 8 fe _ oeressresd) ap eet | awjouread asa | “> ect | gt | sz | Let | 621 | 91 | eon | (otl| sor | 21 |} 20 Seon eo ates (-ysod--ue) que] ; ap | pen | eer lor oer | me | Sry |6cr | eer | tr | rr “7s. sau[OUr Jo Wy8Ue] TOT | *S | = | gor hi [ctl| °° | 9-90 | 69T | eer abo see tees (4sod-"yuw) euro jo yApeoerg | “Z Se GL |. GE | = me rei |) ee |] 2° Sle eee srrrseeeees gutta JO 9USIOH | “T metas “mul “CuLUL mocgeas paged macgons mbedaed “Utd mecgees “ut “uu | “Us |) WS | 8 h 9 te Y 3 x ele ese “MVE Widd y— VY ‘snunzou snoaypdosdomvg JO 4399} JO SUO!SUSULIP SUIATH)—P WAV, PRIMATES FROM MADAGASCAR. 87 Description of the Skull of Mesopropithecus pithecoides. The main features of the skull of Mesopropithecus may be briefly summed up. ‘The brain-case is rounded and proportionately more voluminous than that of Palwopro- pithecus, and in its general conformation closely resembles that of Archewolemur. The frontal region is prominently convex, and: the orbits comparatively small and forwardly directed. The muzzle is broad and truncated. The zygomatic arches are high and widely spreading, giving the skull a rounded form not seen in the present Indrisine. The profile is even more Ape-like than that of Archwolemur (text-figs. 14 62 IIS) eyncl Jel, OCG) Text-fig. 14. Profiles of (A) Mesopropithecus and (C) Archwolemur compared on the one hand with (B) Jndris and on the other with (D) an Old World Monkey. The most satisfactory method to be followed in describing the skull of Mesopro- pithecus will be to compare it on the one hand with Archwolemur, and on the other with its nearest extant allies Propithecus, Indris, and Avahis. Such a comparison at once reveals the fact that the fossil form is robuster and more massive in every way than the recent Indrisine. Not only is it absolutely larger in most of its dimensions, but the various crests and processes are much more strongly developed. The comparison also shows in some respects a closer resemblance to Indris than to Propithecus. A quite young skull of Mesopropithecus also bears a striking likeness to Avahis in general conformation (text-fig. 15). 88 DR. H. F. STANDING ON SUBFOSSIL The Occipital Region —The occiput of Mesopropithecus is but little inclined from the vertical in the adult skull, though a young specimen just cutting its canines has a much inclined occiput, the plane of the foramen magnum making an angle of but 35° with the horizontal plane. In the adult the conformation of the occipital region shows a striking resemblance to that of Archeolemur edwardsi. The foramen magnum is circular or slightly Text-fig. 15, Skull of young (A) Mesopropithecus compared with that of (B) Avahis. elongated in a horizontal direction. The occipital condyles are widest near their upper extremity, the line of greatest transverse convexity here running obliquely outwards and upwards. In a young skull the median convexity for the lodgement of the central lobe of the cerebellum is strongly marked as in recent genera, but in the adult the condition approaches that of full-grown specimens of Archwolemur edwardsi. The paroccipital processes also resemble those of Archwolemur, being blunt and pyramidal in form. ‘The lambdoidal crest is prominent, running backwards in a horizontal direction in a manner closely analogous to the condition seen in Archeolemur. The supraoccipital enters to but a small extent into the roof of the skull, resembling in this the majority of the higher Apes. Among the existing genera, Avahis comes nearest to the fossil in this respect. The lambdoidal suture has a feeble anterior convexity, the curvature being approximately parallel to the lambdoidal crest. The Parietal Region—In the parietal and frontal regions further resemblances to Archeolemur are noticeable. The brain-case is globular or rather pyriform; it is more elevated than in either Indris or Propithecus and narrower in the postorbital region. The parietals form a more considerable part of the roof of the brain-case than in either of these recent genera. The upper lines of attachment of the temporal muscles are either closely approxi- mated along the middle line, or actually meet, forming a more or less strongly marked sagittal crest. This crest is continued forward as far as the coronal suture, when it PRIMATES FROM MADAGASCAR. 89 divides into two curved ridges, which spread widely to form the posterior edge of the raised frontal convexity, and are continued with slight interruption to form the outer edge of the postorbital bar. In adult animals the median suture is closed along the whole length, in the parietal, frontal, and nasal regions. ‘The spheno-parietal suture is similar in length and position to that of Propithecus. The outlines of the squamosal and the extent to which it rises up along the skull-wall closely resemble the condition observed in the modern genera. The Frontal Region.—The form of the frontal and orbital regions of Mesopropithecus sharply differentiates the genus from the extant Indrisine, while closely allying it with the Archeolemurs. Different skulls show considerable diversity as to the elevation and amount of convexity of the frontal region between and behind the orbits. In the case of the specimen figured on Plate XXI., this part of the skull is raised so as to be actually higher than the central parietal region, but in every instance the interorbital part of the frontal bone is much more sharply curved down- wards than is the case with the extant members of the suborder, thus giving to the profile a very Ape-like aspect. A frontal sinus is present, comparable in extent with that found in Jndris and Propithecus. A noticeable feature of Mesopropithecus, and one which we have seen to exist in all the recently extinct subfossil Lemuroids, is the narrowing in of the postorbital region of the frontal bone. ‘The extent of this constriction does not, however, in Mesopropithecus appear to be so great as materially to affect the frontal lobes of the brain. There is, in fact, a very close resemblance between the brains of Mesopropithecus and Propithecus *. The Orbital Region.—Compared with the existing Indrisine the orbits of Mesopro- pithecus are proportionately smaller and more forwardly directed, the postorbital bar is stouter and more distinctly triangular in section, and the malar region while not so deep and strong as in Paleopropithecus is considerably more developed than in Propithecus or Indris. In the case of the last two genera the orbit has so great a lateral extension and the zygoma so flattened a curvature that when the skull is viewed from the front the zygomatic arch is hardly visible. In Mesopropithecus, on the contrary, the zygomata spread widely away from the brain-case beyond the external orbital rim, in this respect again closely resembling Archwolemur (text-fig. 16). The development of the malar region in conjunction with the elevated frontal and steep facial angle must have given to Mesopropithecus a more Monkey-like face than that even of Archeolemur. * See Appendix by Elliot Smith, infra, p. 168. 90 DR. H. F. STANDING ON SUBFOSSIL Text-fig. 16. EWS lax _ Ww) RS seit TQ Front view of skulls of (C) Mesopropithecus and (E) Archwolemur, compared with (A) Propithecus and (B) Jndris, and also with (F) Orang-utan and a nocturnal New World Monkey (D) MNyetipithecus. PRIMATES FROM MADAGASCAR. 91 In the following table the transverse diameter of the orbit of Mesopropithecus is compared with the corresponding dimensions of Archwolemur, Propithecus, Indris, and Avahis respectively. This is also expressed as a percentage-index of the bi-zygomatic breadth of the skull in each case. Archeolemur, Mesopropithecus. Indris. Propithecus. Avahis, Diaimeterrotoulitaeaeni ey anes 29°5 21 23 21 15°5 Orbital index: [> Glam 1100 ] . oF 25 373 342 41-1 i-zygomatic breadth The orbital rim of J/esopropithecus is less raised and everted than in the two modern genera most nearly resembling it in size; but the postorbital bar is, as stated above, much stouter and stronger. ‘There is the same turning in and extension of the sharp posterior margin of both the frontal and malar portions of this bar as occurs in Pateopropithecus and Archwolemur. The infraorbital foramina occupy a similar position to that seen in Propithecus. The shape of the lacrymal bone bears a considerable resemblance to that of Paleo- propithecus, though it has a smaller extension within the orbit and does not completely enclose the lacrymal fossa. The large size of the raised inner margin of the orbit is again another character recalling Palwopropithecus. ‘The position of the lacrymal foramen can be best described as being in line with the sharp inner anterior margin of the orbit, though the fossa itself is almost entirely within that margin. There is within the orbit a fronto-maxillary suture separating a small os planum from the lacrymal. ‘The posterior portion of the maxilla forms the floor of the orbit, but does not extend backwards so far as in the case of Archwolemur, being intermediate both in extent and inclination between the condition of this latter and that seen in Propithecus. The Nasal Region.—The size and shape of the nasal bones strongly recall the same region in Archeolemur, though, as before stated, the median suture is entirely obliterated in full-grown specimens, The upper part of the fused bones forms an upwardly curved suture with the frontal, the lower outer extremities of this curve nearly coming into contact with the anterior angle of the lacrymal as in Paleopro- pithecus. The lateral margins of the nasals gently curve outwards, this curvature being accentuated towards their anterior extremity in a manner exactly analogous to that noticed in Archwolemur (text-fig. 17). ‘The lower external margin here makes a very short suture with an upward extension of the premaxilla. This condition may be seen in Propithecus, though apparently not in Jndris or Avahis. We have seen that in Paleopropithecus this upward extension of the maxilla is external to the anterior margin of the nasal bones. The premaxille are firmly fused along the middle line almost to the alveolar margin, VOL. XVIII.—Part 11. No, 5.— May, 1908. ) 92 DR. H. F. STANDING ON SUBFOSSIL The median incisors have thus been brought into close proximity, in all probability, indeed, meeting in the middle line. Unfortunately the incisors are missing in all the specimens, but the alveoli give us indications for determining their position and direction. ‘This region of the palate is more strongly developed than in any of the Text-fig. 17. a A Cc a € ie Nasalia of (a) Archceolemur edwardsi and (6) Mesopropithecus pithecoides, compared with each other and with those of (c) Propithecus and of three Monkeys (d) Pithecia leucocephala, (e) Pithecia rufiventer, and (f) Chiropotes ater. extant members of the subfamily, and it is very probable that the front teeth still functioned as true “incisors,” being used for biting and nibbling, as in the case of Archwolemur. The nasal aperture is almost circular in form in the one skull where the margins of the encircling bones are intact. The Temporal Region.—As stated above, the zygomatic arch spreads widely away from the skull, the malar process taking a somewhat upward direction as in Archeolemur. This malar element does not extend backwards along the whole lower border of the arcade, though forming a longer horizontal suture than in Palwopro- pithecus. ‘The anterior part of the arch has a much smaller vertical extension than the posterior, the maximum and minimum heights being respectively 16:5 mm. and 7mm. ‘The whole arcade has a far greater thickness than that seen in any of the extant genera, its conformation, indeed, presenting a strong contrast with these latter and closely allying it with Archwolemur. ‘The zygomatic process of the squamosal bears some resemblance to that of Propithecus, but it is again more massive in every way and differs also from the latter in having a much greater forward extension of the horizontal plate which connects the skull-wall and the vertical portion of the arcade. This fact, combined with the very convex curvature of the whole arch, has given at this point a nearly circular horizontal section to the temporal muscle, again recalling the condition of Palwopropithecus and Archeolemur vather than that of Propithecus and Indris, where the temporal fossa is elongated in a direction parallel to the long axis of the skull. The sharp superior margin of the zygomatic arch is continuous by a long flat PRIMATES FROM MADAGASCAR, 95 horizontal ridge with the flange-like lambdoidal crest. In Propithecus a line may be traced running horizontally backwards above the auditory orifice to join the superior occipital ridge, suggesting the former existence of a similar crest. In Mesopropithecus large auditory bulle exist, similar in shape and proportions to those of recent Indrisine, and there is in old skulls a fusion of the postglenoid process with the bulla, The Basilar Region.—The obliteration of the sutures observed in the roof of the cranium is again seen in the case of the bones forming its floor. In the oldest skulls hardly a trace remains of the occipito-sphenoid, sphenoid, maxillo-palatine, or maxillary sutures. The external pterygoid lamine are widely everted at their posterior margins ; the glenoid fosse are broad and deep, resembling those of Palwopropithecus. The postglenoid process is broad and strong, and has also a greater vertical extension than in Propithecus. In the palatal region we again see evidence of the stouter and more robust make of the fossil form as compared with all its existing relatives. The posterior margin of the palatines is three or four times as thick as in Propithecus and Indris. ‘The central portion is depressed (in the direction of the vomer) and in young animals is produced into a short postpalatine spine. The forward extension of the palatine is intermediate in extent between the condition seen in Propithecus and Indris. The interior part of the palate is much broader than in any of the recent representatives of the subfamily, the distance between the internal margin of the anterior pair of premolars being in excess of that between the first molars. In all recent genera the dental series converge towards the anterior part of the palate; in Mesopropithecus they are almost parallel. Dentition.—The dentition of Mesopropithecus so closely resembles that of Propithecus and Jndris that a short comparison with these modern genera will suffice (see Pl. XXI.). As already stated, the inner pair of incisors in Mesopropithecus, while larger than the outer, were more nearly equal than in Jndris. So far as it is possible to judge from the size, position, and direction of the alveoli, the incisors of Mesopropithecus were intermediate in function between those of Archwolemur and Propithecus. ‘There has evidently been some variability as to the position of the external pair of incisors— in one case a short diastema occurring between this tooth and the canine, while in another skull the alveolus is confluent with that of the canine. The same variability in the position of the outer incisor is seen in Palwopropithecus. The canine in Mesopropithecus is a large and powerful tooth, having a horizontal extension of nearly a centimetre and a breadth of 4:5 mm. It has a large internal talon overlapped by the anterior margin of the second premolar. ‘This latter and the following tooth bear a strong likeness to the corresponding teeth of Palwopropithecus, though the greater prominence of the backwardly sloping central external ridge and the development of the small labial cusps ally these teeth even more closely to those of Propithecus. 02 94 DR. H. F. STANDING ON SUBFOSSIL The first and second molars by the strong development of their acute internal cusps are again closely related to the three recent genera, though, on the other hand, the flattening down of the small lingual cusps of these teeth recalls the condition of Paleopropithecus. The third molar is much reduced as in all the Indrisine, though its antero-internal cusp is less vestigial than in most of the other genera. Text-fig. 18. Mandibles of (A) Mesopropithecus pithecoides and (B) Archwolemur edwardsi, viewed from above, and compared with those of (C) Indris and Old and New World Monkeys (D) Colobus and (1) Pithecia. Turning to the mandible, we find here again a condition intermediate between that of Archwolemur and Propithecus. The massive character of the whole jaw, the great vertical height of the body of the rami, the strong development and almost vertical direction of the anterior margin of the ale, the form and size of the coronoid process, PRIMATES FROM MADAGASCAR. 95 the broad condyle, the wide-spreading angle of the lower margins of the mandibular rami carrying the lower edge of each ramus outwards to such an extent as to cause the posterior molars to overhang within the jaw, such are some of the characters in which Mesopropithecus approaches Archwolemur and in most of which also it resembles the ordinary Simian type of mandible (see text-fig. 18, where the mandible of Mesopro- pithecus is compared with those of Colobus and Pithecia). The mandibular symphysis has about the same relative length as in Propithecus. The lower incisors are missing, but the alveoli show them to have been somewhat rounded in section and not laterally flattened as in the existing genera. ‘The outer pair were larger than the inner and the alveoli of the inner pair are decidedly anterior in position to the outer, herein again resembling Archwolemur. All four incisors appear to have been somewhat procumbent, the angle approximating to that seen in Propithecus. The anterior premolars are missing in our specimens, but the remaining teeth bear a very close resemblance to those of Jndris and Propithecus. The first premolar is much everted at its anterior margin. The three molar teeth call for no special notice, except to state that they are in general stouter and broader than the corresponding teeth of Indris and Propithecus and in their conformation show affinities with both these genera. TaBLE 5.—Showing affinities of M/esopropithecus with recent Indrisine and with Archeolemur. Archceolemur Mesopropithecus Propithecus edwardsi. prthecoides. diadema. Constriction of postorbital region of frontals Obliteratronvorusuturesseeere eel aelieociee: Presence of strongly developed crests .......... Broadvoceipital condyles 5.5 sass ss nese Sizerandyshapelorsbulllacws. rrr corer eie cielerairel el SHAD Ol CHOW Goo cocooncdbe Woo phe Mamaaelon Vestige of postorbital septum ................ Nizenolzyoomabicranchesmumrnrrr Tier ier Positionvof lacnymalifossaya.ciec eae ae Hacialvangle nuns Wewenn retake, a eieponils cers sleaialewatesis Procumbency of lower incisors ...............- Mentaliformiulagey vera eyecare oc aessiaeun ey seaaiev ese Parallel disposition of dental series ............ Contact of upper median incisors ....... glbedone SINOANETS OH WN 5 a00060d0050090000000000 Shapeloimasal Suen ry merrier iiceiicertercr or: Angle of symphysis of lower jaw .............. Shape of condyle of mandible ................ * ee ROH: 96 DR. H. F. STANDING ON SUBFOSSIL TasLe 6.—Giving dimensions of four skulls of Mesopropithecus pithecoides. 1. 2. 3. | 4, 1. | From lower margin of foramen magnum to alveolar border of pre-|} ™m. | mm. | mm | mm. MAK atc a awe eae oe LE ene ei reiee 85:5) 81 87-5 2) |. Maximum\lengthyofiskull nesses: er ene aeneeae re Coenen: ae 101 94°5) 101 3. | From lower margin of orbit to upper margin of foramen magnum 58 69 67:5) 70:5 4, | From lambdoidal crest to fronto-nasal suture ............e-eeee eee 63 (75) | 75 81 5. | Length of palate along median line.............. cece ee ee eect eens as 40 [35] | 40-8 6. | From lower margin of foramen magnum to posterior edge of palate....| 33 45 | [46] | 47 7. | From lower margin of orbit to alveolar border of premaxille ........ é!0 33 33 33°5 8. | Length of series of molars and premolars ..........+eeeceeeereees 32 [32] 34 34 9. | From anterior surface of postglenoid process to posterior edge of third molar ieee acc ci vceaetere ne mre ina eer eicmiine eee ee 19 30 | [28] | 380 LOM EWerticaludiametermotsonbitsmerienerer ricer eee error 18-9 18 20 Ha | sebransversendiameter olorbitsmey seria eiicrnetieeinr ler U7 O° 18 20 12. | Maximum vertical diameter of brain-case (external) ................ | 83:5] 38 37 39°5 i 3ea eebi-zyeomaticibreadthy external) maser te ceieicume teeter: | 57-5!) 76 72 80 14, | Maximum external breadth of brain-case ..........0. 0.00000 eee ee 42 48 48:5| 49-2 15. | Breadth between styloid processes ........-...--00--eeeerere eee | 26 [29] | [28] | 29:5) 16. | Transverse diameter of foramen magnum ............--++ eee eee 14 14°5 | 15 12 | 17. | Vertical diameter of foramen magnum ............--+.--+-++-+s6 HA 5) alee 12 12 18. | Distance between orifices of auditory meatus...............---00-- 36 36 35 35:5 | 19, | Breadth of skull at level of postorbital constriction ................ Dy 30 29 27-5 90, | Breadth between lacrymal foramina «...........-.--0--e esse eeee 14 185] [19] 18:8 21, | Breadth of palate between third molars (internal) ...............+45 15 [19] | 20 19 92, | Breadth of palate between antero-internal margins of canines ........ AG [20] | [16:5] 20 CoM eMaxinmummlenothyotmasaliaherentiaineerrri rir ecicr trick ei cir ue] i) ea) eal) 2a Oe Maximum bread thvoranasaliageyanreweeyeimensetyiel siento lite ern [9] 15 bis 16 Oren eMEinimumybreadthvormasaliagenn errr vsti aiid etliare ec 6 8 [8] 9:5 96, | Maximum width (height) of zygomatic arch .............. 0+ ese e ee! 10 7 17-5 TaBLeE 7.—Giving dimensions of upper teeth of Mesopropithecus pithecoides. TS] 28 Soames | mm mm mm mm. Walled (@Hteishtrol canines cya ler ricrle eiareucrciontoraractciereuireirciackarstelcrala= : ae Se 14 OF |) Wrekin Ge Cavan Ch) Go oocosooooacos0ddn00000 0000 000000 30 ab Oa 8 Su lotaldensthvotsmolars!- ymin ei eeicienorn eo: 21°56 | 21-6 | [25] | 22 go = a 4 [Dee piemilet (a ien ceenena) meme 2 |) cel p, | Secondprebolat | yoann ranerone) ee | Be | Smua [ence |. oe mn ic. |) 75| 78 q Send aclet | ye 212k eeemmne Ce ours ae B.(Mthied moles | yrenmin’ 2 a anamna nme Jes| 8 si le PRIMATES FROM MADAGASCAR. 97 Text-fig. 19. IN Restoration of skull and mandible of Mesopropithecus pithecoides. Subfamily ARCH HOLEMURINA. Subfossil ape-like Indrisidee which retain the largest number of ancestral pithecoid characters. They retain the “primitive” Primate dentition in the upper jaw, though there is reduction of the third molar. The lower jaw is massive and has many Simian features; it has lost the canine of the permanent dentition. Upper incisors large and functionally important. Lower incisors show incipient tendency to pectinate disposition, which is characteristic of existing genera. Lower first premolar caniniform. Skull pyriform with narrow postorbital frontal region, long widely curved zygomata, and large auditory bulla. Lacrymal fossa on or near orbital margin. 1.2: ¢. 1: pm. 3: m.3 i 22) cOsppmia sms 3) Dental formula : Synonym: Nesopithecine (Major). Genus ARCHAHOLEMUR (Filhol). (Plates X VI.-XX.) Definition. —Subfossil ape-like Lemuroids with pyriform brain-case, showing marked postorbital frontal constriction. Occipital and sagittal crests frequently developed and interorbital frontal region prominently convex. Incisors and _ premolars functionally important, the latter having a blade-like disposition in both upper and lower jaws. Lower first premolar caniniform ; third molar reduced in both jaws. i, 2: ec. 1: pm. 3: m.3 i. 2: ¢. 0: pm. 3: m. 3° Synonym: Nesopithecus (Major). Dental formula: It is not my intention to give a full description of the fine series of skulls of 98 DR. H. F. STANDING ON SUBFOSSIL Archeolemurs which have been found at Ampasambazimba, since both Dr. Forsyth Major and Dr. Grandidier have memoirs in hand on this subject. It will be necessary, however, for the purposes of comparison, to point out the salient features of the two species represented in the collection of the Malagasy Academy. I shall also in the present section give a table of the chief dimensions of the complete series of 10 skulls from Ampasambazimba, since the whole material will probably be no longer con- veniently available for purposes of comparative study. It should be pointed out, in the first place, that, as in the case of Palwopropithecus, there is a great variety in the size and proportions of this series of skulls. The task of classifying them has been especially difficult from the fact that although various characters may be pointed to as constituting marked differences in different individuals, these individuals cannot be classified as showing simultaneously the same set ot variations, and, moreover, intermediate types occur even in the case of the most extreme divergences. As an example of this difficulty we may consider the condition of the third molar. The reduced triangular shape of this tooth has been used by Dr. Grandidier as forming the chief criterion for the determination of the genus Bradylemur. \n the series under consideration three skulls show a distinctly triangular form of this tooth, two have large quadritubercular third molars, while four show yarious conditions intermediate as to size and shape between the first and second groups. Again, the attempt was made to use the facial angle as a distinguishing character, and it was found that seven skulls had a long, more or less pointed muzzle, so that a line tangential to the alveolar margin of the premaxille and the frontal convexity between the orbits formed approximately an angle of 60° with the plane passing through the crowns of the molars and premolars. In the case of three other skulls the line so drawn was more abrupt, giving an angle approaching more nearly to a right angle. Again, the proportionate length of the series of molars as compared with the premolars has been calculated for the whole series of skulls; this expressed as percentage-ratios gives numbers varying from 98:4 to 85°6. In one case, however, the general aspect of the skull is so strikingly different in many important respects from all the rest of the series that it seems justifiable to make a specific distinction in this instance, and the name platyrrhinus is proposed for this skull. Its distinguishing features will be enumerated after a brief general description of the other members of the series. I have, through the courtesy of Dr. Smith Woodward, had the opportunity of comparing these latter skulls with the type specimen of Archwolemur edwardsi in the British Museum and have no hesitation in referring them to this species. Description of Skull of Archeeolemur edwardsi (Fithol). (Plates XV1.- XVIII.) The skull of Archwolemur edwardsi bears a close general resemblance to that of Archeolemur majori, but besides its greater size and more robust proportions, it differs from this best-known species in several important particulars ; these will be referred to in detail during the course of the following description. PRIMATES FROM MADAGASCAR. 99 Occipital Region.—The plane of the occiput is somewhat less inclined than in the case of A. majori. ‘The occipital crest is in most specimens very strongly developed, rising up in a flange-like projection nearly 1 cm. in height. It is continued as far as the auditory meatus. In some skulls a raised line connects the occipital crest with the posterior and upper margin of the zygomatic arch in a manner suggestive of the condition seen in Mesopropithecus, though the two bony flanges are not confluent as in the latter genus. ‘The position of the occipital condyles does not materially differ from the condition observed in A. majori, though they tend to occupy a larger part of the circumference of the foramen magnum and to be somewhat broader in their upper portion in the larger species. Parietal Region.—In old skulls of A. edwards? the brain-case has the appearance of being more depressed and relatively less voluminous than in the smaller species, but probably this is in part due to the great development of the sagittal crest and the very prominent frontal convexity which many of these skulls show. A small raised triangular area is present on the supraoccipital, and from the apex of this triangle the sagittal crest runs forward as a prominent blade-like ridge (Plate XVII.). Frontal Region.—About 1 cm. in front of the coronal suture the edges of the sagittal crest rapidly diverge to form the posterior margins of the prominent frontal convexity above referred to. This feature forms one of the most marked differences between the two species. ‘The central frontal region of A. edwardsi behind and above the orbits is raised far above the level of the brain-case, forming a very prominent ‘“‘forehead.” ‘This feature occurs, as we have seen, in some skulls of Mesopropithecus. The profile of A. edwardsi is thus in this region more steep than that of A. majori. An extensive frontal sinus exists beneath this convexity, though it does not appear to encroach upon the brain-substance as in the case of Megaladapis. The steep backwardly sloping edges of this frontal convexity accentuate the appearance of extreme narrowness in the postorbital region. This, again, is a feature that occurs in Mesopropithecus. The frontal bone has the appearance in A. edwardsi of extending backwards to a greater extent along the skull-wall than in the case of A. majori; but measurements along the sagittal suture show that this is more apparent than real, being due to the anterior position of the frontal convexity to which reference has been made. Nasal Region.—The median frontal suture is obliterated in its upper part, but can be traced even in old skulls in the region between the orbits. The median nasal suture persists in every case, as also does the fronto-nasal suture. The nasal bones of A. edwardsi agree essentialiy in form and proportionate dimensions with those of A. maori. The upper external contours of the nasals slightly diverge till the fronto- maxillary suture is reached, below this point in some specimens a slight consiriction occurs, when the bones gradually broaden out again to meet long narrow upward projections of the premaxillz, the broadest part of the bone being quite aear its anterior extremity. The free anterior margins when intact follow first a very short VOL. XVIII.—PakT 11. No. 6.— May, 1908. P 100 DR. H. F. STANDING ON SUBFOSSIL forwardly directed course and then converge at a very obtuse angle, finally forming an acute indentation as they meet each other along the median line (text-fig. 20, A). Text-fig. 20. Showing form of nasals of (A) Archwolemur edwardsi and (B) A. platyrrhinus. Orbital Region.—The orbits of A. edwardsi by their size and position closely resemble those of A. majori, though they are in many cases somewhat more forwardly directed, owing to the more forward position of the orbital process of the frontal. The shape of the orbital contour varies considerably, not only in its general proportions, but in regard to the exact position and direction of the prominent marginal convexity, and at times the absence of such a definite raised rim has made it difficult to estimate the real diameter of the orbit. One important and interesting feature of the postorbital bar is the fact that its posterior edge shows a sharp inner blade-like projection, in many cases so thin as to be quite translucent. On some skulls a raised line continues this edge for some distance on the imner frontal wall of the orbit. The probable significance of this feature is explained in the last section of this memoir. The lacrymal foramen is more distinctly within the outer margin of the orbit than in the case of A. majori, though this is not due to any material difference in relative position, but to the more definite and prominent development of the anterior margin of the orbital rim. Temporal Region.—The zygomatic arch is in general more massive than the corresponding structure in A. majori, though there are great variations in its size and robustness, one skull showing a maximum vertical height of over 16mm. ‘The malar has a great vertical depth below the orbit, a feature which we have seen to be still more accentuated in the skull of Palwopropithecus. The lower margin of the zygomatic process of the malar is directed abruptly upwards, forming an angle of 45° with the horizontal plane. | The inner contour of the squamosal element of the zygomatic arch is not so evenly curved as in A. major7, its anterior border forming, as seen from above, a straight line 1 cm. long at right angles to the long axis of the skull (Plate XVIL.). Basilar Region.—The auditory bulle are large and prominent, but show considerable variety in shape, size, and relative position. In one skull a deep cylindrical channel exists between the roughened surface of attachment of the rectus anticus major muscle PRIMATES FROM MADAGASCAR. 101 and the summit of the bulla, in other cases the condition resembles that seen in A. majori. Very great differences exist in the proportions of the various regions of the skull, the extreme examples diverging so widely as certainly to justify their classification as distinct species did not intermediate forms exist which make such a differentiation impossible. We have seen that the genus Palwopropithecus presents the same phenomenon, ‘This extreme variability would appear to be a mark of species approaching extinction. Dentition.—The great variability in the dentition has already been referred to. The occurrence of a triangular third molar was, as above stated, considered by Dr. Grandidier of sufficient importance to form a new genus, Bradylemur; but we have seen that in point of fact even specific distinctions cannot be based on the character of this tooth (see Plate XVI.). The first and second molars also show considerable variations in form. In some cases the transverse furrow which separates the tooth into anterior and posterior halves is continued by a strongly marked vertical groove, which again is confluent with the interstice between the anterior and posterior roots of the tooth. In some specimens, again, this vertical furrow is entirely absent, the tooth presenting one unbroken convex internal surface. Here, too, however, an intermediate form exists, so that it is impossible to use this character for purposes of specific distinction. Reference has been made to the varying proportions and the length of the series of molars and premolars respectively; the fact of this variation will be apparent if expressed in tabular form thus :— Length of Length of Ratio of molars three molars. three premolars. to premolars, mm. mm. LA elanpemmreencdce Ae Oho ee mencatctat ROT Wecene ay. 98 PO aileaenene sir umenras PH (SY, enn napa QO Ualaceteentat 1:10 S) wiepaaateeccanarane ee OION. © Sveaenh ss 205s | ls pedaemas 90 AN a citaisigaiatiostar OA nanan atk PA) Bee tesoanon as 91 Dita deepest aa aan oe QO: MPT canoes 2 ONE a Tere aten 92 Gites HON eres Oy ei oy Re aatere. st ishing aaeNunyend 94 Cetra sroiatshe sare ai OAM AUS cis: Sele ae CASEI saknodbon 85 bo) ihe a wea aurea PALO W el rad eas PHPAG, | eeasecvob 89 DER eto nat 2025 a ssacucs Pisktovel pe amoadhoee hil The amount of backward extension of the whole dental series again is subject to much variation, the third molar in some instances almost abutting on the pterygoid fossa, while in other cases a distance of more than a centimetre intervenes. The interval between the third molar and the auditory bulla varies in different specimens to the extent of nearly 15 mm. ‘The posterior margin of the palate has a deep U-shape, its central portion being so far forward that it is generally opposite the first molar. The individual teeth closely resemble those of A. majort, the chief variations occurring in the molars. ‘The median pair of incisors are stronger and larger in proportion P2 102 DR. H. F. STANDING ON SUBFOSSIL than even those of A. majori and are more vertical in position, bearing indeed a very close resemblance to those of many of the Cercopithecide (text-fig. 21). The outer pair of incisors are small and at their inner margins are turned in to such an extent as Text-fig. 21. Upper incisors of (a) Archeolemur edwardsi compared with those of (b) an Old World Monkey (Macacus cynonologus) ; ¢ and d, the same, as seen from below. to be slightly overlapped by the median pair, in this condition also not greatly differing from Papio and Macacus. The incisors of the lower jaw of A. edwardsi were apparently more cylindrical in shape and less inclined than those of A. majori, though varying degrees of inclination may be observed in different specimens. Text-fig. 22. Showing different stages in evolution of pectinate or scraping instrument formed by the lower incisors :—A. Palaopropithecus maximus; B. Archwolemur edwardsi; C. A. majore ; D. Propithecus; BE. Indris; F. Lemur. These lower incisors are much more massive teeth than those of Hadropithecus, and their worn surfaces show clearly that they were still directly opposed to the upper incisors, and were used for biting just as in the case of the Cercopithecide. It should be noticed in this connexion that A. majori has apparently gone further in the direction of that specialisation of the lower incisors which in the case of the extant genera of the Indrisinz and Lemurine has resulted in the production of a tool of pectinate form used for scraping and rasping and also for combing the fur. ‘The different steps in PRIMATES FROM MADAGASCAR. 103 the production of this peculiar modification are seen in the accompanying diagram (text-fig. 22). Dr. Grandidier has in his thesis on fossil Lemurs pointed out the various simian characters of the mandible of Archwolemur. These need not here be recapitulated. M. Grandidier regards these simian characters as evidences of specialisation, and apparently inclines to the belief that they are the result of convergence, though he frankly confesses he is unable to explain their occurrence. I think there is much more reason for regarding these simian features as general ancestral characters and Text-fig. 23. ff il Ni AN) 7 SN \ == Cranium and mandible of Archeolemur edwardsi, restored. the condition of the recent genera of Malagasy Lemurs as specialised. But for the full discussion of this question the reader is referred to the last section of this memoir. ‘The very Ape-like character of this genus may be clearly seen from the accompanying restoration of the skull and mandible (text-fig. 23). ARCHAOLEMUR PLATYRRHINUS (Standing). (Plate XIX.) It was stated above that one of the skulls of ssouyoryy | \ CF Bee te EF ood * qjp¥eaq - Aostout [euto}xXmy | “f GSI. oon oe wee eee Il BC Oe ed wanes qqsteq, 9 o90 S| 6 68 68 9:8 68 G8 @sf, |p on pera | : : 5 is 6 8 Cg LG 68 9.8 so lo + yysa9y aepOU pang, | °9 ra e.g 6 ce c.g 6 8.6 ma 6 Pa “+ aapeaaq : ee eg ou dio g 6 ag 2.6 ; ot eG SET coe Hp E (I) SeAES Ht A 6. 6 8-6 OL 6 G6 6:6 G6 G.8 8-6 “ Uypeeaq : 8 6 8 8-6 6-8 6 or | 6 [or] st eeces hu) s10 ae and me 8 G8 Or 6 G6 6-6 86 OL 8-6 Sep a ees co aypeedg : 0.6 int ol Oo | 66 | ei | 2B or or | aor Pe ce ee ccqeomnenel ora, | ig G8 8 9-11 8 g-01 OL GB G8 CG ee ee oe 8 GH HOE OL 6-8 8-01 9.01 OL 6-6 .TT oe yy Sey ae puoseg | *z G8 te Or a II 6-01 GIT GOL IL GIL Be eee IE ond : 2 GI ban Il at GALL GI CLI C.9T Cait GT serene “gy Steq f apjouerd ysanq | *T Ey pe a ee = ae sh a= — n 5 ‘OT 6 8 74 9 G v § 6 ae aa = —_— — — on . =) HIGIGNV I — 7 mM A iS) 8-16 86 §-L6 96 4G &@ 96 FG G-66 GG “+ scepourord Jo 492] [eIOL | “CL | oF LY BE See ae F : GL [cts esoUHOrT 2 $-9 G $3 se 3 g G.G yypreiq | aosrout yetoyvy | “TT = [8-F] I OL LAL @8 ay sro a fut} SL tag G zee seoTOIy A Or 8:8 1-8 : G6 ae ae : rrresesees HapYaIq: — AIOSTOUL ULIPETL | “OT x cea BL oct FI 206 ove : IL o 300 56000000 qys1oy G: 8 8: 8 13. , 6 G8 Fh 8:8 $3 : see MEDEERGL Cs ‘ : = fal GL gL 9.9 a) 9 9-9 G9 Pl ieee |e eee conse epeueyy on A) Fy OL G-OL 6-8 G6 OL 8 F-0L OL 9-01 8-6 ygpearq \ ; 5 6 6 6 84, 9-1 18 fi Sept 8 Z.8 Gg yy 8u9T LeTOUL puodag | “g a Lor | GIt | 80L | 201 II OL I iat I 01 MPLA gepoor yall | *L 3 6 C.OL 8-8 G.g 9.6 G.g 6 8.8 2.6 6 OG000000 600000 es Yqysuey 5 II 8-61 IL IT L-GL ¢-01 CIT GIL ; IL OSL9ASUBLY ee 2 dpa . a aK OE 8 | 18 or | 08 | ¢6 | 98 | o6 | 96 ee ee ele |? 6h 8-8 Gs GBT 6 G.), 8 G.8 8-8 6 Be OSIVASULAY) TP BEIQ F ad 1-6 GIL SS. OL GL 8-01 8-6 IL GI (OH) fated [teecs tees es are ae el ATTOMIcd Puosesg | “G G8 6-8 6 6 8-9 6 6 OL 6 2 + (QsqoAstvay eee 2 ual OL GIL 16 L-01 eI BG 8-01 G-01 él It (qsod—que) yySuey f TVIOMONS IH | “Y [ee] 9G FF, GZS G.GG G.8S GPS BRE GZ (irze__ |eseopeeeeeecsacseoasBonHacceAacons “++ gavjout Jo yysuer Ton | °g II GL ee TOL Il aay OL (BIE SL OL > (qsod-yux) eutnes Jo [ypVorg | *Z cee II a el at a: al et | ort | ST ee “egqurauo JO SIH | “T Bee eee es ee ee oe 5 : ee Or 6 8 4 9 iG v tes 6 aD ee er VITIXVIL—'V ‘ISpLOMpA MNWAOMYILPY JO YI} JO (SoLOWL[[LU UL) SUOISUEIIP SULATQH— ‘OT AAV], 106 PRIMATES FROM MADAGASCAR. 107 the skulls of 4. edwardsi, and the facial angle is consequently less steep. The nasals are short and broad (text-fig. 20, B, and Pl. XIX. fig. 3), and the upward extension of the premaxilla has a stout and outwardly curved margin. The nasal orifice has here a wide circular form, differentiating it materially from that observed in all the other skulls of this genus. The postorbital bar is very massive. ‘The external surface is broad and inwardly inclined, so as to form a partial postorbital cloison (Pl. XIX. fig. 2). The orbital margin is distinguished from that of A. edwardsi by being sharply defined on its upper and outer border; the orbits are thus not only actually smaller, but are more tube-like in form than in either of the other species. ‘The interorbital breadth of the new species is also greater than in either of those previously described. The incisors are missing, but their alveoli are not far in advance of the anterior margin of the nasals; the muzzle has thus a truncated appearance. ‘This feature, in conjunction with the small frontal convexity and absence of prominent sagittal crest, gives to the skull a very different profile from that of either of the other species (compare text-fig. 23, p. 103, with text-fig. 47, A, p. 161). Family LEMURID. Subfamily MEGALADAPIN &. Giant subfossil Lemuroids from Madagascar. Skull long and narrow. Brain-case depressed; brain showing evidence of retrogression. Orbits small and tubular. Muzzle elongated. Zygomata high and strongly developed. Mandible massive. 1.0: ¢. 1: pm. 3: m.3 1.2: ce.1: pm. 3: m.3° Note.-—The Megaladapine in many essential features so closely resemble the Lemurine that it seems unnecessary to rank them as a separate family. Dental formula: Genus Mreauapapis (Forsyth Major). Dejinition.—Gigantic subfossil Lemuroids from Madagascar. Skull long and narrow, with marked postorbital constriction, vertical occiput, and strongly developed sagittal and lambdoidal crests. Facial portion elongated, with great interorbital breadth and large aérial sinuses. i.0: ¢. 1: pm.3: m.3 i.2:¢. 1: pm.3: m. 3° Note.—Among ne one foul at Ampasambazimba are several skulls of a new species of Wegaladapis to which the name of Megaladapis grandidiert has been given. Before describing the distinctive features of these skulls it may be useful to recall in some detail the main characteristics of the genus which are common to all the species now known. : Dental formula: All three species agree in having a low flattened brain-case and a long narrow skull. The cerebral convolutions are feebly marked; there is a notable reduction of the VOL. XVIII.—PART 11. No. 7.—May, 1908. Q 108 DR. H. F, STANDING ON SUBFOSSIL frontal lobes; and the hemispheres leave the cerebellum uncovered. The occiput forms a plane at right angles to the axis of the skull and the occipital condyles are vertical. The facial region is greatly developed, large frontal and maxillary sinuses occurring in the adult. The face, indeed, is rather suggestive of a Ruminant than of a Lemur. The lines of attachment of the temporal muscles form a sagittal crest, as in some of the fossil Lemurine and Indriside. There are also strong lambdoidal crests continuous with the high zygomatic processes of the squamosal. One peculiarity in which the Megaladapine differ from almost all the recent and fossil Lemuroids is their great interorbital breadth. This breadth may be conveniently expressed in the form of an index, thus: interorbitel breadth uO maximum length of skull Measurements taken on a series of Lemuroid skulls, recent and fossil, give the following results :— per cent. Chiromys madagascariensis . . . . ». .. . ~ SII Meyaladapis grandidiert . . - . . « « - «+ » 282 UndishOnevicaugaty sain tine nn ee Eee OLS, IPOWEUGD COURT 5 0 nn oo eo 0 8 USS PAGES UIP 0 6 6 Bo oe a go, 1S Mesopropithecus pithecoides . . . . ~~... ~. 178 Paleopropithecus mawimus. . . . . »« »« «. « « 178 JC GO), EB B80 8 a oo og BPA IZGPOCKGOED UW 6 8 6 6 6 8 G6 4 6 LARS) Lemurs variusmer mimic mie ies Li ire n ectnne eto Archeolemur edwardsi . . . . . « « « « . 18D IYRWGOP 9 0 8) 6 8 8 0 8 8 oe po 6 Miles IGA SB oe) op 6 ol Be 6 6 oe) 6 OS It is interesting to note that a quite young skull of Megaladapis grandidiert has an interorbital index of only 18-1, in this respect ranging itself with the various sub- divisions of the Indrisine. It is thus evident that the great interorbital breadth of the adult Megaladapis is a secondarily acquired character. The orbits in all three species are small and tubular and are directed outwards, the plane of the orbital rim being nearly vertical. The lacrymal foramen is situated without the orbit or in line with the orbital margin. No postorbital wall occurs, and the hinder edge of the postorbital bar is rounded. ‘The nasalia are strongly developed and at their anterior extremity are (in the two species where this part of the skull is known) prolonged in a downward direction into a curved beak-like form. Auditory bullae are present, but in none of the species take the prominent inflated form so characteristic of the majority of Lemurs. The zygomatic arches are high and strong, contrasting with the light thin arcades of existing Lemurs; and they do not curve widely away from the skull, as is the case with Palwopropithecus and Mesopropithecus. PRIMATES FROM MADAGASCAR. 109 The postorbital region of the frontal is curiously constricted. This, as we have already had occasion to remark, is a feature recurring in all the subfossil Lemuroids. The frontal region of the brain in the adult Megaladapis is itself still further constricted by the presence of large aérial sinuses. The lower jaw is deep and massive. Its inferior outline forms in all three species a nearly straight line. The mandibular suture is quite obliterated at an early age. The alveolar margin of the molars and premolars is strongly curved upwards anteriorly (text-fig. 24). i.0: 1: pm.3: m.3 The dental formula of Megaladapis is ;5—.4- pa Sein? if we adopt the theory that in the lower jaw the canine functions as an incisor and that the apparent canine is a premolar. This formula, with the exception of the absence of upper incisors, corresponds with the “primitive” Primate formula as found among the modern Lemuride and the Cebidie of the New World. Professor Duckworth states that the upper jaw of Megaladapis insignis (= edwardsi) carries vestigial incisors. In none of the skulls of Megaladapis grandidiert which 1 have examined is there any trace of these, unless, indeed, some slight depressions in the alveolar margin of one specimen represent the scars where such incisors have fallen. Dr. Forsyth Major has pointed out the resemblance between the teeth of Megaladapis and those of Adapis magna. The tritubercular type of molar in the upper jaw is most nearly represented among modern Lemurs by Lepidolemur and Chirogale. The third molar of the mandible of Megaladapis is very strongly developed, having a large talon. In this respect this genus contrasts with other Malagasy fossil Primates and with modern Lemurs, especially the Indrisine, where, as we have seen, this tooth tends to disappear. In all three species of Megaladapis there is a diastema between the canine (or apparent canine) and the succeeding premolar in both upper and lower jaws. Synonymy.—The existence of remains of extinct Primates in Madagascar was first made known by Dr. Forsyth Major, who, in 1894, published a memoir in the ‘Transactions of the Royal Society,’ entitled “On MJegaladapis, an Extinct Gigantic Lemuroid from Ambolisatra.” In this memoir a description is given of a skull brought from the south-west of Madagascar by Mr. Last. ‘To this the name Megaladapis madagascariensis was given. In 1899 M. G. Grandidier described and figured in the ‘Balletin du Muséum d’Histoire naturelle de Paris’ three teeth of a gigantic Lemuroid on which he founded the genus Peloriadapis. In the following year a description by Dr. Forsyth Major of a fragment of an upper jaw in the British Museum appeared. To this the name of Megaladapis insignis was given. : Again, in 1906 a memoir published by Dr. Lorenz y. Liburnau, of Vienna, gives a Q2 110 DR. H. F. STANDING ON SUBFOSSIL Text-fig. 24. Vhs Profiles of three species of Megaladapis:—A. M. madagascariensis ; B, M. edwardsi; C. M. grandidiert. Drawn to the same scale and about 2 actual size. PRIMATES FROM MADAGASCAR. iil detailed description of several skulls brought to Europe by the collector Sikora. These he refers to the same species as that described by Grandidier in 1899, though he distinguishes two varieties as Dolichocephalus and Brachycephalus. Dr. Grandidier himself, who has seen the London and Vienna specimens, does not hesitate to refer them all to one species, identical with that first described by himself. He concludes, however, that in view of the complete series the points of distinction differentiating these fossils from MJ. madagascariensis no longer justify the retention of the genus Peloriadapis, and he gives the name J. edwards? to all these remains. MEGALADAPIS GRANDIDIERI (Standing). (Plates XXIV.-X XVI.) The new species of Megaladapis discovered at Ampasambazimba differs in several important particulars from the two species previously known. In size the skull of this new species agrees in general with Wegaladapis edwardsi, though in certain of its dimensions it surpasses the latter. The maximum length of the skull, measured from the occipital condyle to the anterior border of the nasal bone, is 292 millimetres, this being the average for three perfect skulls, the largest of which, however, measures 312 mm. The average maximum length for four skulls of M. edwardsi in the Vienna Museum is 291 mm., though three of the measurements are apparently only approxi- mate estimates owing to the imperfect condition of the skulls. One of the most striking characters of the new species has already been referred to in the general description of the genus, viz., the great interorbital breadth. In M. grandidieri this measurement exceeds by some 30 mm, the corresponding dimensions Text-fig. 25, Anterior part of mandible of Megaladapis grandidiert, showing spout-like projection carrying procumbent incisors. of M. madagascariensis. The bi-zygomatic breadth slightly exceeds that of M. edwardsi. One of the most noticeable features of M. grandidieri is the curious lengthening out of the anterior part of the muzzle. Correlated with this is a long diastema of some 30 mm. between the canine and the first premolar in the upper jaw. A similar interval exists in the mandible, the anterior part of which is drawn out into a shallow spout-like extension which carries the first premolars and the procumbent canine and incisors (text-fig. 25). 112 DR. H. F. STANDING ON SUBFOSSIL The median line of the palate forms a double curve, the part between the molars and premolars being convex, while the anterior part is concave, forming an angle of 150° with the posterior part. When the lower jaw is apposed to the upper a wide space is left between the alveolar margin of the premaxille and the lower incisors. This feature considered in connexion with the concavity of the palate and the deep trough-like mandible suggests that the animal possessed a thick fleshy tongue. The teeth of both the upper and lower series bear a strong resemblance both in size and conformation to those of I. madagascariensis, The most noticeable difference in size as compared with WM. edwards? is seen in the second and third molars. A reference to the appended comparative table showing the dimensions of the teeth of the three species will make this apparent. Taste 11.—Giving comparative dimensions of teeth of three species of Megaladapis. M. madagascariensis.| M. grandidierc. M. edwardst. Upper Jaw. | M. 3. Greatest length ............ 17°8 19 (7) 26-6 (5) 5 ue mebreadtiertcuctaselcterts 146 16:3 (7) 22:2 (5) M. 2. ay LenYSHN Go ccoK000d a0 17 18:1(7) 24 (5) ss oy) RERUN coccnc0deK6 15:8 13:2 (6) 20°8 (5) M. 1. Pane lengthy yn ct aeiion!: 13:5 15:1 (7) 19:3 (3) 4) fy UREAC No oo on bode oc 12-7 12:8 (7) 16 (2) P. 3. py He So cocoocnago 10 10°7 (s) 11:3 (3) 43 WREKIN GS 606008.0.00 9°7 10-2 (5) 13°3 (3) 13 ho Peleng Liye ery crite 10:3 10:4 (6) 11:8 (4) 3 “ee iubreadtieyae mrciscccces 8:5 9°3 (6) 10°5 (4) 125 Als om USGUN. isos o600 0000 ss 12:1 (6) 12°3 (4) 4 Hy ibreadthienw erence as 6°7 (6) 7°8 (4) Lower Jaw. M. 3. Greatest length ............ 23:5 26:1 (s) 37°5 (4) FS WRENN conD 600000 i 11 () 16:5 (4) M. 2. eee lenothurrrr rrr irat 15°5 18:1 (7) 24:8 (3) 3 Fae umibreadthveratrvenentcren a: 11:2 (7) 14:7 (3) M. 1. ea erlengthimscrnetse nent 13 14:2 (s) 19 (2) a si tbread thiv yj. jratevacuen: a 9:3 (8) 12°5 (3) P. 2. syaulength emcee seis | S6 11:1 (7) 13:9 (2) a | WREAGUN GY! oda6. coe uk ae 71(7) 8°7 (2) Blt seens vlenethyiesideateceeerenl| Be 10-5 (7) 12 a m5 ibreadtheeirercaereicncre ie 6:5 (7) 6:5 The small figure in parentheses indicates the number of specimens measured, PRIMATES FROM MADAGASCAR. 113 There are considerable variations in the shape and pattern of the molars. In some specimens the posterior margin of the third molar is nearly straight, the tooth thus having a distinctly triangular form. In other cases this margin shows a distinct tendency to the development of a postero-internal cusp (see Plate XXVI.), thus giving to the tooth an irregular rhomboidal outline. ‘The backward extension of the dental series also varies greatly. In some skulls the centre of the posterior margin of the palate reaches almost to the second molar, while in other specimens the palatal margin is posterior to a line joining the hinder edges of the third molars. Some of the crania show a curious lateral torsion of the long axis, the facial portion being deflected to the extent of several millimetres from a line bisecting the posterior part. This peculiarity is well seen in the skull figured on Plate XXVI. In his memoir on Megaladapis madagascariensis Forsyth Major says, “it is to be anticipated that skulls of young specimens will bear a much closer resemblance than the adult to the existing Lemuride”; and he instances several particulars in which this will probably be the case, viz., a more rounded cranium, a brain-cavity relatively and possibly absolutely larger than in the adult, a relatively shorter facial portion, and Text-fig. 26. Profile of cranium and mandible of young specimen of Megaladapis grandidiert. x 3. the absence of postorbital prolongation of the frontals. He also anticipates that the constriction of the olfactory fossa, due to the development of aérial sinuses (so marked a feature of the adult) will be wanting in the young animal. The occurrence of a quite young specimen of Megaladapis grandidiert among the fossils exhumed at Ampasambazimba makes it possible to test the validity of these forecasts. In the skull in question the permanent canine and the second molar have not yet appeared below the alveolar margin, but are visible deep in their sockets. A reference to the figure of this skull (text-fig. 26) will show that in its general aspect it conforms much more nearly to the type of modern Lemurs than is the case with the adult. ‘The brain-case is rounded and relatively much more voluminous, 114 DR. H. F. STANDING ON SUBFOSSIL The maximum breadth in the case of the young animal is approximately one-third of the maximum length of the skull, while in the adult it falls short of one-fourth. The face, as might be anticipated, is of relatively small dimensions in the younger skull, the nasals only occupying about 31 per cent. of the superior contour of the skull, as compared with 41:5 per cent. in the full-grown animal; the lengthening out of the maxilla as adult age is reached occurs at both ends of the dental series. At the posterior end space has to be found for the permanent molars, while at the anterior portion the diastema between the first premolar and the canine, which in the young specimen under consideration measures 15 mm., increases to double that length as growth advances. Text-fig. 27. i TZ: Yi LZ y 4 WWW if WY y A f eo PASS Restoration of skull and mandible of Megaladapis grandidieri. x 4. As already stated, the olfactory fossa of the Megaladapine is greatly constricted by the lateral development of aérial sinuses. This effect extends to the cerebral fossa itself with an accompanying reduction of substance of the frontal lobes. This constriction is already visible in the immature skull, though it is not yet advanced to the extent noticed in the adult. The backward lateral extension of the frontal and the upward extension of the squamosal are not appreciably different from the condition of these bones in the fully developed skull. There is, as might be expected, a total absence of crests; the two superior lines of attachment of the temporal muscles occupy the position seen in recent Indrisine. The frontal suture can still be traced, though already fused, and the interorbital convexity is almost as prominent as in the adult. In spite of the relatively smaller interorbital breadth already noticed, PRIMATES FROM MADAGASCAR. 115 the orbits are directed more outwardly and upwardly than in full-grown specimens, and are relatively larger in the young skull. An index showing the vertical orbital diameter as compared with the maximum length of the skull gives, in the case of the young animal, 16-4 per cent. This feature again brings it nearer to the recent Lemurs. The contour and proportions of the supraoccipital bone very closely resemble the condition seen in recent Indrisine. An almost perfect mandible was found accompanying the skull, and in this, again, there are certain features of interest. The symphysis is already firmly fused, with hardly a trace of the suture. The anterior part of the mandible, although sloping forward considerably, does not yet show the strange spout-like prolongation so characteristic of this species of Megaladapis. A wmilk-incisor is still present, and occupies a less procumbent position than do the incisors of the full-grown animal. Its form does not at all recall the incisors of the adult, being flattened in a direction TasLe 12.—Showing dimensions of the skull in the three species of Megaladapis. | M. grandidiert. | M. edwardsi. | M, madagascariensis. | \— | | SKULL. mm. mm. mm, From inferior border of foramen magnum to anterior TPIS OF joRSAEPAED ~ Goooscacocudcobeuedo dS | 266 230 215 Maximumilenethiof skull i emeseseectciecicieciele 292 288 ? From anterior margin of orbit to upper border of NOMIC WANN so g5c00b000000b00000000000 168 156 142 From posterior border of palate to anterior margin OLiprem axel ay ovens secre) stasis cenit we 130 | 92 u | Total length of dental series.................... 137 | 126 ? | | Length of series of molars and premolars.......... 82 | 94 75 Miameterotsorpits year tees | 33 36 26 Maximum height of skull with jaw in position ... | 173 192 2 a vertical height of brain-case .......... | 67 89 68 B bi-zygomatic breadth ................ 1127 150 100 is external breadth of brain-case ........ 70 57 58 Transverse diameter of foramen magnum.......... 26 18 17 Vertical diameter of foramen magnum............ 24 25 20 Maximum breadth of skull between middle of orbits. 135 110 111 Distance between lacrymal foramina ............ 70 67 63 Breadth of palate between postero-external angles of (WERE FNOEME 56 dc cccdoqogbcDcd0ba000duu0a G0 69 90 56 Angle of intersection of planes through orbital margins. 90° 70° 100° Manvisre. Distance from condyle to summit of symphysis .... 230 192 ? Herchtiohsymphysismremeeter ter tir kertteliry tier 86 70 ee Length of series of three molars ................ 58 80 53 VOL. XVIII.—PART ir. No. 8.— May, 1908. R 116 DR. H. F. STANDING ON SUBFOSSIL at right angles to the alveolar margin. A diastema of 9 mm. has existed between the first (apparent) milk-premolar and the milk-canine. The lower profile of the mandible is not straight, but shows a gently convex curvature. Appendicular Skeleton.—-The humerus of M. edwardsi has been found, as also that of WM. madagascariensis, A fragmentary ulna and a femur of the larger animal are figured by Lorenz in his recently published memoir. The resemblance of the humerus to that of the Anthropoid Apes is noted by Dr. Grandidier, though he attributes the similarities to mere convergence. Of Megaladapis grandidieri a considerable number of humeri and radii and one imperfect ulna haye been discovered along with several metacatpals and phalanges. But a discussion of the significance of the characters of the appendicular skeleton of Megaladapis is deferred to a later section. Taste 13.—Giving dimensions of five skulls of Megaladapis grandidieri. 1. 2, 3. 4. 5. mn. | mm. | mm. | mm. | mm. 1. | From lower margin of foramen magnum to alveolar border of WUMEETIED ooobeosoosocoonccNddcoG sscuoDoD Da DOONNC 276 | 241 ets BS 280 oe eMaximumylenethyotskulligerrrtcther riko tr enrvlnieet rar 295 | 268 eo oe 312 From lower margin of orbit to upper margin of foramen MMA saooooonpoevoDeDOU Sooo OdHOOdOaooOUDDCONDME 171 | 163 BiG es 172 4. | From lambdoidal crest to fronto-masal suture ............-- 172 | 157 Be O60 205 5. | Length of palate along median line .............00+.0005- USL MOS] || NAA 6 139 6. | From lower margin of foramen magnum to posterior edge of WHIES cogodeoodobooc coud DED OoEDGOOOODOO DODO DOOONS 153 | 145 20 a6 158 7. | From lower margin of orbit to alveolar border of premaxille..| 130°5\[103-5]| 125 |[120] | 183-5 8. | Length of series of molars and premolars ..........+-.0+- Ux) Gil 78:2| 78:2) 87 9. | From anterior surface of postglenoid process to posterior edge ofsthindgmolareegretien create oriirsihornercie 116 98:5] 2. .. |[116°5] OMI aVierticalldiameteroflorbita Per peer er hmee cere aence eee. 35 | 33 | 34 | 31-2| 325 eee transverse idiametersoMonbibsaaplr rrr r erik 35 32°8| 32 31:8} 33 | 12. | Maximum external height of brain-case ..............05-- 69 64 ie ee 66°5 IMi3un/eBi-zycomatic breadthy (@xternal)) Myarenaeiterie slates tei 2 |), 28} ve se 132 14, | Maximum external breadth of brain-case ............-+-: 68 ORB 6.0 ts 72 15. | Breadth between styloid processes .............---+--+-- [66] | 71 0.0 ae [65] 16, | Transverse diameter of foramen magnum .............-.. 27 2079) |e si 25 17. | Vertical diameter of foramen magnum ...........-...-+; 25 24 ; 25 18. | Distance between orifices of auditory meatus ...........0+- 78 7 a a 79 19S Breadth of/skullvat levelvofiorbitsy-rye-ireieleieieborntoreeeriorne 382°5| 384 35 26 38 90. | Breadth between lacrymal foramina..............00.e+e0- “ie 59 76 66:5) 80 |91, | Breadth of palate between third molars (internal) .......... 39°8| 38 41°38} 40:8} 38 | 22. | Breadth of palate between antero-internal margin of canines...) .. aie D0 Si 46 Of | Maximumebreadthvofmasalian Were eocatickitrenilnkecn i: Siliciay lear 29:2) 28:2) [36] OA ENEmimume bread thos wasaliaueeme arc irtiee rtecrlach nr acnetne DEN oo 18 17°5| 22 25. | Maximum width (height) of zygomatic arch ...........--- 34:5} 31:2 PRIMATES FROM MADAGASCAR. a7 Tasie 14.—Giving dimensions (in millimetres) of mandible of Megaladapis grandidieri. 1 2. 3 Il. || Bom, Gomalylle tho thos Or WMGSOS Gocogccooooaonsacdse0ae oo0c0008 [235] | [243] | a Eomucondyloytorsummiutrofsymphysislesmeere ee selec eee 227 =| 234 3. | From lower end of symphysis to a line joining inferior angles ........ 182 | 180 4 PELOMNCOronoldyprocessstojimteriormanel esi vjeace ee eile acerca 1085 5. | Maximum distance from condyle to inferior angle ................ 87 87 6) |) Length of series of molars/and premolars) ...........-.....+0+-+0- 84:5) 742) 82:5 7. | From posterior edge of last molar to postero-inferior angle .......... 108 | 121 8. | From condyle to posterior edge of last molar .......5...........-.- IKE |} ata Q; |} engin Ge aymTplayS: cooodoccsoocaocoonAeauc seo en ud Oo moUOdUOS 86 88:5 NOM Distance) between! coronoid processes @-4.-64-4-.-++2-4020+ 0s 40. [52] iL, |} Disiames loomed ConGhyles cosccoagcocovode ebongdoob0GdoKes He bo 50 2a istancenbetweensimtenormane esata are ecient ene ate 97 13. | Breadth between rami behind third molars ...................... 34 41 14. | Space occupied by dental series from median incisor to third molar ..| 61 545] 59 15, || Specs Ceenjmes! ly wiv jomamolens codcooscoasnvdacgoocgsco000000 23°5) 20 23°5 IG; ||) Win@kagss OF yen ex level or slash NONE 555050506000 00d000000000Ke 10 WN GPG 17. | Height of jaw at level of first molar (including tooth) .............. 62:5] 60 [50] Subfamily LE MURIN &. Genus Lemur. Of the existing Malagasy Lemuroids (if we exclude the aberrant Chiromys) the members of the genus Lemur bear the least resemblance to the higher Primates. At the same time their affinities with the scattered members of the family of Lemuride in the continental areas of the Old World are obvious. Many of these sporadic genera of Lemuride show evidence of specialisation. As illustrating this statement reference may be made to Loris, Nycticebus, and Tarsius. “How far the similarities between the various genera are the result of convergence or specialisation along similar lines due to similar habits and conditions, it would be difficult at present to determine. There is evidence in the case of the Indrisine group of a strong tendency to the production of some of the distinctly “ Lemuroid”’ features among the surviving members of a family whose extinct representatives were evidently descended from ancestral forms with strong Simian affinities, if not, indeed, from progenitors which can only be described as Apes. A comparison of Jndris, Propithecus, and Avahis with such forms as Archewolemur and Hadropithecus leads to the belief that the long muzzle, flattened brain-case, weak upper incisors, and frail mandible are in the case of the extant genera secondarily acquired. Although in the case of the Lemurine the evidence is as yet incomplete, no RQ DR. H. F. STANDING ON SUBFOSSIL 118 “OT t IT G-96 | ~~ GIT | 7" GLI | 7 G-6 : GPL | L 8 IL | 4:8 6-9 4 OL | 6-6 ial GT “ST | “OL | \sdacBasvaceeansqueondobasonsvun 60 a ai Fay ee eee be oon ay awpour para | “| ast | Lt |e2t it | at nna} TORE PHOORS | epee Maal) oa Gat an sole alt OE mew) seomerd pow: |p 2 aa ver 6 Ne eel aejouerd puosag |e ob avjomedd uLdojluruvs Jo Yypeetg | *Z oes me Ss Be er fteereteeeeteeeeeeeeeees reommaad muTOFTUTUB Jo ISOM | “7 9 ty Vv € 6 ae sivpouleid jo WUE] [eIOT, | “OT ee eater ABOU PAT, | ‘6 7 aa eat avout puoaag | “g eat | ert | Sat ay TiN || a ee tf ERO HEE IE ee | CO ee rece emg tee) Cee cata Uy eee ee rac OBESE | S| ale [ER oon RRIMARRRE) ovomat as] oo | G-6G 0g |[8.0¢] tresesereeees guvrou JO WSUeT [vO | “S ¥1 L-OL [¢.61] Lor] eee e cee sccecsccesssctese . eulueo jo WIpratg, % oz | Ze re wes ectereeeeeeseeonsseses guttre) Jo 4USI0Fy | *T = SS | ss 2 |e 59) g v $ xo "UE ‘VITIXVIN—' V -daypypunsb sidopnyobayy Jo Yjee} Jo (SoAjoWT][IM UI) SUOTSUSWMIp SuTALQ— GT ATAV], PRIMATES FROM MADAGASCAR. 10) fossil forms having come to light, as in the case of the Indrisine, proving conclusively that the existing genera are allied to obviously pithecoid forms *, still even in their case analogy would suggest that they also are descended from Ape-like ancestors, The undoubted degeneracy of the allied Megaladapine and the fact that certain members both of Malagasy and continental genera appear to retain Simian characteristics which have disappeared in the Lemurs properly so called (e. ., the postorbital septum of Tarsius, the shorter muzzle of Hapalemur, the proportionately larger incisors of Perodicticus and Nycticebus), would suggest that in the case of the Lemurs also many of the so-called “ Lemuroid ” characters may be the result of convergent specialisation. Dr. Elliot Smith informs me that the histology of the brain-cortex of certain of the Lemuride supplies a complete link with some of the South-American forms, while “ the furrow-pattern of Perodicticus is an almost exact copy of that of Pithecia.” There is thus a strong presumption that the various genera of the Lemuride as well as the New World Monkeys have had a common origin ; that, in fact, the Lemuride are, in general, specialised and more or less degenerate decendants of ancestors more Simian in character. This subject will be further discussed in the concluding sections. In the Ampasambazimba deposit eleven perfect or almost perfect skulls of Lemur have been discovered. ‘These are evidently closely allied to the species described by Dr. Filhol from Belo and named by him Z. insignis. I have given in the ‘ Bulletin de Académie Malgache’ my reasons for differentiating the two specimens there briefly described from. L. insignis. In view of the complete series of skulls, one of the characters on which I based my distinction—viz. the differences in the line of the facial profile—has less weight than [at first attributed to it, since considerable variation occurs in this respect throughout the series. The broad oval shape of the palate of ZL. insignis, however, especially in the region of the premolars, is not reproduced in any of the new specimens—the anterior part of the palate in all of them showing more resemblance to the narrow elongated form seen in existing genera than to the broad comparatively ovoid form of L. insignis. But since the variations of form among the specimens found at Ampasambazimba are very considerable, I do not feel it safe to speak too dogmatically as to the value of the specific distinctions made among such a series of skulls all obviously closely allied, especially in view of the great variations which we have seen to exist in the case of the subfossil genera already described. The same remark will apply to the specific distinction which I make in the case of one of this series of skulls, in which the occur- rence of certain distinctive characters appears to justify a separate notice. For this latter skull I propose the name of Lemur majori. * The evident resemblance of the humerus of Lemur jullyi to that of Archeolemur referred to in the following section goes far towards furnishing such evidence, as also do certain characters of the skull of Lemur jullyi to which attention is called in the present section. 120 DR. H. F. STANDING ON SUBFOSSIL Lemur JuLiyi (Standing). Description of Skull of L. jullyi. (Plate XXVIL.) Nine of the eleven skulls of Lemur in the collection of the Académie Malgache form a series which, while presenting many individual variations, can best be described under one specific name. One of this series I have already figured and briefly described under the name L. jullyi. A lengthy description of this species is not necessary for my present purpose, the series of photographs given on Pl. XX VII. with the table of dimensions at the end of the present section will sufficiently indicate its main features. The existing species which comes nearest to this subfossil form is Z. varius, and the species, as already pointed out in my preliminary notice, is intermediate between Text-fig. 28. L. insignis and L, varius. A comparison with the latter shows it to be larger and more massive in every way. Nearly all the full-grown specimens possess powerful ridges and crests, the sagittal and lambdoidal being specially prominent. Occipital Region.—The condyles are much broader than in the extant species, a line of chief convexity running obliquely outwards and upwards and separating two distinct planes of contact with the atlas, an upper posterior and lower antero-external one, in a manner closely analogous to that seen in Mesopropithecus and Archeolemur. The plane of the occiput is much larger than in the recent species, though this condition is partly due to the prominence of the lambdoidal crest already alluded to. The paroccipital processes are large and prominent (text-fig. 28). Parietal and Frontal Regions——The brain-case is pear-shaped, presenting in every case the curious frontal constriction which we have seen to be so marked a feature of the Malagasy subfossil Lemuroids. The supraorbital frontal ridges are prominent, though slightly less so than in adult specimens of Z. varius. The central frontal region differs greatly in appearance, however, from the corresponding part of L. varius, owing to the fact that the sagittal crest divides at or near the coronal suture and is continued by two prominent diverging ridges which are confluent with the outer posterior edge of the postorbital bar. These anterior and posterior paired ridges thus enclose a diamond-shaped more or less flattened region similar to that observed in Megaladapis, and evidently analogous to the raised frontal region existing in Archeo- lemur and Mesopropithecus (text-fig. 29). PRIMATES FROM MADAGASCAR. 121 Nasal Region.—The facial profile varies much in different specimens, in some cases being decidedly steeper than in others. In the type specimen of LZ. jullyt the profile forms a nearly straight line from the coronal suture to the tip of the nasals. In others the median suture of the nasals forms a concave curve. The distance between the inner pair of incisors is longer than in the case of L. varius, and these appear to have been functionally less important than in the recent species. The position of the alveoli shows a close resemblance to the condition of LL. insignis. Text-fig. 29, Vertex view of raised frontal area of : (A) Mesopropithecus pithecoides, (B) Archcolemur edwardsi, and (C) Lemur jullyt. The Orbital Region.—The postorbital bar is triangular in section. The posterior surface is of variable breadth (in some cases its upper portion exceeding 5 mm.), and shows the sharp inner edge which has been referred to in the case of Mesopropithecus, Paleopropithecus, and Archeolemur. The resemblance between this region and the corresponding part of Jesopropithecus appears too striking to be due to mere convergence. Temporal Region.—The strong and massive character of the fossil is especially apparent in the zygomatic arch, which is much higher and stouter than in any extant Lemur. ‘The base of the zygomatic process of the squamosal is especially broad and strong, in this again resembling all the forms previously described. Dentition.—The description given of the dentition of the type specimen of ZL. jullyi in my preliminary notice of this species will apply with hardly any modification to all the specimens examined. The internal collar of the third molar is in some cases rather more perfectly developed than in the skull originally figured in the ‘ Bulletin’ of the Académie Malgache, thus bringing it slightly nearer to the condition of L. insignis, It should also be here remarked that the imperfect skull described as L. maziensis must, in view of the complete series now available for comparison, be classed with the species now described. The fact that the mandible of these extinct Lemurs was much more massive than that of any of the existing species must be correlated with the more massive teeth, strong sagittal crest, and probably also the narrow postorbital frontal region, which we have seen to be invariable features of all the extinct Lemuroids hitherto described. 122 DR. H. F. STANDING ON SUBFOSSIL Lemur Masorr (Standing). (Plate XXVIII.) One of the series of skulls of the genus Lemur referred to above is distinguished from all the others by the possession of several characters, no one of which alone perhaps would constitute specific distinction, though the simultaneous occurrence of these comparatively numerous variations in one specimen would seem to justify its reference to a distinct species. It will not be necessary after the description given above of L, jullyi to do more in the present case than point out where this new species differs from the former. We may notice first the very voluminous brain-case. Dr. Elliot Smith, who has examined a cast of the cranial cavity of this specimen, estimates the weight of its brain when living at 65 grammes as compared with 33 grammes, the average weight of three specimens of the brain of Z. varius. This figure differentiates it from all the other specimens in the collection of the Académie, with the possible exception of one skull. As the teeth of this last specimen are all missing, it is not possible to be certain of its exact position. A second striking feature as compared with the cranium of L. gully and also that of Z. insignis is the entire absence of crests. That this feature is not due to immaturity is evident from the worn condition of the molars. The large size of the canines would, moreover, indicate that the individual was a male, so that the absence of crests is not merely a sexual variation. Not only is there no sagittal crest, but the two superior edges of the area covered by the temporal muscles do not quite meet along the sagittal suture. This is in striking contrast with the skulls of Z. jwllyt, where this crest is frequently very prominent and much roughened. Apart from this absence of crests the general contour of the skull is not strikingly different from that of Z. jullyi, though its greater size is apparent. Turning to the dentition three peculiarities are observable. The third premolar differs from the corresponding tooth of L. jullyi by the possession of a well-developed internal collar similar to and almost as strongly marked as that seen in the first and second molars. The second molar, which in all the specimens of LZ. jullyi has a large antero-external projection overlapping the first molar, is in LZ. majori entirely destitute of this blade-like projection, the tooth presenting the appearance of having had this prominent corner completely filed away. A third feature, which though perhaps hardly by itself of classificatory value, is the fact that the whole series of molars and premolars shows an entire absence of that close overlapping of each tooth by its successor which is a marked character both of L. insignis and L. jullyi. The shape of the palate is intermediate between that of the two species just named. The appended tables giving the principal dimensions of the whole series of skulls will in conjunction with the photographic reproductions on Plates XXVII. and XXVIII. preclude the necessity of further detailed description. 123 PRIMATES FROM MADAGASCAR. 8 aL 9g Or 02 6-36 G18 [¢.¢7] [st] LI Ze 9-4 [9-62] GBF GS ZS G.Gg IF CF FS 1g GR [os] 9¢ L8 Los] B-GSL. G-60T ‘OT BS Ia G2 lew 8 (g] |¢8 8 6.6 OL |¢-6 OL IL |¢8 OL IL 66 Ks Tg +g ce | [rel | 9¢ Ge oe 6 6 8 |¢-6 8 14.6 ia 03 Zo 1G-6L |¢-61 IG GS 0 oe 6-86 |G-GS |¢-16 |G-8S 13-66 be Tomle:ceall 8-08 | 626 86 | S-FS 6 8Z 86 | a1 Gly |GZr |¢.cP 9F |Z-9F oF : LY Hal WSChe esse Nia GES oe : GL LSl |@SL |¢.eT |¢-St GT cl ge ial Te 0g 0S |208 | 6-6 6 a 1B 6h_ |¢-Lh |G.87 |¢.8F o¢_ |¢-2r | [orl | te [¥-89] |[g-g2] | °° oP re) | °° pe a G68 | G-LF 68 |G-L6 |4-6r |¢-68 |¢-6¢ OF & #6 ¥ 18-86 |G-E% 19% | T-¢¢ a G13 GB 1G i NOs Ids GBC oe ecg |9-78 16-4 wal 8e FE |3-Ge nO 8-0F | 6-IF or | Lor] |8-0F |¢-1F OF ae [tr] | 6F OF |L-SF |¢.9F | ¢-eF Lt Bo G.9G | ¢.gg gg 8g I cede || (egl} | °° [te] | ¢¢ \e-1¢ |¢-1¢ eg |¢.9¢ |G.9¢ 22 8-38 eg 18 &8 es &8 98 &8 LL 2 9L 0g 08 | 22 | (oell |) 2 (c-prtl|! “Got || Sil | eon | 9c I¢:611 | Fel | SIT G-IOL | S0L|/¢-FOL | 60L | ott | For | [zZ0T]| oot 6 8 "2 9 ¢ = 2 g 3 ‘T el[eseu Jo Yypeemq WNIT, Bleseu JO [Apeald WNMIxeyy BITBSBU JO YISUST UINMIXe TL "souLueo 0 SUISIBIT [BUIOZUI-0109UB Us8A4eq oWvyed Jo YypeoIg reteees (remrequt) SIB[OU PALY} Woaajeq oyeyed Jo yypRoig Totes esse sss BUlUIBLoy [eMAIOB] MeaAMjeq YIpBeIg UOTPILI4SMOD [VqIq.1oysod fo [oasy 4e T[NYs Jo Ypeorg Trt s sess sss * sngvour A10zIpne Jo saoytio usaMjoq eoueqsI(y "77" UNUSBUT WSMIBLO JO IOJOWILIP [VOT4AS A. “* WNUSBUL USTIBIOJ Jo IaJoWBIP OSIOASULL, “7 +" sassooord ptop{qs ueeMyoq Wypeolg 9880-UIBIG JO YIpvoiq [VUIOJXO VINUIIXL]L “cess (7eUIe9xe) yyprorq oevUo0sdz-1g 98B0-UleIq JO FYSIOY [eUIO}Xe WINWIXY yy "co 555" SHTGIO JO IoJOUVIP OsIOASUBL, “25 "** S1IqIO JO IeJOUIVIP [BOT}IO A. scot * rerouL paryy fo espe ot1eysod 03 sseoo1d proueSysod jo oovyangs xo1eyue WoL ° "tT" "* siejowmerd pue srepour jo setsos jo yySuaT wypxemerd Jo Lep10q IBpooATe 0} 4Iqzo Jo WISIN IaMOT WHOL eee eee + aqered jo ope norroyeod 0} WNUSeM UOUTeLOF Jo ULSIeU IOMOT WIT Be eS yer ** outy uerpou Suope aqered yo yySueT emyns [eseU-oF 1075 09 sex TeplopquIe] WOIT pears wWnUseUL UoeureL0y Jo ursivut Joddn 0} 4tqao Jo ulS1em IaAMO] WoL, slag ae “** [NYS Jo WSU] UNMIXe yy see ee ee we e]ixeureid Jo Topiog IB[OOATe 0} WNUSEM UOUTBIOJ Jo UISIVU IOMOT WOT sete eee ee oer ee nee Ad Tiss Moa qi | [LON] alofowu unwary Jo [[nys euo pue [IT-8 3 9-1 ‘son | ahyynf unwary Jo stays voy jo (saxjouyprU UL) SUOISHDUMIp SUTALQ—"gT ATAVI, VOL. XVIII.—PART 1. No. 9.—May, 1908. DR. H. F. STANDING ON SUBFOSSIL 124 GPL [z-0¢]|¢-21 | [et] | st rele |e ic: eee = bb ee Fe GS-81 GG [¢-cT] G.GG G-06 1-86 als ee ed (43003 Surpnya -UI) Iejou 4suy Jo Jodo, ye Mel Jo 4YSIO “+++ yefouL 4sIyp Jo JoAoT ye MeL Jo ssouHOTyT, “rrsesees+ sreromord 0M4 Aq petdnoso sovdg o000 050006 ++ +s+++ sqerour eerqy Jo yySueT * *SIB[OUL PITY} pUTYog Tuva WOOAJZoq odTIeIST(T tressstsss go8ue IOLAyUI WeaMyoq eouL4st(T trrrerressssss gatfptoo uesAyoq eouvzsi(T De es sasso00i1d PloUoI0D WADA oq GdUe{SI(T Dees e recess eeeeee es grekydurds Jo yySu0Ty * Teo 4se] Jo eSpe 1o11e4sod 0} o[Apuoo WOT esos ere stor er eor ee ee oe eset 9]SUB LOTLE}Ut -o010480d 03 rejotm 4sv] Jo espa aot10}sod wor] +++ srejomerd pue sivjou Jo seltos Jo YySUe'T weet we eee SeCICECONC ROC EON nn CnC nn CSC aC g[sue IOWOJUL OF ep Apuoa Wood goeySIp TANUILXB JL eget so[sue AOTLOJUT OF ssooo1d plouolod WoL Pe ee ee So]SUe IOTIOJUL Surutof our @ 0} sishydurds Jo pue Jomo, WOT ++++++ stskydurss Jo yrmuns 0} eyApuoo Wor, “OT ‘LNUWATT JO SeTAIPULL VOE}ITY} Jo (soTJOTUT][IUA UT) sUOTsSUEWIP SUIATO—")T ATAVT, PRIMATES FROM MADAGASCAR. 125 TaB.e 18.—Giving dimensions (in millimetres) of teeth of Lemur jullyi and Lemur majori [No. 7]. aa A.—MAXILLA. | | | | | 1 3 4 5 ie 8. | 11.* 1. | Height of canine ....... SS SOM RAD ORE oe oa || “oe o6 O° on 13 16 | 2. | Breadth of canine (ant.-post.) .............. 90 B06 O00 56 10 9:5 3. | Total length of molars .................... 22:8 | 22:2 |[21-2]| 21-8 | [21] | 20-5 | 20:8 | | sell Death {length (ant.-post.).......... [62]| 58] 68] 6 | [65)]| [6] | [63 ey | BNNs joemnole | breadth (transverse) ........ [42]} 29] 33) 38] [8-8]| [42)| [4-5] 3 length (ant.-post.) ........ 75 78 8 78 8 7-8 67 D, || Seganel pomole: { breadth (transverse) ...... 6:2 6:5 6 5:5 6:2 6 65 : length (ant.-post.) ........ GS) Ce) GS) we} (Tos we 8 G; |) Liste) ppxounollon { breadth (transverse) ...... 8 | 82) 52] 78] [781] 75| 89 length Creda uel, Ma Oke 9 | 87] [8] | 88] 85] [rs] 7 | i | Sissh mole: { breadth Aye te ator. wane Of | G2 | (eR! Oat |) Ceo | (Szall Os | EHS oo aroimoaeene als 66 oo ae 7:8 eu 8 7:8 78 7 7-5 | Baaeccondmolar { preadthiaes ane ae es Mk 97| 89| 99] 9 92] 95) 9-9 5 lene Thee Be UU OREN 7 | Gel @ | Gel tl | 6 | Ge OL | dines ea: { breadinenmes eee eral 7 | 68) 721 7 | [65]| 65) 6-9 | B.—MANDIBLE, 1 2 3 seni bee eve To | Is || FSG OF COMO Soooccogccccpeesvo0HuGde 75 | 85 | Ze | Wraachin GF GMIN® goaoocooddooogs00G0000000 ees as a 8 is) |B. | Second premolar {ess 20 TTIIIT) be Waal ae fn | | Ae Tonthe yet Whe Sr ule See eet enol iene. Wega chhied premolar { Breadthie ce ena : 5) BO o. | 86] 6 | | Tenetie ch Wee ce eee cane 85| 85) 85| 9 9 9 a | B. | Dinse melen \ breadihea tite eee wate ee oe 52) 55| 52] 6 6.) 6 6s lengths tere. he Suen 95| 72) 82] 89! 85] 35| 8 G | Sevoad mols: { Dread Meee eee oc eh SO Gol S| 6G | 6 | 6 | se Espa lenathreen tne Shere carr. 8 72| 65| 79| 8 8 8 | Go, | Mae aos tier LO SERIO) Bits, 6 Go 26) 6 | B lB | a6 | * These numbers correspond to skulls in Table 16. 126 DR. H. F. STANDING ON SUBFOSSIL THE APPENDICULAR SKELETON. Fully to describe and discuss the fine series of limb-bones and pelves which have been found at Ampasambazimba would require much more time and space than is now at my disposal. It will be useful, however, by way of forming a basis for future comparison, to give detailed measurements of the greater number of these specimens, and at the same time to call attention to some of the most striking and interesting of them, and very briefly to discuss the affinities of such as appear to throw further light on the problems discussed in the present memoir. One preliminary remark must, however, be made. It was stated in the introductory section that no single skeleton has been found complete in the fossil-diggings at Ampasambazimba, but that the bones are in general scattered without order or connection. This fact makes the task of identification and description specially difficult, and I wish it to be understood that I do not regard all the determinations here made as absolute. They must be subject to revision as the material for a more complete comparative study is accumulated. In several cases, however, analogy with extant forms and the relations already proved to exist among the subfossil Lemuroids whose skulls have been described will form a valuable guide, while inferences as to the probable mode of life of certain of the peculiarly specialised genera will furnish further indications. THe Humervs. Megaladapis.—Seven perfect humeri which must be referred to the genus Megala- dapis have been found. The dimensions of these are given in Table 19, Nos. 1—7 (p. 132) (three are shown in text-fig. 30). A comparison of the various details of this series of bones will show how greatly they differ in some of their principal dimensions. Thus the humerus figured as No. 4 in the table (C in text-fig. 30) is shorter by 4 cm. than No. 1. That this is not a mere sexual difference is shown by the fact that the shorter bone is much the more robust of the two. Had these two bones alone been found I should not have hesitated to refer them to two distinct species, and, indeed, it is quite possible that the smaller bone may be the humerus of M/. madagascariensis ; but in view of the fact that several humeri occur of intermediate length, and one even still shorter, it seems better for the present not to refer these bones to several new species but to wait for a fuller knowledge of other parts of the skeleton. The general features of the humerus of Megaladapis are already known. ‘The long straight cylindrical shaft, the large rounded head, the broad distal end, the disposition of the tuberosities, the clearly marked bicipital groove, strongly recall the human humerus. We shall presently see that the ulna of Megaladapis also has certain features resembling the condition observed in Man and the higher Apes rather than the ordinary ‘‘ Lemuroid” type. Dr. Grandidier regards these Simian resemblances PRIMATES FROM MADAGASCAR. 127 merely as secondarily acquired through the adoption of a similar mode of life. This assumption, of course, implies that the arm-bones of the ancestors of Megaladapis had characteristic “ Lemuroid” features which have since been lost. That such may have Text-fig. 30. ——S —— £4 Y SS Z = ===> = == f! yy Hic Tt) ) \ \\ 3 !) i —~ DR Humeri of Megaladapis. x been the case is indeed possible, but seems an unnecessary assumption. ‘To take one instance: the characteristic curvature of the humerus of the true Lemurs appears to me much more likely to be secondarily acquired, depending on the peculiar use of the 128 DR. H. F. STANDING ON SUBFOSSIL fore limbs necessitated by the disproportionately long legs of the extant members of the genus. ‘That their curvature is not a primitive character seems the more probable since it is hardly noticeable in Archwolemur, Mesopropithecus, Palwopropithecus, or indeed in the recent Indrisine. It is interesting to observe that, although some of the characteristic features of the humerus of the genus Lemur are absent in that of Megaladapis, the affinities are nevertheless rather with the Lemurine than with the Indrisine. The size, shape, and position of the lesser tuberosity, the form of the trochlear and capitular surfaces, the curvature of the broad flange running down to the external epicondyle—these and other features confirm the evidence already afforded by the skull that Megaladapis belongs rather to the Lemurine than to the Indrisine group of Malagasy Lemuroids. Dr. Lorenz suggests that the Megaladapis was probably arboreal in its habits, clinging to the trunks of trees and crawling along the branches. It hardly seems to me likely that so large and massive an animal, as its skull and limb-bones show it to have been, could in this manner conveniently reach either the leaves or fruits of forest-trees. Its very massive jaws and strong cusped teeth seem, however, to be adapted for feeding on the large hard fruits which occur abundantly in the forest débris at Ampasambazimba. But fruits are found on the ground in profusion at certain seasons in the Malagasy forests; at the same time its long arms would enable it to hang on to the branches of trees, serving in fact the same purpose as those of the Gorilla and Chimpanzee. Paleopropithecus.—One nearly perfect and two fragmentary humeri (the former marked No. 9 in Table 19) must be referred to Palewopropithecus (text-fig. 31). This bone is of great interest when considered in connection with the supposed mode of life of the animal. We have seen reason for suspecting that Palewopropithecus was aquatic, and several features of the humerus under consideration appear to confirm that view. The bone is very short and massive, as the annexed figure will show. ‘The ridges and surfaces for muscular attachment are large and prominent, and the distal end is of extraordinary breadth. The very large roughened entepicondyle indicates the possession of powerful flexor muscles of the hand. Of its affinity with the Indrisine there is strong evidence: the very prominent deltoid ridge with clearly marked surface of attachment for the great pectoral and deltoid muscles, the prominent backwardly curving internal epicondyle retracted from the articular surface, the regular cylindrical form of the trochlear surface—these and other features show its family-connection with the recent Indrisine. One peculiarity of this humerus is worthy of notice as bearing on the mode of life of the animal. The limited extent of the spherical surface for the articulation of the head of the radius shows that quite full extension of the forearm has been impossible. This condition might well be due to the fact that the fore-limb was probably mainly used in swimming. Archeolemur.—tThe bones marked 10 to 14 of Table 19 all appear to be referable PRIMATES FROM MADAGASCAR. 129 to the genus Archewolemur. In view of the extraordinary diversity which we have seen in the skulls of this genus, it is not surprising that there should be considerable variations in the size of these bones. Without attempting to determine with any certainty the various species to which these humeri should severally be referred, it seems probable that the long, somewhat slender one numbered 13 may be the humerus of Archeolemur edwardsi, the somewhat shorter and stouter ones (11 and 12) those of A. platyrrhinus ; while the one indicated in the table as 14, which though Text-fig. 31. Humerus of Paleopropithecus maximus. 0 much more slender than the preceding is of the same general type, may possibly be the humerus of Archwolemur majori, though no skull of the latter species occurs in the collection *. It will not be needful to describe in detail this series of bones. Their most interesting feature is the fact that they show many points of agreement with the humerus of the members of the genus Lemur. The proximal end of the humerus of * None of the arm-bones of Avchwolemur in our collection appears to equal in size those figured by Dr. Grandidier in his recent memoir as belonging to Bradylemur robustus, so that this form will stand as a fourth species of Archwolemu. 130 DR. H. F. STANDING ON SUBFOSSIL Archeolemur may be said to possess features intermediate between Lemur and Propithecus. Tf, indeed, a robust specimen of the humerus of L. jullyi or LL. majori be compared with those of Archwolemur, the general agreement of the two bones is unmistakable (see text-fig. 32); there are at the same time certain features which show a distinct affinity with the Indrisine type of humerus. These become more apparent if the more slender bone which we have provisionally referred above to Archeolemur majori be compared with that marked 15 in the table. This latter is apparently the humerus of Mesopropithecus, and a brief description of it will now be given (text-fig. 33). Mesopropithecus.—This bone is somewhat longer and considerably more robust than the humerus of Propithecus. Its upper portion is more flexed than the latter bone, Text-fig. 32. Text-fig. 33. Humeri of (A) Archeolemur edwardst Humerus of and (B) Lemwr jullyi, compared. Mesopropithecus pithecoides. being intermediate in this respect between the modern Lemur and Propithecus. ‘The head is more spherical (7. ¢., less ovoid in form) than is the case with either of the extant subfamilies. The shape and position of the greater tuberosity more nearly resembles that of Archwolemur and Propithecus than that of Lemur, while the lesser tuberosity is intermediate between Archwolemur and Propithecus, the groove which separates this tuberosity from the head of the humerus being less marked than in the extant genus, though more prominent than in either Archwolemur or the Lemurs, whether recent or subfossil. The “deltoid” crest bears much resemblance to that of the slender humerus of Archeolemur referred to above, the roughened surface for muscular attachment PRIMATES FROM MADAGASCAR. 131 on the outer side of the bicipital “lip ” being closely similar in the two bones, though there is in Mesopropithecus a short prominent ridge about 1 cm. in length running down from the lower angle of the greater tuberosity, which evidently has its homologue in the well-marked crest which in the fossil “ Lemurs” is continued downwards in a spiral course to join the lower end of the deltoid ridge. The distal end of the bone is imperfect, but its trochlear surface apparently bears more resemblance to that of Propithecus than to that of Archewolemur. The long, projecting, backwardly-curved outer epicondyle is, on the other hand, strongly suggestive of the corresponding part of the humerus of Archwolemur. Lemur.—The humeri, the dimensions of which are given in Table 19, Nos. 16-26, are all referable to the genus Lemur, They are all very similar in size and type, as might be expected from the similarity which we have noticed in the series of skulls of this genus. A detailed description of this bone will not be necessary for my present purpose. It is interesting to note in passing, however, that its straighter shaft tends (text-fig. 32, B) to approximate it to the humerus of Archewolemur. Tue Rapivs. Appended is a table (No. 20) giving dimensions of several types of radius occurring in the collection from Ampasambazimba. Only two of these will be referred to in the present section—those of Megaladapis and Paleopropithecus. Megaladapis.—The radius of Megaladapis (Nos. 1-6 in Table 20) is a long, some- what slender, curved bone (text-fig. 84). The curves of its head correspond exactly with the capitular surfaces of the humeri already described, and with the lesser sigmoid cavity of an ulna to be presently referred to. Its neck is somewhat slender, its tubercle long and much roughened, showing the possession of a powerful biceps muscle. The shaft is widely curved, its centre being some 2 centimetres distant from a chord joining its two extremities. With the exception of this greater curvature and its mere slender proportions, the radius of Megaladapis bears a considerable resemblance to that of Man. The interosseous edge has a regular curvature which hardly deviates from a plane bisecting the radial head and passing through the styloid process. The three articular facets for the ulna, scaphoid, and semilunar bones are very similar in shape and disposition to the condition seen in Man. ‘The styloid process is less prominent, and the diverging lines of the inner and outer edges of the distal end of the bones are less abruptly curved than in the human radius. It is not, of course, intended to imply by such a comparison that there is any close connection between such an animal as Megaladapis and Man, but merely to point out that types which by their brain-development are removed as far as possible from each other in the Primate series have both retained similar features in their limb-structure. At the end of his text-book on ‘ Morphology and Anthropology,’ Professor Duckworth concludes that the balance of evidence indicates that, were the material still available, VOL. XVIII.—PART 11. No. 10.—May, 1908. T 132 DR. H. F, STANDING ON SUBFOSSIL TasLe 19,—Giving dimensions 8 humeri of Megaladapis [1-8]. i humerus of Palwopropithecus [9]. 4 humeri of Archeolemur edwardsi [10-13]. ilo 2. 3. 4. 5. 6. 7. 8. 9: 10. {. | From the summit of head to capitulum../321 |819 (295 |280 |274 |294 |2945] .. | 210-5 | 152 2. | From the summit of head to trochlear SULLACOM Ue CaN TMCNET ate ee 321 817°5 |297:5 |277:5 |274 | 293-5 | 293 ee) 20 Smee als2 3. | Maximum breadth of distal end [be- ' tween epicondyles]................ 61:5 | 60:5 | 59 62:2 | 59°5 | 60:5 | 60 60 45 41-2 4, | Maximum distance between greater and NessersubenGsitlesier acini ote 42 40:5 | 41:5 | 465 | 42 40°5 | 41:5 Ey. 50 32:2 5. | Vertical diameter of head ....-.0..... SRY || 37 40 39 89°5 | 38 88:5 se 37 26 §. | Transverse diameter of head.......... 86°5 | 36°5 | 38:5 | 35:5 | 37 36 38°5 Bie) || DS; cS Minimum antero-posterior breadth of thetdiaphysis maar cicaerchierciee 18 18:5 | 19:5 | 21 18:5 | 21 18 19 195 | 15 8. | Maximum thickness of external condyle.) 21 22:2 | 21:5 | 24 22 22:5 | 22 23 19 14°5 9. | Maximum thickness of trochlez ...... 18:5 | 19:5 | 21 20°5 | 21 21 21 20:2 | 19:2 | 155 10. | Width of inferior surface of capitulum..| 14 14:5 | 15-2 | 15 15:5 | 145 | 13:5 | 16 10:2 9 11. | Minimum transverse breadth of diaphysis.. 125 | 125 | 15 16 15:5 | 17-5 | 15:5 | 15:5 | 105 | WE 12. | Width of inferior surface of trochlea....| 21 22°5 | 21-8 | 21 19 18 21 18-5 | [168] 14-5 13. | Minimum thickness of trochlea........| 18:5 | 13:2 | 15 16 16 Wee) 54 15 19 22:5 the descent of Man would be traceable by a line collateral to that of the Simiide from a Tertiary ‘‘ Lemuroid” ancestor. ‘To connect such ancestor with progenitors of a degraded form like Megaladapis may at first sight seem unreasonable, but we have evidence which apparently warrants our connecting the Lemurine with such ape-like forms as Archeolemur. If, then, we are correct in classing Megaladapis among the Malagasy Lemurine, there is a strong probability that its ancestors were once as far advanced as Archwolemur in brain-development. The occurrence in the ancestry of the Hominide and Simiide of forms as low in the scale as the ape-like progenitors of Archwolemur and Hadropithecus is at all events only a question of greater or less remoteness in geological time; and if it is necessary to trace back the lines of descent of such nearly related forms as the Hominidee and Simiidie to middle or early Tertiary times to arrive at their common ancestry, we see at any rate how in | i 4 - PRIMATES FROM MADAGASCAR. 13} (Sis) (in millimetres) of 1 humerus of Archeolemur majori (2) (14). 1 humerus of MWesopropithecus pithecoides [15]. 11 humeri of Lemur [16-26]. idle 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. | 22. 23. 24, 25. 26. 155 164:5 | 147 148 |[182] | 123°5 | 120 1235 | 125 120/122 | 120-2 | 119-5 | 124 122 126 153:5 | 164 147-5 | 147-2 [132] | 123-2 | 121 122°5 | 124 120°5 |121-7 |120-2 |119°5 |124 | 199-5 11926 495 | 41:5 | 40:5 | 34 39 | [84:5]| 33:5 | 345 | 82:5 | 34:8 [84] 30 30°5 | 30-2 | [29:5]| 29:8 31:5 |} 34 33:5 | 26 25:5 | 24:2 | 24 26 21:5 | 23 23 21 21 22 24:8 | 22°5 22:5 | 23 23-5 | 18 18:8 | 17-5 | 19 Tse) || GH 2) ales) 1) TA} 16 S|) AGS | ALY 24 QACOnI 22:5) |) 25) | Ak7 1S Siloam eels iG: OmellG-5iaimleyebil) alle 15 IUG}95) |} SS} |) IL7/ Olson sion We Ioy 0) | 2 Tes) |] ales) 4) aOR} |) Lae} | Lal lh Di Os ra eel 10:7 | 10 145 | 145 | 12:5 | 12 10 iL [9°5]| 10:2 | 10 10 10 10°5 8:8.| 10 [10] 10:2 | 14:5 | 14-7 | 13 12 | 10 | 12:5 | 12:2 | 12 11 Ikea): |} IHL} ]) - aa 10 IUCR) || ORES. || AO) 9:5 | 10:5 | 105 9:5 | [9:5]) 9:5 9°5 9 8 9°5 8 85 75 7:8 85 | [8] U2) Ako) 19 8 6 10°5 9 9 8 65 8 9:5 7 7 9 8 | | [19]. | [20:5]| [17-5]) 15-7 | [11-5]) 145 14 14:5 | 13°5 | 13 Iss || Tales || SHS |) EY |) 1183 12:8 22:8 | 26 22 7 | BH | GHB | AO 88] 9 9 9 2 | 8 92} 88] 8 their case similarity of conditions and mode of life has retained with but little alteration characters derived fram such common ancestry. That Megaladapis, whose size has corresponded with that of Man and the Anthropoid Apes, should have retained with little alteration certain features of its limb-bones which connected it with these highest representatives of the order, seems to me a more probable explanation than to maintain that such features have been secondarily acquired. Paleopropithecus.—This radius is numbered 7 in Table 20. It is from the right side. A reference to text-fig. 85 shows it to be a very stout, much curved bone. The concavity of the head corresponds with the curvature of the capitulum of the humerus of Paleopropithecus described above, though the latter bone is from the left arm. The circumference of the head also fits the lesser sigmoid cavity of an ulna which will be referred to under the next heading. The posterior edge of the circumference of T 2 DR. H. F. STANDING ON SUBFOSSIL 4 G9 184 |Gr 169 | 2 g L \¢-L | G8 9 |¢-2 9 |TF |L9 8 G FL | SL | stsdqderp Jo yy prs osxoasuery wnuMTOTyY | “9 #9 [1:9 |9¢ |¢¢ |¢-2 | [ol |¢-2 Iez |¢-9 | 9-2 9 |¢8 |G9 |8-8 OE | OE | Bit eee iiesee ee BSSOJ PLOULSIS IOSso] Jo YySUET | “¢ Fat GUse IGHSIE ee |Geore | ell | ie ae Wait RIE AL CHIEN The RAE GHC IO kes NIC ee sae eee eos 7+ dy sy 0} WOURLOI[O JO 4IMMUINS WoIy YAsuUET | “F gee || ie | BE BL GL TALI SL IGE BEE IGE | OL | GaAL GS c eee > FAL | PAL WE HS Poe ee eee oleracea BAS] PplOWSIS ToyBaIS Jo YISueyT WNUIXe]T |e 9 eee LONG) S dS} : G z aEGse G Ga ier sae on OSes “++ 90RyINS IB[NIYAIV [BIPVl LOTIayUT ay} 0} ssooord proydjs woz TPT |‘S eg ei) Corll 6c " | Pll] °° |S TST) S-6S1|¢-6ST) * ie See OGGe eee ke eens “+ sseaoad prop {ys 0} WOULOI[O WOT, YSU] WNUIXe]T | *T ere | Me | He SS | SAG | ie wig me Oe BP ey A Se Re ‘RUPN UI9}YS19 Jo (SaAJOWI[ [LU UL) SUOISUSWIP SULIAIN— [ZG WAV], 8:8 + |G > |¢0b |e) 6 1¢6 |le-trllls-10 |¢-70 | lotl) ¢1 \i¢-er | = See LisIel |soeeee eee eee “* goRpIs IB[NIIPAS [VISIP JO ssouUyIIy} WNUIIX]T | °G OL |S-0T |8-8 |G:8 |S: |8-01 | IL |¢-1L /@-EL |] 1] FL 19:9 | €21¢-66| 36] 16] 86} SB) °°” peeYy Jo JojowErp UMUTIxeHy | 7 AGL || 2° GaP " |G-91 | FL 1G¢-6T 1491 | 61 |/9-8e |G6s |S-1E |1¢-66 | TE | - Siig Feecase GLC were as Spa tae AS oovyins IejMoyse [VIstp Jo YJpvorq wnw1xeyq |e e.9¢ |¢-1¢ |9-e¢ 1¢-16 |6-93 |e |¢90 | es] oa} ro) 22] Fo lore | ap | oF |¢-9F leer | Scr | > Aqtsoroqny peydiorg 07 peoy wor | -G Bet] °° |Shl| ~° |S-FFT] SSI] GET] 9ST| EFI/¢-TET] SET/S-EST] ¢-s6c] sel * “+ | ++ | g.99z|" + ssoooad propéys 07 pray moay YASUO | *] HS PVA 1 OS | Se SAS I ele ee Wy ae Ee He os) Sp te a Ss | é © [2] snaayjedoudomjog Jo sniper auo ‘[gt-g]| mper snoaurypeostut weseTa ‘[9-{] sedvpmyvbazy Jo upea xis Jo (SoOUN|[TUA UL) SUOISUOUIIp SuLALQ— QZ TTAVT, PRIMATES FROM MADAGASCAR. 135 the head is considerably raised, while the small articular facet on the anterior border is depressed, thus causing the plane of the proximal end of the bone to lie somewhat obliquely with reference to the long axis. The tubercle is wide and has a deep oblique depression. The interosseous edge is less sharp than in Megaladapis and its lower Text-fig. 34, Text-fig. 35. Radius of Radius of Megaladapis grandidiert. X 3- Palewopropithecus maximus. x3. fourth is much roughened. The anterior surface is deeply concave. The outer edge of this concave surface is clearly defined by a ridge of regular curvature running from the lower end of the tubercle nearly to the carpal end of the bone. The central part of the posterior surface is broad and flat, but its outer edge does not present the sharp angular ridge seen in Megaladapis owing to the very convex transverse curvature of 156 DR. H. F. STANDING ON SUBFOSSIL the outer surface of the bone. ‘The outer edge is continued down to a prominent tubercle (the “ middle thecal tubercle”’ of human anatomy), but in the bone under consideration it is a broad prominent ridge of about 1 inch in length. The two facets seen on the inferior surface of the radius of Megaladapis are here represented by a single concavity of roughly semicircular outline which extends across the end of the bone from the styloid process to the lower edge of the sinus lunatus. THe ULNa. The two ulne referred to in the previous section will alone be here described. ‘That figured (No. 1 in ‘Table 21) is the ulna of Megaladapis (text-fig. 36). Its carpal end is imperfect, but its length estimated from its corresponding radius is approximately 290 mm. ‘This bone is slender and much curved, its curvature being slightly less than that of the radius. The proximal end is remarkable for its curious general resemblance to that of the Hominide. The greater sigmoid cavity is broad and deep. It is less distinctly divided into two lateral halves than in the human ulna. The coronoid process is high and somewhat oblique in direction and internal in position. The olecranon does not extend backwards beyond the sigmoid cavity in a long process as in the majority of the Lemuroids and Monkeys, but its condition closely resembles that of Man. Its lower surface is rounded off, being convex both transversely and longitu- dinally. The shaft of the bone has a somewhat rounded anterior ridge, the surface external to this ridge being concave, while the internal one is convex. The posterior surface is broad and flat. ‘The external boundary of this under surface is a long sweeping curve which can be traced from the lesser sigmoid cavity to the broken distal end of the bone. Between this curve and the outer edge of the anterior surface is a narrow external surface whose sides run nearly parallel at a distance of 3 or 4 mm. apart along the lower three-quarters of the bone. The ulna of Paleopropithecus is a short stout bone, broken and apparently abnormally enlarged at its distal end (text-fig. 37). Its proximal end bears a curious resemblance to that of the ulna of Megaladapis just described. The greater sigmoid cavity is more distinctly divided into two trochlear surfaces of articulation, and the coronoid process is somewhat broader in the transverse direction. The shaft of the bone is rather stouter in actual dimensions than that of Megaladapis, the under surface being much more rounded in section. The angles limiting the various surfaces are less clearly defined, and some of the areas for muscular attachment are rougher in the shorter bone. It may be of interest to compare the radio-humeral index of Megaladapis and Palewopropithecus with that of the Simiide and Hominide. This index is derived from the formule Index = 20 In the case of Meyaladapis the average lengths of three humeri and three radii PRIMATES FROM MADAGASCAR, 137 are taken, The figures for Man and the Anthropoid Apes are quoted from Duckworth’s ‘ Morphology and Anthropology’ :— Ralzopropithecus). .))) 0)) ean Gore Miesaladapishme-.) sin syienencr ay Gore Gorilla sma ey acalin non i ieee Op Clnimaypemvay g 5 5 5 5 60 BERD SMM," ig. Te ag lolol er Tel ak ot) ORR? JSIODAWES 5 5 5 0,0 a0 Wee lelormimes 5 > 5 oo 6 oo th OL Text-fig. 36. Text-fig. 37. Ulna of Ulna of Megaladapis grandidiert, x3. Palwopropithecus maximus. X 3. It is thus possible with the material above described to construct the arm of both of these giant Lemuroids, Unfortunately the same cannot be said of the lower limb. In view of the large number of individuals both of Megaladapis and Paleopropithecus represented by the remains exhumed at Ampasambazimba, the almost total absence of 138 DR. H. F. STANDING ON SUBFOSSIL the large bones of the lower limb of these animals is very remarkable. One perfect and one incomplete femur have been found resembling in their general appearance the femur of Megaladapis edwardsi figured by Dr. Lorenz in his recent memoir on this fossil; but these bones are much too small to belong to the gigantic Megaladapis described in this memoir. In text-fig. 38 are figured two femurs which I am not able satisfactorily to refer to any of the crania described in this memoir. Text-fig. 38. | A. Femur of type referred to by Grandidier and Ameghino: still undetermined, B. Femur of Megalindris gallientt. x3. The femur marked A closely resembles one figured by Grandidier and referred by him to an extinct Edentate which he named Bradytherium madagascariense. Ameghino, however, in speaking of this femur, thinks it hardly possible to attribute it to a Sloth. He says “Ce groupe [the Sloths] s’est constitué a une époque trds récente, quand la communication entre l'Afrique et l’Amérique du Sud n’existait plus, ou était incompléte, représentée par une chaine @iles. Des animaux arboricoles, les plus lents que l’on connait, et qui ne s’éloignent jamais de la localité qui les a vu naitre comment auraient-ils pu passer au continent Africain et arriver jusqu’a Madagascar? Le fait est tellement improbable quil parait presque impossible.” His conclusion as to the character of the bone itself is that while it bears a distant PRIMATES FROM MADAGASCAR. 139 resemblance to the femur of Bradypus it more closely resembles that of the Lemurs. He says: “‘ The heavier and more massive appearance of this bone compared with that of extant Lemuroids is easily explained by the very large size the fossil species must have attained to; it is also to this same cause that the small height of the great trochanter must be attributed. I think, then, that the femur described as belonging to a Sloth of the family of the Bradypodide belongs to an extinct Lemur of great size, probably to Megaladapis madagascariensis or to a related species.” The very curiously flattened shaft of both these femurs seems to me of interest in connexion with the supposed aquatic mode of life of some of the extinct Lemuroids before described, as this flattening of the femoral shaft is a feature which may be observed in aquatic mammals belonging to various orders. The femur of Palo- propithecus ingens figured by Dr. Grandidier in his memoir on ‘Extinct Lemuroids’ presents the same peculiarity. Text-fig. 39. 2 , i | ‘| (‘— SS Os innominatum of Paleopropithecus maximus (?), The ischiae portion has been broken and re-set during life. I propose provisionally to denote the second of these femurs (text-fig. 38, B) by the name of Megalindris gallienii after the distinguished founder of Te Académie Malgache. The dimensions of this femur are given in Table 22, No. 28. The subjoined tables include series of tibize and scapule and pelves, but the only bone to which any reference is here necessary is the 0s innominatum shown in text-fig. 39 VOL, XVIII.—ParT 11. No. 11.—May, 1908. U 140 DR. H. F. STANDING ON SUBFOSSIL TaBLE 22.—Giving dimensions (in millimetres) | eee ic ewMe haber ech) Gh | WO. | Rt, | | | 1. | Length between summit of head and lower surface of internal condyle ..| 146 | 150 | 150 | 145 | 1585) 148-2) 153 | 189°8/ 151 | 149 | i44 2. | Width of bone above third and | lesser trochanters ....,........- 27 29 30 29 29:°5| 32°8| 32:5) 31 30°5| 29°38) 33:5 3, | Minimum width ..css.4 ose yeu e- 12. |] 13:2) 15 13:5| 13 182) G6 25)| 32) lS 71 ele Sh ale Ab || Jalen emake So oodooscedoonsa ac 14 14:8) 14 15:8} 142) 149] 175} 13:8) 14 15 13-2 [ete aWadthyofeneckwuemry ari tunelstachtittes 10°38} 18:5) 13 13 13:5} 13 14:5} 12°38) 11:8] 13-8} 12:2 | 6. | Antero-posterior diameter of head....| 16°38) 18:2} 19 18:5} 19 18 20:5) 18 18:7} 18 ey 7. | Transverse diameter of head ........ 165] 185) 188} 18:2] 188] 18 198} [18] 17-7} 18 17 | 8. | Angle made by axis of head with | Prats OF SME oodooagooosacccgde 36°°8| 41° | 40%5| 40°7} 41°8) 42° | 42° | [42°] | 40°5] 41°35) 33°°5 9. | From summit of head and upper surface of third trochanter........ 47 47 49:2} 47 45°5| 48:2] 54:9] 45:2) 48 48:9} 43-5 (10. | Length of digital cavity............ 20 18:8} 22 21 19 21 21:3) 22 20 20 17 11, | Maximum width of digital cavity....} 10-5 7 a 7 68) 7:2) 95 SH] 55\| SR 12. | Thickness of femurbetween the condyles; 13:5) 14:8) 18 16 AON 5 Sh 1G:2tl el Gn 15 15 14:8 | 13. | Thickness of femur at narrowest part OCH AMIE Dogopoodocodsaoonoboo~ 10 WS) BH NES 12 12:8) 12 115) 11:8; 11:8 14. | Antero-posterior width of external face of external condyle...,...... 20°83) 23:°5| 25°38) 25:2) 24:8) 25:8) 26 |[22-4]| 22:8) 24 23 15. |To maximum width between outer edges of posterior surfaces of condyles.| 24:8} 27°8| 29-8] [27] 29°5| 29:2) 30 | [32] 27:2) 27 25:2 16. | Length from great trochanter to | internal condyle)... .:............- 149-8} 154 | 154-2) 147-5) 153 | 154:8) 157-2) 154 | 155-8} 152°5) 148 |17. | Maximum width of great trochanter..; 33-9) 34:9] 86-8] 36:2} 35:2) 86:2) 41:8} 38:5) 351] 35:5) [325] 'aBLE 23.—Giving dimensions (in millimetres) 1 2 3 4. 5 6 dey Maximum\don ethane scrim tbe criemcnerteeters cusiicere | 166 | 159-5 | [142:5] | 187:8) 128 | 125-2 2. | Maximum diameter of proximal end ...............005 | 39°5| 86:8) 18:2 (17-8]) 27:8) 27 3. | Antero-posterior diameter at level of anterior tuberosity 28 25°8| 15 13:8) 21:8) 23:8 4, | Minimum antero-posterior diameter of head ............ [24] | 24:5) 13°8 14 16°8| 19 5. | Maximum transverse breadth of distal end.............. 22 23:2 | [13-2] 13 18 16°8 6. | Antero-posterior diameter of middle of shaft ............ 12-2} 10:8} 58 58 9 8:8 7. | Height of malleolus (perpendicular).................-5- 8:5} 10 5:5 7 6°8 of twenty-eight femora. - PRIMATES FROM MADAGASCAR. 141 | | 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. | 24. 25. 26. 27. 28. | 137 | 188 | 135:2| 142 | 144-2) 151 | [210] | 150°8 [129:9]| 102 | 183 | 190:5) 176-8) 190 185°2 29-8| 26:8| 27:5 21:2| 26 | 30 | 23:9] 19-5| [22] | 17 | 10 | 29 | 33:8] (26) | 35 | [28-8] 49 14-2) 13:9) 12:5) 11 13:1} 1389] 10:8 9:9} 11-5 8 65| 17 18 17 17:8| 18 26 15:2} 15:8) 138 14-2) 16 15:2 10 12:8 8 7 17 1775| 18 17 17:8] 29:5} 12:2 9:3} 10 10:2) 13:5) 12 gone 58 4:8} 12 15 16°38; 13:8) 13:5) 20:8 19°8)| 17:1} 16:5) 17 17 19 10:8) 15 10:5 7:8 23°5| [19:2]) 23:5] 22:2) 34 19:8} 17:3) 16:8; 17 17-2) 18 11:8) 15:2) 10-2 8 2 23°5| 23°65) 23 23 348 45° 41° 41°5 | 40° 42° 40° | [52°] | 44°5) 41° 38°°5 | 38° 35°-3| 39° 37°°2| 38°5 31°°8 dt 42°38} 41:8) 40 46°38) 49°2) .. 82:2 | [42:5]| 27:8) 17:2| 66 68 64 65:2} 645) 81:5 eineeta |) 168) 17 | [18] | 19:5) [25:5]| 23:5) 23 | 12:8)| 9 | 29 | [si] | 27:5) 31 | 29 | [25] a6] sl & | ies) es OF ia | GS Be) eS) Si aS abs) a eh) es |) i 14-8) 14:2} 18:8] 13 13:2} 14:8) 15°8) 12 9:8; 175) 19 18°5| 19 21 11:8) 11 11:5] 105) 11:2} 11°5 9°83 78} [125] 8 6 16:5) 17°83) 13:8) 15:5] 16 18 22:2| 225) 20:9) 21-8| [22] | 229) 282 19 15:8| 28-8] [30:5] [27] | 31:8 (44) Q7-1| (27:5]) 16 | 24 | 129] | 27 | 242) 17-5 [14] | 14 | 37-5] [40°81] [87] | 43:5 52 [139-8}) 140 | 139-5) 144-8 |(147] 156 | 204 | 154:8] .. [134] | 104 187-5) 195:2| 180:5| 196 Pn (SA [36-8]/ 31 | 32 | 33 | [84] | 345 27 | 34:5] [28°5]) 16:8| 525] 52 | 50:5| 52:8] [463]] 58:5 | of twenty-one tibie. | 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. | 20. 21. 127 | 130°5| 119-3) 121-5| 128-2/ 128 | 119-7] 122-8] 128 | 118-8| 132 | 1515) 92 | 185 | 164:8 [25] | 28-7| [24] | 27-5| 265) 26 | 26 | 25 | 23 | 24 | 29 | 33 | [10] | 245] 23 [19] | 21 | (2051 20-2) 20 | [20-5] 21 | [22] | 22 | 20 | 24 | 215) 51) 88) 82 [15] 16 17 16:5; 16 |) A 19 16 19:2) 19-5 3:5 | 11:8 15:8} 16 15°5| 17:8) 15:8) [17-5] 17 19 145} 13 185) 208, 85, 15:2) 15 8 10 8:8} 10 8-2 (a) Spit} ILL) 8:2 7) 8:7 | 8:2 4:8) 13 12:8 75| 62° 78| (7) | 6 Bi) CG | 68) So) Fel IE | 142 DR. H. F. STANDING ON SUBFOSSI: Taste 24.—Giving dimensions (in millimetres) of scapule. ako 2. 3. ep | 1. | Length of axillary border ..........2---+eeeee seen eee 63°5 70°5 ye 50 | | 2, | Length from spine to summit of acromion .............-| 74:5 | 745 sc 68:8 | 3. | Length of coracoid border to end of glenoid cavity ........ 775 78:5 a 69 | 4, | Length of vertebral border ..........0+ sess eee eee eens 47-5 48°5 66 60 | 5. | Length of spine at its base ......- + sees ee ewer eee ees | 63:8 | 635 ae 86 | | 6. | Vertical diameter of glenoid cavity .....- see eee ence ees 19-2 19 | ay 24 7. | Length of coracoid procesS ...... ee eee ee ee eee eee eee 17 00 24:8 22 | 8. | Maximum transverse diameter of glenoid cavity .......... isp lsl 11-2 13 16 9. | Breadth of coracoid process at its base........-+-.++eeee 9°5 a6 13 11 (Nos. 1 & 2 in Table 25, p. 144). ‘This elegantly formed bone apparently does not belong to any of the animals described in this memoir, The acetabulum shows a large depression for the insertion of a ligamentum teres, but the only two femora the heads of which at all agree in size with this acetabulum (viz. those just referred to above) are apparently quite without any corresponding imprint for the round ligament. That this pelvis belongs to an animal of “ Lemuroid” type there is no doubt from the elegant everted hook-like form of the upper part of the ilium. It is too large to correspond to the femora of Archwolemur, several of which occur in the Collection. It may possibly be the pelvis of Palwopropithecus, though it appears somewhat slender for so massive an animal. COMPARATIVE SURVEY AND CONCLUSIONS. It will be useful in conclusion to take a general comparative survey of the groups described in the preceding sections, with a view of ascertaining what conclusions may be legitimately drawn towards the solution of the problems proposed in the intro- duction of this memoir. Briefly stated these problems are as follows :— (1) What relationship do the recently-discovered subfossil Malagasy Lemuroids bear to each other and to the existing Lemuride and Indrisinz ? (2) What conclusions may be drawn from a study of these subfossil forms as to the origin of the Lemuroid fauna of Madagascar ? (3) In particular, what light, if any, do these subfossil Lemuroids throw on the probable course of development of the Old and New World Monkeys? PRIMATES FROM MADAGASCAR. 143 In endeavouring to answer these questions there are certain considerations which might be expected @ priori to offer some guidance as to the directions in which a solution of these problems might be looked for. We might begin by asking the question: What would be the probable course of evolution of a group of Primates isolated since early or middle Tertiary times, confined to a limited area, and removed from competition with the dominant Mammalian groups which at a later date spread over the African Continent ? Various analogies suggest a probable answer to such a question. ‘The Malagasy flora (in which, according to Baron, three-fourths of the species and one-sixth of the genera are peculiar to the island) would lead us to expect a considerable amount of specialisation and divergence from the most nearly allied continental forms. Again, even a cursory view of the Malagasy representatives of certain other Mammalian orders—e. g. the Insectivora and Rodentia—not to mention the non-Mammalian groups—would fully confirm this surmise. Now it is evident that the course of evolution which has pre-eminently characterised the Primates on the great land-areas has been one of brain-development. We shall, indeed, probably be justified in maintaining that it has been by the acquisition of superior brain-power rather than by superior strength that the members of the _ order as a whole have been able to escape extinction. The removal of a group of Primates from connexion with extensive continental areas, and their isolation in a large island, would probably in course of time so modify the conditions of existence as to eliminate much of the struggle which had hitherto been one of the main factors of the brain-development of the race. One might therefore expect an arrest of such development, and possibly even a “ retrogressive evolution ” leading to actual degeneracy and loss of brain-substance. Now the evidence brought to light by recent discoveries does in fact substantiate these @ priori surmises. In the existing Lemurs of Madagascar we have a highly specialised group many of which have widely diverged from their nearest representatives in other parts of the world. Chiromys may be taken as an extreme instance of this specialisation. In the case of the Indrisine we have been able definitely to connect the existing genera with various extinct subfossil forms which in many of their characters give striking evidence of descent from Ape-like ancestors. These pithecoid subfossil Lemuroids themselves, however, show signs of specialisation and retrogressive changes, in some cases carried even further than in the extant genera. Now in order to bring into clearest relief the fact that these Malagasy genera living and extinct of the Indrisine group do, considered as a whole, retain many charac- teristics which undoubtedly point to descent from Ape-like ancestors already possessing most of the features which we associate with the true Monkeys, I propose to discuss 144 DR. H. F. STANDING ON SUBFOSSIL TasLE 25.—Giving dimensions (in millimetres) 1 2 3 4. 5 6 7 8 9: 10. 1. | Maximum length of os innominatum ...... 245:5| 241 ae 185 2. | From summit of ilium to superior edge of acetabulum rrr ieetametastmr eect 156-5) 146 | 140 | 117 | [95] is [86:5] [87-5] 8. | Maximum diameter of acetabulum ........ 48 47°5| 86°5|) 25 24:5) 25:5) 22 19 Pils} || DS 4, | Minimum diameter of body of ilium (above acetabulum) eeyeppariclerrietelomlreneiretort-ier- 30 28 20 25 21:5] 25:2) 17 14 16 16 5. | Breadth of ilium at level of iliac spine...... 87 83 ap 82:8 | 6! | Width of sacrallsurface!.. 002.0. 0.-s+---.- 82:5} 30°5| 25 26:5; 18 25 16:2) 14:5) [16] | [16-5] 7. | Distance from pubic suture to ischial (MMOH? so ovoaccoonvoouooddo0aoeod 109 briefly a considerable number of the characters enumerated by various writers as differentiating the ‘“‘ Lemuroidea ” from the “ Anthropoidea.” Several preliminary observations should, however, be made before entering on this comparative survey :-— (1) It should be clearly realised that all the subfossil Lemuroids hitherto described although extinct are of very recent date, and from the biological and geological points of view are the contemporaries of the existing genera. (2) In considering the characters of a group of animals, especially where marked specialisation and retrogression have occurred, it will not always be possible to distinguish between ancestral and acquired characters. (3) When any doubt exists we are on the whole more likely to be correct when we interpret as vestigial rather than as recently acquired any suggestions of Simian affinity *. (4) It is in regard to brain-development and such structures as depend upon brain- development that we are likely to find the least satisfactory evidence as to close actual affinity with the higher Primates ; but, on the other hand, it is here that we shall find the clearest evidence of degeneration. (5) Where all the Malagasy genera agree in possessing in various degrees some character which would ordinarily be described as “ Lemuroid,” it may be often possible * It should be noted that it is some of the evtinct genera of Lemuroids which show the greatest number of “‘Anthropoid” characters; and it would seem morc reasonable to regard these features as ancestral than to consider them as adaptive modifications which have arisen by convergence in Madagascar. We have seen, on the other hand, that the specialisation in certain directions to which apparently the recent Lemurs owe their survival (e.g. the procumbent pectinate disposition of the lower incisors) is just such as has tended to disguise their close relationship to the Simian type. PRIMATES FROM MADAGASCAR, 145 of twenty-seven pelves. 11. 12. 13. 14. 15. 16. 17. 18 19. 20. 21. 22. 23. 24. 25. 26. 27, [91] | 80-5 [66-5] [72] | [87] | [88] | [70] [84] | [74-5] [55] | 525] [52:5] [53] 45-2 20:2) 24 23 20 22°8| 22 23 29°35] 14 145; 13:5] 14 138) 14 13 11 9°5 15 16 16:8; 14 175} 18 17 13 10 12:5} 11 85} 10 11 10 6 55 [26] | [22:5]) [17-5] [23-5] [15] [16] | [14:5]| [16] | [15] | (15) | 18 | 15:2) [17-5] 8 | [10] | [125] [8] | 85) (10) | 4:5) [5:8] | 32 31 24 to name some genus or species of Monkey where a similar character occurs. Should such examples prove to be chiefly drawn from one or two families, the presumption will be strong that they are not mere examples of convergence, but point to community of origin (see text-fig. 48, p. 152). To begin, then, our comparative survey :— The Brain.—A comparison of the various subfossil Malagasy Lemurs reveals one feature which is common to all of them without exception, and which has therefore been repeatedly referred to in the previous descriptive sections, namely, the curious narrowing of the frontal region of the skull immediately behind the orbits. The annexed figure (text-fig. 40) shows the outlines of these various skulls all reduced to a common unit length for convenience of comparison. Examples of the nearest extant genera and the skull of a New World Monkey are also appended. If it is asked what has caused this curious narrowing of this region of the skull, it is only possible to offer one or two suggestions. It is noteworthy that it is in all the extinct forms that this peculiarity is most marked. Dr. Elliot Smith points out in his report on the brain-casts of Mesopropithecus, Palwopropithecus, and Lemur major?, which is published simultaneously with this memoir, that it is the frontal lobes of the brain which are the latest to be evolved, and it is this region which is the first to go when retrogressive evolution occurs. May not the loss of intelligence which this degeneracy connotes be among the causes of the disappearance of these forms? It is also perhaps not a mere coincidence that these forms all have relatively more massive mandibles than the extant genera. ‘The presence in nearly every case of a sagittal crest and widely curved zygomata implies the possession of powerful temporal muscles. Is it not possible that the mechanical pressure of these muscles 146 DR. H. F. STANDING ON SUBFOSSIL Text-fig. 40. Showing vertex views of five genera of subfossil Malagasy Lemuroids compared with recent Lemurs and a New World Monkey :—A. Archwolemur edwardsi; B. Mesopropithecus pithecoides; C. Proptthecus diadema; D. Lemur jullyi; FE. Hapalemur griseus; F. Lemur varius; G. Megaladapis grandidicri ; H. Oolobus; 1. Paleopropithecus maximus. These drawings are all reduced to unit length for convenience of comparison. Note the constricted postorbital region in all the extinct forms. PRIMATES FROM MADAGASCAR. 147 may have had its share in accelerating the process of brain-degeneration which had already set in?* Postorbital Wall——With the narrowing of the postorbital frontal region has probably come the almost complete disappearance of the postorbital septum. In the various families of Monkeys the frontal postorbital region is generally so broad and the orbits so comparatively small that the space to be filled in by the septum between the brain-case and the orbital bar is only a small fraction of the circumference of the eyeball—_that is to say, a quite short outgrowth of the malar and frontal bones is sufficient to bridge across the vacant space between the orbital bar and the brain-case. The accompanying text-figure (41), which represents a transverse (nearly horizontal) section through the orbit of Chiropotes, will show how small the actual septum may be. In the case of the Malagasy Lemuroids which we are considering the narrowing of the skull has carried the frontal, orbito-sphenoid, and alisphenoid far away from the Text-fig. 41. f | post- f| orbital septu m Transverse (nearly horizontal) section through the orbit of Chiropotes. postorbital bar, so that supposing a septum did exist in a position analogous to that which it occupies in the Apes, it would have to wall in a space which may be expressed as equivalent to from 70 to 90 per cent. of the transverse diameter of the orbit. That a more or less completely ossified septum did exist, however, in the ancestors of the Indriside seems probable, for there appear to be distinct traces of it both on the outer, upper, and occasionally inner edges of the posterior orbital contour. Attention has been called, in the case of Archwolemur, Mesopropithecus, and Palcopropithecus, to * Tam aware that this theory jvould imply the transmission to offspring of a character acquired during the life of the individual; but it has yet to be satisfactorily shown that a cause acting during long geological ages may not be efficient where experiments extending over a comparatively few generations give purely negative results. ; + In certain skulls of Archwolemur edwardsi there is a raised line on the inner wall of the orbit in exactly the position where the septum would abut on the skull-wall were it continuous behind the eyeball. VOL. XVIII.—PART I. No. 12.—May, 1908. x 148 DR. H. F. STANDING ON SUBFOSSIL the broadening of the internal posterior edge of the orbital bar in both a vertical and horizontal direction. This inner edge is often so thin as to be quite translucent. The extent of this apparently vestigial septum is disguised by the fact that the postorbital bar is triangular in section. If the actual internal measurement from the orbital rim to this sharp edge be expressed as a fraction of the transverse diameter of the orbit, in the case of Archwolemur platyrrhinus it is 47:8 per cent., in Mesopro- pithecus 36°8 per cent. Similar measurements taken in a horizontal direction from the inner anterior margin of the orbit to the posterior edge of the postorbital septum, and expressed as a fraction of the transverse diameter of the orbit, give in the case of Chiropotes 25 per cent., and in the case of Nyctipithecus 47 per cent. The Occipital Region.—That the arrest of brain-evolution and actual loss of cerebral substance in the Malagasy Lemurs should have entailed other retrogressive Text-fig. 42. Inclination of the occipital plane of young Mesopropithecus (A) compared with that of Alouatia belzebul (B). changes is not to be wondered at. The vertical occiput which is seen in the adult Megaladapis and Palwopropithecus is probably an instance in point. It is note- worthy in this connexion that the young of all the Lemurs appear to have the occipital plane much more inclined than the adult. That a vertical occiput may, however, be a secondarily acquired character depending on a massive mandible or on some peculiarity of the vocal organs is seen in the case of Alowatta (see text-fig. 42). Long narrow occipital condyles have been mentioned by certain writers as a Lemuroid feature. We have seen in the case of several of the subfossil forms described in this memoir that the condyles are broad and rounded in shape, con- forming rather to the “* Anthropoid ” than to the “ Lemuroid ” type (see text-fig. 43). A conspicuous convexity for the lodgement of the vermis of the cerebellum is found in the extant Lemurs, and has been mentioned as a characteristic feature. This is absent, at any rate in the adult, in Archewolemur edwardsi, Paleopropithecus, and Mesopropithecus. PRIMATES FROM MADAGASCAR, 149 The great extent to which the supraoccipital enters into the skull-roof has been urged against the association of the Lemurs with the higher Primates; but, as we have seen in both Paleopropithecus and Mesopropithecus, the lambdoidal suture is but very little removed from the occipital crest, and examples may be found among the Text-fig. 43. Showing form of foramen magnum and comparative size and shape of occipital condyles in Lemurs and Monkeys:—A & B. Archeolemur edwardsi; C. Archeolemur platyrrhinus; D. Propithecus; E & F. Paleopropithecus ; G. Mesopropithecus; H. Lemur jullyi; 1& J. Megaladapis; K. Papio; L. Colobus. Text-fig. 44. Showing extent to which occipital bone enters into cranial root in:—A. Mesopropithecus pithecoides ; B. Lepidolemur; C. An Old World Monkey (Cercopithecus dentz). Monkeys in which the extension of this bone along the median line of the skull-roof is quite as great (text-fig. 44). Frontal Region—The small backward extension of the frontal bone along the skull-wall is another character said to differentiate the Lemurs from the Apes; but in this case, again, our fossils furnish instances in which this feature will compare x2 150 DR. H. F. STANDING ON SUBFOSSIL with certain of the South-American forms, as the accompanying figure shows (text- fig. 45). It should, moreover, be noted that the obliquity of the orbits is apt to disguise the extent to which the frontal in reality enters into the formation of the brain-case. Again, it has been urged that the horizontal direction of the facial extension of the Text-fig. 45. Showing backward extension of frontal of Palwopropithecus compared with that of Chiropotes. Text-fig. 46. Facial angle of (A) Alouatta belzebul compared with that of (B) Mesopropithecus pithecoides. frontal offers no analogy to the condition seen in the Monkeys. We have found, howeyer, in the case of Archwolemur, and especially of MJesopropithecus, that this region of the frontal is by no means horizontal, but forms a true “forehead.” In text-fig. 46 the profile of Mesopropithecus is compared with that of Alowatta, the facial PRIMATES FROM MADAGASCAR. 151 angle being indicated in each case by lines tangential to the forehead and median incisors. Nasal Region.—The length of the nasals is another supposed distinguishing feature of the true Lemurs. There can be no question that the lengthening of the muzzle and the concomitant development of the nasal fossa mark one of the directions in which specialisation has taken place among the Malagasy Lemurs. An extreme instance is offered by the newly-discovered species of Megaladapis. ‘The most Ape- like of our fossils however, especially Archeolemur platyrrhinus, have comparatively short and broad nasals, somewhat larger it is true than those of the Cebide, but not greatly differing from these latter either in shape or size (see text-figs. 17 & 20). That the mere lengthening-out of the muzzle can in no sense be regarded as differentiating the Lemurs from the Apes is at once apparent on comparing such Malagasy forms as Archwolemur platyrrhinus with some of the Cercopithecide (sce text-fig. 47). Orbital Region.—We have already noticed one feature of the orbital region, but may briefly allude to two or three others in this place. The size and obliquity of the Text-fig. 47. Truncated muzzle of Archwolemur platyrrhinus compared with the elongated facial portion of skull of Papio. orbits of the Lemurs have been referred to as Lemuroid features. In both these respects some of our fossils range themselves rather with the Monkeys than with the recent Lemurs. Again, the large extension of the lacrymal into the face, the small degree to which this bone enters into the orbital wall, and the position of the lacrymal fossa with its duct opening without the orbit, are further so-called Lemuroid features. It is the exception, however, among the subfossil forms which we have been studying for the lacrymal foramen to open distinctly on the face, and in many cases it is as much within the orbit as in certain genera of Monkeys. It is true that the lacrymal bone never extends so far within the orbit in either extant or fussil Malagasy Lemurs as to meet the ethmoid—there is always, so far as I have been able to observe, a fronto-maxillary suture of longer or shorter extent,—but in Palwopropithecus and 152 DR. H. F. STANDING ON SUBFOSSIL Archwolemur there is a very considerable downward extension of the lacrymal in a manner very analogous to the condition seen in certain American forms (text-fig. 48). Dr. Forsyth Major has in this connexion made a laborious comparison of many hundreds of skulls of Lemurs and Monkeys, and even before the discovery of the new forms described in this memoir had already come to the conclusion that neither this character nor the position of the lacrymal foramen can be used as differentiating Text-fig. 48. Showing shape and extension of lacrymal bone within the orbit and position of lacrymal foramen in various Lemurs and Monkeys:—1. Young Palwopropithecus; 2. Archaolemur edwardsi; 3. Mesopropithecus pithecoides ; 4. Mycetes palliatus; 5. Callithrix personata; 6. Propithecus edwardst. characters (see text-fig. 48 *). Forsyth Major has also shown that even in the extant Indrisine the ethmoid enters to a small extent into the orbital wall, while some non- Malagasy genera have a large os planum. There is no doubt that the large size of the orbits of many Lemurs is a character due entirely to their nocturnal habits. This feature is absent in some of our fossils, but appears in some of the Lemur-like South-American nocturnal Monkeys (text-fig. 49). Temporal Region.—The high, strong zygomatic arch of the majority of the fossil Malagasy Lemuroids has certainly no close analogy with the ordinary condition of this region among the Apes. Is it not probable, however, that this feature is in relation with the heavy jaws, stout temporal muscles, and vertical occiput which have been referred to above? In the higher Apes the expanding of the cranial vault and shortening of the region between the foramen magnum and the posterior margin of the palate has brought about a corresponding shortening of the zygomatic arch. * Nos. 4, 5, 6 are taken from the illustrations to the study by Dr. Forsyth Major above alluded to. PRIMATES FROM MADAGASCAR. 153 That a vertical occiput and heavy mandible may even in a Monkey be associated with long stout zygomata is seen in the case of the “ Howlers” of South America (text- fig. 46, A). The Basilar Region.—The presence of prominent auditory bulle on the base of the skull is considered by some writers as a distinctive Lemuroid feature, but in Paleopropithecus this character is entirely absent. As previously stated, this is probably a secondarily acquired character, and cannot be urged as an argument for Text-fig. 49. Front view of skull and mandible of Nyctipithecus, showing large Lemur-like orbits (¢f. text-fig. 16). specially connecting Palwopropithecus with the Cercopithecide and higher Apes. But in Megaladapis the bulle are by no means prominent; and since the two American families both exhibit this feature it can hardly be urged as a peculiarity of the Lemurs. The Mandible.—Several characters of the mandible of Archwolemur and Mesopro- pithecus have been referred to in the detailed description of these fossils as allying them with the Apes, and need not here be recapitulated. Reference should be made, however, to one or two features urged by Milne-Edwards and Grandidier as differentiating the mandible of the Indrisine from the higher Primates. They state that the coronoid process is much higher than the condyle. ‘This is of course true in the case of the three existing genera; but a comparison with the mandibles of the various fossil relatives of the Indrisinee and with the American Monkeys shows that this character cannot be relied upon as a differentiating feature distinguishing between the Malagasy Lemuroids and the Monkeys (text-fig. 50). The Dentition—TVhe definite association of the recent Indrisine through Mesopro- pithecus with the Archeolemurine disposes of many of the objections raised against the Simian affinities of the extant genera, since the dentition of Archwolemur and Hadropithecus is distinctly Ape-like in character. It is just on the question of the dentition, however, that there seems least prima facie reason for maintaiuing the 154 DR. H. F. STANDING ON SUBFOSSIL Text-fig. 50. Showing comparative height of condyle and coronoid process in mandibles of various Lemurs and Monkeys :— A. Paleopropithecus maximus; B. Archewolemur majori; C. Semnopithecus; D. Nyctipithecus ; EK. Pro- pithecus ; F. Hapalemur. Text-fig. 51. yess Al AS SIS/ AS SS “is A SiS is 12 pmip--2 RSIS IS |S c 1 1 H 3 pm {2 f m 1 mf/g 3 31 6 A Showing progressive deviation from the “ generalised” Primate formula among the various genera of Indriside :—A. ‘‘ Generalised ” formula as seen in mandible of Pithecia; B. Mandible of Archeolemur ; C. Mandible of Paleopropithecus; D. Mandible of Chiromys. On the right the same facts are shown in diagrammatic form. PRIMATES FROM MADAGASCAR. 155 close connexion between the living and extinct genera, which practically all the other features of their skulls bear witness to. It is desirable to examine this question closely. The following considerations may help us to a clear understanding of the subject. There is, according to Milne-Edwards and Grandidier, evidence in the milk-dentition of the mandible of Jndris and Propithecus that the original formula of these genera Was j,2:¢.1:p.3: m8, Which is the typical ‘ generalised” dentition of the Primates, retained by the Lemuride and the Cebide. It is considered that by the non- replacement of one (the second) premolar and the canine the adult dentition is arrived at. There is here, at any rate, evidence of specialisation. Now - : ; ss Cae . t ry i ‘ A ' sn: 5 4 ‘ e 3 wrely Ms y 5 a ev nek s 2 i x ¢ ny ‘ a) cE oe r" * ; eee = xe f i y a ‘ i: S eee: iS 5 eas ‘ c mi x. ~ 5 : : P i = é 5 \ = . PLATE XL SUBFOSSIL PRIMATES FROM MADAGASCAR. % PLATE XI Profile view of skull and mandible of Palwopropithecus maximus A : ‘ <8 dam -o4 otdoosoezayq Wwopuoy ae — oe a “SNWIXVW WONG a a SNOAHLIGOUdOaV1TVd ‘Tep ‘eTreeg TEDL Siem bee ‘neice ea PLATE XIl. Profile view of six skulls of Palwopropithecus maximus. AH. Searle, del PALAEOPROPITHECU MAXIMUS. Trams Xoo Soo V MU FC MM London Stereoscopic Co ip. PUATE XML SUBFOSSIL PRIMATES FROM MADAGASC PLATE XIII. Vertex view of six skulls of Paleopropithecus maximus. X LEP ray lng dk A an “SNWIXVW SNOAHLIPGOUGOaAV1Vd dun 04 ear SO0310. + Wo puo' E T°P eoica-ilal HY Es) coe Here es) ee PUARE, Xv Basilar view of six skulls of Palwopropithecus maximus. lq imap AH. Searle, del MAXIMUS. PALAEOPROPITHECUS erat SUBFOSSIL PRIMATES FROM PLATE XV. Lateral and dorsal views of mandibles and posterior view of skulls of i is Paleopropithecus maximus. Slightly reduced. . London Stereoscopic Co. imp AH. Searle, del MAXIMUS. PALA OPROPILAE CUS i, xv 192 SUBFOSSIL PRIMATES FROM MADAGASCAR. PAG AVE GV EL: Archeolemur edwardsi (Filhol). Bigemle Profile of skull and mandible. Figs. 2-4. Views of mandibles showing variations in dentition. Figs. 5-8. Dentition of upper jaw showing transition from triangular to quadri- tubercular form of 5rd molar; with variations in size and shape of Ist and 2nd molars and premolars. Natural size. st, Newman imp. VULY XV We TAT DX ie) rand Loot. Foo Vo A.H.Searle delet lith. ARCHABROLEMUR HDWARDSI. ny A ; 4 ~ f . 7 ; < net SUBFOSSIL PRIMATES FROM MADAGASCAR PLATE XVII. Vertex view of six skulls of Archwolemur edwardst. -ISQUVMGH YUNWAIOAVHOUV dar 09 erdoosoesayq) wopu07 Tep ‘eTreag # We eh UMN AL MMU PPK PORNO: PGA OX WOU, VOL. XVIM.—PART 11. No. 18,—ay, 1908. PLATE XVIII. Profile view of six skulls of Archwolemur edwardsi. x ee “ISQCUVMaAA UNWAIOAVHOUV ‘dum 1249) atdoososza3q UopuoT, cooly x MAY Abo MME 92°F 72 PY TSP eTteeg Hy oe emer. ‘SUBFOSSIL PLATE XIX. Bares olen View of skull in profile. Front view of skull, Mar tigennts oF Pie faa Geran Doe FeV UM LLL A.H.Searle del.et lith. West, Newman imp. ARCHAEOLEMUR PLATYRRHINUS. TILT Gr SUBFOSSIL PRIMATES FROM PLATE XX. Basilar view of five skulls of Archeolemur edwardsi (Nos. 1-5) and one skull Archeolemur platyrrhinus. xX }. “SNNIHYMALV Td “V ‘9 “ISQUVMGE UNNATIOAVHOAUYV 'S-I du ‘09 etdoosoesayg wopuoTy ‘Tep eo ica=i-al AW KN Ble MME 9% °F PIR POY UAE trtss ony tt ee Gs io oy Hat aa 202 SUBFOSSIL PRIMATES FROM MADAGASCAR. AWE Xxexel: Mesopropithecus pithecoides (Standing). Fig. 1. View of skull and mandible in profile. Figs. 2 & 3. Views of mandible. Fig. 4. Base of skull showing dentition of upper jaw. Figs. 5 & 6. View of occiputs of two adult skulls. All natural size. Dra Lol Lee OME CW. A.H.Searle del.et lith. West,Newman imp. MOF SOIMROMULEUAGUS IWIMMSNs COMI S. sa ORCI, VOL. XVIU.—ParT 11. No. 19.—May, 1908. 204 SUBFOSSIL PRIMATES FROM MADAGASCAR. PLATE XXII. Mesopropithecus pithecoides (Standing). . Vertex view of young skull showing sutures. . Profile of young skull, . Base of young skull (canine just appearing). . View of occiput of young skull. . 5-7. View of vertex, profile, and base of adult skull. ‘The occiput of this skull is shown in Plate XXI. Fig. 6. zee igs] oa de dQ ag m oo pO Fr} me JQ eg go ier) All natural size. rans Look Hoo. Vol SVU LE XMM. A.H.Searle delet lith. WHS) SOMROWIIMEla CWS ~ JZ Isls COD Ia S, Sard eee 206 SUBFOSSIL PRIMATES FROM MADAGASCAR. PLATE XXIII. Skulls of Archwolemur edwardsi (Filhol) and Mesopropithecus pithecotdes (Standing) compared. Figs. 1-3. View of vertex, profile, and occiput of Archeolemur edwardst. All } natural size. Figs. 4-6. View of vertex, profile, and occiput of Mesopropithecus pithecoides. All slightly reduced (¢ of natural size). Compare also Plates XVII. & XVIII. with Plates XXI. & XXII. Srant Dool So. Vel WUE XX A.H.Searle delet lith. West, Newman imp. 1-3. ARCHAEOLEMUR EDWARDSI. A= IMs SOMO LISI CUS) ILDISL Ia CO iD In Ss r ( ee ae ‘ Ta A Ae ain ike eae ‘ | PLATE XXIV. | Profile of skull and mandible of Megaladapis grandidier “IMHIGIGNVHD SIdVGVIVOAW duit 0¢ etdoosoesayq wopuoT Top ‘etreeg Py Ht x i Roy PE ne, peepee MO oe 1% SUBFOSSIL PRIMATES FROM MADAGASCAR. PLATE XXV. Profile and vertex views of Megaladapis grandidice ri. x: Sramdxool Soc Vl AV TC AAV cal London. Stereoscopic Co. imp AH. Searle, del. GRANDIDIERI ME GALADAPIS SUBFOSSIL PRIMATES FROM _ PLATE XXVI. @ Basilar view of skull and top view of mandible of Megal Rear ft lek a rick A UD Coty a da London Stereoscopic Co. imap. AH Searle, del. GRANDIDIERI ME GALADAPIS by et aks ete an PLATE SX Ain, PLATE XXVIII. Three basilar, three vertex, and two profile views of skulls of Lemur ju Lyi. Tramsxool Soe Vl USC NAVI AH. Searle, del. London Stereoscopic Co. mp LEMUR JULLYI vee Nie PI OCT, PLATE XXVIII. Trandaool, Soc Vl AU ICL AVM AFH. Searle, del London Stereoscopic Co. mp LEMUR MAJORI. +h ie TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON (continued). To Fellows. To the Public. - ese. de Ou Stun VOLUME XYV. (1898-1901, containing 52 Plates) . Price5 15 6 7 14 0 Part 1. (1898, containing 8 Plates) . . . . . Bn Ola dele 30 » 2. (1899, contaming 5 Plates) . . . . . oe OnreOnan() 012 0 » 9%. (1899, containmg 4 Plates) . . .. . BO. 8 10 TORE O » 4. (1899, containmg 4 Plates) . ... . Pee Ose) (0) . O12 O » 9. (1900, containmg 24 Plates) . . . . . Pea A AG} ae on LOF. eon (LO0Is contamina, 4cPlates)\e si 5 610.75 ea On Mlseke Oplae0) » ¢. (1901, containing 3 Platesand Titleand Index), 0 9 O 012 0 VOLUME XVI. (1901-1903, containing 88 Plates) . Price5 8 O . 7 4 0 Part 1. (1901, containing 6 Plates) . . . . . a ) Ne 1 4 0 aes (LOOM comtamminenoeelates) mw see ser ys Ones 015 0 eo (hoOls cont aimmeansO\ Plates) a shee sere @ 159 ed 0) > 4. (1902; containme 5 Plates) . . . . ee One Ie 012 0 » 0. (1902, contaming 4 Plates) . . .. . a Wiley 20) 018 0 pe On (1902 Jcentaiminer, oePlates)) =<) 0 he ope @) alia) TslieO se te (L902; contaminey debates). 8 a so et Od 5 6 Ori -@ » 8 (1908, containing 38 Plates andTitle and Index),, 013 6 018 O VOLUME XVII. (1903-1906, containing 41 Plates and 50 Text-figures). . . . . . . Price5 18 6 718 0 Part 1. (1908, contaiming 5 Plates)... . . Se la een O. 110 0 io (OOS: coutammayes Plates) 2 <).. s 3, O1S) 26 018 0 » 9 (1904, containing 13 Plates) . Pte dlre es 219) 110 O 4c (LOWS; cdutammme 8 Plates). Sse. 1G, ok 2G 110 0 » 5. (1905, contaiming 50 Text-figures) . . Ba Orla uO Te2Ou20) » 9. (1906, containing 12 Platesand Titleand Index),, 1 2 6 110 0 VOLUME XVIII. Parr ise 90d eontamune) OUPlates)" sso.) 6 erice OF LooUnes 2. 1 0 0 2. (1908, containing 19 Plates and 63 Text-figs.) , 110 0... 2 ce) So S 3) CONTENTS. II. On recently discovered Subfossil Primates from Madagascar. By Hersert F. Sranpina, D.Sc. (Leeds), M.Sc. (Vict.)—With an Appendix: On the Form of the Brain in the Extinct Lemurs of Madagascar, with some Remarks on the Affinities of the Indrisine. By G. Exuior Suirs, W.4., W_D., FBS, Professor of Anatomy, School of Medicine, Cairo. (Plates X.—XXVIIL; Dextstigs, AOS) coe ae a ee arate ci 5 ae mean ence THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tue scientific publications of the Zoological Society of London are of two kinds—* Proceedings,” published in an octavo form, and “ Transactions,” in quarto. According to the present arrangements, the “ Proceedings”’ contain not only notices of all busi- ness transacted at. the scientific meetings, but also all the papers read at such meetings and recom- mended to be published in the “ Proceedings” by the Committee of Publication. A large number of coloured plates and engravings are issued in the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Among such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “ Proceedings”’ for each year are issued in four parts, in the months of June, August, October, and April, the part published in April completing the volume for the last half of the preceding year. The pagination is now consecutive throughout the issue for the year, so that the year and page are the complete reference. The “‘ Transactions’ contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are issued at irregular intervals. Fellows and Corresponding Members, upon payment of a Subscription of £1 Is. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1881, if they exceed the value of five pounds. Fellows also have the privilege of subscribing to the Annual Volume of the ‘ Zoological Record’ for asum of £1 10s. (which includes delivery), payable on the lst July in each year; but this privilege is forfeited unless the subscription be paid defore the 1st December in each year. Such of these publications as are in stock may be obtained at the Society’s Office (8 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, H.C.), or through any bookseller. P. CHALMERS MITCHELL, May, 1908. Secretary. ia aa nat ata, ~ Ty TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. To Fellows. To the Public. 25 kp Oh £ s. d. VOLUME I. (1833-1835, containing 59 Plates). . Price 313 6 . . . 418 O* VOLUME _ II. (1835-1841, containing 71 Plates). . 5» 4 00... 5 6 6* VOLUME III. (1842-1849, containing 63 Plates). . 5 38 8 3... 411 O* VOLUME IV. (1851-1862, containing 77 Plates). . , 6 20... 8 2 6% VOLUME _ V. (1862-1866, containing 67 Plates). . ,, 5 4 3 SOs LOO VOLUME VI. (1866-1869, containing 92 Plates) 05355 00 bho 07a cia Loar 0 VOLUME VII. (1869-1872, containing 73 Plates). . ,, 10 4 0. . .1312 0 VOLUME VIII. (1872-1874, containing 82 Plates). . ,, 9 8 8 : ea cesl oeleleen VOLUME IX. (1875-1877, containing 99 Plates). . , 12 16. . : 16 2 0 VOLUME X. (1877-1879, containing 95 Plates). . ,, 10 0 38 ...18 7 GENERAL INDEX, Vols. I. to X. (1833-1879) . ,, 0 7 6 0 10 VOLUME XI. (1880-1885, contaiming 97 Plates). . ,, Qe OR ee eee 16 0 VOLUME XII. (1886-1890, containing 65 Plates). . ,, DiS GO ie eae Nee ee) VOLUME XIII. (1891-1895, contaming 62 Plates). ._,, Oyatotin oie snes eaten alll eG) VOLUME XIV. (1896-1898, containing 47 Plates) . A Bebop OM tea a snoring hee) * No copies of these volumes remain in stock. Continued on page 3 of Wrapper. III. An Account of the South-American Cheliferine in the Collections of the British and Copenhagen Museums. By C. J. Wirn.* Received January 2, 1907; read April 23, 1907. [Puates XXIX.—-XXXI.; Text-figures 64-84. ] AT the suggestion of the Director of the British Museum, I have worked out the collections of Chelonethi belonging to that institution; this paper deals only with the South-American part; for showing me this confidence I ask Sir E. Ray Lankester to accept my best thanks. I am also obliged to Dr. C. F. Meinert, the Curator of the Arthropods of the Copenhagen Museum, for letting me work out its rich collections of South-American Cheliferine Sim., partly collected by Dr. Meinert himself in Venezuela. I most heartily thank, too, Mr. Kdw. Ellingsen, of Krageré (Norway), for placing several of his original specimens of South-American species at my disposal, as well as Mr, A. S. Hirst, who was kind enough to assist me with the literature in several ways. Copenhagen, November 1906. GENERAL REMARKS, The author who has done more than any other to advance the study of the Cheliferine from South America is L. Balzan. His researches have converted this region from practically “terra nuda” to one of the best-explored continents with respect to the grsup in question. Later K. Ellingsen published a number of papers describing imperfectly known and new species. He has also done much to increase our knowledge of the geographical distribution of these animals, and by practical analytical keys has made the study and determination of the species a fairly easy task. As the collection at my disposal was very rich (at least 300 specimens, comprising 36 species, of which 13 were new—about 45 species having been previously described), and included a large percentage of the known species, as well as several new ones, and as 1 found that important groups of organs were scarcely mentioned in the somewhat imperfect descriptions given in Balzan’s earlier papers, I thought it would be very useful to work out a kind of monograph. ‘The richness of the material also made it possible for me to define more sharply than in a recently published paper the limits of the systematic groups, most of which in future will probably be recognised as natural genera f. * Communicated by Dr. W. T. Carman, F.Z.S. + To two papers of Alb. Tullgren (23-24), in which he deals with South-American Chelonethi, I have not been able to pay due attention because they were published during the printing of this paper. VOL. XVIII.—PART 11. No. 1.—October, 1908. 2H 218 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA Some of the species established may possibly be identical with forms described by Banks from North America; but if this proves to be the case I may be excused by the fact that the descriptions of that author, at least so far as the False-Scorpions are concerned, are generally too imperfect for even approximate identification. The South-American and perhaps all Cheliferine are divided into at least four groups; the reasons why I have not given these groups the rank of genera are, that one of them may be subdivided, that some interesting species were not at my disposal, and that a comparative investigation of the male organs ought to be undertaken before definitely settling the case. Key to the Groups of Cheliferinze (Simon). a. Fingers of the palps without accessory teeth. Galea better developed in females than in males. a‘, Femora of the first pair of legs with slightly developed articulation between the two portions, with the posterior condylus placed higher than the anterior; flagellum consisting of four hairs; g with sternal spines; distinct transverse grooves ; obtuse or clavate hairs ; generally real eyes; “tactile” hair of tarsus IV. well removed from base; claws and subterminal hair simple . . II. Group of Chelifer subruber Sim. 6. Femora of the first pair of legs with well-developed articulation and with posterior condylus placed lower than anterior; ¢ never with sternal spines. a. Real eyes; obtuse or clavate hairs ; “ tactile’ hair of tarsus 1V. far removed from the base ; distinct transverse grooves ; genital area of g with long ram’s-horn-shaped organs; coxe 1V.(¢) with coxal sac, and its tergites often with lateral keels ; claws and subterminal hair generally with teeth. . . . . . . I. Group of Ch. caneroides L. b?. No real eyes; pointed almost simple hairs; ‘‘ tactile” hair of tarsus IV. basal; transverse grooves generally indistinct or wanting; genital area of ¢ without ram’s-horn-shaped organs ; coxe 1V. without coxal sac and tergites without lateral keels ; claws as well as subterminal hair always simple. IV. Group of Ch. birmanicus Thor. ). Fingers of the palps with accessory teeth. Femora of the first pair of legs with well-developed articulation between the two portions and with posterior condylus placed lower than anterior. III. Group of Ch. cimicoides ¥, «a, “Tactile ” hair of tarsus 1V. more than 4 removed from base ; tibia 1V. without dorsal median “ tactile” hair; tibia of the palps without posterior “ tactile’ hairs in middle; galea larger in ¢ than in 6. a‘, Four hairs in flagellum; real eyes; abdominal tergites with 30 short clavate hairs along hindmost margin and 8 in front. IIL a. Subgroup of Ch. rudis Balz. IN THE BRITISH AND COPENHAGEN MUSEUMS. 219 6°’. Three hairs in flagellum; generally ocular spots; abdominal tergites with not more than 6 hairs in front of the marginal row. a’. Hand at least 1:1 as long as the tibia; transverse grooves generally wanting; obtuse or pointed hairs. III. Subgroup of Ch. subrobustus Balz. 6°. Hand less than 1:1 as long as the tibia; transverse grooves always) prescntaur-W unl onn ella SubsrouphomChsnrcolonbalz. &°. “Tactile” hair of tarsus IV. + removed from base; tibia IV. with dorsal median “ tactile” hair; tibia of the palps with two posterior “tactile” hairs; no sexual difference in galea; hairs pointed and almost simple. . .. . . . . . . IIId. Subgroup of Ch. argentinus Thor. In the following are set forth a few remarks about the specific characters. ‘The structure of the eyes is, of course, of some importance, but undue value should not be given to this character, as variations are sometimes found within the same species, at least between males and females. The shape of the cephalothorax, or more correctly the cephalothoracic shield (for instance, its length compared with its breadth), the structure of the transverse grooves, of the granulations and hairs are, of course, of some importance. The development of the galea is of no small value within certain groups, but its sexual and often important individual variations must necessarily be taken into account ; the flagellum is of more generic than of specific value. The shape of the serruda and lamina interior shows specific differences, but as the individual variations are often well pronounced, and as they are rather difficult to examine, I think that they are generally not worth describing. Jn the description of the palps not only is the shape of the joints important, but also the proportions between their length and breadth. The relative proportions of the different joints are of the greatest importance. As far as the hand is concerned it is necessary sharply to distinguish between the breadth and the depth, to give their proportions to each other, and to compare them with the length of hand and finger. Not only should the structure of the usual hairs be mentioned, but also the arrangement of the longer and more slender “tactile” hairs. Valuable characters are often found in the position of the tactile hairs as well as in the number and arrangement of accessory teeth and “ spots” or ‘‘ sense-spots,” in spite of often considerable variation within the same species. The cove, especially those of the fourth pair, often provide characters of great interest and should always be examined, the sexual difference, of course, being taken into consideration. In most previous descriptions the /egs are scarcely mentioned, and the workers on the systematics of this group have not at all realised the importance of these appendages, not only in the definition of groups of species (cf. the structure of the first pair of femora, of the claws and subterminal hair, and the position of the 2H 2 220 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA “tactile” hair of tarsus 1V.), but also in distinguishing nearly related species (cf. for instance Ch. javanus Thor. and Ch. plebejus With). As a whole, the value of these organs can scarcely be too strongly emphasised, not only when our purpose is the mere determination of a species, but also when we wish to understand its real systematic position; of course, it is not of much value to examine the legs in situ: the first and the fourth pair on the right or left side should be dissected off and measured exactly. The colour of the palps and dorsal surface of the body is sometimes of value, that of the rest of the body is non-important in the description of species. Finally, it should be emphasised that it is of the greatest interest to determine the sex of the specimen to be examined. So far as the terminology, measurements, &c. are concerned, I refer to my earlier publication (cf. 20. pp. 56-58). I. Group of CuELirer cancrores L. Femur of the first pair of legs with wide oblique articular cavity. Chela of the palps without accessory teeth. Real eyes. “ Tactile” hair of tarsus IV. well removed from base. 3 with genital plates of peculiar structure, with long ram’s-horn- shaped organs, with coxal sac, and with posterior margin of the coxa concave. (The claws are generally provided with teeth and the subterminal hair is often of peculiar shape. Two distinct transverse grooves, of which the posterior is sometimes the broader. Hairs obtuse or clavate. 3 generally with tergal lateral keels.) For the characteristics of this group I refer to my earlier paper (cf. 20. p. 152) and to the above diagnosis; I shall only add that the subterminal * hair is sometimes completely simple (cf. the descriptions). The flagellum consists of three hairs, and a sexual difference is found in the galea. ‘Twenty species at least of the great number of Chelifers described as having real eyes may with more or less certainty be referred to this group. From Asia: Ch. bicarinatus Sim.; Ch. bifissus Sim. (?); Ch. bisulcws Thor. (?) ; Ch. borneonensis Kll.; Ch. depressus C. K.; Ch. hians Thor. ; Ch. mortensenti With ; Ch. superbus With; and Ch. amurensis Tullgr. From Africa: Ch. seulpturatus Lew. ; Ch. socotrensis With ; Ch. mucronatus Tuller. ; and Ch. torulosus Tuller. From Europe: Ch. cancroides L.; Ch. disjunctus L. K.; Ch. hermanni Leach ; Ch. lampropsalis L. K.; Ch. maculatus L. K.; and Ch. meridianus L. K. From North America: Ch. biseriatus Bk. and Ch. scabiculus Sim. * T hope that it will meet with general approbation to designate ‘‘ the terminal lateral hair” subterminal in the future. IN THE BRITISH AND COPENHAGEN MUSEUMS. 221 From South America: Ch. imperator, sp.n.; Ch. rex, sp. n., and Ch. macropalpus Tuller. From Australasia (Hawaii): Ch. bifissus Sim. As species of this group have never previously been described from South America (the domestic species Ch. cancroides, of course, excepted), the occurrence of two curiously developed species is of peculiar interest. Synopsis of Species *. a. Claws of fourth pair of legs with teeth. Tergites of ¢ with lateral . keels, and the distinctly gaping fingers of the palps without any process on the inner margin of the movable finger. a’. The hand as long as the finger. The tarsus of the first pair of legs in ¢ dorsally without terminal projection . . . . . . . I. Ch. caneroides L. b'. The hand at least 1-4 as long as the finger. Femur 8°5 (¢) or 5°8 (?) as long as wide. The tarsus of the first pair of legs in 3 dorsally with terminal projection Spe OUST woes 2. Ch. imperator, sp. n. 6. Claws of fourth pair of legs without teeth. Tergites of g with- out lateral keels, and the slightly gaping fingers of the palps with a distinct process on inner margin of movable finger. The femur is 4°7 as long as wide. The tarsus of the first pair of legs in ¢ without any terminal projection dorsully. . . . . . . . . 38. Ch. rex, sp. n. 1. CHELIFER cANcrorDeEs L. From South America this widely distributed species has been recorded from Cape Horn (Simon, 6. p. 36), from Tierra del Fuego (Simon, 12. p. 167), and from Para (Ellingsen, 18. p. 16). 2. CHELIFER IMPERATOR, sp. n. (Plate X XIX. figs. 1 a-2.) Male. Cephalothorav.—One pair of distinctly convex eyes. Cephalothorax distinctly longer than wide behind. Two prominent very deep transverse grooves, which are almost straight, are present. Cephalothorax uniformly and distinctly granular, but also provided on each side with about 12 small tubercles, which each bear a single short clavate hair. Abdomen (Plate XXIX. figs. 1 a-b).—The abdomen is fairly elongated and depressed, in connection with the cephalothorax showing an elongated egg-shaped circumference. All tergites except the first two with distinct longitudinal line. From the first to the eighth the abdominal tergites possess, like the second thoracic, /ateral keels, which are * Chelifer macropalpus Tullgr. (23. pp. 26-29, fig. 1) from Haiti is nearly related to Ch. imperator, sp. n., but scarcely identical with it on account of smaller size, less slender tibia of the palps, and different structure of claws (posterior claw of first pair of legs possesses a tooth). 222, MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERIN A blunt anteriorly, but backwards and upwards produced into a fairly acute stylet; the whole structure appears, in the median segments at least, similar to a hatchet ; the keel of the ninth segment is only slightly developed and that of the tenth just indi- cated (fig. 1 a). Skin everywhere distinctly granular. ‘Tergites bear in a row along the hindmost margin from 12-14 clavate hairs in addition to a single lateral hair in front of the row, which together with the outermost of the row is placed on the lateral keel; the second to the tenth tergites bear also a median hair in front of the row on each side, as well as a single hair between the two mentioned, with the exception of the second and third segments. ‘The eleventh tergite bears in addition to eight hairs, placed in two rows, about 100 small round spots, perhaps seats for minute fissures, and so does the corresponding sternite. Antenne.—The terminal hair extends distinctly beyond the slender cylindrical galea, which possesses two almost obsolete teeth just beyond the middle and three more developed terminally. Palps (Pl. X XIX. figs. 1 c-d).—The mavzill@ are granular all over. Trochanter and femur of palps are distinctly granular, the latter joint less distinctly so ventrally ; the two following segments only slightly granular, except in front laterally ; in addition to the usual granulations the trochanter bears posteriorly some more rounded small tubercles as well as about five larger conical eminences ; on the anterior surface of the femur about 25 triangularly shaped tubercles are observed, while the tibia shows traces only of similar structures. Hairs are short; those of trochanter as well as of anterior surface of femur and in a less degree of tibia are slightly clavate, while others are more or less obtuse except pointed ones of fingers; the tubercles just mentioned always possess a single hair. The trochanter, which has a long and very well-defined stalk, is 2°3 as long as wide ; it is wider than femur or tibia and twice as long as chela is wide ; the anterior outline beyond the stalk is slightly convex and then almost straight, while the posterior is produced just beyond the stalk and then distally becomes a little concave. The femur, which has a short, well-pronounced stalk, beyond which it is eradually, but most distinctly, widened out towards the extremity, is 8:5 as long as wide; anteriorly, beyond the stalk and a short basal elevation, marked off from each other by a low notch, it is slightly concave, while posteriorly beyond the stalk it is slightly convex, and then straight, or even a trifle concave. The ¢bia, which is distinctly shorter and a little wider than the femur, has a short fairly well-defined stalk and is gradually and distinctly widened out towards the tip; it is almost eight times as long as wide terminally; the outline beyond the rather shallow notch of stalk is almost straight and then a trifle concave anteriorly, but posteriorly beyond the small condylus and the elongated badly defined basal elevation it is moderately concave and then convex but for a slight concavity terminally. The chela is seven times as long as wide; the hand, which is much shorter than the tibia and 1:4 as wide, is 4-2 as long as wide, as wide as deep, and 1°5 as long as the finger. The fingers gape very distinctly when IN THE BRITISH AND COPENHAGEN MUSEUMS. 223, closed, as both those movable and, in a less degree, those immovable are concave in the middle and bear on the distal fifth, where they touch each other, a close row of conical teeth. In addition to this the movable fingers possess about five marginal teeth in the middle, and the immovable six teeth in the middle and five more basally. The immovable fingers bear anteriorly about seven “ spots,” arranged as shown in figure 1d, and the movable fingers about 15 arranged longitudinally from base to tip; both fingers appear to have numerous spots on the posterior surface, but details could not be investigated in the single specimen at my disposal. Coxe (Pl. XXIX. fig. 1 6).—The second and third pair of somewhat irregular shape, especially the third pair, which widened out towards the extremity and rather elongated in shape somewhat resemble the fourth pair, which are almost 2°5 as long as wide, very short basally and distinctly widened out towards the end, where they bear a prominent exterior spine. ‘The posterior margins of the fourth pair are distinctly concave in the middle. Coxal sac.—The coxal sac is very small and is far from filling the whole coxa; in structure it is very similar to that of Ch. mortensenti With, as it has a well-developed basal portion, scarcely half as long as the distal; posteriorly the two portions pass into each other, but are separated anteriorly by a deep cleft, continued above and below into a shallow depression. As regards the structure of the funnel, which has some of the fused hairs prolonged into the cavity of the distal sac (cf. 21. pl. vii. fig. 4), I refer to Ch. mortensenii With (20. pp. 46-47). The distal portion of the sac tapers somewhat towards the end, where it bears numerous setigerous cones and ridges, which, towards the base, are most marked on the anterior surface. Legs (Pl. XXIX. figs. 1 e-f).—Hairs short, especially dorsal ones, obtuse or slightly clavate. Subterminal hairs of the first pair bear no teeth; the anterior is shorter and more suddenly curved (fig. 1 f, hz) in contrast to the posterior, which is more slender and only slightly curved (he); the posterior subterminal hair of the fourth pair is slender and moderately curved, with a tiny ventral tooth near the tip. A tarsal ‘* tactile” hair, 4 removed from the tip, seems to be present. The iegs are rather long and slender; the trochanter of the first pair is much deeper than the femur proper, probably in correspondence with the very wide articular cavity; the tibiais much (1:3) longer than the ¢arsus, which is about four times as long as deep in the middle, where it is deepest on account of the slightly convex ventral margin; this joint is terminally, behind and above the articulation of the claws, produced into a small conical eminence (p). The jemur of the fourth pair of legs, which is very suddenly raised beyond the stalk, is 2°8 as deep as long, 1:3 as long as tibia, and 1-4 lower than tarsuslong. The anterior claw of the first leg (fig. 1 f, ci) is abruptly curved, with a median, almost obsolete, ventral tooth, in contradistinction to the posterior claw (cex), which is only moderately curved, without any tooth; the claws of the second pair of legs appear bifurcate, while those of the third and fourth pair possess a rather small anterior tooth. 224 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERIN Colour.—lhe palps are reddish brown; the body is darker brown, with almost black lateral keels. Measurements.—Cephalothorax 110 (0:920); abdomen 2°19 (1°38) mm. Palps: trochanter 0°736 (0°322); femur 2°162 (0°253); tibia 2:070 (0°263); hand 1-541 (0°368), depth 0°370; finger 1-035 mm. Leg I: femur 0°70 (0:190), trochantin 0-130 (0°2138); tibia 0-486 (0-129); tarsus 0-32° (0:099) mm. Leg IV.: femur 0°930 (0°334); tibia 0°699 (0°167); tarsus 0-471 (0:106) mm. Lremale. Cephalothorax.—The transverse grooves, especially the posterior, are less distinctly marked. Aldomen.—Vhe abdomen is less depressed; the second to the tenth tergites with distinct longitudinal line and the first and last tergites with trace ofone. Tergal longi- tudinal keels are wanting. Number of hairs is somewhat greater; in the hindmost margival row, for instance, 14-16 are found. Antenne. The terminal hair extends scarcely beyond the galea, which terminally bear about four short teeth. Palps (Pl. XXIX. figs. 1 g-h).—The mazilie are only distinctly granular laterally. The trochanter, which has a rather short, but well-defined stalk, is only 17 as long as wide, and 1:3 as long as chela is wide; anterior outline is slightly convex, while the posterior, beyond the stalk, is rather suddenly produced into a conical tubercle somewhat dorsaily placed. The femur is only 5°8 as long as wide and is less widened out towards the extremity. The ¢idia is almost as in the male, but only 4:3 as long as wide. ‘The chela is 4:4 as long as wide; the hand is only 2 6 as long as wide, but 1:4 longer than the finger; the fingers (fig. 1h) gape slightly when closed and are provided with marginal teeth from base to tip. The number of “ spots” is almost the same as in the male, but they are somewhat smaller anteriorly; the immovable finger has posteriorly about five “spots,” arranged between two basal and the median tactile hairs, and the movable has about ten ‘‘ spots” arranged in a longi- tudinal area between and beneath the tactile hairs (cf. fig. 1h). Core (Pl. X XIX. fig. 1 7).—As regards the difference from the male, I refer to fig. 17. “ Legs.—The claws of the first pair of legs possess large anterior teeth. The tibia of the first pair of legs is only 1:2 as long as the tarsus, which gradually tapers towards the end, where no dorsal conical process is found, and attains a length 4°5 times its depth. The femur of the fourth pair is three times as long as deep and 1-6 lower than tarsus is long. Colour.—The palps and, more markedly, the body somewhat lighter. bo bo Or IN THE BRITISH AND COPENHAGEN MUSEUMS. Measurements.—Cephalothorax 1:06 (0°88) ; abdomen 1°84 (1°27) mm. Palps: trochanter 0°552 (0°322) ; femur 1:472 (0-253); tibia 1-288 (0-299); hand 1:127 (0°437), depth 0°437 ; finger 0°805 mm. Leg I.: femur 0°638 (0°190), trochantin 0°122 (0°198) ; tibia 0:456 (0:122) ; tarsus ()-395 (0:089) mm. Leg IV.: femur 0°836 (0-296); tibia 0°646 (0°160); tarsus 0°456 (0°122) mm. Materiai.—Of this very interesting species from Brazil I have examined tw~ speci- mens, a male and a female. Remarks.—lIt is not without hesitation that I have referred the two specimens( ¢ @ ) to the same species, as the differences in their whole appearance are so striking, and as the finger of the female is somewhat longer than in the male, while the contrary is generally the rule; but as they were found together, and as great similarity is found in the structures which are not subject to sexual variation, I think that the course I have taken is the best one. They differ from Ch. cancroides L. in the hand, which is much shorter than tibia and longer than finger, together with numerous other characters taken from minute structure of claws, coxal sac, &c. 3. CHELIFER REX, sp.n. (Plate X XIX. figs. 2 a~e.) Male. Cephalothoraz.—One pair of fairly distinct convex eyes. Cephalothorax as long as wide behind. Two not very prominent, almost straight transverse grooves present. Cephalothorax uniformly and distinctly granular everywhere, but besides on each side provided with at least 20 very smail tubercles, each bearing a single very short and distinctly clavate hair. Abdomen.—The abdomen is moderately slender and depressed, with almost parallel sides. All tergites except the hindmost part of the eleventh with distinct longitudinal line. Tergites without trace of longitudinal lateral keels. Skin everywhere distinctly granular with scale-shaped granules. ‘The first to the eighth tergites each bear along the hindmost margin a row of 14 short, distinctly clavate hairs, while the ninth and the tenth bear only ten or twelve; the third tergite bears in addition to those in the row a lateral hair on each side, in front of the row, while all the following segments bear six hairs in front. The eleventh tergite bears about eight hairs and a number of small round spots, and so does the corresponding sternite (cf. Ch. imperator, sp. n., in which the number appeats to be greater). Antenne.—VThe terminal hair extends very markedly beyond the galea, which is short, suddenly attenuated in terminal third, with three short terminal teeth. Palps (Pl. XXIX. figs. 2a-c)—The mawillw are scarcely granular at all in the middle, but fairly distinctly so laterally. The palps, with the exception of the fingers, are distinctly granular, especially the three basal joints laterally and above ; VOL. XVIII.—ParT 11. No. 2.—October, 1908. Zl 226 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERIN2 the trochanter has in addition to the usual granulations small round tubercles as well as three or four larger eminences; the anterior surface of the femur shows traces at least of similar, though much smaller tubercles, which, like those of the trochanter, carry a single hair. Hazrs are rather short; those of the trochanter and of the femur, at least dorsally and anteriorly, as well as those of the anterior surface of the tibia, are more or less distinctly clavate, while a few on the posterior surface of the femur, those on the posterior surface of the tibia, as well as those on the hand, are obtuse with one or several terminal teeth; fingers with completely simple and pointed hairs in addition to the tactile ones. The trochanter, which has a rather short, but very well- defined stalk, is 1:8 as long as wide; it is wider than the femur as well as the tibia and 1:4 as long as the chela is wide; its anterior outline beyond the stalk is moderately convex, while posteriorly it is slightly convex on the ventral surface, but distinctly produced on the dorsal. The femur has a short and fairly well-defined stalk, beyond which it is gradually and fairly distinctly widened out towards the end. It is 4:7 as long as it is wide; anteriorly, beyond the stalk and a scarcely marked short convexity, it is almost straight (or even a little concave), while posteriorly it is slightly convex basally and terminally, but almost straight between. The tibia, which is much shorter and a little wider than the femur, has a fairly long and well-defined stalk and is gradually widened out distally; it is 3-4 as long as broad; the anterior outline beyond the stalk is very slightly convex, while the posterior one beyond the insignificant condylus and elongated basal elevation is at first almost straight and then a little convex. The chela is 4:4 as long as wide; the hand, which is somewhat shorter than the tibia and 1:4 as broad, is almost 2°5 as long as wide, much broader than deep and scarcely 1:1 as long as finger; the lateral outlines are very slightly convex, with the greatest width just beyond the base. The fingers gape moderately when closed and have along the margins a row of conical teeth (fig. 20); the margin of the movable finger has near to the base a ventral elongated elevation (pv) and opposite to this another much more pronounced dorsal one, bearing about five teeth and fitting into a depression or swelling of the immovable finger (figs. 2 0-c, pd), the margin of which is accordingly curved outwards; the immovable finger has just beyond this depression posteriorly near the dentated margin a small tubercle (pz). The immovable finger bears anteriorly six and the movable five “ spots,” while both fingers posteriorly bear three spots, arranged as shown in the figure (fig. 2 ¢). Coxe (Pl. XXIX. fig. 2d)—They are less elongated than in the preceding species (cf. above), especially the third pair, which are as broad as the second interiorly, and not very much widened out; the fourth are long and slender and distinctly widened out from a very short base towards the extremity, where they are scarcely as wide as the third pair; the exterior spine of the fourth pair is represented only by an insignificant short process (s), and the posterior margin is almost obtuse-angled. Coral sac (Pl. XXIX. fig. 2d)—The coxal sac is small, only about one-third as IN THE BRITISH AND COPENHAGEN MUSEUMS. 227 long as the whole coxa and only occupying an inconspicuous portion of it. Its basal portion, which is almost half as long as the distal, seems in structure to be similar te that of Ch. mortensenii With, but has that part of the funnel which consists of fused “cones” very long. The distal portion of the sac is distinctly attenuated from the base, which is wider than the basal portion towards its end; about 15 long and slender setigerous cones, with their hairs directed towards the base and the interior of the sac, are placed along the anterior surface only; the ridges are only poorly developed. Legs (Pl. XXIX. fig. 2 ¢).—Hairs of the dorsal surface are rather short, slightly clavate or obtuse; those of the ventral side are generally long and slender, almost simple or completely so. Of the subterminal hairs only the anterior of the first pair, which is rather straight and apparently with a single tooth, and the anterior of the fourth pair, which is moderately curved and without teeth, are present. A tarsal ‘“tactile” hair 3 removed from the base seems to be present. Legs are fairly long and slender ; the trochantin of the first pair of legs is much deeper than the femur proper, probably in correspondence to a very wide articular cavity; the ¢¢d2a is scarcely 1:1 as long as the tarsus, which is 5°6 as long as deep and gradually attenuated towards the tip from the base, where dorsal and ventral margins are slightly convex. The femur of the fourth pair of legs, which is fairly suddenly raised beyond the stalk, is 2°3 as long as deep, 1°3 as long as tibia, and 1:2 lower than tarsus deep. Claws bear no teeth ; the posterior of first pair much more slender and moderately curved. Colour.—Palps and cephalothorax generally dark reddish brown; abdominal tergites brown. Measurements.—Cephalothorax 0°92 (0°92); abdomen 1°73 (1:15) mm. Palps: trochanter 0°506 (0°280); femur 1:081 (0-230); tibia 0°897 (0-265); hand 0°828 (0°368), depth 0°330; finger 0°782 mm. Leg L: femur 0°559 (0°160), trochantin 0:114 (0:182); tibia 0-580 (0-129); tarsus 0:357 (0099) mm. Leg IV.: femur 0°737 (0°327); tibia 0°585 (0°152); tarsus 0°403 (0:098) mm. Female. Cephalothoraz &c.—The cephalothorax is a little longer than wide; transverse grooves, especially hinder, less distinct ; the abdomen is more cylindrical in shape. Antenne.—The terminal hair scarcely extends beyond the galea, which has a few terminal teeth. Palps.—the trochanter is 1:6 as long as wide and of a less slender appearance ; the other joints scarcely different from corresponding ones of the other sex. ‘The arrangement of the spots anteriorly only differs from the male in mere details; the immovable finger has the fourth “spot” placed beyond the median hairs and the two following close together, while the movable has only three spots, two basally and a 212 a 228 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINZ single one more apart; posteriorly the three “spots” on the movable finger stand more apart. The fingers do not touch each other in the middle when closed, but leave a narrow space between. Coxe.—The fourth pair are very similar to those of the preceding species, but even shorter (cf. fig. 17). Legs.—Subterminal hair of the first pair is rather suddenly broken or curved without any tooth anteriorly, while posteriorly it, as in the fourth pair, is moderately curved. The tarsus I. is much more slender, being 5°7 as long as deep; the femur of the fourth pair of legs is 2°5 as long as deep and 1-4 lower than tarsus long. The posterior claw of the first pair of legs is more slender and less curved than the other claws, but in a smaller degree than in the male. Measurements.—Cephalothorax 1:08 (1:04); abdomen 2°55 (1°38) mm. Palps: trochanter 0°529 (0°322); femur 1-219 (0°253); tibia 0-989 (0-299) ; hand 0-920 (0-414), depth 0°368; finger 0-898 mm. Leg I.: femur 0°651 (0°175), trochantin 0-122 (0°190); tibia 0-410 (0:114); tarsus 0388 (0:067) mm. Leg IV.: femur 0-820 (0°330); tibia 0:646 (0-167); tarsus 0-471 (0-084) mm. Material.—Of this species I have examined five males and one female, as well as an immature specimen from Brazil. Remarks.—That the female and the males belong to the same species I regard as certain. I am not quite convinced that the males are quite full-grown, as the first abdominal tergite is longitudinally divided as in the female, while the male of Lophochernes has generally the first two tergites undivided ; but as well-marked sexual characters are present in the cox, tarsi, and maxillar chele, I think they are probably mature. This species, which is most remarkable by the curious structure of the male chela, is easily distinguished from Ch. inperator, sp. n., as well as from Ch. cancroides L., by the unarmed claws. II. Group of CHELIFER SUBRUBER Sim. Articular cavity between the two parts of the femur of the first pair of legs not very wide and only slightly oblique, with posterior condylus placed more ventrally than anterior. Fingers of the palps without accessory teeth. Real eyes or ocular spots. “Tactile” hair of tarsus IV. never basal. Male with genital plates of simple structure and its median sternites with differently developed spinous areas. (Claws simple and subterminal hair never with teeth. Two distinct transverse grooves. Hairs slightly or strongly clavate. Flagellum consists of four hairs.) The characteristics of this group are in the main as set forth in my earlier publication on this theme (cf. 20. p. 133). ‘The eyes are real or represented by ocular spots; sometimes we find eyes in the females, but ocular spots in the male—for instance, in Oh. longichelifer Balz. The cephalothorax is more or less granular, with two IN THE BRITISH AND COPENHAGEN MUSEUMS. 229 distinct transverse grooves. The tergal sclerites are also granular, with more or less distinetly clavate hairs along the hindmost row in addition to hairs (generally six) in front of the row in the median segments at least. ‘The genital area of the male is very similar to that of Ch. subruber Sim. (ef. 20. p. 121, pl. iti. fig. 76); the anterior plate, which is wider than long, has a posterior marginal seam, and is shorter, as well as narrower, than the posterior plate; a great number of accessory club-shaped glands are present. The sternites of the males are beset with short spines, arranged in well- defined areas; the number of the spines on each segment, the shape of the area in which they are placed, and the number of the sternites, which are adorned with such areas, vary considerably in different species: in Ch. murrayi Poc. the fourth to the tenth sternites bear numerous spines, and in Ch. canestrinii Balz. only the seventh and the eighth sternites bear areas of a somewhat remarkable structure (cf. Ellingsen, 32. pp. 13-14; With, 20. pp. 129, 158, & 162, tab. iii. figs. 76 & 86; also Pl. XXIX. figs. 3a, 5a, & 6a)*. The galea in the female of the South-American species is long and slender, distinctly extending beyond the terminal hair and provided with about five terminal branches, while in the male it extends scarcely at all beyond the terminal hair and its teeth are more or less obsolete; in the two Asiatic species examined the galea has a somewhat different structure (cf. 20. tab. iii. figs. 7d & 8c-d). The flagellum always consists of four hairs (cf. 20. p. 158). The palps are generally longer and more slender in the males than in the females, but seldom in so marked a degree as in Ch. longichelifer Balz.; the fingers never bear accessory teeth and seldom gape distinctly in the male. The articulation between the two parts of the femur of the first pair of legs is very similar to that of Ch. subruber Sim., but seems to be somewhat better developed (cf. 20. p. 37, figs. da, 6). The articular cavity is not very wide, slightly oblique, with the anterior and posterior condyli opposite to each other, or the latter more dorsally placed. The subterminal hair has no tooth, and the dorsal “tactile” hair of the fourth pair of legs is median or placed beyond the middle ; the proximal joints of the legs are more or less granular, and the claws are always simple. The following 14 species may with certainty be referred to this group, but without doubt the number will be largely augmented by described as well as undescribed species, when naturalists, working at this group, have realised the value of a more thorough investigation. I am disposed to think that the group of Ch. subruder Sim. is the central one of the whole subfamily, but before settling the question definitely a comparative-anatomical investigation of the male organs of the different types of Chelifer Geott., as well as of the Garypide Hauns., is absolutely necessary. From South America: Chelifer canestrinii Balz.; Ch. eailimanus Balz.; Ch. insignis, * Alb. Tullgren (23. p. 33) writes about Ch. torulosus Tullgr., which belongs to the group of Ch. can- croides 1..:—** Die vordere Ventralplatten scheinen daneben eine transversale Area kurzer Stacheln (wie Ch. angulatus Ell.) zu besitzen.” 230 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA sp. n.; Oh. longichelifer Balz.; Ch. nobilis, sp. n.; Ch. proximus Ell. (*); Ch. rufus Balz.: Ch. satanas, sp. n.; Ch. segmentidentatus Balz.; Ch. emigrans Tullgr. From Europe: Ch. subruber Sim. From Asia: Ch. subruber Sim.; Ch. murrayi Poc. From Africa: Ch. angulatus Ell.; Ch. simoni Balz.; Ch. tenuimanus Balz. (?); Ch. angustatus Tullgr.; Ch. concinnus Tullgr.; Ch. facetus Tullgr.; Ch. lamellatus Tullgr. ; and Ch. termitophilus Tullgr. From Australia: Ch. subruber Sim. Synopsis of Species *. a. Hand 1°] as long as finger or shorter. a. Second thoracic tergite with two white spots. Ocular spots. Chela is 3°5 as long as wide; hand at least 1°] as long as tibia. Femur IV. 2:5/as long as\deep.. - 5 9.4. 6. Ch: nobihs) spon. d)- b'. Second thoracic tergite unicolor. a>, Real eyes. Chela at least 3:4 as long as wide. d tergites not serrated and fingers of palps scarcely gaping. Femur IV. 2-7-8 as long as deep. a’, Fingers 1:25 as long as hand, which is distinctly shorter than the tibia. Maxille almost smooth. . . . . 4. Ch. proaimus Ell. é°. Fingers not more than 1:1 as long as hand. a’. Tibia 1:2 longer than the hand. Maxille distinctly granular all over. Femur 5-4 (g:) or 4 (?) as long as wide and the hand as long as finger. Hairs distinctly clavate. VII.-IX. sternites (¢ ) with spinous areas not circular and with spines apart. . . . . . 5. Ch. rufus Balz. 6'. Tibia as long as hand. Maxille almost smooth. Femur 3:7 as long as wide and hand about 1:1 as long as finger. Hairs slightly clavate. VII.—-VIII. sternites (¢) with a single one or a pair of circular areas with spines (?) close together Gen ga ee 7. Ch. canestrinii Balz., 3. B. Ocular spots. Chela 2°8-3 as long as wide. Tergites of ¢ with hindmost margin serrated, and fingers of palps distinctly gaping. Femur IV. 2-22 as long as deep. a’. Trochanter beneath with long conical process . . 10. Ch. satanas, sp.n., 3. b°. Trochanter beneath without a conical process. . 9. Ch. segmentidentatus Balz. b. Hand at least 1:2 as long as finger. a’, Femur with basal anterior apophysis; hand 1:1 as wide as tibia and 1:7 aslong as finger . . . ... . =. ~~ ~ 412. Ch. extlimanus Balz. * Chelifer emigrans Tullgr. (23. pp. 35-86, fig. 5), from Brazil, certainly belongs to this group. It is similar to Ch. insignis, sp. n., by number of tergal hairs in front of row, but differs by wider abdomen and more slender palps. IN THE BRITISH AND COPENHAGEN MUSEUMS. 231 6°. Femur without basal apophysis. a’. Median abdominal tergites with 8 hairs in front of row. Real eyes. Chela is 3 as long as wide; hand is 1°3 as long as tibia and 1°3 as long as finger. . . so 0 6 0 0 dil, Gh, CASIFIS) S05 Woy 2 o. Medien abdominal tergites with 6 mi or less in front of row. . Femur is 6°3 as long as wide and distinctly widened out towards the extremity. Hand is 1‘5 shorter than tibia and i-byaslongeasiuhe timceny. mere le eee Oa Calongrchelsfermbalzs avs b°. Femur not more than 4 as long as wide. Hand 1:2 shorter than tibia and 1°3 as long as finger or less. a. Femur at least 3°5 as long as wide and tibia at least 3 as long as wide. Femur IV. at least 2°5 as long as deep. a’. Chela at least 3°5 as long as wide. Hand only 1:2 as long as finger. a’. Hand as long as tibia. Hairs slightly clavate. Second thoracic tergite unicolor. Realeyes. 7. Ch. canestrinii Balz., 3. 6". Hand much shorter than tibia. Hairs distinctly clavate. Second thoracic tergite with two white spots. Ocular spots . . . 0.0 0 | Gh ClO, MAWHOS, S95 Wo, @ o 6". Chela 3°1 as long as wide. Hand Nt 3 as long as finger, and 1:1 shorter but 16 wider than tibia. Ocular SPOLSH(G Madeeanees) /vsihiruerh disap . . . . 8. Ch. longichelifer Balz., 2. 6°. Femur 3°2 as long as wide and slbie about 2°5 as long as wide. Chela 2°8 as long as wide; hand 1:2 as long as finger. Femur LV. 2:2 as long as deep. Ocular spots. 9. Ch. segmentidentatus Balz., ? . 4, CHELIFER PROXIMUS Ell. 1905. Ellingsen, (19) pp. 324-326. This species agrees in almost every respect, so far as Mr. Ellingsen’s description goes, with Ch. nobilis, sp. n., except that it is somewhat smaller, that the fingers are 1-25 as long as the hand, instead of nearly equal to it, and that the second thoracic tergite has no white spots.—Argentine. 5. CHELIFER RUFUS Balz. (Plate XXIX. figs. 3a-e; text-fig. 64.) 1890. Balzan, (10) pp. 431-482, tav. xv. figs. 15-15.0. 1891. Balzan, (11) pp. 533-534, pl. xi. fig. 25. 1902. Ellingsen, (15) pp. 158-159. 1907. Tullgren, (23) p. 37. Male. Cephalothorax.—Fairly distinct, moderately curved. Real eyes. Cephalothorax distinctly longer than wide behind. ‘wo fairly distinct, almost straight transverse 230, MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA grooves. Distinctly and coarsely granular all over; strongly clavate hairs, somewhat similar to scales of a butterfly, found in numbers. Abdomen (Pl. XXIX. fig. 3a).—All tergites longitudinally divided, but the first three very indistinctly so. Sclerites as well as lateral dorso-ventral margin distinctly granular. About 10 strongly clavate hairs are placed along posterior margin, and 4-6 in front of row on the hindmost (VI.—X.) tergites; XI. tergite bears in addition to two rows of clavate hairs a pair of not very long simple “tactile” hairs (cf. Ellingsen, 15. p. 158). About genital area, cf. p. 229. The IV.-VI. sternztes are longitudinally divided and so are sometimes the X.-XI.; the VIII. sternite is undivided or shows a trace of longitudinal division in anterior fifth only; the VII. is either completely divided or shows only trace of division anteriorly and so generally does the IX. sternite. The sternites are smooth in the middle, but granular laterally. The VIII sternite is longer than any of the others and has in the middle an almost semi-ellipsoidal pale area, which has its anterior convex margin fairly well marked off from the surrounding darker skin, and its hinder margin, which is a trifle convex, fairly well raised and distinctly chitinised; this area occupies 1 of breadth of the sternite and # of its length. It possesses in addition to the usual hairs about 50 short spines, placed within well-marked rings, crowded in the middle but standing more apart laterally. The VII. and IX. sternites show areas of similar appearance, possessing about 15 or 20 spines respectively, but less distinctly limited laterally ; but these areas almost disappear in specimens in which the corresponding sternites are completely divided, and the spines are then placed interiorly near the posterior margin in each half of the sclerite (fig. 3a). In this species it thus seems evident that the spinous area forms part of the sclerite and does not stand between its two halves. Antenne.—The terminal hair extends beyond the short acute galea without distinct teeth (cf. Ellingsen, 15. p. 159). Palps (Pl. XXIX. fig. 3 6).—The mazille are distinctly granular all over and so are the palps with the exception of the fingers. Hairs strongly clavate and rather short on trochanter and femur, less distinctly so on tibia and hand. The trochanter, which has a short, but well-defined stalk, is 1:5 as long as wide; the anterior outline is distinctly convex, while posteriorly it is produced ventrally, and even more so dorsally, so that it appears bigibbose. Jemur has a short, not well-limited stalk, beyond which it is distinctly widened out towards the tip. It is almost 5 as long as wide; anteriorly beyond a very short basal elevation it is very slightly concave; posteriorly it becomes gradually convex just beyond the stalk and towards the termination, but between it is almost straight. ‘The tibia, which is distinctly shorter and wider than the femur, has a short, very well-defined stalk and is 5°5 as long as wide; the anterior outline beyond the stalk is slightly convex, with the greatest curvature near the base, terminally it even becomes a trifle concave ; posteriorly beyond the low condylus and fairly well- marked basal elevation it is slightly concave and then terminally moderately convex. IN THE BRITISH AND COPENHAGEN MUSEUMS. 233 The chela is 3°7 as long as wide; the hand, which is distinctly narrower than the trochanter is long, is 1:2 shorter but 1:4 wider than tibia, is almost 1:8 as long as wide, distinctly wider than deep, and scarcely as long as fingers, which only gape a trifle, when closed. The immovable finger possesses anteriorly only a single ‘‘ spot,’ placed just beyond and beneath the third tactile hair, which is placed almost in the middle; the immovable finger has three spots posteriorly and the movable two only, arranged Text-fig. 64. as figured (fig. 3d). Coxe (text-fig. 64)—The second pair are distinctly widened out from the rather short interior margin and become rather UZ 3% b*. Palps only distinctly granular late- rally, with obtuse hairs; hand wider than deep. Second thoracic tergite WMG? 6 bo 6 ee oo 8 Ble Cha m@onesan) Sim, 6". Femur 2-2 as long as wide; palps in- distinctly granular and hairs obtuse. a”, Femur not attenuated and minutely granular ; tibia 2 as long as wide; chela is 2°7 as long as wide, and hand 1-4: as long as finger. Femur IV. 2:5 aslong-asdeep. . . . =. . . «| 284 Ch. subrotundatus Balz- 5°. Femur attenuated and almost smooth; tibia 2°2 as long as wide; chela 3 as long as wide and hand 1:1 as long as finger. Femur IV. 3:2 as long as deep. a. Single transverse groove . . . . 383. Ch. celerrimus, sp. n. ob. Two transverse grooves . . . . 32. Ch. brevifemoratus Balz. a. Subgroup of CHELIFER RUDIS Balz. Flagellum consists of four hairs. Sexual difference observed in galea. 30 short clavate hairs along hindmost margin of tergites and 8-10 in front. Proawimal joints of palps short and clumsy and chela long and slender. Only a single species known, to the description of which I refer. 13. CHELIFER RUDIS Balz, (Plate XXX. figs. 10 af.) 1890. Balzan, (10) pp. 423-424, tav. xiv. figs. 9-9 ce. 1891. Balzan, (11) p. 548. Male. Cephalothorax (cf. Pl. XXX. fig. 10 ¢).—At a distance from the front margin at least equal to their diameter, are two white ocular spots; they are of moderate size and as a whole not well separated from the surrounding parts of the integument, even showing traces of granulation; on account of these features they are designated ocular spots, though they are similar to real eyes in the anterior part, which, at least on the left, is somewhat arched and prominent like a true lens. The cephalothorax, which is much wider behind than it is long, tapers gradually towards the eyes, but becomes rather suddenly attenuated in front. Two rather indistinct transverse swéures, of which the posterior is straight, while the anterior appears slightly 254 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINZE curved backwards in the middie, are present; the second tergite is longitudinally divided similarly to the abdominal ones. The cephalothorax is distinctly granular all over, or, more correctly, is covered with a mosaic of small rounded plates at least on the head and first tergite. A large number (about 100) very short, distinctly clavate hairs are found, of which four somewhat longer are piaced along the front margin and about 20 along the hinder margin of the cephalothorax. Abdomen.—The abdomen is broad and flattened and has all the tergites longi- tudinally divided; the sclerites are minutely granular, and bear a very large number of rather short distinctly clavate hairs along the hindmost margin, viz. about 20 in the first segments, 30 in the median, and 24 in the tenth; in addition to these the median and posterior segments possess two lateral hairs in front of the row and three more median hairs in each half, the number varying somewhat, as it is not always easy to realise the position of a hair. No “ tactile” hairs seem to be present on the eleventh segment. The genital area is not very conspicuous; the anterior plate is granular, with a number of hairs, of almost equal length to those of the posterior plate, with many hairs in front. Antenne.—The terminal hair extends very much beyond the short and clumsy attenuated galea, which shows traces of a few terminal teeth (cf. Balzan, to. fig. 9c). The flagellum in the examined specimen consisted of only two long hairs, of which the anterior was marginally serrated, but it was probably imperfect. Palps (of. Pl. XXX. figs. 10 a-b, d—e).—The maville, which, as shown in the figure, are almost triangular, are almost smooth. The palps show a mosaic of small plates similar to the cephalothorax, which is, however, rather indistinct beneath and on the hand and is completely wanting in the fingers. Numerous very short hairs are found, which vary in form from distinctly clavate on the trochanter to very slightly clavate on the hand. The trochanter, which has a very short, well-defined stalk, is 1:3 as long as wide; its anterior outline beyond the stalk is almost circular, while ventrally the posterior surface is produced into a short somewhat conical protuberance, and the dorsal into a larger more rounded one, giving a bigibbose appearance. The femur has a short, very well-defined stalk, beyond which it is suddenly widened out and then distinctly attenuated towards the end. It is basally much wider than the trochanter and terminally only slightly so, and is 1°9 as long as wide; the anterior outline just beyond the stalk is suddenly produced into a somewhat triangular process, which merges gradually into the almost straight more distal portion; the posterior surface beyond the stalk is very distinctly and almost perpendicularly raised, then almost straight, and at last slightly convex. The tibia, which has a rather short, but exceedingly well-marked stalk, is scarcely longer, but distinctly wider than the femur, and is 1:8 as long as broad; the anterior outline is first, just beyond the stalk, strongly convex and then distally almost straight; posteriorly, beyond the basal elevation, which is scarcely marked off distally but basally is separated by a deep IN THE BRITISH AND COPENHAGEN MUSEUMS. 259 notch from the prominent condylus, it is at first almost straight and then slightly convex. The chela, which is 2°8 as long as wide, is scarcely 1-2 as broad as the trochanter is long; the hand, which is distinctly longer and 1°3 as broad as the tibia, is 1:7 as long as wide, 1:3 broader than deep, and 1:5 as long as the fingers, which gape slightly when closed. Both fingers bear a number of accessory teeth distally arranged, as shown in the figures (figs. 10 ab); the immovable finger has 9 posteriorly and 3 anteriorly, while the movable has 11 and 5 respectively. The number of “spots” anteriorly on both the movable and immovable fingers is five and the same is shown in 9? number is found on both posteriorly ; the arrangement of these ‘“ spots the figures. Coxe (cf. Pl. XXX. fig. 10¢).—The coxe are very characteristic, being very elongated. ‘The first pair are at least twice as long as broad and of almost equal breadth throughout; the two following pairs are triangular, having the interior side very short and being widened out towards the end. The fourth pati again are trape- zoidal, being distinctly broader than the third and scarcely enlarged towards the extremity ; the interior margin is much shorter than the posterior, slightly concave one and is well separated from it by the obtuse-angied postero-interior corner. Legs (cf. Pl. XXX. fig. 10 f).—Proximal joints with scale-shaped granules. The legs bear dorsally short or rather short, more or less distinctly clavate hairs; those of the ventral surface are longer and-slightly clavate or pointed; no “ tacéile” hairs observed. The legs are rather short and clumsy; the trochantin of the first pair of legs has a very wide articular cavity and is somewhat deeper than the tibial part; the tibia is 1:2 longer than the tarsus, which is only 3°2 as long as deep. The femur of the fourth pair of legs is 2:1 as long as deep, 1°1 as long as tibia, and scarcely 1*1 lower than tarsus is long. Colowr.—The palps and the cephalothorax are light brown; the abdominal sclerites are more yellowish brown. Measurements.—Cephalothorax 1°265 (1:495); abdomen 3°450 (2°300) mm. Palps: trochanter 0°690 (0°529); femur 1:173 (0-610); tibia 1-196 (0-667) ; hand 1°380 (0°815), depth 0-635; finger 0°920 mm. Leg I.: femur 0°745 (0-296), trochantin 0°175 (0°315) ; tibia 0°532 (0-198) ; tarsus 0-448 (0°140) mm. Leg IV.: femur 1-034 (0:02); tibia 0°911 (0-254); tarsus 0-532 (0-178) mm. Female. Cephalothoraz (Pl. XXX. fig. 10c).—Two real eyes are present of a somewhat peculiar shape, being rather low behind, but well curved and very prominent in front. Abdomen.—tThe longitudinal line is very broad in the median segments, decreasing towards the extremities, especially in front. 256 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA Antenne.——The galea extends somewhat beyond the terminal hair and seems to bear a few short teeth in the imperfect specimen examined (cf. Balzan, 1o. fig. 9¢, 2 ). The flagellum consists of four hairs, of which only the anterior has marginal teeth (of. Balzan, fig. 9 a). Pailps (Pl. XXX. fig. 10 d)—The palps are only slightly different from those of the male in the proportions between the joints. The fem is only 1-8 as long as wide, and much more attenuated, terminally being scarcely as wide as the trochanter ; the tibia is only 1:7 as long as wide and more strongly convex anteriorly; the chela is 3 times as long as wide; the hand is only 1:2 wider than the tibia and 1:4 longer than the fingers, which do not gape when closed. ‘The number of the accessory teeth is the same as in the male and the marginal teeth are much better developed. ‘The number of the “spots” is almost the same as in the male, but the arrangement shows a few interesting modifications, the most important being that the immovable finger has a single “spot” anteriorly behind the basal hair, and posteriorly five “spots” close together and a single much larger one, standing more apart, while the movable finger possesses two pairs of spots, one behind the other, and a single spot more apart. Cove (Pl. XXX. fig. 10.e).—The fourth pair are less slender and the posterior margin only a little longer than the interior, Legs (Pl. XXX. fig. 10f).—The legs perhaps less slender, especially the tarsus of the first pair of legs, which is only three times as long as deep. Colouwr.—The colour is distinctly darker: dark reddish-brown palps and dark brown cephalothorax and abdominal sclerites. Measurements.—Cephalothorax and abdomen as in the male. Palps: trochanter 0647 (0°506); femur 1:012 (0°575); tibia 1:035 (0:598); hand 1-226 (0°690), depth 0°540; finger 0°897 mm. Leg I.: femur 0°661 (0°266), trochantin 0-170 (0°289); tibia 0:494 (0-182) ; tarsus 0-414 (0-137) mm. Leg IV.: femur 0-950 (0°437); tibia 0-850 (0°220); tarsus 0°502 (0-170) mm. Material &c.—Of this very interesting species I have only examined two specimens, viz. a female collected by E. v. Benzon in Monte Rita (Brazil?) and a male collected by Dr. W. Sorensen in Argentine (Riacho dell’ Oro). Balzan’s specimens are from Paraguay, living under the bark of trees. Remarks.—The species described is certainly identical with Balzan’s, in spite of slightly different proportions of the joints of the palps, somewhat smaller size, greater number of hairs in front of the row and in the row on the median tergites; a rather curious difference is found in the flagellum, which has the three posterior hairs simple, not dentated as in Balzan’s specimens. i) or “I IN THE BRITISH AND COPENHAGEN MUSEUMS. 6. Subgroup of CHELIFER SUBROBUSTUS Balz. Flagellum consists of three hairs. Cephalothorax generally with some ocular spots and without transverse grooves or with a median one. The proximal joints of palps short and clumsy in contradistinction to long and slender chela, which has hand much longer than tibia. Tarsal tactile hair at least a half removed from the base. With the very scanty material at my disposal it is very difficult to mark this group sharply off from the following, but I am, nevertheless, practically convinced that it is quite a natural one. The shape of the cephalothorax, generally without transverse grooves, the contrast between the slender chela, with hand which is distinctly larger than the finger and much longer than the tibia, and the clumsy femur, together with the elongated coxee, give these species a peculiar appearance, somewhat similar to that of Chelifer rudis Balz.* The three previously known species referred to this group were originally described as belonging to Atemnus Can., but | think that the presence of accessory teeth on fingers of palps and the position of the tarsal “tactile” hair, well removed from base, make it necessary to place these species apart from Ch. nidificator Balz. &c. 14, CHELIFER SUBGRACILIS, sp.n. (Plate XXX. figs. 11 a-0.) Female. Cephalothorax.—Two fairly distinct ocular spots. The cephalothorax, which is somewhat longer than wide, has the membranous cucullus which covers the base of the antennz very well developed; only the anterior transverse groove, which is slightly curved backwards in the middle and fairly prominent, is found. ‘The integu- ment is minutely, but distinctly granular, except in the hindmost part of the second tergite ; the hairs are short and slightly clavate. Abdomen.—The abdomen, which is not very slender and somewhat depressed, has all the tergites except the hindmost part of the eleventh divided by a wide longitudinal band. The sclerites are indistinctly granular with scale-shaped granules; and the tergites bear from 12-16 generally rather short and slightly clavate hairs along the hindmost margin, in addition to two lateral ones distinctly, and two median ones, very indistinctly, in front of the row. A moderately big round spot, covered with numerous short hairs, is observed in front of the vulva. The dorso-ventral integument is beset with densely placed, parallel, undulated, longitudinal ridges. Antenne.—The rather slender galea, which possesses six short terminal branches, scarcely extends beyond the terminal hair. Palps (Pl. XXX. figs. 11 a—b).—The maaille are almost smooth in the middle, but granular laterally ; the trochanter, femur, and in a less degree the tibia, are distinctly * Marked differences from this species are, however, found in number of tergal hairs, &c. VOL. XVIII.—PART 111. No. 6.—October, 1908. 26 258 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA granular above and laterally, but almost smooth below, while the hand is partly quite smooth and partly indistinctly granular. The hairs of the three proximal joints are short and obtuse, while those of the hand ave more simple. ‘The trochanter, which is as long as wide, is anteriorly regularly and distinctly convex in the middle and terminates with a slightly marked convexity; posteriorly it is suddenly produced into a rounded protuberance and dorsally into a rather low, somewhat conical protuberance. The femur has a very short and well-defined stalk, beyond which it is distinctly attenuated towards the extremity. It is 2°2 as long as wide; both margins are almost straight beyond the stalk, from which the anterior outline is slightly raised, while the posterior is very distinctly and abruptly so. The ¢2é¢a, which has a short and well- defined stalk, is a little longer and wider than the femur, and 2:1 as long as wide ; the anterior outline is basally moderately convex and then very slightly concave; the posterior outline beyond the basal elevation, which is separated from the prominent condylus by a deep notch, but which is not limited distally, is first almost straight and then moderately convex. ‘The chela, which is about 3:5 as long as wide, is scarcely wider than the trochanter is long; the hand, which is at least 1-1 as long as the tibia, but scarcely 1-2 wider, is about twice as long as broad, almost 1:2 as wide as deep, and 1-4 as long as the fingers, which are much longer than hand is wide and do not gape when closed; the exterior outline of the hand just beyond the stalk is very abruptly produced and then becomes almost straight, while the interior is more gradually convex; the dorsal and ventral margins are almost straight. Both fingers bear anteriorly 2 accessory teeth near to the tip, while the number posteriorly is much greater, viz. about 8 in the immovable and 5 in the movable finger in the distal half. Anteriorly the immovable finger has only a single “spot” placed between the tactile hairs; posteriorly it has a single “spot” placed between the three basal tactile hairs which stand close together, and two just beyond the median tactile hair; while the movable finger has two “spots” at least, placed between the basal and the median tactile hairs, and a single one beneath and in front of the latter. Coxe .—The coxe show perhaps greatest similarity to those of Ch. rudis Balz., but differ by their less elongated shape; the second and the third pair are distinctly widened out towards the extremity and are of somewhat triangular appearance, as the inner margins, especially those of the second pair, are short. The fourth pair, which are somewhat trapezoidal and scarcely widened out towards the end, have the inner margin, which is about as long as the somewhat concave hinder margin, passing gradually into the latter. Legs.—The proximal joints are indistinctly granular. The hairs of the dorsal side are short and obtuse or dentated ; those of the ventral are longer, pointed, and more or less simple. ‘The tarsal “tactile” hair is two-thirds removed from the base and as long as this distance. The legs are fairly long and slender, and have the trochantin IN THE BRITISH AND COPENHAGEN MUSEUMS. 259 of the first pair much deeper than the tibial part, while the tibia is 1:3 longer than the tarsus, which is 3:3 as long as deep. ‘The femur of the fourth pair of legs is 2°4 as long as deep, 1:2 as long as the tibia, and 1:3 lower than the tarsus is long. Colour.—The palps are dark reddish brown; the cephalothorax, which has a some- what lighter hinder margin, is brown and so are the tergal sclerites. Measurements.—Cephalothorax 1-104 (0-989) ; abdomen 2°80 (1:73) mm. Palps: trochanter 0°506 (0°355); femur 0:897 (0°410); tibia 0-910 (0-437); hand 1:035 (0500), depth 0-414; fingers 0:713 mm. Leg I.: femur 0°623 (0-217), trochantin 0-114 (0:236); tibia 0-456 (0-137); tarsus 0°342 (0-105) mm. Leg IV.: femur 0°874 (0365) ; tibia 0-707 (0-198); tarsus 0-426 (0-129) mm. Material.—Of this species I have examined a single female from Brazil mounted together with Ch. imperator, sp. n., Kc. This species seems to differ from the three following by the presence of a fairly distinct median transverse groove; it differs besides from Ch. gracilis Ell. by the granular cephalothorax, the slightly clavate hairs, and by the hand, which is scarcely 1-1, not 1-5, as long as the tibia; from Ch. robustus Balz. by the less slender femur, being 2°2 instead of 2°7 as long as wide, and more slender chela, being 3:5 instead of 3 as long as wide; from Ch. subrobustus Balz. by much more clumsy tibia and more slender chela, which is 3:5 instead of scarcely 3 as long as wide and has only a few spots anteriorly on the movable finger. 15. CHELIFER GRACILIS Ell. 1902. Hilingsen, (15) pp. 148-149. Cephalothorax smooth, without transverse grooves, and with short simple hairs. Femur, which is distinctly attenuated, is anteriorly proximally slightly convex and then distinctly concave and posteriorly “réguliérement et médiocrement convexe ”’ ; ; the hand, which is some- ? tibia with strong stalk and “ presque régulicrement ovale’ what wider and 1-5 as long as the tibia, is 1:5 as long as the finger.—Hcuador. This species is distinguished from the others of this group by the very long hand. 16. CHELIFER ROoBUSTUS Balz. 1890. Balzan, (10) pp. 418-420, tav. xiii. figs. 6-6 b. 1891. Balzan, (11) p. 548. Cephalothorax minutely granular, without transverse grooves and with short dentated hairs. The femur, which scarcely seems to be attenuated, is 2°7 as long as wide; the tibia, which is 2°3 as long as wide, has outlines slightly convex; the hand, which is scarcely 1:1 as long as, but 1:4 as wide as, the tibia, is 1-6 as long as the finger; the chela is about 3 as long as wide.—Paraguay and Brazil. 2Na 260 MR. ©. J. WITH ON SOUTH-AMERICAN CHELIFERINZE 17. CHELIFER SUBROBUSTUS Balz. (Plate XXX. figs. 12 a-0.) 1891. Atemnus subrobustus Balzan, (11) pp. 512-518, pl. 9, figs. 3-3 a. Cephalothorax.—The cephalothorax, which is about as long as wide, has no transverse grooves, and is minutely, not very distinctly granular. Abdomen.—The hairs are moderately long, pointed or obtuse. Antenne.—The terminal hair extends slightly beyond the rather short and clumsy galea, which is exactly like Balzan’s figure 3 @ and with six short branches. Palps (Pl. XXX. figs. 12 a—b)—The mazille are smooth. ‘The palps are minutely granular, with the exception of the almost smooth chela; the dorsal and lateral surfaces of the trochanter and femur are fairly distinctly granular, while the lower surfaces and the tibia are less distinctly so. The hairs seem to be rather short and slightly obtuse. The trochanter, which is about 1°5 as long as wide, is anteriorly strongly convex and posteriorly produced into a rounded protuberance; the low dorsal tubercle is somewhat conical. The femur has a very short and well-defined stalk, beyond which it is somewhat attenuated. It is about 2:1 as long as wide; the anterior outline is slightly convex and then concave, while the posterior is abruptly raised and then almost straight. ‘The ¢ibia, which has a very short and well-defined stalk, is longer and wider than the femur and scarcely twice as long as broad ; anteriorly it is distinctly convex and then a trifle concave, while the posterior outline, beyond the well-marked basal elevation, which is separated by a deep notch from the prominent condylus, is first almost straight and then moderately convex or obtuse- angled. ‘The chela, which is about 3 as long as wide, is 1-1 broader than the trochanter is long; the hand, which is 1:1 as long as, but 1:3 wider than, the tibia, is almost 1°8 as long as wide, 1°3 broader than deep, and about 1:4 as long as the fingers, which are much (1°3) longer than the hand is wide and gape a trifle when closed ; the hand beyond the stalk is very abruptly raised posteriorly, but anteriorly, as well as dorsally and ventrally, more gradually convex, and then almost straight (cf. Balzan’s good figure 3 of palps, pl. 9). Anteriorly no accessory teeth were observed, but posteriorly about 15 in both fingers (figs. 12a—-b). The immovable finger bears anteriorly about eight “‘ spots” arranged as figured and posteriorly a similar number, while the movable finger has none. Coaxw.—The second and third pair are longer and more slender than those of Ch. subgracilis, sp. n., but less so than those of Ch. rudis Balz. ‘The fourth pair are somewhat triangular, as the inner somewhat shorter margin merges into the hinder and as they are widened out towards the extremity. Legs.—Only the femur of the fourth pair distinctly granular, the other joints smooth or indistinctly so. The hairs of the dorsal side of the legs are short, obtuse, and with a few teeth, while those of the ventral side are long, pointed, and more or less simple. A median tarsal “tactile” hair seems to be present. ‘The legs are rather short and clumsy; the trochantin of the first pair is deeper than the tibial part and the IN THE BRITISH AND COPENHAGEN MUSEUMS. 261 tibia is 1°3 as long as the tarsus, which is 3°5 as long as deep; the femur of the fourth pair of legs is scarcely 2°2 as long as deep, 1:2 as long as the tibia, and 1-1 lower than the tarsus is long. Colour.—The palps are reddish brown. Measurements.—Body about 4 mm. long without antenne. Palps: trochanter 0°483 (0°335); femur 0°805 (0°380); tibia 0°851 (0-437); hand 0-943 (0°540), depth 0-437 ; finger 0:690 mm. Leg I.: femur 0:600 (0°205), trochantin 0-091 (0-215); tibia 0:410 (0-126); tarsus 0-312 (0-091) mm. Leg IV.; femur 0°798 (0°372); tibia 0-661 (0:182); tarsus 0°410 (0°122) mm. Material.—Of this species I have examined a single damaged specimen ( ? ) from New Granada (Keyserling Coll.); the typical specimen is from Venezuela. Remarks.—That the described specimen belongs to Ch. subrobustus Balz., I regard as quite certain, as the differences are only trivial; it differs from Ch. robustus Balz. by the less slender femur and tibia and from Ch. gracilis Kil. by the granular cephalothorax and less slender hand. ¢. Subgroup of CHELIFER BICOLOR Balz. Flagellum consists of three hairs. Cephalothorax generally with ocular spots and two transverse grooves. The hand seldom much longer than the tibia. The tarsal “tactile” hair is more than one-fourth removed from base. This group contains species which in many respects differ greatly among themselves. It includes, in fact, all those which cannot be naturally placed in any of the three preceding groups. Nevertheless, several of the species show close relationship to each other: Ch. bicolor Balz., Ch. subrudis Balz., &c., are nearly related to each other, showing similarity in the structure of the male genital organs, in the shape of the palps, and the colour of the second thoracic tergite; Ch. patagonicus Tullgr. and Ch. ellingsenit, sp. n., are nearly related to each other as well as to Ch. elegans Balz., referred to the following subgroup. 18. CHELIFER ARMIGER, Balz. 1891. Trachychernes armatus Balzan, (11) pp. 527-528, pl. 10. fig. 18. Cephalothorax minutely granular, with two transverse grooves. Abdomen with long pinnate hairs along hindmost margin and six in front. The hairs of palps dentated or pinnate; the femur at least 5 as long as wide; tibia, which is about 2°5 as long as wide, has moderately convex outlines. Hand, which is scarcely longer than finger, of a most remarkable shape, being provided with three conical protuberances.—Amazons and Peru. 262 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERIN 19. CHELIFER ECHINATUS Ell. 1904. Ellingsen, (16) pp. 2-4. Cephalothorax strongly granulose, with two strong transverse grooves. Abdomen with fairly long strongly clavate hairs. Hairs of palps posteriorly short, strongly clavate, anteriorly very long, somewhat thickened; the femur is 2°5 as long as wide; the ¢ihra with long stalk and strongly convex outlines; fingers 1°5 as long as hand, which is 1°5 as wide as tibia.—Patagonia. 20. CHELIFER FOLIOSsuS Balz. 1890. Balzan, (10) pp. 427-428, tav. xiv. figs. 12-12 ¢e™. 1891. Balzan, (11) p. 548. Cephalothorax minutely granular, with two distinct grooves. Abdomen with granular sclerites and with clavate hairs. Hairs of palps long and strongly clavate; the femur is about 3 as long as wide; the ¢idza, which is 2:2 as long as wide, has lateral outlines moderately convex. The hand, which is distinctly longer and 1:2 wider than the tibia, is about 1-2 as long as fingers.—Paraguay. 21. CHELIFER GERMAINII Ballz. 1890. Balzan, (10) pp. 424-426, tav. xiv. figs. 10-10 ec. 1891. Balzan, (11) p. 548. 1905. Ellingsen, (17) pp. 1-3. Two distinct eyes. Cephalothorax minutely granular, with two grooves. Abdomen with granular sclerites and distinctly clavate hairs. Palps granular, with distinctly clavate hairs; the femur is 3:1 as long as wide; the ¢2bia, which is 2°4 as long as wide, has lateral outlines moderately convex. The hand, which is somewhat longer and 1:4 wider than tibia, is 1-2 as long as finger. ‘The second thoracic tergite is, according to Ellingsen, whitish with a dark spot in middle.—Matto Grosso and Ecuador. Balzan and Ellingsen refer this species to their subgenus Chelifer; on account of the structure of flagellum, galea, and colour, I think it is most naturally placed in this group in spite of real (?) eyes. Until the structure of fingers and legs has been studied, the question cannot be settled definitively. 22. CHELIFER ANTILLARUM, sp.n. (Plate XXX. figs. 13 a—.) Cephalothorax.—Vhe ocular spots are very indistinct. ‘The cephalothorax, which is somewhat longer than wide, has the anterior almost straight transverse dine fairly distinct, in contradistinction to the shghtly concave, almost obsolete, posterior one. The integument is distinctly granular, at least in front, and bears rather short, distinctly clavate hairs. IN THE BRITISH AND COPENHAGEN MUSEUMS. 263 Abdomen.—tThe long and slender abdomen has apparently all the tergites, except part of the eleventh, longitudinally divided. The sclerites have rather indistinct scale- shaped granulations, and the tergites bear along their hinder margin 8-10 more or less Jong and slender distinctly clavate hairs, in addition to two lateral hairs distinctly in front of row and two median indistinctly so. The eleventh sternite has a pair of rather short “tactile” hairs; similar ones probably found on corresponding tergite. Antenne.— The fairly slender galea, which has about six terminal teeth, extends slightly beyond the terminal hair. Palps (Pl. XXX. figs. 13 ¢-b).—The mazille are almost smooth. The palps are indistinctly granular and striated above, below smooth but not polished. The hairs, which vary much in length, are more or less strongly clavate, at least in the three proximal joints, and basally in the hand, with the exception of a few which are only dentated ; the clavate ones of the hand pass gradually into the pointed and more or less simple ones of the fingers. The trochanter, which is 1-4 as long as wide, is moderately convex anteriorly and slightly produced posteriorly ; dorsally it is prolonged into a fairly high, somewhat rounded protuberance. The femur has a short and well- defined stalk, beyond which it tapers towards the extremity. Itis 2°3 as long as wide ; anteriorly it is just beyond the base a trifle produced and then almost straight, and posteriorly the outline beyond the rather abrupt basal convexity is almost straight and then slightly convex. The tibia, which has a short not very well-defined stalk, is as long as and a little wider than the femur, and just twice as long as broad; the anterior outline beyond the stalk is rather abruptly and strongly convex and then in a slight degree concave, but the posterior beyond the very small condylus is almost straight and then moderately convex. ‘The chela, which is 2°6 as long as wide, is about 1:1 as broad as the trochanter is long; the hand, which is scarcely as long as the tibia, but 1:3 broader, is 1-5 as long as wide, almost 1-1 as wide as deep, and 1-3 as long as fingers, which are somewhat longer than the hand is wide and which do not gape when closed ; the lateral and the dorsal outlines are moderately convex, while the ventral is only slightly so. Anteriorly the immovable finger has a single accessory tooth distinctly beyond the middle; posteriorly both fingers possess a single distal tooth (fig. 13 0). Anteriorly the immovable finger has the median tactile hairs placed the one above the other, and somewhat behind the lower there is a large “spot”; posteriorly both fingers have three spots, arranged as shown in figure. Coxw.—The fourth pair are trapezoidal, only in a slight degree widened out towards the end, much longer than broad, and have the interior margin shorter than and fairly well separated from the posterior. Legs.—The proximal joints are smooth. The long hairs are dorsally in the fourth pair of legs strongly clavate and slightly curved, and in the first pair only slightly convex; the ventral hairs are pointed and more or less simple. The tarsal “ tactile”’ hair a little more than two-thirds removed from the base and as long as the distance to 264 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA the tip. The legs are short and clumsy, and the ¢rochantin of the first pair of legs is scarcely deeper than the femur proper, while the tibia isa trifle longer than the tarsus, which is 3:6 as long as deep; the femur of the fourth pair of legs is 2°5 as long as deep, 1:2 as long as the tibia, and 1-7 lower than tarsus long. Colour.—The palps are reddish brown; the cephalothorax and abdominal sclerites are yellowish brown. Measurements.—Cephalothorax 0°483 (0°437); abdomen 1:°380 (0-782) mm. Palps: trochanter 0°230 (0-161); femur 0414 (0°175); tibia 0-414 (0-207); hand 0-407 (0:265), depth 0:245 ; finger 0°299 mm. Leg I.: femur 0°274 (0°099), trochantin 0°061 (0-099); tibia 0°198 (0-072); tarsus 0-190 (0-055) mm. Leg IV.: femur 0°319 (0:129); tibia 0-266 (0°076) ; tarsus 0220 (0-065) mm. Material.—Of this species I have examined a single female, infected with black round bodies, perhaps spores of fungi: collected in St. Vincent by KE. Simon. Remarks.—This species, which is well characterised by its small size, the long, strongly clavate hairs of the palps, and the short distinctly convex tibia, appears to be similar to Ch. foliosus Balz., which, however, has the palps much more slender, the femur being, for instance, almost three times as long as wide and of almost double the size. 23, CHELIFER SUBRUDIS Balz. (Plate XXX. figs. 14 af.) 1891. Trachychernes subrudis Balzan, (11) pp. 321-322, pl. 10. figs. 13-13 0. Male. Cephalothorax.—Small distinct ocular spots are present. The cephalothorax, which is much longer than wide, has a median fairly distinct transverse groove slightly curved backwards in the middle, and shows a trace of a hindmost one. ‘The integument is minutely granular, and the hairs are short and obtuse. Abdomen.—tThe long slender abdomen is almost three times as long as wide and has all the tergites, with the exception of the first and the eleventh, longitudinally divided. The sclerites are distinctly granular and with from 10-14 fairly long and slightly clavate or obtuse hairs along the hindmost margins as well as generally four hairs in front of the row. The eleventh tergite has a pair of “tactile” hairs. The genital area has a long anterior and a short posterior plate, with a wide transverse split between. Antenne (cf. Pl. XXX. fig. 14 @).The terminal hair extends in a slight degree beyond the rather short and clumsy galea, which has six short distal branches. ‘The Jlagelium consists of three hairs, of which the anterior has several marginal teeth and the two behind a few only (fig. 14 a). IN THE BRITISH AND COPENHAGEN MUSEUMS. 265 Palps (of. Pl. XXX. figs. 14 6-c).—The maaille are smooth and glossy; the palps are minutely and not very distinctly granular above, below very indistinctly; the short hairs are obtuse or dentated. ‘The trochanter, which is 1:5 as long as wide, is anteriorly moderately convex and posteriorly very slightly produced; dorsally it is very slightly raised (in contradistinction to Balzan’s description, but similar to his fig. 13). The femur hasan extremely short stalk, beyond which it is distinctly widened out to a point somewhat behind the tip. It is twice as long as wide; the anterior outline is at first very slightly convex and then distally concave, while the posterior and especially the dorsal outlines (on account of the very short and clumsy shape it is very difficult sharply to distinguish between these two margins) are at first rather abruptly convex, then very slightly concave, passing again finally into a slight convexity. The tibia, which has a short well-marked stalk, is somewhat longer and broader than the femur, and scarcely twice as long as broad; it is much deeper than wide; the anterior outline, beyond the stalk, is abruptly convex, almost protuberant, and then almost straight ; the posterior outline, beyond the fairly prominent condylus and slightly marked basal elevation, is at first for a short distance almost straight and then slightly convex; the dorsal outline is slightly produced and the ventral is moderately convex. ‘The chela, which is 2°7 as long as wide, is scarcely 1:1 as broad as trochanter is long; the hand, which is about as long as but 1:5 wider than the tibia, is 15 as long as wide, but only 1-1 as long as deep, almost 1-4 deeper than wide, and 1:2 as long as fingers, which are 1-1 shorter than hand deep and do not gape when closed; the lateral outlines of the hand as well as the ventral are gradually and moderately convex, while the dorsal out- line just beyond the stalk is very high and very steep, even slightly inclining backwards and then slightly convex. The fingers bear anteriorly a few accessory teeth distally, and posteriorly a number from the middle to the tip; the number and arrangement of “ spots” could not be investigated on account of the black colour of the chela. Coxe.—The second and the third pair are slightly enlarged, the former being very slightly narrowed basally. The fourth pair are quadrangular in shape, slightly widened out, and with the inner margin somewhat shorter than the slightly concave hinder, not well limited from it. Legs (of. Pl. XXX. figs. 14 e-f).—The proximal joints are very indistinctly granular only. The fairly long hairs are dorsally dentated and obtuse (sometimes slightly clavate), but ventrally pointed and more or less clumsy; a short tarsal “tactile” (2) hair, scarcely three-fourths removed from base and a little longer than its distance tothe tip. The legs are very short and clumsy ; the tibial part of the femur of the first pair of legs is a little deeper than the trochantin, and the tibia is distinctly (1:1) longer than the tarsus, which is scarcely 3 as long as deep. The femur of the fourth pair of legs is 2:2 as long as deep, 1:3 as long as the tibia, and 1:2 lower than the tarsus is long. VOL. XVIH.—Part 11. No. 7.—October, 1908. 20 266 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA Colowr.—The palps are dark brown, and so is the cephalothorax, while the abdominal sclerites, placed within white skin, are light brown. Measurements.—Cephalothorax 0°851 (0°690); abdomen 3:100 (1-035) mm. Palps: trochanter 0-460 (0-322); femur 0°713 (0°368); tibia 0:736 (0°591); hand, 0-759 (0°506), depth 0-690; finger 0°621 mm. Leg I: femur 0-502 (0°220), trochantin 0-091 (0°213); tibia 0°557 (0°157); tarsus 327 (0:114) mm. Leg IV.: femur 0-737 (0°334); tibia 0-578 (0-198) ; tarsus 0:410 (0-144) mm. Female. Antenne (Pl. XXX. fig. 14 a).—The galea is a little longer and extends in a slight degree beyond the terminal hair. Paips (Pl. XXX. figs. 14 b-c).—The palps are only in a very slight degree different ; the ventral outline of the tibia is slightly and the dorsal fairly distinctly produced (fig. 14.¢); the hand is only 1-4 as long as wide and scarcely 1'1 as long as deep, scarcely 1:3 deeper than wide. Of “ spots” at least three are found at the base of the movable finger anteriorly. Cove (Pl. XXX. fig. 14d).—The fourth pair are distinctly different from the corresponding pair of the male, being as wide basally as terminally, and with the inner margin separated from the hinder by a distinct obtuse angle and much longer; the whole coxa trapezoidal. Legs (Pl. XXX. figs. 14 e-f).—The legs are perhaps a trifle less slender. Colour.—Body and palps a little lighter. Measurements.—Cephalothorax 0-989 (0°805); abdomen 4:25 (1:75) mm. Palps: trochanter 0-506 (0°345); femur 0-782 (0°391); tibia 0-305 (0-414); hand 0:828 (0-598), depth 0-782; finger 0:°690 mm. Leg I.: femur 0°532 (0236), trochantin 0:099 (0:228); tibia 0:380 (0°167); tarsus 0°350 (0°122) mm. Leg IV.: femur 0°810 (0368); tibia 0-616 (0-220); tarsus 0-433 (0-160) mm. Material &c.—Of this species I have examined a female collected m Caracas by Mr. Stroll and a male collected at Los Tejas (Caracas) by Dr. Meinert. Balzan’s specimen is from Venezuela. Remarks—The insignificant differences found between Balzan’s description and my specimens are scarcely worth recording, except that he writes “ Digiti . . . manu multo breviores.” This species may easily be distinguished from all other species known to me by the curious palps, especially the excessively deep chela. In these, as well in other respects, it shows a certain similarity to Ch. bicolor Balz., from which it, however, differs by colour and shape of coxe and legs. bo (or) ~I IN THE BRITISH AND COPENHAGEN MUSEUMS. 24, CHELIFER BICOLOR Balz. (Plate XXX. figs. 15 a-e.) 1891. Trachychernes bicolor Balzan, (11) pp. 524-525, pl. 10. figs. 16-16 a. Male. Cephalothorax.—Small, fairly distinct ocular spots are found. The cephalothorax, which is slightly longer than wide, has two fairly distinct, almost straight transverse grooves. ‘The integument is distinctly granular, except on the two white spots of the second tergite, and bears a number of short, slightly clavate hairs. Abdomen (Pl. XXX. fig. 15¢)—The abdomen is long and slender, and all its tergites are longitudinally divided; the dark spots of the tergites are granular, and from 10-12 moderately long, obtuse or slightly clavate hairs are placed along the hinder margin of the tergites, which in addition possess two lateral and two median hairs in front of the row, the last-mentioned not very distinctly so. The eleventh tergite has a pair of rather short “tactile” hairs. All the sternal sclerites are well developed, and all but the eleventh longitudinally divided by a broad band. The genital area had the anterior plate much longer than the posterior (fig. 16 c¢). Antenne.—The terminal hair extends distinctly beyond the very short, somewhat pointed galea, which has a tiny median as well as two terminal teeth. Palps (Pi. XXX. figs. 15 a—b).—The maaille are almost smooth in the middle, but laterally granular; the palps are minutely granular, except the fingers and the hand below; the hairs of the proximal joints are obtuse and dentated, rather short, and passing over into the moderately long, simple, or almost simple of the hand and fingers. The trochanter, which is 1:5 as long as wide, is anteriorly moderately convex, while posteriorly it is slightly produced beyond the stalk; dorsally it is preduced into a somewhat conical protuberance, only slightly pronounced. The femur has a short and well-defined stalk, beyond which it is distinctly widened out to a point somewhat beyond the middle and then becomes slightly attenuated. It is 2°3 as long as wide; the anterior outline is slightly convex and then concave, while the posterior outline is rather abruptly convex beyond the stalk, in the middle slightly concave, and then a trifle convex. The ¢idia, which has a rather short and fairly well-marked stalk, is a little longer and wider than the femur and 2-2 as long as wide ; anteriorly beyond the stalk it is convex and then slightly concave terminally, while the posterior outline beyond the condylus is almost straight, as the basal eleva- tion is very insignificant, and then slightly convex. The chela, which is about 2°4 as long as wide, is 1-4 broader than the trochanter is long; the hand, which is as long as but 1:5 wider than the tibia, is 1-4 as long as broad, but only 1-1 longer than deep, is 1-2 deeper than broad and about 1-4 longer than finger, which is a trifle longer than the hand is broad; the lateral outlines are moderately convex and so is the ventral, but the dorsal is almost semicircular. The number of accessory teeth seems to be fairly constant, viz. anteriorly about five on both fingers and posteriorly about ten. 202 268 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA The number of “spots” shows rather marked variations, at least anteriorly, where I have examined a good many specimens, but their arrangement is nevertheless fairly characteristic; the immovable finger possesses anteriorly, in the area between the four tactile hairs and near to it, from 15-20 “spots,” while the movable finger has two or three; the immovable finger has posteriorly about six “spots” basally and a single median one, and the movable finger five in the basal third. Cove (Pl. XXX. fig. 15¢).—The coxz of the second and third pair are both widened out towards the extremity; the second pair are rather suddenly restricted towards the base and consequently the interior margin is shorter than that of the third pair. The fourth pair, which are widened out towards the end, have the short interior margin gradually merging into the posterior, almost straight one. Legs——The proximal joints are indistinctly granular. Hairs of dorsal side slightly clavate or obtuse and rather short; those of ventral surface are longer, pointed, and almost simple or completely so; the tarsal “tactile” hair is at least three-fourths removed from the base and distinctly longer than its distance to tip. The trochantin of the first pair is scarcely as deep as the femur proper, and the tibia is almost 11 as long as the ¢arsus, which is 3°9 as long as deep. The femur of the fourth pair of legs is 2°2 as long as deep, 1-1 as long as the tibia, and 1°5 lower than the tarsus is long; the tibia of the fourth pair of legs is somewhat club-shaped, being distinctly convex terminally, not as usually in the middle. Colour.—Palps dark brown, almost black, with the stalks and extremities lighter ; the head, first thoracic tergite and a median spot on the second are blackish, while the lateral parts of the second tergite are white. ‘The abdominal tergites have on each side a black spot near to the middle, a brown one more laterally placed, and a brown band along the front margin of the tergite, connecting the two; the rest of the dorsal surface is white; the sternites are brown and divided by a longitudinal white band. Measurements.—Cephalothorax 0°759 (0°736); abdomen 1:886 (0°989) mm. Palps: trochanter 0°368 (0°240); femur 0°690 (0°299); tibia 0-713 (0-322); hand 0-713 (0°515), depth 0-635 ; finger 0°529 mm. Leg I.: femur 0-418 (0°167), trochantin 0-076 (0-160); tibia 0°319 (0-110); tarsus 0°296 (0-076) mm. Leg IV.: femur 0°532 (0:243) ; tibia 0°464 (0:°129); tarsus 0°357 (0:099) mm. Female. Abdomen (Pl. XXX. fig. 15 e).—The number of hairs in the hindmost row of the tergites is somewhat greater (from 12-14) and the median hair is placed more distinctly in front of the row. The abdominal sclerites are not well pronounced ventrally; a large number of hairs in an elongated area in front of the sexual area (fig. 16 ¢). Antenne.—The galea extends slightly beyond the terminal hair and has three pairs of distal short branches (cf. Balzan, fig. 16 a). IN THE BRITISH AND COPENHAGEN MUSEUMS. 269 Palps (Pl). XXX. fig. 15 d).—The proximal joints of the palps scarcely differ, but the chela is distinctly different, being 2°6 as long as wide; the hand, which is 1°5 as long as wide and 1:4 as long as deep, is only 1:1 deeper than broad and 1°5 longer than the fingers, which are somewhat longer than the hand is broad; the dorsal outline is distinctly convex, but far from being semicircular. The number of accessory teeth seems to be somewhat smaller; the difference in arrangement of spots is not greater than that between different specimens of the male. Coxe (Pl. XXX. fig. 15 ¢e).—The fourth pair are distinctly wider, with the inner margin only a little shorter than the hinder. Legs.—The legs seem to be somewhat longer and more slender, the femur of the fourth pair being, for instance, 2°6 as long as deep and 1-6 lower than tarsus is long; the tibia of the fourth pair has the usual shape and is four times as long as deep. Colour.—The palps and the cephalothorax are somewhat lighter and so are the brown areas on the abdominal tergites, which are somewhat larger; the median and lateral dark spots are not only fused in front but also behind, so that a dark area with a white spot in the middle is established. Measurements.—Cephalothorax 0:713 (0°736); abdomen 2°875 (1:265) mm. Palps: trochanter 0°368 (0-240); femur 0°718 (0-310); tibia 0°736 (0°335); hand 0-782 (0°506), depth 0°565; finger 0°529 mm. Leg I.: femur 0:456 (0°175), trochantin 0:084 (0°182); tibia 0°342 (0-118); tarsus 0:319 (0-080) mm Leg IV.: femur 0°631 (0°243) ; tibia 0547 (0:137); tarsus 0°395 (0°104) mm. Material &e.—Dy. Meinert collected from 20/vi—1/ix., 1891, numerous specimens (5 3,10 2) of this pretty species in Venezuela, where it seems to live under the bark of trees; Dr. Meinert at least mentions localities where they have been collected under bark, while it is not found amongst the numerous specimens which were captured by sifting. A single male carried a small worm in its large chela. Balzan’s specimens are from Petare and Caracas. Remarks.—This species is well characterised by its colour and deep chela. 25. CHELIFER ALBOMACUALTUS Balz. (Plate XXX. fig. 16a; text-fig. 67.) 1891. Trachychernes albomaculatus Balzan, (11) pp. 526-527, pl. 10. fig. 17. Male. Cephalothorax.—Ocular spots very indistinct. ‘The cephalothorax, which is distinctly wider behind than it is long, has two fairly distinct transverse grooves, of which the anterior is curved backwards and the posterior forwards in the middle. The integu- ment all over is strongly and coarsely granular with rounded or pointed granules, and provided with moderately long, distinctly clavate hairs. Abdomen.—The abdomen, which is only a little longer than wide, is flattened and of 270 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA a very shining appearance ; the tergites are longitudinally divided with the exception of the eleventh. All the tergites are strongly granular all over, and have along the posterior margin about 12 long, strongly clavate hairs in addition to two lateral ones, but they have scarcely any median hairs in front of the row. T he “ tactile” hairs of the eleventh tergite are missing in the specimens examined. Antenne.—The terminal hair extends in a very high degree beyond the short galea, which is rather clumsy, somewhat attenuated, with three short terminal teeth. Palps (Pl. XXX. fig. 16a; text-fig. 67).—The maaille are eranular, most distinctly laterally; the palps are distinctly granular above with the exception of the hand ; the ventral side is less distinctly granular and the fingers are smooth. The hairs of the femur and the trochanter, especially Text-fig. 67. anteriorly, are long, strongly clavate, somewhat curved, and directed forwards; the hairs of the two following joints increase in length and slenderness towards the fingers, in which the pointed hairs are completely simple or provided with one or several median teeth. The trochanter, which is 1:5 as long as wide, is strongly convex anteriorly, but posteriorly slightly produced ventrally ; dorsally it is produced into a fairly high rounded protuberance. ‘The femur has a fairly long and well-defined stalk, beyond which it is widened out to a point somewhat behind the tip and then is a little attenuated. It is 2°7 as long as wide; the anterior outline is very slightly convex and then concave, while the posterior is gradually and moderately convex beyond the stalk and then with a moderate concavity or even incision, distally passing into a slightly marked convexity, which is again terminally followed by a concavity. The tibia, which has a long and fairly well-marked stalk, is almost as long as the femur, but distinctly wider and 2-2 as long as wide; anteriorly, beyond the stalk, it is fairly distinctly convex and terminally slightly Ch. elbomaculatus concave, while the posterior outline beyond the fairly pronounced ieee) Go Cel ain condylus is almost straight and then slightly convex. The chela, ng ey ee which is about 2°5 as long as wide, is 1-3 wider than the trochanter is long; the hand, which is almost as long as but 1:4 wider than tibia, is almost 1:5 longer than both wide and deep, a trifle wider than deep, and 14 as long as fingers, anterior view. xX 44. which are a little longer than hand is broad; the lateral outlines as well as the ventral are moderately convex, while the dorsal one is very distinctly convex. ‘The fingers gape a trifle when closed, and each bears anteriorly two or three accessory teeth. Posteriorly each finger has about eight accessory teeth. ‘The number of the “spots” and the arrangement are a little variable in the specimen, but the arrangement in the main is the same as that in the figures; anteriorly the immovable finger has a single “spot” behind the basal hair, four beyond and behind the median hair, and three in IN THE BRITISH AND COPENHAGEN MUSEUMS. 271 the middle along the lower margin; while the movable finger has four in a longitudinal row; posterioriy, the movable finger has about four in the middle, and the immovable four basally and a single median one more apart. Coavw.—The coxe are very similar to those of Ch. bicolor Balz., but the second pair are less restricted basally, and the fourth pair are scarcely widened out towards the end. Legs.—The proximal joints are granular. The hairs of the dorsal side are long and distinctly clavate in the fourth pair of legs, moderately so in the first pair; the hairs of the lower surface are pointed and simple or almost so; a rather short tarsal “tactile” hair ~ removed from the base is found. The legs are fairly long and slender; the trochantin of the first pair is distinctly deeper than the femur proper and the tibia is a trifle shorter than the tarsus, which is 4°6 as long as deep. The femur of the fourth pair of legs, which is 2°9 as long as deep, is 1+1 as long as the tibia, and 2 lower than the tarsus long. Colowr.—The palps are dark brown with the stalk and the extremities lighter, and so is the cephalothorax, except two white spots on the second tergite laterally. The abdomen is dark brown with the exception of a broad longitudinal band, which is widest in the median segments, and consequently appears to be somewhat zigzag, and a spot on each half of a tergite; the colour of the band and the spots is white or yellowish. Measurements.—Cephalothorax 0°851 (0°920) ; abdomen 1°265 (1°150) mm. Palps: trochanter 0°368 (0°253); femur 0°745 (0°276); tibia 0°736 (0°335) ; hand 0:713 (0°483), depth 0°460; finger 0-506 mm. Leg I.: femur 0°464 (07148), trochantin 0°114 (0160); tibia 0°334 (0:106) ; tarsus 0:°350 (0:076) mm. Leg IV.: femur 0°593 (0°205); tibia 0°517 (0°129); tarsus 0-410 (0:095) mm. Female. Antenne.—The galea extends beyond the terminal hair, and possesses three pairs of distal short branches. Palps.—The proximal joints scarcely differ ; the hand is comparatively wider, namely 1°5 wider than the tibia; it is 1-6 longer than deep, and wider than the fingers are long. ‘The “spots” in the main are arranged as in the male, the most important difference being that the movable finger has only three “spots” anteriorly; in one specimen I found two spots behind the basal hair anteriorly. Legs.—The legs are perhaps a little more slender, the tarsus of the first pair of legs being, for instance, 4°9 as long as deep and the femur of the fourth pair 3-1. Measurements.—Cephalothorax 0°920 (19104); abdomen 1-840 (1:610) mm. Palps: trochanter 0°460 (0°310); femur 0:897 (0°330); tibia 0°897 (0°400); hand 0:8972(0°598), depth 0°552; finger 0-075 mm. 272 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERIN A Leg I.: femur 0°532 (0160), trochantin 0:122 (0-178); tibia 0°403 (0-114); tarsus 0:410 (0°084) mm. Leg IV.: femur 0°714 (0:228); tibia 0-631 (0°137); tarsus 0°471 (0°108) mm. Variation —A female from La Guayra differed from the described specimen by its smaller size, its less slender palps and legs, and by the colour of the abdomen, which had the longitudinal band of equal breadth throughout and the lateral spot better pronounced. Material &e.—Of this species I have examined two males, six females (of which one carried 2 bundle of eggs), and two young animals from New Granada (Keyserling Coll.). In addition to the above-mentioned badly-preserved specimens, I have had at my disposal a female collected by Dr. Meinert in the month of July on a road, La Guayra. Balzan’s specimens are from Venezuela. Remarks.—The described specimens differ in a slight degree from Balzan’s description by the colour and the proportion of the joints of the palps. It is easily distinguished from Ch. bicolor Balz. by the shape of the femur of the palps with its strongly clavate hairs, and by its colour. 26. CHELIFER CRASSIMANUS Balz. (Plate XXX. figs. 17 a-b; text-figs. 68 & 69.) 1890. Balzan, (10) pp. 421-428, tav. xiv. figs. 8-8 c. 1891. Trachychernes crassimanus Balzan, (11) p. 548. 1905. Ellingsen, (18) pp. 138-15. 1907. Tullgren, (23) pp. 71-72. Cephalothorax.—Fairly distinct ocular spots or perhaps real eyes, as they show traces of a curved lens. The cephalothorax, which is distinctly longer than wide, has two prominent, deep, almost straight, transverse grooves. ‘The integument of the head, the first thoracic tergite, and the black spot in the middle of the second are distinctly and coarsely granular, and provided with rather short and moderately clavate hairs, somewhat inclined. Abdomen (Pl. XXX. fig. 17 a).—The fairly long and slender abdomen has all the tergites, with the exception of the hindmost part of the eleventh, divided by a broad longitudinal band. All the tergites are distinctly granular, and the median bear from 12-16 fairly long and moderately clavate hairs (fig. 17 a) along the hindmost margin, in addition to two lateral and two median hairs most distinctly in front of the row. The eleventh tergite bears a pair of short “ tactile” hairs. Antenne.—The galea extends beyond the terminal hair, and bears about eight short distal branches. Palps (Pl. XXX. fig. 17 6; text-figs. 68 & 69).—The maaill@ are slightly granular in the middle and behind, but very distinctly so laterally and in front; the palps are distinctly and coarsely granular all over, with the exception of the almost smooth fingers. The palps bear many rather short or moderately long hairs, which are IN THE BRITISH AND COPENHAGEN MUSEUMS. 273 moderately clavate on the femur and slightly clavate or obtuse on the two following joints ; those of the fingers are not very long, pointed, and generally broken, with one or a few median teeth. The trochanter, which is 15 as long as broad, is anteriorly moderately convex, but posteriorly moderately produced and dorsally prolonged into a fairly deep and well-marked, somewhat conical eminence. The femur has a fairly Text-fig. 68. Text-fig. 69. VA AN /f i \ \ a Ch, crassimanus Balz., 2 .—Chela of right palp in anterior (a) and posterior (6) view. 44. long and well-marked stalk, beyond which it is widened out to a point somewhat behind the tip and then is in a slight degree attenuated. It is about 2:5 as long as wide; the anterior outline is very slightly convex and then concave, while the posterior is rather suddenly convex just beyond the stalk, almost straight in the middle and slightly convex distally. The t2bia, which has a moderately short and well- marked stalk, is a little longer, but distinctly wider than the femur, and 2:2 as long as wide ; anteriorly, beyond the stalk, it is distinctly convex and then slightly concave, while the posterior outline, beyond the small condylus and insignificant basal elevation, is almost straight and then moderately convex. The chela, which is 2°8 as long as wide, is only 1-1 broader than the trochanter is long; the hand, which is as long as, but 1:4 wider than the tibia, is 1:5 as long as wide, a trifle deeper than wide, and scarcely 1°2 as long as the fingers, which are much longer than hand is deep and do not gape when closed; the lateral outlines are distinctly convex, and so is the dorsal, the latter is basally more suddenly produced, while the ventral side is only slightly convex. ‘The fingers bear three or two accessory teeth anteriorly, and about ten posteriorly. ‘The immovable finger has three spots anteriorly behind the basal hairs and three in a longitudinal row beyond, and the movable finger has none, while VOL. XVIII.—ParT 111. No. 8.—Octoder, 1908. 2P 274 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERIN Z posteriorly the immovable finger has ten spots and the movable about twelve in the middle. Coave.—The coxee are almost the same as in Ch. bicolor Balz., the only difference being that the third pair have the anterior and inner corner produced into a short blunt process (cf. fig. 15 e). Legs.—The proximal joints are granular. The hairs of the dorsal surface are moderately long and slightly clavate or obtuse, those of the ventral side are generally pointed and simple, seldom dentated and obtuse. A tarsal “tactile” hair, a little longer than its distance to the tip and three-fourths removed from the base, is found. The legs are fairly long and slender; the trochantin of the first pair is much deeper than the femur proper; the tibia is of almost equal length with the ¢arsus, which is five times as long as deep. The femur of the fourth pair of legs is almost three times as long as deep, 1°1 as long as tibia, and about 2:2 lower than the tarsus is long. Colour.—The palps are dark brown and so is the cephalothorax, with the exception of the second tergite, which only has a small triangular black spot in the middle. The abdominal tergites are light brown with the longitudinal band and a spot on each side yellowish. Measurements.—Cephalothorax 0-920 (0:805) ; abdomen 2-070 (1:380) mm. Palps: trochanter 0°483 (0°322); femur 0°805 (0:345); tibia 0°828 (0°391); hand 0°828 (0°552), depth 0°565; finger 0-713. Leg I.: femur 0°578 (0°175), trochantin 0:122 (0:195); tibia 0-433 (0-114); tarsus 0426 (0:084) mm. Leg IV.: femur 0:707 (0:243) ; tibia 0°646 (0:140); tarsus 0°517 (0-095) mm. Variation.—The other specimen which was examined only differed by the number and arrangement of the “ spots”; anteriorly the immovable finger has either one or two “spots” behind the basal hairs, and two behind and one beyond the terminal hair; posteriorly the most important differences are that the immovable finger has only two “ spots” beyond the median hair, and that the movable seems only to possess five spots. Material.—Dr. Meinert obtained two females by sifting in the month of August, at La Moka. This species had previously been collected in Paraguay and Brazil (Matto Grosso). Remarks.—It is with the greatest hesitation that I have referred the above-described form to Ch. crassimanus Balz., as the hand is only 1:4 instead of 1:6 as wide as the tibia, and 1:2 instead of 1:4 longer than the fingers ; from Ellingsen’s description it differs besides by the maxille, which are not smooth, and by the tibia, which is not nearly smooth below. The only one of the mentioned characters which would justify the establishment of a new species is the longer fingers; but as the fingers of the male are only a little shorter than the hand, I think it better not to establish a new species. From the two preceding species, Ch. crasstmanus Balz. differs by shape of the femur and its hairs as well as by its colour. IN THE BRITISH AND COPENHAGEN MUSEUMS. 275 27. CHELIFER MEINERTH, sp.n. (Plate XXX. figs. 18 a-g.) Female. Cephalothorax.—Two rather indistinct ocular spots. The cephalothorax, which is as long as broad, has a peculiar shape, as it is almost of equal breadth behind the median groove, but in front distinctly attenuated towards the front margin; two fairly prominent and deep, almost straight transverse grooves are found. The integument is strongly granular, with prominent, either rounded or conically pointed granules everywhere except on the two white spots of the second thoracic tergite, and it is provided with a number of short, fairly distinctly clavate hairs. Abdomen (Pl. XXX. figs. 18 ab, g).—The abdomen is obovate in shape, and is, in this female, which is distended with eggs, almost as wide as long. The distance between the seventh and eighth sclerites, for instance, is, in this specimen, longer than the sclerites, and the distance between the two lateral parts only slightly narrower than each half. All the tergites except the last are longitudinally divided by a broad band; their sclerites are distinctly granular. The hairs, which anteriorly are short and distinctly clavate, but posteriorly longer and only moderately so (figs. 18 a—6), are found on the median tergites numbering 10-12 along the hindmost margin; the median pair of these hairs, which are placed outside the dark portion of the tergite, are set scarcely in front of the row in contradistinction to a lateral hair on each side. The eleventh tergite has a pair of rather short “ tactile” hairs. The hairy spot in front of the genital opening is triangular in shape (fig. 18 4). Antenne (Pl. XXX. fig. 18 ¢).—The galea, which has about six short distal branches, extends scarcely beyond the terminal hair (fig. 18 ¢). Palps (Pl. XXX. figs. 18 d-f)—The magille are smooth in the middle, but laterally, anteriorly, and even interiorly distinctly granular; the palps, with the exception of parts of fingers, are distinctly and coarsely granular. The palps bear a number of more or less short hairs, which on the trochanter and the femur are strongly clavate, while the two following joints bear more or less distinctly clavate or obtuse hairs; those of the fingers are generally short, pointed, and broken, with a single or a few median teeth. The trochanter, which is about 1°3 as long as wide, is moderately convex anteriorly, but posteriorly it is rather suddenly produced, dorsally only slightly so. The femur has a rather short well-defined stalk, beyond which it is of almost equal breadth throughout (broadest in the middle). It is 2:4 as long as wide; the anterior outline is a trifle convex and then concave, while the posterior, beyond the rather sudden basal convexity, is almost straight and then a trifle convex. The ¢ié¢a, which has a rather short and well-marked stalk, is as long as and a little wider than the femur, and twice as long as broad; anteriorly, beyond the stalk, it is suddenly and distinctly convex and then slightly concave, while the posterior outline, beyond the small condylus and insignificant basal elevation, is almost straight and then moderately convex. ‘The chela, which is 2°5 as long as wide, is 1°3 as broad as the 2P 2 276 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA trochanter is long; the hand, which is a little longer, but 1:5 wider than the tibia, is 1-4 as long as wide, a little deeper than wide and 1:3 as long as fingers, which are a little longer than hand deep, and do not gape when closed; the lateral outline, as well as the dorsal which is more suddenly produced, is distinctly convex, while the ventral is only slightly so. Both fingers bear anteriorly two distal accessory teeth (cf. fig. 18 e), while the immovable finger has posteriorly 9 teeth and the movable as many as 12, placed from near the base to almost the tip (cf. fig. 18f). The immovable finger possesses anteriorly three ‘‘ spots” behind the basal hair, six in the area between the four “tactile” hairs, and four in a longitudinal row near to the lower margin, while the movable finger has about eight in a longitudinal row from the base to near tip (fig. 18 e); posteriorly the immovable finger has about ten and the movable eight “ spots” arranged as shown in fig. 18 f. Cove (Pl. XXX. fig. 18 g).—The coxe are very similar to those of Ch. bicolor Balz. (ef. fig. 15 e), the only difference being that the third pair are basally restricted, and that these, as well as the second pair, show traces of an anterior blunt basal process: the fourth pair are besides more slender. Legs.—The proximal joints are only very indistinctly granular. ‘The hairs of the dorsal surface are moderately long, and distinctly clavate in the fourth pair of legs, but slightly clavate or obtuse in the first pair, while those of the ventral surface are pointed and more or less simple. The tarsal “tactile” hair, which is distinctly longer than the distance to the tip, is three-fourths removed from the base. The legs are fairly long and slender; the trochantin of the first pair is much deeper than the femur proper ; the tibia is a little shorter than the tarsus, which is 4°5 as long as deep. ‘The femur of the fourth pair of legs is 2°7 as long as deep, 1:1 longer than the tibia, and almost 1:9 lower than tarsus long. Colour.—The three proximal joints of the palps are yellowish brown with the basal and terminal part lighter; the chele are dark brown, and so is the cephalothorax with the exception of two white spots, which im the middle are separated by a median quadrangular dark spot and laterally bounded by a narrow dark stripe. The tergal sclerites of the abdomen are brownish, and the membranes between are yellowish. Measurements.—Cephalothorax 0°690 (0°690) ; abdomen 1:725 (1°495) mm. Palps: trochanter 0°335 (0°207); femur 0°598 (0°253); tibia 0°598 (0°285) ; hand 0°621 (0-437), depth 0-437; finger 0-495 mm. Leg I.: femur 0°395 (0°132), trochantin 0-084 (0°140); tibia 0°296 (0-091); tarsus 0:312 (0068) mm. Leg IV.: femur 0°532 (0:200); tibia 0°479 (0-114) ; tarsus 0:372 (0-080) mm. Material.—Of this pretty species I have examined a single female, collected by Dr. Meinert in the month of January on a road “ Old la Guayra (Caracas); I have named this species after Dr. Meinert, who collected such a number of these animals in Venezuela. IN THE BRITISH AND COPENHAGEN MUSEUMS, MT Remarks.—The described species is in most respects very similar to the preceding, but it seems, nevertheless, as far as can be judged with only a single specimen at one’s disposal, to be well characterised by the shorter and more distinctly clavate hairs especially of the femur, by the shorter finger, by the longitudinal row of spots anteriorly on the movable finger, and by the second thoracic tergite, which has two white spots. From Ch. bicolor Balz. and Ch. albomaculatus Balz. it differs in colour and in the shape of the femur and hairs. 28. CHELIFER SUBROTUNDATUS Balz. (Plate XXX. figs. 19 a-c.) 1891. Trachychernes subrotundatus Balzan, (11) pp. 522-528, pl. 10. figs. 14-l4a. Female. Cephalothorax.—Ocular spots almost obsolete. The cephalothorax, which is of almost equal length and width, is provided with two transverse grooves, of which the anterior is fairly prominent and slightly curved backwards in the middle, while the posterior is indistinct. The integument is distinctly and coarsely granular, except on the two large white spots of the second tergite, and is provided with short obtuse hairs. Abdomen.—The abdomen is fairly long and slender, and has all its tergites except the eleventh longitudinally divided by a broad band. The sclerites are granular, and the tergites bear 8-10 hairs along the hindmost margin as well as a lateral hair on each side, but there is no median hair in front of the row; the hairs are slender, fairly long or long, obtuse or even slightly ciavate. The eleventh tergite has a pair of rather short tactile hairs. ‘The abdomen has behind the first pair of sclerites and in front of the vulva an area with about 30 short hairs placed close together. Antenne.—The short galea, which has only about six distal teeth, scarcely extends beyond the terminal hair. Palps (Pl. XXX. figs. 19 a-c)—The maxille are smooth in the middle and granular laterally; the palps are fairly distinctly granular, but the lower side, the hand, and of course the fingers are less distinctly so, and are, indeed, more or less smooth. ‘The fairly long hairs are obtuse in the trochanter and femur, but pointed and dentated on the hand. ‘The trochanter, which is 1-5 as long as wide, is moderately convex anteriorly, but posteriorly is scarceiy produced below, and dorsally has a very low rounded protuberance. The femwr has a short and well-defined stalk, beyond which it is somewhat widened out and then in a slight degree attenuated. It is 2-2 as long as wide; the anterior outline is slightly convex and then concave, while the posterior, beyond the not very sudden basal convexity, is almost straight and then a trifle convex. The tibia, which has a short well-marked stalk, is almost as long as and somewhat broader than the femur, and twice as long as wide; anteriorly, beyond the stalk, it is strongly convex and then slightly concave, while the posterior outline beyond the prominent condylus and insignificant basal elevation is at first almost 278 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA straight and then moderately convex. The chela, which is about 2°7 as long as wide, is almost 1°3 as broad as the trochanter is long; the hand, which is somewhat longer and scarcely 1°5 wider, is 1:6 as long as wide, a little wider than deep, and 1:4 as long as the fingers, which}are distinctly longer than the hand is wide and do not gape when closed; the lateral outlines are moderately convex, while the dorsal outline beyond the stalk is at first abruptly convex and then slightly so. The immovable finger bears anteriorly, as well as posteriorly, three accessory teeth, while the movable finger has two only. The number of “ spots” is small, for the immovable finger has only four and three, anteriorly and posteriorly respectively, while the movable finger has only a single one on each side; they are arranged as shown in figures (figs. 19 b-c). Coace.—The second and the third pair are very similar to those of Ch. bicolor Balz. (cf. fig. 15 e), but seem to be comparatively shorter; the fourth pair have the com- paratively shorter inner margin merging into the hinder without any limitation, and are slightly widened out towards the extremity. Legs.—The proximal joints are only indistinctly granular. The fairly long hairs are dorsally slightly clavate, obtuse or dentated, and ventrally pointed and more or less simple; the tarsal “‘¢actile” hair is two-thirds removed from, and much longer than distance to, tip. The legs are moderately long and slender; the trochantin of the first pair is scarcely deeper than the femur proper, and its tibia scarcely as long as the tarsus, which is 3°9 as long as deep. The femur of the fourth pair is 2°5 as long as deep, almost 2°2 as long as the tibia, and 1-7 lower than the tarsus is long. Colour.—Palps are reddish brown ; cephalothorax dark brown, with the exception of the yellow second tergite, which has a quadrangular black spot in the middle. Tergal sclerites brown. Measurements.—Cephalothorax 0°621 (0°600) ; abdomen 1-610 (1-081) mm. Palps: trochanter 0-276 (0-184); femur 0°481 (0°227); tibia 0°483 (0-240); hand 0:552 (0°345), depth 0°335; finger 0:375 mm. Leg I.: femur 0°312 (0°122), trochantin 0-076 (0:122); tibia 0-228 (0-089); tarsus 0:236 (0-061) mm. Leg 1V.: femur 0:410 (0-167) ; tibia 0°350 (0-106); tarsus 0:281 (0-072) mm. Material —Of this species I, have examined a single female, collected by Dr. Meinert on aroad in the month of July, La Guayra (Caracas). Balzan’s specimens are also from Venezuela. Remarks.—That the animal described really belongs to Ch. subrotundatus is some- what doubtful, as the latter has the cephalothorax longer than wide, the femur of the palps only twice as long as wide, a prominent basal elevation on the tibia posteriorly, and the hand basally much more abruptly convex, especially anteriorly ; the two last- mentioned structures may perhaps be explained by the fact that the chela has not been figured in a true dorsal view, and that they have been somewhat exaggerated by the artist as well as the engraver. IN THE BRITISH AND COPENHAGEN MUSEUMS. 279 29. CHELIFER PLUMOSUS, sp.n. (Plate XXX. figs. 20 a-e.) Femate. Cephalothorax.—Large rather indistinct ocular spots. The cephalothorax, which is a little longer than wide, has two prominent, almost straight, transverse grooves. The integument of the head and first thoracic tergite as well as of the small black spot on the second tergite is strongly granular all over with rounded or pointed granules and is provided with short distinctly clavate hairs. Abdomen (Pl. XXX. fig. 20 e)—The fairly long and slender abdomen has all the tergites with the exception of the eleventh longitudinally divided by a broad band. The sclerites, which on the first tergite are represented by a mere transverse line, are distinctly granular. From 12-14 distinctly clavate hairs, short in front and fairly long behind, are placed along the hindmost margin of each tergite, which possesses besides a median hair, placed within the longitudinal band, slightly in front of the row, and a lateral one distinctly so placed, at least in the median segments. The eleventh tergite as well as the corresponding sternite bears a pair of “ tactile ” hairs. The hairy spot in front of the vulva is rather small and insignificant (fig. 20 e). Antenne.—The rather short galea extends a trifle beyond the terminal hair and has six rather short distal branches, decreasing towards the end. Palps (Pl. XXX. figs. 20 a-c)—The maville are smooth in the middle, laterally granular ; the palps are very distinctly granular with the exception of the more or less indistinctly granular lower surface and completely smooth fingers. ‘The short hairs of the three proximal joints are anteriorly more or less distinctly clavate, but posteriorly obtuse or dentated, sometimes similar to those of the hand; these, which are so numerous that the hand appears somewhat shaggy, are slightly curved with the end directed forwards, or, more correctly, the distal half makes an obtuse or even a right angle with the proximal; the basal portion is without any teeth, while the distal has about ten short teeth and consequently has a somewhat plumose appearance (fig. 20d). The trochanter, which is 1°5 as long as wide, is anteriorly moderately convex and posteriorly fairly distinctly produced, but very slightly so dorsally. ‘The femur has a short and distinct stalk, beyond which it is of almost equal breadth throughout, but for the slightly attenuated terminal part. It is 2-3 as long as wide ; the anterior outline, beyond the stalk, is very slightly convex and then concave, while the posterior is rather suddenly convex and then almost straight. The ¢ib¢a, which has a moderately long and well-marked stalk, is a little longer and broader than the femur and 2°2 as long as wide ; anteriorly it is, beyond the stalk, moderately convex and terminally a trifle concave, while the posterior outline, beyond the fairly big condylus, is at first almost straight and then moderately convex. The chela, which is 2°4 as long as wide, is 1-4 as broad as the trochanter is long; the hand, which is of almost equal length to and 1:6 wider than the tibia, is 1-4 as long as wide, scarcely 230 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA 1-1 as deep as wide, and 1-5 as long as the fingers, which are a trifle shorter than the hand is broad and scarcely gape when closed; the lateral outlines of the hand are distinctly convex, and so is the dorsal, which is, however, more suddenly raised beyond the stalk, while the ventral is only slightly convex. The fingers bear anteriorly only two accessory teeth each, while the number posteriorly is about ten. Anteriorly the immovable finger has basally between the tactile hairs about four “ spots” and farther down three much smaller ones, while the movable finger has seven in a longitudinal row from near the base to near the end; posteriorly the immovable finger has eight “spots” in two groups and the movable finger has seven spots (cf. figs. 20 b-c). Cove (Pl. XXX. fig. 20¢).—The third and the second pair are similar to those of Ch. bicolor Balz., the latter being distinctly narrower at the base; the fourth pair are rather peculiarly shaped, being slightly pointed interiorly and in the front, and scarcely widened out towards the extremity, with the inner margin, which is the shorter, distinctly merging into the hinder. Legs.—The proximal joints are very slightly granular. The rather short hairs are either moderately or slightly clavate, or obtuse, dorsally, in the fourth and first pair of legs respectively; those of the ventral side are pointed and more or less simple ; a tarsal ‘‘ tactile” hair, much longer than the distance to the tip, is two-thirds removed from the base. ‘The legs are not very slender and the trochantin of the first pair is deeper than the femur proper, while its tibia is a trifle longer than the tarsus, which is almost four times as long as deep. ‘The femur of the fourth pair is about 2°3 as long as deep, scarcely 1-1 as long as the tibia, and 1:6 lower than the tarsus is long. Colowr.—The palps are light brown with the chela much darker; the head, first tergite, and a tiny median spot of the second tergite are light brown, darker than the proximal joints of the palps; the dorsal side of the abdomen has brown sclerites and the skin between white; the white transverse band in the middle of the body encompasses almost the whole second thoracic tergite and the first abdominal with the exception of a transverse darker line. Measurements.—Cephalothorax 0°647 (0:621) ; abdomen 1:265 (0°828) mm. Palps: trochanter 0°276 (0°184); femur 0°506 (0°217); tibia 0-515 (0°230); hand 0-529 (0°375), depth 0:400; finger 0:368 mm. Leg I.: femur 0°327 (0°125), trochantin 0-070 (0:133); tibia 0:251 (0-084) ; tarsus 0°256 (0°061) mm. Leg IV.: femur 0-418 (0°182); tibia 0°388 (0°106); tarsus 0-288 (0:068) mm. Material.—Of this beautiful little species I have examined a single female sifted in the month of August by Dr. Meinert, at La Moka. Remarks.—The described species, which may be easily distinguished from all other species of this group by the singular feathery hairs of the hand, is probably nearly related to Ch. subrotundatus Balz. In the shape of the hairs of the hand it has some similarity to Ch. armiger Balz., but in other respects scarcely any. ? IN THE BRITISH AND COPENHAGEN MUSEUMS. 281 30. CHELIFER NITIDIMANUS Ell. (Plate XXX. fig. 21 a; text-fig. 70.) 1905. Ellingsen, (18) pp. 11-12. Female. Cephalothorar.—Large rather ‘indistinct ocular spots (cf. Ellingsen, 18.). The cephalothorax, which is distinctly longer than wide, has two distinct transverse stripes, which are slightly curved backwards in the middle. The integument is fairly distinctly granular with rather flat granules, with the exception of the almost smooth lateral white spots of the second tergite; the short hairs are slightly clavate. Abdomen.—tThe fairly long and slender abdomen has ail the tergites except the eleventh divided by a distinct longitudinal band. The sclerites are granular with scale-shaped granules, and bear along the hindmost margin of each pair six or eight fairly long slightly clavate hairs as well asa lateral one in front of the row (cf. Ellingsen) ; the eleventh tergite possesses a pair of “tactile” hairs. The hairy spot in front of the vulva is very insignificant. Antenne.—The fairly slender galea, which has at least six distal branches, extends slightly beyond the terminal hair. The flaged/wm consists of three hairs. Palps (Pl. XXX. fig. 21a; text-fig. 70).—The maxille are completely smooth; the trochanter and femur are distinctly granular and so is the tibia, but less distinctly ; all these joints are smooth below, and so is the hand all over except anteriorly at the base of the fingers (cf. Ellingsen). The long hairs are either slightly clavate, as on the inner surface of the femur, or obtuse and dentated. ‘The trochanter, which is about 1:6 as long as wide, is anteriorly moderately convex and posteriorly as well as dorsally slightly produced. The femur has a short and well-marked stalk, beyond which it is of almost equal breadth throughout, only very slightly attenuated towards the end. It is about 2°2 as long as wide; the anterior outline is very slightly convex and then a trifle concave, while the posterior is rather abruptly convex beyond the stalk, then straight and again ter- minally a trifle convex. The ¢idia, which has a moderately long and well-marked stalk, is a little longer and wider than the femur Oh. nitidimanusEll., and about twice as long as broad; the anterior outline is fairly @. Chela of right palp in anterior view. x 77. Text-fig. 70. distinctly convex with a slight terminal concavity, while the posterior, beyond the moderately big condylus, is basally straight and then regularly and moderately convex with the most raised part a little beyond the middle. The chela, which is three times as long as wide, is almost 1-2 as broad as the trochanter is long; the hand, which is scarcely longer than the tibia, but 1:4 as wide, is almost 1-7 longer than broad, about 1-1 as deep as wide, and about 1:2 as long as the fingers, which are distinctly longer than the hand is deep and do not gape when closed; the lateral as well as the dorsal outlines of the hand VoL. XVIUI.—Part I. No. 9.—October, 1908. 2Q 282 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA are moderately convex, while the ventral is almost straight. Both fingers possess anteriorly two accessory teeth near to the end, and posteriorly a few distally. The immovable finger is anteriorly adorned with about nine basal ‘‘ spots” densely crowded (cf. text-fig. 70) and has posteriorly three between the three basally placed tactile hairs ; while the movable has only two spots, placed basally. Coxw.—The second pair are only slightly narrower at the base; the fourth, trapezoidal pair, are much longer than wide, scarcely widened out towards the extremity, the inner margin is fairly well distinguished from the slightly concave and much longer hinder margin. Legs.—The proximal joints are slightly granular. The fairly long hairs are slightly clavate or obtuse dorsally ; those of the ventral side are pointed; the tarsal “ tactile ” hair, which is as long as its distance from the tip, is two-thirds removed from the base. The hairs are moderately long and slender, and the trochantin of the first pair of legs is scarcely deeper than the distal femoral part, while its tibia is as long as the tarsus, which is 3-7 as long as deep. ‘The femur of the fourth pair of legs is 2:2 as long as deep, 1-2 as long as the tibia, and 1:6 lower than the tarsus is long. __ Colowr.—The palps, the head, and first thoracic tergite as well as a semicircular spot of the second, and the tergal sclerites are yellowish brown, while the rest of the thoracic tergite and the interstitial parts are yellowish or white. Measurements.—Cephalothorax 0:529 (0-437); abdomen 1150 (0°690) mm. Palps: trochanter 0:228 (0:137); femur 0°395 (0°179); tibia 0°410 (0-190); hand 0-438 (0°266), depth 0°296; finger 0°360 mm. Leg I.: femur 0:251 (0°106), trochantin 0°053 (0:106); tibia 0-182 (0-072); tarsus 0°182 (0°049) mm. Leg IV.: femur 0°319 (0144); tibia 0-274 (0-084); tarsus 0:228 (0-061) mm. Material.—. Simon collected a single female in St. Vincent, whilst Ellingsen’s specimen was from Para. Remarks.—1 have referred this specimen to Ch. nitidimanus KHll., in spite of its small size, lighter colour, less strongly convex outlines of tibia and hand, and shorter fingers. By its smooth hand it is distinguished from at least most South-American species. 31. CHELIFER MICHAELSENI Sim. (Plate XXX. figs. 22 a—c.) 1902. E. Simon, (14) p. 44. 1904. E. Ellingsen, (16) pp. 4-5. Cephalothorax.—Ocular spots (?) rather indistinct; as their chitin is transparent, without granulation and slightly convex, they may perhaps be designated real eyes. The cephalothorax, which is somewhat longer than wide, has two transverse grooves, a wide and shallow anterior one almost straight and very prominent, and a posterior distinctly curved forwards in the middle and less prominent. The integument is IN THE BRITISH AND COPENHAGEN MUSEUMS. 283 distinctly granular on the head and first tergite, but less distinctly so on the second tergite and is provided with short obtuse hairs. Abdomen.—The fairly long slender abdomen has apparently all tergites except the first and eleventh longitudinally divided. The sclerites are indistinctly granular, and about ten rather short and slightly clavate hairs are placed along the hinder margin of each tergite in addition to two lateral ones in front of the row. No “tactile” hair was observed in the eleventh segment in this badly preserved specimen. The sexual area is very prominent, with the anterior genital plate, which has its prominent hinder margin concave, longer than the posterior; chitinous apparatus as well as the accessory glands, visible through the transparent skin, very similar to those of Ch. cimicoides ¥. Antenne.—tThe rather short galea scarcely extends beyond tne terminal hair and has about six short terminal branches. Palps (Pl. XXX. figs. 22 a-c).—The maaille are smooth in the middle, but granular laterally; the three proximal joints are distinctly granular on the anterior surface, but more or less indistinctly so everywhere else ; the hand seems to be granular only anteriorly and posteriorly, elsewhere being smooth. The fairly long or long hairs are obtuse and dentated or even slightly clavate. The trochanter, which is 1:5 as long as wide, is anteriorly moderately convex and posteriorly as well as dorsally slightly produced. The femur has a fairly short and well-defined stalk, beyond which it is of almost equal breadth throughout, but for the slightly attenuated terminal part. It is 2°4 as long as wide; anteriorly it is slightly convex and then concave, but posteriorly beyond the not very abrupt basal convexity almost straight and then terminally slightly convex. The ¢édia, which has a moderately long and well-marked stalk, is a little longer and wider than the femur, and 2°2 as long as wide ; the anterior outline is moderately convex, while the posterior, beyond the slightly marked condylus and obsolete basal elevation, is almost straight and then slightly convex. The chela, which is three times as long as wide, is 1°2 broader than the trochanter is long; the hand, which is about as long as, and 1:4 as wide as, the tibia, is 1-6 as long as wide, a trifle wider than deep, and scarcely 12 as long as fingers, which are much longer than the hand is wide and gape slightly when closed ; the lateral as well as the dorsal outlines are moderately convex, while the ventral is very slightly so. Anteriorly both fingers bear about six accessory teeth placed close together just beyond the middle, while the movable finger has four posteriorly and the immovable three ; marginal row of teeth complete in both fingers. Anteriorly eight spots, placed in two groups (fig. 22 6), are found on the immovable finger and only a single one basally on the movable finger, while, posteriorly, the immovable finger is adorned with seven spots and the movable with three only (fig. 22 ¢). Coww.—The coxe are yery similar to those of Ch. celerrimus, sp. un. (ef. below), 2Q2 284 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA but the second pair show traces of a basal narrowing, while the fourth pair are trapezoidal and of almost equal width throughout, with the inner margin much shorter and fairly well separated from the hinder one. Legs.—The proximal joints are almost smooth. Dorsally the fairiy long hairs are obtuse or dentated, ventrally pointed, and almost or completely simple. ‘The tarsus of the fourth pair of legs possesses a ‘‘ tactile” hair two-thirds removed from the base, and behind that, and removed one-third from the base, another large articular cavity ; in the specimen examined no terminal “ tactile” hair was observed at the tip of the femur and tibia as in Ch. celerrimus, sp. n. (ef. figs. 23 d—-e). The legs are long and slender, and the trochantin of the first pair is much deeper than the femur proper, while its ¢arsus, which is 4:6 as long as deep, is 1:1 longer than the tibia; the femur of the fourth pair of legs is 2°8 as long as deep, 1*2 as long as tibia, and 2 lower than tarsus is long. Colour.—The palps and the cephalothorax yellowish brown, the latter the lighter, with the chela more reddish brown. The abdomen is very pale, almost white. Measurements.—Cephalothorax 0°759 (0°690) ; abdomen 1°610 (0°897) mm. Palps: trochanter 0°345 (0°230); femur 0°647 (0°26) ; tibia 0°667 (0°299); hand 0°667 (0°414), depth 0°391; finger 0-575 mm. Leg I.: femur 0°418 (0:129), trochantin 0:091 (0°144); tibia 0°312 (0-091); tarsus 0°342 (0-073) mm. Leg IV.: femur 0°570 (0°205); tibia 0°464 (0°125) ; tarsus 0°410 (0-088) mm. Material &e.—Of this species I have examined a single male from Rio Janeiro, collected by the Danish ‘ Galathea’ Expedition; it has previously been recorded from Tierra del Fuego and Patagonia. Remarks.—The specimen examined differs from Ellingsen’s description (Simon’s description of the typical specimens was not at my disposal) in mere trivialities, viz., ocular spots fairly distinct instead of wanting, second transverse groove convex instead of straight, better-marked granulations, and comparatively longer femur, and con- sequently it is most naturally referred to the same species. ‘This species bears in the structure of the legs, &c., great similarity to Ch. celerrimus, sp. n., but is easily distinguished by the femur, which is scarcely attenuated. 32. CHELIFER BREVIFEMORATUS Balz. 1890. Balzan, (10) pp. 420-421, tav. xiv. figs. 7-7 c. 1891. Balzan, (11) p. 548. 1902. Ellingsen, (15) pp. 156-158. 1905. Ellingsen, (19) p. 324. This species is very nearly related to the following, to which I refer, as I have there enumerated the differences. It has been recorded from Paraguay and Ecuador. IN THE BRITISH AND COPENHAGEN MUSEUMS. 285 33. CHELIFER CELERRIMUS, sp. n. (Plate XXX. figs. 23 a-c, ¢; Plate XXXI. fig. 23 d.) Cephalothoraz.—No ocular spots were observed. The cephalothorax, which is distinctly longer than wide, has a median fairly prominent transverse groove, which is slightly curved backwards in the middle; there is no posterior groove. The integument is minutely not very distinctly granular and the short hairs are slightly obtuse. Abdomen.—The fairly long and slender abdomen has all the tergites, with the exception of the first and part of the eleventh, divided longitudinally. The tergites have smooth sclerites, and from 14-16 long, pointed, and terminally dentated hairs along the hindmost margin, in addition to four hairs in front of the row; these hairs increase in length from the base to the end of the abdomen, and from the middle towards the side in each tergite. ‘The eleventh tergite and the corresponding sternite have each two pairs of “ tactile” hairs. Antenne.—The galea, which extends somewhat beyond the terminal hair, has in the distal half two diverging branches, bearing respectively one and three branches (somewhat similar to Balzan’s figure 7 a (10)). Palps (Pl. XXX. figs. 23 a—-c)—The mavzillw are almost smooth beneath; the palps are minutely and very distinctly granulated, the inner surface having the granulations most pronounced, the lower surface being almost smooth. The hairs are fairly long or long, are provided with a few teeth, and are scarcely pointed on the proximal joints, but distinctly pointed and almost simple or completely so on the distal joints. The trochanter, which is 1-5 as long as wide, is very slightly convex anteriorly ; posteriorly it is slightly produced, and dorsally it is prolonged into a rounded, rather low protuberance. The femur has a short well-defined stalk, beyond which it tapers distinctly towards the end. It is about 2:2 as long as wide; the anterior outline is slightly convex and then concave, while the posterior is very abruptly convex beyond the stalk, straight in the middle, and then terminally a trifle convex. The ¢é¢a, which has a moderately long and well-marked stalk, is a little longer and wider than the femur, and is almost 2:2 as long as wide; the anterior outline is moderately convex with a scarcely marked terminal concavity, while the posterior, beyond the fairly big condylus and small but distinct basal elevation, is first almost straight and then slightly convex. The chela, which is three times as long as wide, is scarcely broader than the trochanter is long; the hand, which is about as long as the tibia, but 1-5 wider, is 1-6 as long as wide, scarcely deeper than wide, and 1-1 as long as the fingers, which are much longer than hand is deep and scarcely gape when closed; the lateral outlines are moderately convex, while the dorsal and ventral are only slightly so. The fingers each bear anteriorly a single accessory tooth, while posteriorly the movable finger has two and the immovable finger three teeth anteriorly. Hight “spots” are found in the basal half of the immovable finger, 286 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA while posteriorly the immovable finger has four and the movable only one “ spot” (of. figs. 23 b-c). Coxe.—The coxe are distinctly different from those of Chelifer bicolor Balz.; the second and the third pair are both widened out towards the end, especially the latter, as the former is much broader than this at the base and not suddenly narrower; the fourth pair are almost wider basally than distally, quadrangular, with the inner margin well separated from, and shorter than, the hinder one. Legs (Pls. XXX.-XXXI. figs. 23 d-e).—The proximal joints are completely smooth ; the fairly long or long hairs are pointed and completely simple or almost so; dorsally just behind the tip of the femur a very long and slender hair is observed in the fourth pair of legs; a similar hair, but even longer, is placed exactly in the same position in the tibia. The tarsus IV. has the “ tactile” hair placed nearer to the base than to the tip, and it is a little longer than the latter distance (cf. fig. 23 e); between the “tactile” hair and the base another wide articular cavity is observed (cf. above). The legs are very long and slender, and the trochantin of the first pair of legs is much deeper than the femur proper, while its tarsus, which is about five times as long as deep, is 1:1 longer than the tibia. The femur of the fourth pair of legs is 3-2 as long as deep, 1:1 as long as tibia, and 2°3 lower than the tarsus is long. Colouwr.—The palps, cephalothorax, and tergal sclerites are light brown, but of somewhat different shades. Measurements —Cephalothorax 0°552 (0°487) ; abdomen 1:541 (0°897) mm. Palps: trochanter 0266 (0-175); femur 0:410 (0-190); tibia 0°426 (0:198); hand 0:433 (0:266), depth 0°266 ; finger 0°385 mm. Leg I.: femur 0-296 (0084), trochantin 0:076 (0-099); tibia 0:228 (0-068); tarsus 0°251 (0:050) mm. Leg IV.: femur 0:380 (0-118) ; tibia 0-342 (0-084); tarsus 0°266 (0-061) mm. Material.—Of this species I have examined a single female with a bundle of eggs collected by Mr. H. H. Smith in the month of March, in a halfshady place in a mixture of earth and old manure, at Balthazar (Windward), Island of Grenada. Remarks.—This species is, as far as can be judged from Balzan’s very imperfect description, in most respects similar to Ch. brevifemoratus Balz. (1890, 10. pp. 420- 421, tay. xiv. figs. 7-7 c); but as it has no trace of ocular spots or second transverse groove, as it has smooth tergal sclerites with long and pointed hair instead of “setole .... corte, troncate all’ apice et dentate,” and as the palps are not granular all over, without “peli quasi clavati,” and the inner outline of the tibia is moderately, not strongly, convex, I prefer to establish a new species, especially taking into consideration that the above-described form is at least not identical with one from Ecuador with “de poils claviformes,” which Ellingsen (15. pp. 156-168) has referred to Ch. brevifemoratus Balz. It differs from Ch. michaelsenti Sim. by the attenuated femur of the palps. IN THE BRITISH AND COPENHAGEN MUSEUMS. 287 In this place I shall mention a specimen, probably a young animal, from Bogota (Keyserling Coll.), on account of an abnormality (?) in the number of the tactile hairs of the fingers. The immovable finger had only three “ tactile” hairs anteriorly and the movable only three posteriorly (cf. fig. 24a, Pl. XX XI). 34, CHELIFER PATAGONIcUS Tuller. 1900. Tullgren, (13) pp. 155-157, figs. 1-5. Cephalothorax “finely punctuated,” with two transverse grooves. Abdomen with slightly clavate hairs. alps finely reticularly granular, clothed with short simple toothed hairs; femur about 3 as long as wide and tibia 2°3; the hand is 1:4 as wide as the tibia and as long as the finger—Patagonia. 35. CHELIFER ELLINGSENIIL, sp. n. (Plate XXXI. figs. 25 a-e; text-fig. 71.) Cephalothoray.—The ocular spots are very indistinct, sometimes even absent. The cephalothorax is distinctly longer than wide, and is provided with two almost straight, broad, transverse stripes or grooves, of which the anterior is much more prominent than the posterior. The integument is minutely, but distinctly granular, with pointed granules at least laterally in front of the posterior transverse groove; the hairsare very short and obtuse. Abdomen.—The fairly long, slender abdomen seems to have all the tergites, with the exception of the first and eleventh, longitudinally divided. The sclerites appear almost smooth, and the tergites have along the hindmost margin from 10-14 short, obtuse o1 slightly clavate hairs in addition to four in front of the row on the median segments. Antenne.—The galea is rather short with about six distal branches or teeth, and extends in a slight degree beyond the terminal hair. Palps (Pl. XXXI. figs. 25 a-c).—The maaille are indistinctly granular in the middle, but more distinctly so laterally; the palps are distinctly granular laterally, especially anteriorly, almost smooth dorsally and completely so ventrally. The short or very short hairs are indistinctly pointed, with a median and a terminal tooth. The trochanter, which is 1-8 as long as wide, is, anteriorly beyond the stalk, moderately convex, and posteriorly is produced into a rather short round protuberance, which together with the fairly high and better-pronounced one of the dorsal surface gives the posterior surface a bigibbose appearance. The femur has a fairly long, not well- defined stalk, beyond which it is gradually and distinctly widened towards the middle and then terminally a little narrower again. It is three times as long as wide; the anterior outline, beyond the stalk, is first very slightly convex and then concave, while the posterior is smoothly and gradually convex basally, but almost» smooth in the middle, and beyond a trifle convex again. The t¢ibca, which has a very long and slender, but not well-defined stalk, is a little shorter, but wider than the femur, and is 288 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA 2°6 as long as wide; the anterior outline, beyond the stalk, 1s moderately convex and then terminally a trifle concave, while posteriorly beyond the condylus and insignificant basal elevation it is completely straight and then distally slightly convex. The chela, which is 3°3 as long as wide, is scarcely as broad as the trochanter is long; the hand, which is about as long as, but 1°4 wider than the tibia, is almost twice as long as broad, about 1-1 wider than deep, and 1:3 as long as the fingers, which are much longer than the hand is wide, and gape a trifle when closed; the lateral outlines are moderately convex, the anterior and the dorsal most distinctly so, and in a less degree the ventral. Both fingers bear anteriorly three accessory teeth, while posteriorly the immovable finger has four teeth and the movable finger five. The nwmber of “ spots” is extremely small, as the movable finger does not seem to possess any, in contra- distinction to the immovable finger, which anteriorly has two groups of three spots close together and Text-fig. 71. posteriorly only two close together (figs. 25 b-c). Coxe (text-fig. 71).—The second as well as the third pair are distinctly widened out towards the extremity ; at the base they become suddenly narrower, especially the second pair. The fourth pair are quadrangular, scarcely widened out towards the extremity, and have the ellingsenii, sp. n., 9. Fourth hinder, slightly concave margin well separated from pair of coxe. x 44. the inner and almost 1°5 longer. Legs (P|. XXXI. figs. 25 d-e)—The proximal joints are only indistinctly granular. The hairs of the dorsal side are rather short, or short and pointed, with a median and a terminal tooth; the hairs of the ventral side are longer and more or less completely simple. The tarsal ‘‘ tactile” hair was missing in the specimen described, but in another a very long one placed somewhat behind the middle and as long as the distance to the thigh was observed. The legs are extremely long and slender, and the trochantin of the first pair is much deeper than the femur proper, while the tibia is about 1:1 as long as the ¢arsus, which is 5°8 as long as deep; the femur of the fourth pair of legs is 3°3 as long as deep, scarcely 1:1 as long as the tibia, and at least twice lower than the tarsus is long. Colowr.—The palps are light or dark reddish brown, and the body is brown. Measurements.—Cephalothorax 1:0355 (0:920); abdomen 2°415 (1:245) mm. Palps: trochanter 0°552 (0°310); femur 1:012 (0°335); tibia 0-989 (0°375); hand 0°989 (0°529), depth 0°483 ; finger 0:736 mm. Leg I.: femur 0°676 (0-175), trochantin 0:144 (0-190); tibia 0°532 (0-114); tarsus 0°486 (0°084) mm. Leg IV.: femur 0-950 (0°288); tibia 0-874 (0-167); tarsus 0:585 (0-114) mm. Variation.—The four specimens examined (probably all females) were almost com- pletely alike in all respects, the most important differences being that the colour of IN THE BRITISH AND COPENHAGEN MUSEUMS. 289 the one female (from New Granada) was much darker and that the transverse stripes of another specimen (from Bogota) were more prominent; the accessory teeth and “ spots” differ only in mere trivialities. One startling variation, however, was observed, namely, in the position of the tarsal “tactile” hair, for in the two specimens from New Granada this was placed distinctly behind the middle, but in those from Bogota placed beyond the middle. Material—I have examined two females of this species from New Granada and two from Bogota ( @ 2), all four belonging to the British Museum (Keyserling Coll.). Remarks.—TVhis species seems to be very similar to Ch. patagonicus Tuller. (ef. 13. pp. 155-157), but as it has the hand 1-3 instead of as long as the fingers, and narrower instead of much wider than the trochanter is long, and as also minor differences were observed, I do not think it correct to refer it to this species. To Ch. elegans Balz. (11. pp. 520-521, pl. 10. fig. 12) it bears also great similarity, but may be distinguished by the shorter hairs of the femur, which is not attenuated and less convex anteriorly, as well as by the less strongly convex outlines of the tibia. d. Subgroup of CHELIFER ARGENTINUS Thor. Flagellum consists of three hairs; galea almost alike in 3 and 2. Cephalothorax with or without ocular spots, and with distinct median and indistinct posterior transverse stripe. Tibia of palps with two posterior “tactile” hairs in muddle; usual hairs stiff, pointed, and almost simple. <<‘ Tactile” hair of tarsus IV. 4-4 removed from the base ; in the middle of tibia a long, slender, completely simple hair. Ocular spots are generally well developed, but real eyes are never found. The cephalothorax, which is seldom distinctly granular, but generally almost smooth, has the anterior transverse stripe broad and fairly distinct, while the posterior is rather indistinct or even wanting. The abdomen, which has all its tergites except the first and last one longitudinally divided, has along their hindmost margin a variable number of more or less long, stiff, pointed, and almost simple hairs in addition to a median and a single or two lateral hairs on each side in front of the row, in the median -segments at least. The tenth tergite as well as the sternite has a pair of “ tactile” hairs, while the eleventh has two. As far as the more detailed arrangement of the hairs is concerned, I refer to the description of Ch. similis Balz. and Ch. argentinus Thor., as I think that it is the preservation of the specimens only which has made it impossible to find a similar arrangement in all the species of this group. The genitai area of the male seems to be alike in all the examined species and to be very characteristic. The anterior and posterior plates are of almost equal length, the former having a median concavity fitting into a pronounced convexity on the posterior plate; through the skin accessory glands exactly like those found in Ch. cémicoides F. as well as an undivided elongated testis can be seen in transparent specimens. The galea varies from a very simple form with few teeth to a large profusely branched VOL. XVIII.—PaRT U1. No. 10.—October, 1908. 2k 290 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERIN Z organ; but in neither case is any sexual difference observable (cf. text-figs. 72 & 73). The flagellum consists always of three hairs. The palps are generally polished, and granular only on the anterior surface. In addition to the usual fairly long, stiff, pointed, and not completely simple hairs, very long slender and completely simple hairs, Text-figs. 72 & 73. Ch. macrochelatus Tom. Right galea of Gg and 9. x 200. are present, viz., a single ventral hair on the trochanter in the middle; on the femur posteriorly a single hair in the middle and two terminal hairs of which the ventral is the longer; the tibia has two such hairs posteriorly in the middle, one more proximal nearer the ventral margin, and one more dorsal nearer the dorsal margin ; the hand has basally one of these hairs posteriorly and another ventrally, and beyond the middle a single ventral hair. The fingers bear in addition to the usual twelve tactile hairs on the immovable finger a single long and slender hair dorsally in the middle, and on the movable finger two beyond the middle. All the above-mentioned hairs were only observable in a few forms, but their fragility taken into consideration, I think we are right in regarding them as constant within this subgroup. ‘The fingers in some species gape considerably in the male, and are provided with a very varying number of accessory teeth, which are, however, always more numerous posteriorly than anteriorly. The number of spots is variable, but is generally great, and anteriorly behind the two basal “ tactile” hairs of the immovable finger there are at least a few. The cove are generally very similar to those figured of Ch. argentinus Thor., and are in some degree sexually different (text-figs. 74 & 75). The legs bear, in addition to the usual fairly long stiff hairs, very long and slender, most often completely simple hairs, a single ventral one on the trochanter and trochantin, terminally a single one, not always completely simple, dorsally on the femur, as well as a single one on the tibia, somewhat beyond or behind the middle. The tarsal “tactile” hair is {-} removed from the base and about as long as the tarsus. The legs are fairly long and slender; IN THE BRITISH AND COPENHAGEN MUSEUMS. 291 the femur is not very much longer than the tibia, and distinctly (1-4-2) as low as the femur is deep; the tibia of the first pair is generally 1:2 as long as the tarsus. In two species I have found curiously developed teeth on the claws, the presence of which does not necessarily show that the two species are closely allied (cf. below). s Sob Ch. argentinus Thor. Coxe of g and @. »x 37. 2 ANS | This subgroup is in the main identical with Balzan’s “Lamprochernes Tém.,” with the important exception that those species which he mentions from Asia and Africa must be expelled and be referred to the group of Ch. birmanicus Thor., which in external appearance is very similar to the Ch. argentinus 'Thor., but has scarcely any claim to nearer relationship. All the species which I have examined of this group go naturally together; of those which I have not examined I refer without hesitation Ch. ovatus Balz. to it, while the position of Ch. elegans Balz., Ch. venezuelanus Balz., and Ch. brasiliensis Dad. is somewhat doubtful. ‘The species of this group fall again into minor divisions: Ch. communis Balz. and Ch. intermedius Balz. stand closely together; Ch. nodulimanus Tom. and Ch. nitidus Ell., two species of very different appearance, are closely connected with each other by a number of intermediate forms (of. Ellingsen, 18. p. 6). As the species of this subgroup are very variable and often difficult to determine, a few words may perhaps be added about the specific characters. In the first place, the sex of a specimen must be taken into account both in description and in examination, as the shape of the palps generally varies considerably in male and female. The trochanteral protuberances are of great importance; the width and the depth of the hand should always be sharply distinguished from each other, and are with profit compared to each other as well as to the length of the hand and finger. The number of accessory teeth and “spots” are of value, but the variations (which are often considerable) should be taken into account. ‘The proportion between the joints of the legs is, of course, of importance. 2R2 292 MR: C. J. WITH ON SOUTH-AMERICAN CHELIFERINA Synopsis of Species. a, Claws without teeth. a’. Femur of the palps almost 3 as long as wide. a’. Trochanter strongly bigibbose. ‘Tibial stalk rather short. Galea with teeth only airs Seiten Pee ten Rs . Trochanter only strongly gibbose doraye Tibial stalk extremely long and slender. Galea branched 6’, Femur of the palps 2°5 or less as long as wide. a. Movable finger with “spots” anteriorly. Femur about 2 as long as wide. a‘. Hand 1°8 as long as finger, which is shorter than hand wide 6 : : . Hand 1:4-1°5 as ibe as singer: hid is at Nee a 1 as long as hand wide. a’, Tibia less than 2 as long as wide. Chela 2°7 as long as wide. Hand wider than deep. Ghat eine 6°. Tibia at least 2 as long as wide. Chela 3 as long as wide. Hand at least as deep as wide . 6°. Movable finger without “ spots” anteriorly. a°. Hand 1°8 as long as finger, which is shorter than hand is wide ; be, en 1°5 as long as ifnwer . Femur as well as tibia okay (3) or about 2 as long as wide . : Binllieaictiey We 6’, Femur as well as tibia more “pe 2 as long as wide. a’. Finger 1:4 as long as hand wide. Hand not more than 1:2 as long as finger. Tibia of ¢ scarcely produced. Small species. a. Chela 3:3 as long as wide. Hand 1°1 as wide as deep. Finger 1:5 as long as hand wide. Galea with teeth. Tibia 0°647 mm. b°. Chela 3 as long as wide. Hand less than 1° 1 as wide as deep. Finger 1°4 as long as hand wide. Galea branched. Tibia 0°815 mm. 6°. Finger not more than 1:3 as long as hand aides Larger species. Tibia of ¢ generally produced. a, Hand a trifle wider than deep (2) or scarcely deeper than wide (g). Granulation of cephalo- thorax wanting or indistinct. Tibia of ¢ slightly produced. a. Smaller species. Long hair on palps. Less branched galea. Tibia 0:828 mm. 6". Larger species. Moderately long hair on palps. Strongly branched galea. Tibia 1:02 mm. 36. Ch. elegans Balz. 37. Ch. venezuelanus Balz. 39. Ch. ovatus Balz. 41. Ch. intermedius Balz. 42. Ch. communis Balz. 39. Ch. ovatus Balz. 38. Ch. subovatus, sp. nu. 43. Ch. nitidus Ell. 44. Ch. similis Balz., 3. 44, Ch. similis Balz., 9°. 45. Ch. argentinus Thor. IN THE BRITISH AND COPENHAGEN MUSEUMS, 293 6°, Hand as deep as wide (9 ) or at least 1°3 as deep as wide(¢). Granulation of cephalothorax fairly distinct. Tibia distinctly produced in ¢. a, Finger 1:2 (¢) or 1:05 (¢) as long as hand deep. ‘Tibial protuberance not very long. Fingers gape moderately when closed. ‘Tibia NSS renym, | 5G . . . . 46. Ch. macrochelatus Tom. bz seHineger well ion tee eae deep. Tibial protuberance long and conical. Fingers gape considerably when closed. Tibia 1:91 mm. . 47. Ch. nodulimanus Tom., g. Ge SUM MIME RCMP aE eUN sesh ei corti a cag) tate ek re 40. Ch. brasiliensis Dad. b. Claws of the fourth pair of legs with big median teeth. a, Hand 1:4.as deep as wide and 1°3 as long as finger, which is a little shorter than hand deep. ‘Tibia with long slender stalk and with conical protuberance. Femur exteriorly shaggy, with long slender simple hairs. . . . . . . 48. Ch. cervus Balz., g. 6%, Hand 1°] as wide as deep and 1°5 as long as the Gants cian is 1:2 as long as hand deep. Tibia with short stalk and not conically produced. Femur with usual hairs posteriorly. . 49. Ch. depressimanus,sp.n., 2 . 36. CHELIFER ELEGANS Balz. 1891. Lamprochernes elegans Balzan, (11) pp. 520-521, pl. 10. figs. 12-12 @ Cephalothorax has a single transverse groove and is minutely granular. Galea with short branches. ‘The palps are granular. ‘The érochanter is strongly bigibbose ; the fenwr, which is 2-9 as long as wide, is posteriorly beyond stalk gradually convex. The tibia, which has a rather short and well-defined stalk, is 2-4 as long as wide, and is anteriorly moderately convex; posteriorly, beyond the prominent condylus which is separated by a deep incision from the insignificant basal elevation, it is first almost straight and then moderately convex. The hand is 1°4 as wide as tibia and 1:5 as long as the finger, which is 1:3 as long as hand is wide. As none of the usual slender hairs on the palps are observed in Balzan’s figure, and as the usual hairs are rather short, we cannot regard the position of this species within the subgroup of Ch. argentinus Thor. as secure; a new examination of the palps, but especially of the legs, will be necessary ere the question can be settled. Perhaps it is related to Ch. ellingsenit, sp. n., and Ch. patagonicus Tullgr.; regarding the differences I refer to the description of these species.— Venezuela. 37. CHELIFER VENEZUELANUS Balz. 1891. Lamprochernes venezuelanus Balzan, (11) p. 518, pl. 9. figs. 9-9 a. Cephalothorax minutely granular with distinct transverse stripe. Galea is well- branched. The trochanter is strongly produced dorsally; the femur, which is 3 as 294 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA long as wide, is rather gradually convex posteriorly; the ¢id¢a, which has a very long and slender stalk and is 2°3 as long as wide, is anteriorly slightly convex and posteriorly beyond the elongated basal elevation first concave and then moderately convex. Hand is 1-2 as wide as tibia, but 1-1 as long, and 1°4 as long as the finger, which is almost 1-4 as long as hand is wide. The position of this species cannot yet be regarded as secure.— Venezuela. 38. CHELIFER SUBOVATUS, sp.n. (Plate XXXI. figs. 26 a-f.) Male. Cephalothorax.—Ocular spots rather indistinct. ‘The cephalothorax, which is some- what longer than wide, has the anterior broad transverse stripe fairly prominent and straight, like the rather indistinct posterior one. ‘The integument is almost smooth or completely so, with rather long hairs. Abdomen.—The abdomen, which is long and slender, has all the tergites, with the exception of the first and hindmost part of the eleventh, longitudinally divided ; the sclerites are smooth and have along the hindmost margin from 12-14 not very long hairs, and in front of the row in the median segments two lateral and a single median hair, as far as could be observed; last segments with “ tactile” hairs (cf. below). Antenne.—tThe slender and fairly long galea, which extends distinctly beyond the terminal hair, has distally about six branches, decreasing towards the end. Palps (Pl. XX XI. figs. 26 a-d).—The palps are only minutely and not very distinctly granular posteriorly in the trochanter and anteriorly in the three following joints. The hairs are fairly long or long; the usual slender and completely simple hairs are partly missing in specimens examined. ‘The é¢rochanter, which is 1-6 as long as wide, is anteriorly slightly convex, and posteriorly produced into a short rounded pro- tuberance; dorsally it is prolonged into a deep, somewhat conical and rounded protuberance, and its depth is lower than that of the femur. The femur has a short and well-defined stalk, beyond which it is distinctly attenuated. It is scarcely twice as long as wide; the anterior outline is moderately convex and then slightly concave, while the posterior, beyond the stalk, is very abruptly convex and then almost straight ; dorsally it is even more suddenly raised and then very slightly convex to the end. The tibia, which has a rather short and well-defined stalk, is a little longer and wider than the femur, and it is 1-8 as long as wide; anteriorly it is very strongly convex, almost semicircular in the middle, and scarcely concave terminally; posteriorly, beyond the prominent condylus and low, but fairly distinct, basal elevation, it is almost straight and then terminally distinctly convex; the dorsal outline is only slightly convex, while the ventral is moderately so. The chela, which is 2°8 as long as wide, is about 1:1 broader than the trochanter is long; the hand, which is a little longer and 1°3 wider than the tibia, is 1°8 as long as wide, about 1:1 wider than deep, and almost 1-4 as long as the fingers, which are almost 1-2 longer than the hand is wide, and scarcely IN THE BRITISH AND COPENHAGEN MUSEUMS. 295 gape when closed; anteriorly the outline is only slightly convex, and so are the dorsal and ventral outlines, while the posterior is moderately so. Anteriorly the immovable finger has four terminal accessory teeth and posteriorly seven, while the movable finger has three and ten respectively. The immovable finger has anteriorly four or five “spots” behind the basal tactile hairs, and from two to four between and beyond (figs. 26 c-d), and posteriorly about four beyond the basal hairs, while the movable finger has no spots either anteriorly or posteriorly. Coxw.—The fourth pair are of a somewhat triangular appearance, as the angle between the inner and the hinder margin is very obtuse; the former of these margins is at least twice as long as the latter. Legs.—The hairs are fairly long, and in addition to the usual slender and completely simple hairs, one as long as the terminal hair of the femur is placed dorsally just behind the tip of the femur of the fourth pair of legs. The legs are rather short and clumsy ; the tibia of the first pair of legs is 1-2 as long as the tarsus, which is 4:4 as long as deep; the femur of the fourth pair of legs is 2°4 as long as deep, 1-1 as long as the tibia, and 1:4 lower than the tarsus is long. Colour.—The palps and the anterior portion of the cephalothorax are reddish brown, while the rest of the body is brown or yellowish. Measurements.—Cephalothorax 0:805 (0°690); abdomen 1°840 (1:035) mm. Palps: trochanter 0-437 (0°276); femur 0-713 (0°368); tibia 0-736 (0°405) ; hand 0782 (0°483), depth 0°457; finger 0°575 mm. Leg I.: femur 0-517 (0°175), trochantin 0-091 (0-182); tibia 0-410 (0-118); tarsus 0°354 (0:076) mm. Leg IV.: femur 0669 (0°278); tibia 0-608 (0-164); tarsus 0°395 (0-106) mm. Female. Abdomen.—The abdomen is much longer and more slender. Palps (Pl. XXXII. figs. 26 e-f).—The dorsal protuberance of the trochanter is some- what lower and more rounded. The femur, which is distinctly twice (2:1) as long as broad, appears much more slender and has the posterior as well as the dorsal outlines more evenly convex beyond the stalk. The édia, which is just twice as long as broad, appears much more slender and has the anterior as well as the posterior outlines less distinctly convex. ‘The chela is 2:7 as long as wide, and the hand as well as the fingers are shorter and more clumsy than in the male; hand is namely 1-4 broader than tibia, 16 longer than wide, distinctly 1-1 wider than deep, and 1:5 longer than the fingers, which are only 1-1 longer than the hand is wide. Cove.—TVhe fourth pair of trapezoidal appearance, as the angle between the interior and posterior margins is much better defined; the posterior margin is even shorter. Legs.—The legs are perhaps a little more slender, the tarsus of the first pair being, for instance, 4°5 as long as deep. 296 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERIN A Measurements.—Cephalothorax 0°805 (0°690); abdomen 2°530 (1:035) mm. 0:851 (0°510), depth 0-460; finger 0°575 mm. Leg I.: femur 0°547 (0°182), trochantin 0-091 (0°190) ; tibia 0-433 (0°118); tarsus 0:357 (0:079) mm. Leg IV.: femur 0°725 (0°296); tibia 0°623 (0-167); tarsus 0°418 (0-114) mm. Variation.—A single male from the same locality as the typical one differed by the very prominent basal elevation of the tibia. Material.—I have examined four specimens (2 ¢ and 2 2 ) from Argentine, 2 speci- mens {¢ and 91%), collected by Dr. W. Sérensen under bark at Riacho dell’ Oro, and two others (¢ and 12), sifted by Dr. Meinert, at La Moka, in the month of August. Kemarks.—This species seems to be nearly related to Ch. ovatus Balz., but differs by its much smaller size, by the tibia of the palps, which is less strongly convex, and by the hand, which is only about 1°5 not 1°8 as long as the fingers, which are distinctly longer, not shorter, than hand is wide. From Ch. argentinus Thor. and other species, in which the movable finger of the palps has no spots anteriorly, it differs by less slender palps. From Ch. communis Balz., as well as from Oh. intermedius Balz., it differs by the absence of spots on the movable finger; from the former of these also by the tibia, which has a short stalk and is only twice or less as long as broad, and from the latter by the tibia of the male, which is not produced anteriorly and dorsally. 39. CHELIFER OVATUS Balz. 1891. Lamprochernes ovatus Balzan, (11) p. 519, pl. 9. figs. 10-10 a. This large species may perhaps most easily be recognised by the following structures : the galea is moderately branched ; the femur of the palps is about 2°3 as long as wide ; the tibia, which has both lateral outlines equally and distinctly convex, is twice as long as wide, and 1:3 narrower than the hand, which is as wide as the finger is long and 1-8 longer.— Venezuela. 40, CHELIFER BRASILIENSIS Dad. 1889. Daday, (9) pp. 238-24, tab. 1. figs. 5 & 15. Cephalothorax smooth and polished, with single transverse stripe; the galea is well branched ; the fib¢a is rather suddenly convex anteriorly and shorter than hand, which is about as long as finger. A re-examination of the typical species is necessary, as the description is too imperfect to be of any value.—Brazil. IN THE BRITISH AND COPENHAGEN MUSEUMS. 297 4]. CHELIFER INTERMEDIUS Balz. (Plate XXXI. figs. 27 ad.) 1891. Lamprochernes intermedius Balzan, (11) pp. 515-516, pl. 9. figs. 6-6 a. 1902. Chelifer rotundatus Ellingsen, (15) pp. 151-152. 1905. Chelifer intermedius Kllingsen, (18) pp. 8-10. 1907. Chelifer intermedius Tullgren, (23) pp. 52-53, figs. 18 af. Mate. Cephalothoraz.—Ocular spots fairly distinct. The cephalothorax, which is some- what longer than wide, has the anterior, almost straight, transverse stripe broad and fairly prominent, in contradistinction to the rather indistinct posterior one. ‘The integument is smooth, somewhat polished in front of the median stripe, and provided with fairly long hairs. Abdomen.—The abdomen, which is fairly long and slender, has all the tergites, except the first and the last, longitudinally divided; the sclerites are almost smooth and have along their hindmost margin in the median segments from 14-16 long and moderately long hairs, in addition to two lateral and a single median one on each side in front of the row. Number of “tactile” hairs on last seginents is probably as in Ch. similis Balz., but could not be investigated in the specimen examined. Antenne.—The galea, which is fairly long and slender, extends distinctly beyond the terminal hair and has from 6-10 branches differing in length. Palps (Pl. XXX1. figs. 27 a-d).—Anterior surfaces of the joints minutely and not very distinctly granular. The hairs are long or even very long. The trochanter, which is about 1:5 as long as wide, is slightly convex anteriorly, and posteriorly is produced into a short basal protuberance; dorsally it is prolonged into a deep conical and somewhat rounded protuberance, and its depth is less than that of the femur. The femur has a short and well-defined stalk, beyond which it tapers towards the extremity. It is about 1-9 as long as wide; the anterior outline is slightly convex and then concave, while the posterior beyond the stalk is very abruptly convex and then almost straight; dorsally it is somewhat more regularly, though more strongly, convex beyond the stalk, and then slightly convex, most raised towards the middle, from which it gradually slopes towards the end. The tdbia, which has a rather short and well- defined stalk, is a little longer and wider than the femur, and is 1°8 as long as wide; anteriorly it is very strongly convex, almost semicircular in the middle and then terminally very slightly concave; posteriorly beyond the rather small condylus and low ill-defined basal elevation it is almost straight and then strongly convex ; the ventral margin is gradually and strongly convex in contradistinction to the dorsal, which is somewhat produced in the middle (figs. 27 6-d). The chela, which is about 2°6 as long as wide, is about 1:3 as broad as the trochanter is long; the hand, which is distinctly longer and 13 wider than the tibia, is 1:6 as long as wide, only a trifle wider than deep, and 1°5 as long as the fingers, which do not VOL. XVIlI.—PaRT 11. No. 11.—October, 1908. 28 298 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERIN A gape when closed and are scarcely 1°] as long as the hand is wide; the hand is moderately convex anteriorly and dorsally, but only slightly so posteriorly and ventrally. Anteriorly the fingers bear only a few accessory teeth distally (fig. 27 ¢) and posteriorly about ten. ‘The immovable finger has anteriorly from 11-18 “‘ spots” arranged some- what differently in the various specimens, in two groups (cf. tig. 27 ¢), and the movable finger has from three to five spots in the basal half; posteriorly the spots in the single specimen examined are placed as shown in figure (fig. 27 d). Coxe.—The fourth pair are somewhat triangular, as the hinder and inner margins, which are of about equal length, are not well distinguished from each other, the angle between being very obtuse. Legs.—Vhe hairs are moderately long. ‘The legs are rather short and clumsy: the tibia of the first pair of legs is 1-2 as long as the ¢arsus, which is four times as long as deep; the femur of the fourth pair of legs is 2°3 as long as deep, 1-1 longer than the tibia, and 1-6 lower than the tarsus is long. Colour.—The palps are dark reddish brown; the cephalothorax and the abdomen brownish. Measurements.—Cephalothorax 0-805 (0°735); abdomen 2°530 (0-966) mm. Palps: trochanter 0°395 (0°265); femur 0-690 (0°368); tibia 0-725 (0°391); hand 0-820 (0°529), depth 0-515; finger 0°555 mm. Leg I.: femur 0°471 (0°175), trochantin 0-084 (0-179); tibia 0°357 (0-114); tarsus 0-304 (0-076) mm. Leg IV.: femur 0°588 (0°258) ; tibia 0°532 (0°144); tarsus 0°365 (0:095) mm. Variation.—The specimen which Mr. Ellingsen kindly sent me for examination differed by the less slender abdomen and more powerful palps, but scarcely in any character of importance. Female. Cephalothorax &c.—Vhe second transverse stripe is obsolete ; the abdomen in the specimen examined is rather clumsy, being dilated with eggs. ‘The arrangement of the hairs is in the main as in Ch, similis Balz. ‘The galea of the antennze is almost exactly similar to Balzan’s figure 8 a, pl. 9. Palps—The trochanter has the dorsal protuberance somewhat lower and more rounded and is 0:7 as long as wide. The femur, which is comparatively more long and slender, viz. 2°1 as long as wide, is posteriorly as well as dorsally less abruptly convex. The tibia, which is about 1:9 as long as wide, is anteriorly less strongly convex, and has posteriorly the condylus as well as the basal elevation more prominent. The chela, which is 2°7 as long as wide, is only 1-1 broader than the trochanter is long ; the hand is 1:7 longer than deep and 1:1 wider than deep. Coaw.—The fourth pair are trapezoidal, at least as long as broad, with the inner margin well separated from the much shorter hinder one. IN THE BRITISH AND COPENHAGEN MUSEUMS, 299 Legs.—The femur of the fourth pair of legs is 2°5 as long as deep and 1°d lower than the tarsus long. Colour.—The palps are lighter brown. Measurements.—Cephalothorax 0:920 (0°805) ; abdomen 3°5 (2:00) mm. Palps: trochanter 0°552 (0322); femur 0°920 (0-457); tibia 0°943 (0-485); hand 1:055 (0°621), depth 0-565; finger 0690 mm. Leg I.: femur 0-684 (0:220), trochantin 0-114 (0°236); tibia 0°532 (0-157) ; tarsus 0-435 (0°106) mm. Leg IV.: femur 0°836 (0°334) ; tibia 0°760 (0-182); tarsus 0-509 (0-129) mm. Material.—oOf this species I have examined four specimens (3 ¢ and 1 ¢ 2) from Rio Janeiro and a male from Lagoa Santa, collected by Reinhardt; a male collected in the month of December at Los Trinchéras, and the female described collected on a road at La Guayra, Caracas, in the month of July by Dr. Meinert. ‘This species has previously been recorded from Venezuela, Ecuador, and Brazil. Remarks.—Vhat the described males, which I had the good luck to be able to compare with a specimen determined by Mr. Ellingsen, ought to be referred to Ch. intermedius Balz., I regard as certain in spite of their much smaller size ; they differ from Ellingsen’s description (18. p. 9) by the femur, which is scarcely 1-2 instead of 1-5 deeper than wide. I think that the female mentioned is most naturally referred to the same species in spite of the differences enumerated above. 42. CHELIFER coMMUNIS Balz. (Plate XXXI. figs. 28 a—g.) 1880. Balzan, (10) pp. 416-417, tav. xii. figs. 4—4 0. 1891. Balzan, (11) p. 548. 1902. ? Hllingsen, (15) pp. 167-168. 1905. ? Hllingsen, (18) p. 10. 1905. ? Ellingsen, (19) p. 324. Male. Cephalothorax.—Ocular spots not very distinct. The cephalothorax, which is distinctly longer than wide, has two broad, almost straight, transverse séripes, of which the anterior is the more prominent. The integument is smooth and the hairs are fairly long. Abdomen.—The abdomen, which is exceedingly long and slender, has all the tergites with the exception of the first and the last longitudinally divided ; the sclerites are smooth and have along their hindmost margin about 14 hairs in addition to generally two lateral and a single median one in front of the row on each side, at least in the median segments; the arrangement of * tactile’ hairs probably as in variation (cf. below). Antenne.—The rather short galea, which extends only slightly beyond the terminal hair, is bifurcate and provided with about six branches, arranged from base to the tip. 28 2 300 MR. OC. J. WITH ON SOUTH-AMERICAN CHELIFERINE Palps (Pl. XXXI. figs. 28 a-c).--The palps are smooth except on the anterior surfaces, which are minutely and not very distinctly granular. The hairs are fairly long or long. The trochanter, which is 1°6 as long as wide, is almost straight anteriorly, and posteriorly is produced into a short basal protuberance ; dorsally it is prolonged into a very deep, almost triangular, somewhat blunt process, and its depth is only a little less than that of the femur. The femur has a short and well-defined stalk, beyond which it is slightly attenuated towards the extremity. It is 1-9 as long as wide and distinctly (1:1) deeper than wide; the anterior outline beyond the stalk is first slightly convex and then a trifle concave, while the posterior, beyond the stalk, is very abruptly convex and then slightly so or almost straight ; dorsally it is more suddenly and more strongly convex beyond the stalk, and then slightly convex, most raised towards the middle, from which it slopes towards the end. ‘The ¢bia, which has a long, slender, fairly well-defined stalk, is longer but scarcely wider than the femur, and is distinctly twice as long as broad ; anteriorly it is distinctly convex and then terminally a trifle concave, while posteriorly, beyond the fairly prominent condylus and not very pronounced basal elevation, it is first almost straight and then distinctly convex ; the ventral margin is regularly and strongly convex, while the dorsal just beyond the stalk is rather suddenly, but not very distinctly, convex for a short distance and then almost straight, sloping towards the end. The chela, which is almost three times as long as wide, is about 1-1 as broad as the trochanter is long; the hand, which is as long as, but almost 1°3 as wide as the tibia, is almost 1:4 as long as wide, scarcely 1:1 deeper than wide, and is scarcely 1-3 as long as the fingers, which are 1:3 as long as the hand is wide, but 1-2 as long as deep, and which gape a trifle when closed; the anterior outline is moderately convex, but the posterior as well as the dorsal and ventral ones are only slightly convex. Anteriorly both fingers bear about three accessory teeth towards the end (¢f. fig. 28¢), while the number is about ten posteriorly on the immovable as well as on the movable finger. The immovable finger has anteriorly, behind and between the basal ‘“ tactile” hairs, about twelve spots, and beyond about five, while the movable finger has eight spots in the basal two-thirds (cf. fig. 28 f ) ; posteriorly the immovable finger has beyond the basal tactile hairs six and the movable finger four spots, Cove.—The fourth pair are more quadrangular than triangular and have the hinder end inner margins, which are well limited from each other, of almost equal length. Legs.—The hairs are moderately long. The legs are moderately long and slender ; the tibia of the first pair of legs is 1:1 as long as the ¢arsws, which is 4:7 as long as deep; the femur of the fourth pair of legs is 2°6 as long as deep, 1:2 as long as the tibia, and 1:6 lower than tarsus long. Coiour.—The palps and head are reddish brown, the rest is pale brown or yellowish. Measurements.—Cephalothorax 0-920 (0:805) ; abdomen 2:760 (0:989) mm. Palps: trochanter 0-460 (0°305) ; femur 0°805 (0-414); tibia 0°851 (0-414); hand 0°351 (0°515), depth 0°552 ; finger 0-667 mm. IN THE BRITISH AND COPENHAGEN MUSEUMS, 301 Leg I.: femur 0°585 (0-185), trochantin 0-106 (0°198) ; tibia 0°464 (0°129) ; tarsus 0-426 (0:091) mm. Leg IV.: femur 0-760 (0-296) ; tibia 0°646 (0°175) ; tarsus 0°486 (0:114) mm. Female. Cephalothorax.—The posterior stripe is obsolete or very indistinct. Palps (Pl. XXXI. figs. 28 d—q).-—The dorsal protuberance of the trochanter is much lower and more rounded (fig. 28 e). The femur, which is distinctly attenuated towards the extremity, is 2-1 as long as wide; it is less abruptly convex posteriorly as well as dorsally. The tb¢a has the stalk less slender than in the male, and the dorsal and ventral outlines are less curved. The chela is less slender, being scarcely three times as long as wide; the hand is only 1°6 as long as wide, only a trifle deeper than wide, and 1:4 as long as the fingers, which are 1:2 as long as the hand is wide. The variations in number and arrangement of accessory teeth and “ spots,’ which are observed when comparing the figures and previous description, are scarcely greater between males and females than between different specimens of the same sex (cf. figs. 28 f-q). Coxe.—The fourth pair are very similar to those of the male, the inner margin being only about twice as long as the hinder. Legs.—The tarsus of the first pair is only four times as long as deep. Measurements.—Cephalothorax 0-805 (0°690) ; abdomen 3°2 (1°61) mm. Palps: trochanter 0°391 (0°253); femur 0°690 (0°322) ; tibia 0°690 (0°335) ; hand 0:713 (0°437), depth 0°460; finger 0-529 mm. Leg I.: femur 0°441 (0-144), trochantin 0:076 (0°160) ; tibia 0°350 (0-:099) ; tarsus 0-319 (0-076) mm. Leg IV.: femur 0-638 (0°236); tibia 0-509 (0-140); tarsus 0-388 (0°091) mm. Variation.— 3 . Several males collected by Dr. Meinertin La Moka differ from the described one from Bahia by the more slender palps, viz. the femur, which is 2:1 as long as wide, is less abruptly convex posteriorly as well as dorsally ; the tibia is 2°2 as long as wide; the other joints also differ in a slight degree, but with regard to these I refer to the measurements. ‘The legs are in a remarkable degree more slender, the tarsus of the first pair being at least five times as long as deep, and the femur of the fourth pair being almost three times as long as deep and 1-7 lower than the tarsus islong. While almost all the hairs were missing in the described male, this well-preserved specimen has them well-developed and their number appeared to be greater, viz. from 14-15 along the hindmost margin in addition to two lateral and a single median one in front of the row on each side; a few of the hairs along the margin are sometimes placed a little in front of it; this is especially the case with a very long and slender hair, whicn in the hindmost segments becomes a “ tactile’ hair exactly as in Ch. similis Balz., to 302 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERIN A which I refer. The measurements are as follows :—Cephalothorax 0°805 (0°690) ; abdomen 1:040 (0°874) mm. Palps: trochanter 0-483 (0°299); femur 0°805 (0°368); tibia 0°$51 (0°391); hand 0-851 (0-529), depth 0°552; finger 0-647 mm. : Leg I.: femur 0°585 (0°182), trochantin 0-114 (0°196); tibia 0-456 (0:122); tarsus 0-418 (0-080) mm. Leg IV.: femur 0:798 (0-281); tibia 0-669 (0-160); tarsus 0:479 (0°110) mm. 2. A female from the same locality as the above-mentioned male differed from the described female in the following respects. The size is much greater; the shape of and the proportion between the joints of the palps are almost exactly alike, the only differences worth mentioning being, that the ¢id¢a@ is both longer and wider than the femur, and that the hand is scarcely deeper than wide. In contradistinction to the palps, the legs are much longer and more slender, the tarsus of the first pair of legs being almost five times as long as deep, and the femur of the fourth pair being 27 as long as deep and 1°7 lower than the tarsus is long. The measurements are the following :—Cephalcthorax 0-920 (0°805) ; abdomen 4°25 (1-75) mm. Palps: trochanter 0°506 (0°299); femur 0°828 (0°391) ; tibia 0°897 (0°425); hand 0:989 (0°575), depth 0°575; finger 0°690 mm. Leg I.: femur 0°623 (0°185), trochantin 0:114 (0°205); tibia 0-494 (0°135); tarsus 0-441 (0:090) mm. Leg 1V.: femur 0°836 (0°312); tibia 0-722 (0-175); tarsus 0-494 (0-114) mm. Material.—Of this species | have examined a number of specimens, of which only a few are completely similar to those described, most agreeing with the variation. The Danish ‘ Galathea’ Expedition brought home from Bahia the described male and female, the latter carrying a bundle of eggs; Professor Reinhardt collected in Rio Janeiro a single somewhat larger female, which had the tarsus ot the first pair of legs 4-4 as long as deep, and the vulva, wide open, showing two remarkable bodies within. In Venezuela Dr. Meinert from August to January sifted ten specimens (6 2 and4 ¢ ) of very different size, mentioned above under variation ; among these a female with a bundle of eggs in the month of December. From New Granada asingle female was examined (Keys. Coll.). Mr. H.H. Smith collected a single female with a bundle of eggs, in the Old Botanical Garden, Kingstown, Island of St. Vincent, in the month of October in rotten wood, 500 ft., and three females, carrying bundles of embryos, each bundle like a fruit of a Malva, in Bornwood Valley, near Kingstown, 800 ft., Island of St. Vincent, under the bark of rotten log. This species has previously been mentioned by Balzan from Paraguay and Brazil (Matto Grosso), and by Ellingsen from Argentine, Uruguay, Paraguay, and different localities in Brazil. IN THE BRITISH AND COPENHAGEN MUSEUMS. 303 Remarks.—The small female described from Bahia is completely like Balzan’s description as far as it goes, unimportant trivialities in the structure of the galea and less slender femur of palps excepted ; certainly it would not be natural to refer the described male to another species in spite of its larger size, as the differences in the palps and partly the legs are of only sexual importance, especially as they are from the same locality. ‘The difference in shape and in the length of the tarsus of the first pair of legs between the female from Bahia and those from La Moka &c. is certainly great ; but as specimens are found intermediate in size and partly in length of the tarsus I. (in ? from Rio Janeiro 4:4 as long as deep), and as the males from the different localities are even more similar in most respects, | do not think it right to establish a new species. A single badly preserved male, one of Ellingsen’s specimens, differed in several respects from the above ; the palps are much larger—tibia, for instance, 1:012 (0°483),—with the tibia almost produced anteriorly (cf. Pl. XX XI. fig. 28 h) and with the dorsal tubercle of the trochanter much lower and of another shape; the fingers gape moderately when closed and the movable one had no spots anteriorly. On full consideration I am convinced that the examined specimen must be regarded as belonging to another species ; but as only one badly preserved specimen was at my disposal, I do not wish to describe it as a new species, but refer to Ellingsen’s description (15. pp. 167-168), in which he writes “ C’est avec quelque doute que j’ai rapporté cette forme au Chelifer communis Balz.” His identification of the specimens from Brazil with Ch. communis Balz. was certainly justified, but I scarcely think that the “very small” specimens from Argentine (1905, 18. p. 10) are correctly referred to the same species as the large ones from Brazil. It will only be possible to settle the question of the geographical distribution of these two (three P) species and the correct limitation of Ch. communis Balz. by the examination of a very large amount of material. This species differs from Ch. intermedius Balz., chiefly by the tibia, which has a longer and more slender stalk and is at least twice as long as broad, and by the hand, which is wider than deep; the male differs besides by the somewhat triangularly shaped and pointed dorsal process of the trochanter. From the following species, as well as from Ch. subovatus, sp.n., it differs by less slender palps and the movable finger, which has ‘“‘ spots ” anteriorly. 43, CHELIFER NiTIDUS Ell. (Plate XX XI. figs. 29 a-0.) 1902. Ellingsen, (15) pp. 155-156. 1905. Ellingsen, (19) p. 324. Female. Cephalothorax &c.—The smooth cephalothorax, which is somewhat longer than wide, with fairly distinct ocular spots, has two almost straight transverse stripes, of which the 504 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA very broad anterior one is much the more prominent. | Abdomen has along the hind- most margin of each tergite about twelve hairs in addition to a single lateral and a median one in front of the row; the “ tactile” hairs are arranged as in Ch. similis Balz. Antenne.—The very slender galea, which extends distinctly beyond the terminal hair, being almost twice as long as the latter, has about five teeth decreasing in length from the base to the tip. Palps (P|. XX XI. figs. 29 a-b).—The palps are smooth and polished, with very slender hairs. The trochanter, which is 1-5 as long as wide, is slightly convex anteriorly and posteriorly slightly produced ; dorsally it is prolonged into a low rounded protuberance. The femur has a short, well-defined stalk, beyond which it is scarcely attenuated towards the end. It is 2°4 as long as wide; the anterior outline is very slightly convex and then concave, while the posterior, beyond the stalk, is not very abruptly convex, and then almost straight; the dorsal line, beyond the stalk, is moderately convex and then very slightly so. The t2dia, which is as long as and a little wider than the femur, has a short, not very well-defined stalk and is 2:2 as long as wide ; anteriorly it is moderately convex and then a trifle concave, while posteriorly, beyond the not very prominent condylus and almost obsolete basal elevation, it is almost straight and then slightly convex; the dorsal as well as the ventral outlines are slightly convex. The chela, which is 3:3 as long as wide, is about 1-1 broader than the trochanter is long; the hand, which is a little longer and 1:4 wider than the tibia, is 1-7 as long as wide, scarcely 1-1 as wide as deep, and 1-2 as long as the fingers, which are 1:5 longer than the hand is wide and do not gape when closed; the anterior outline is moderately convex, the posterior as well as the dorsal and ventral areas are only slightly so. Anteriorly the immovable finger has two accessory teeth, while the movable has a single one only; the former of the fingers has anteriorly about eleven “ spots,” arranged as shown in figure (cf. above). Coxe &e.—The fourth pair are almost trapezoidal, with the inner margin about 1-5 as long as the hinder, and with the obtuse angle between them not well defined. The legs appear to be long and slender, but could not be thoroughly investigated in the single specimen which was at my disposal. Colour.—Palps light reddish brown. Measurements.—Cephalothorax 0-736 (0°621); abdomen 2°30 (1°38) mm. Palps: trochanter 0°345 (0-230); femur 0°647 (0°265) ; tibia 0-647 (0-285); hand 0-713 (0-400), depth 0°368 ; finger 0-598 mm. Material.By the kindness of Mr. Eliingsen I have been able to examine one of his original specimens ( 2 ); the species has only been collected in Ecuador. Remarks.—For several details I refer to Mr. Ellingsen’s description. This species is well defined from the other species of this group by its very minute size, and slender palps with fingers only a little shorter than the hand. IN THE BRITISH AND COPENHAGEN MUSEUMS. 305 44, CHELIFER SIMILIS Balz. (Plate XXXL. figs. 30 a—f.) 1891. Lamprochernes similis Balzan, (11) p. 517, pl. 9. figs. 8-8 a. Male. Cephalothoraz.—Large distinct ocular spots found. The cephalothorax, which is distinctly longer than wide, has two transverse stripes, of which the anterior almost straight one is very broad and prominent, while the posterior is slightly curved forwards in the middle and not very distinct. The integument is smooth or almost smooth in the middle, but laterally minutely and not very distinctly granular, and appears polished in front of the anterior stripe ; the hairs are moderately long. Abdomen.—The abdomen, which is fairly long and slender, has all the tergites, with the exception of the first and the eleventh, longitudinally divided. The almost smooth tergites have along the hindmost margins from 14-18 hairs, which in the same segment vary from rather short, with a few distal teeth, to very long, almost simple ones, but as a whole increase in length towards the end of the abdomen; in addition to these hairs, a long median as well as a single lateral hair (seldom two) on each side are observed. One of the hairs in the row, which already in the fourth tergite is conspicuous by its length, increases in length and is gradually placed somewhat more in front of the marginal row, towards the tip of the abdomen, so that it becomes in the tenth tergite a completely simple “tactile” hair, placed distinctly in front of the row. The eleventh tergite bears two pairs of “tactile” hairs, like the one above mentioned exceedingly long and slender; the corresponding sternite bears two and the tenth only a single pair of these hairs. Antenne.—the fairly long and slender galea, which extends very much beyond the terminal hair, is bifurcate from the middle, each branch with from 2-6 teeth, almost exactly as figured by Balzan (pl. 9. fig. 8 a). Palps (Pl. XXXI. figs. 80 a—f).—The palps are almost smooth, only the trochanter being rather indistinctly granular posteriorly, the femur minutely, but fairly distinctly so anteriorly, and the tibia, as well as the hand at the base of finger, not very distinctly granular anteriorly ; the hairs are long or very long. The trochanter, which is about 15 as long as wide, is anteriorly slightly convex, and posteriorly produced into a rather low rounded protuberance; dorsally it is prolonged into a deep, conical, somewhat rounded protuberance; the trochanter is much deeper than wide and even deeper than the femur. The femur has a short and fairly well-defined stalk, beyond which it is scarcely attenuated towards the end. It is about 2°3 longer than wide; the anterior outline is very slightly convex and then concave, while the posterior, beyond the stalk, is not very abruptly convex and then almost straight ; dorsally it is rather suddenly convex beyond the stalk, then gradually convex, not gradually sloping to the tip, but terminally with a slightly marked concavity. The ¢ibia, which has a fairly long and not very well-defined stalk, is as long as, but somewhat wider than, the femur VOL. XVII.—PART 111. No. 12.—October, 1908. 2T 306 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA and is scarcely 2'1 as long as wide; anteriorly, beyond the stalk, it is moderately convex and then a trifle concave, while posteriorly, beyond the not very prominent condylus and long, not well-defined basal elevation, it is almost straight, and then terminally moderately convex; the ventral outline is moderately convex, while the dorsal is very slightly so, but with the antero-dorsal surface slightly produced (cf. figs. 30 d-e). The chela, which is distinctly three times as long as wide, is scarcely 1-1 as broad as the trochanter is long; the hand, which is a little longer than, but 1:3 as wide as the tibia, is 1:6 as long as wide, a trifle deeper than wide, and scarcely 1-2 longer than the fingers, which are 1-4 as long as the hand is wide and gape slightly when closed ; the anterior outline is moderately convex, the posterior and dorsal are slightly so, and the ventral is almost straight. ‘The immovable finger has anteriorly five accessory teeth and the movable one three, while both fingers posteriorly possess about 15 accessory teeth (cf. figs. 30 e-f). Anteriorly the immovable finger has from 5-16 “ spots” in two groups (fig. 30 e) and posteriorly five in a longitudinal row, valle the movable finger has only two posteriorly (fig. 30f). Cozw.—Cf. description of following species and text-fig. 74 (p. 291). Legs.—The hairs are long. The legs are long and slender; the tibia of the first pair of legs is about 1:1 as long as the tarsus, which is 4:6 as long as deep; the femur of the fourth pair of legs is 2-9 as long as deep, a trifle longer than the tibia, and only half as deep as the tarsus is long. Colour.—The palps are light reddish brown, the cephalothorax is somewhat darker brown, and the abdomen again paler brown. Measurements.—Cephalothorax 0°966 (0°805); abdomen 2:070 (1:035) mm. Palps: trochanter 0-483 (0-322); femur 0-805 (0°345); tibia 0°815 (0°391); hand 0°851 (0:515), depth 0°529 ; finger 0°736 mm. Leg I.: femur 0°555 (0°180), trochantin 0-091 (0-190); tibia 0-448 (0-114); tarsus 0°403 (0°088) mm. Leg IV.: femur 0:684 (0°240); tibia 0°669 (0°140); tarsus 0-471 (0°103) mm. Female. Palps.—The trochanter is dorsally prolonged into a rather low rounded protuberance. The femur, which is abont 2-4 as long as wide, is more gradually convex beyond the stallx posteriorly and so it is dorsally without terminal concavity. The ¢ibia, which is almost 2°3 as long as broad, has not the antero-dorsal surface produced. The hand, which is 1:4 as wide as the tibia, is 1:8 longer than wide, a trifle wider than deep, and {-3-1:4 longer than the fingers, which are distinctly 1:3 as long as the hand is wide, and which do not gape when closed. The number of accessory teeth is somewhat smaller than in the male, while the “ spots” do not provide any characters of interest. Coae.—The fourth pair are very similar to those of Ch. argentinus Thor. (cf. text- fig. 75), being somewhat trapezoidal with the inner margin almost twice as long as the hinder. IN THE BRITISH AND COPENHAGEN MUSEUMS. 307 Legs.—The legs are perhaps even longer, with the tarsus of the first pair 4°7 as long as deep, and the femur of the fourth pair scarcely half as deep as the tarsus is long. Measurements.— Cephalothorax 0°966 (0-805) ; abdomen 2°645 (1-265) mm. Palps: trochanter 0-460 (0-310); femur 0-805 (0°335); tibia 0°828 (0°368); hand 0°920 (0:515), depth 0°506; finger 0-667 mm. Leg I.: femur 0-570 (0-190), trochantin 0-091 (0-205); tibia 0°464 (0-114); tarsus 0-418 (0089) mm. Leg LV.: femur 0°715 (0°247); tibia 0-705 (0°144); tarsus 0°502 (0°110) mm. Abnormality.The hindmost segments of the described male show dorsally an abnormal structure, similar to that which I have described in Ch. javanus Thor. (31. p. 137, pl. viii. fig. 1@), though of a less complicated structure. The left exterior fourth of the tenth and the eleventh tergites are laterally completely fused without a trace of limitation between; the two tergites are consequently only distinguished from each other in the right three-fourths ; the arrangement of the hairs on the left is somewhat irregular, and the longitudinai line of the tenth tergite is directed obliquely backwards towards the left. Material.—Of this species Dr. Meinert sifted six specimens (2 g and 4 2, one with a bundle of eggs) in the month of August at La Moka, and a single female (without date) as well as a single male at Caracas in the month of July. Balzan’s specimens are from the Amazons. Remarks.—The identification of this species with Balzan’s species I regard as certain, in spite of its larger size and slightly different proportions of the joints of the palps. From Ch. nitidus Ell. it differs by larger size, shorter fingers, and bigibbose trochanter ; as regards the differences from Ch. argentinus I refer to that species. 45. CHELIFER ARGENTINUS Thor. (Plate XXXI. figs. 31 a—f; text-figs. 74-77.) 1877. Thorell, (3) p. 216. 1888. Ch. capreolus Balzan, (7) Pug. iii. pl. 2. figs. 1-4. 1890. Balzan, (10) pp. 414-415, tav. xiii. figs. 3-3 6. 1891. Lamprochernes argentinus Thor., Balzan, (11) pp. 516-517, pl. 9. fig. 7. 1905. Ellingsen, (17) p. 1. 1905. Ellingsen, (18) pp. 6-8. 1905. Ellingsen, (19) p. 324. 1907. Tullgren, (23) pp. 51-52, figs. 12 a-f. Male. Cephalothoraz.—Large ocular spots fairly distinct. The cephalothorax, which is distinctly longer than wide, has the anterior transverse stripe fairly distinct and slightly curved backwards in the middle, while the almost obsolete or rather indistinct 272 308 MR. CO. J. WITH.ON SOUTH-AMERICAN CHELIFERINA posterior one is slightly curved forwards. The integument is polished, and either completely smooth or very minutely granular, at least in front of the median transverse stripe; the hairs are rather short. Abdomen.—The long and slender abdomen has all the tergites with the exception of the first and the eleventh more or less distinctly divided by a fine longitudinal line. The tergites have their sclerites rather polished and completely smooth or slightly shagreened, and the median ones have along their hindmost margins at least 16 moderately long hairs in addition to a single median and generally two laterai ones in front of the row; the arrangement and number of the tactile hairs are as in the preceding species (¢f. above). Anienne.—The galea, which extends very much beyond the terminal hair, is moderately long and bifurcate, each branch with a number of ramifications. Palps (Pl. XXX1. figs. 31 a-b ; text-fig. 76).—The palps are smooth, or almost so, and polished, except the dorsal protuberance of the trochanter and at least the anterior surface of the femur and tibia, which are minutely granular; the hairs are generally long. The trochanter, which is 1:6 as long as wide, is anteriorly slightly convex and posteriorly produced into a basal, rounded, and fairly prominent tubercle; dorsally it is prolonged into a deep, somewhat conical, and obtusely-pointed protuberance, which, as shown in text-figure 76, has rather a peculiar shape and has the point directed upwards and towards the extremity ; the trochanter is much deeper than wide and is as deepas the femur. The femur has a fairly long and well-defined stalk, beyond which it is slightly attenuated towards the end. It is distinctly 2°2 as long as wide; the anterior outline is very slightly convex and then a trifle concave, while the posterior, beyond the stalk, is rather abruptly convex and then almost straight ; dorsally it is gradually but distinctly convex, most raised 4¢ sdisins elk fa crn just beyond the middle and then concave ; it is about as wide as deep. terior view. x 24. The tibia, which has a fairly long and well-defined stalk, is a little longer and wider than the femur and is 2°] as long as wide: anteriorly it is beyond the stalk distinctly convex, and posteriorly beyond the condylus and slightly marked basal elevation almost straight or even a trifle concave and then terminally distinctly convex; ventrally it is moderately convex and dorsally slightly produced, while the Text-fig. 76. Ch. argentinus Thor., go. Tro- chanter and femur antero-dorsal surface has generally a moderately deep and rounded tubercle, but some- times a deeper and more conical one, resembling that of Ch. machrochelatus Tom. The chela, which is 2:8 as long as wide, is almost 1:2 broader than the trochanter is long ; the hand, which is somewhat longer, but 1:4 wider, than the tibia, is about 1:6 as long as wide, scarcely 1-1 deeper than wide, and about 1:3 as long as the fingers, which are distinctly 1-2 longer than hand is wide, but scarcely 1:2 longer than hand is deep, and which gape slightly; the lateral outlines are moderately convex, while the IN THE BRITISH AND COPENHAGEN MUSEUMS. 309 dorsal and ventral ones are less distinctly so. The number of accessory teeth is some- what variable, but is generally large as shown in figs. 51 a—b, viz. eight and six anteriorly, and twelve and fifteen posteriorly, in the immovable and movable finger respectively. ‘The movable finger has anteriorly no “spots” and posteriorly only a few, while the number in the immovable finger varies, but is often more or less similar to that represented in the figures (figs. 31 a—0). Coxe (text-fig. 74, p. 291).—The fourth pair are somewhat triangular, distinctly widened out towards the extremity, and have the hinder margin much longer than the inner and not very well separated from it. Legs (Pl. XXXI. figs. 31 c-d).—The hairs are moderately long. The legs are long and slender ; the tibia of the first pair of legs is 1-2 as long as the tarsus, which is five times as long as deep; the femur of the fourth pair of legs is 2°9 as long as deep, scarcely longer than the tibia and about half as deep as the tarsus is long. Colowr.—The palps and the head are dark reddish brown; the tergal sclerites blackish brown. Measurements.—Cephalothorax 1:150 (0°989) ; abdomen 2°530 (1:265) mm. Palps: trochanter 0°598 (0°368); femur 0-966 (0°435); tibia 1-012 (0°495); hand 1-081 (0-690), depth 0°736 ; finger 0°851 mm. Leg I.: femur 0°722 (0:220), trochantin 0:144 (0°236); tibia 0-593 (0-140); tarsus 0:494 (0°099) mm. Leg IV.: femur 0°851 (0:296); tibia 0°851 (0-170); tarsus 0°585 (0-120) mm. Female. Palps (Pl. XXXI. figs. 31¢-f; text-fig. 77).—The dorsal protuberance of the trochanter is much lower and more rounded (fig. 77); the femur is less Heese 00 abruptly convex as well posteriorly as dorsally, and has no terminal con- cavity dorsally. The tidéa, which is 2-2 as long as wide, is anteriorly moderately convex, and posteriorly only slightly so terminally; the ventral outline is moderately convex and the dorsal is very slightly so, and has the antero-dorsal surface scarcely produced. ‘The chela, which is three times as long as wide, is only 1-1 broader than the NG Sas trochanter is long; the hand is almost 1°8 as long as wide, a trifle eo wider than deep, and 1-4 as long as the fingers, which are about 1:3 longer than the hand is either wide or deep, and scarcely gape when Ch. argentinus closed; the lateral, as well as the dorsal and ventral margins of the Thor. 2. Tro- hand, are less distinctly convex. The numbers of accessory teeth and Peau oe of spots are generally somewhat smaller (figs. 31 e—f). had. ‘arto Thoms 32Oh. Coxe (text-fig. 75, p. 291).—The fourth pair have a quite different shape, being much broader and scarcely widened out towards the extremity, with the inner margin much longer than the hinder and of a somewhat trapezoidal appearance. 310 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA Legs.—Vhe legs differ only in mere trivialities from those of the male. Measurements.—Cephalothorax 1:150 (1°081) ; abdomen 3°52 (1°38) mm. Palps: trochanter 0°575 (0°345); femur 0-966 (0°437); tibia 1:035 (0°460) ; hand 1:127 (0°647), depth 0°635; finger 0°828 mm. Leg I.: femur 0-714 (0°205), trochantin 0°137 (0°225); tibia 0°593 (0°137); tarsus 0479 (0°095) mm. Leg I1V.: femur 0°851 (0-296); tibia 0°358 (0°175); tarsus 0°593 (0:122) mm. Variation.—The numerous specimens of this species which I have examined differ in several respects, viz. size, granulation of the cephalothorax, shape of joints of palps, especially shape of the tibia of the male, &c., but on the whole are as described. Material.—Of this species I have examined a great number of specimens (about 150), most of which were sifted by Dr. Meinert in Venezuela from June to September ; several females collected in August carried their eggs fastened to the abdomen. Other specimens were collected in Mexico, Brazil, and Argentine (Dr. W. Sdrensen ; under bark, Riacho dell’ Oro). It has previously been recorded from Venezuela, Ecuador, Brazil, Paraguay, and Argentine. Parasitism.—A male from La Moka had two minute mites, probably larval stages of a Tyroglyplus, fastened to the trochantin of the first pair and to the articular membrane between the femur and tibia of the fourth pair respectively. Remarks.—Regarding the differences between this species and Ch. macrochelatus'Vom. I refer to that species; from Ch. similis Balz. it differs by much larger size and shorter hairs, as well as by comparatively shorter and less slender palps—for instance, the fingers, which are 1:2 (3) or 1°3 (2), not 1-4, as long as hand is wide. 46. CHELIFER MACROCHELATUS Tém. (Plate XX XI. fig. 32 a; text-figs. 72-73, 78-80.) 1884. Témosvary, (5) p. 20, tab. i. figs. 12-13. 1891. Lamprochernes macrochelatus Tém., Balzan, (11) pp. 513-514, pl. 9. figs. 4-4 a. 1902. Ellingsen, (15) pp. 152-154. 1905. Ellingsen, (17) p. 1. 1905. Ellingsen, (18) p. 6. 1907. Chelifer nodulimanus Tém., pars, Tullgren (23) pp. 46-49, figs. 10 af. Male. Cephalothorav.—Large distinct ocular spots are found. The cephalothorax, which is somewhat longer than wide, has the almost straight broad anterior stripe very prominent, while the posterior is less distinct and is slightly curved forwards in the middle; from the anterior stripe to half distance from the front margin a longitudinal exoove or depression is found; the second thoracic tergite is divided by a longitudinal dark band. Abdomen.—The fairly long and slender, but somewhat depressed abdomen has all 9 IN THE BRITISH AND COPENHAGEN MUSEUMS. SIU the tergites but the first and the eleventh longitudinally divided. ‘The tergites have the sclerites smooth or indistinctly shagreened, and have along their hindmost margin about 18 moderately long hairs in addition to a median and generally two lateral hairs in front of the row on each side; the number and arrangement of “ tactile” hairs are probably as in Ch. similis Balz. Antenne (text-fig. 72, p. 290).—The galea, which extends very much beyond the terminal hair, is very powerful and has numerous branches; it is very variable, as seen by comparing text-fig. 72 with Balzan’s fig. 4a (pl. 9). Palps (Pl. XXXI. fig. 32@; text-figs. 73-79).—The palps are polished and almost smooth, except the shagreened dorsal tubercle of the trochanter and minutely granular surface of the femur and tibia ; the hairs are moderately long or long. Text-fg. 78. "The trochanter, which is 1°5 as long as wide, is anteriorly very slightly convex and posteriorly produced into a basal rounded and fairly prominent tubercle; dorsally it is prolonged into a deep protuberance of a characteristic shape, having on the whole a somewhat triangular appearance, with the obtusely-pointed tip directed somewhat towards the extremity; the outline is obtusely-pointed proximally and almost semicircular distally with the tip well-defined from the lower part; Se, the trochanter is much deeper than wide, almost as deep as long, and deeper than the femur. The femur has a rather short and well- a marked stalk, beyond which it is only slightly attenuated towards the He extremity. It is 2°2 as long as wide; anteriorly it is slightly convex Oh. macroche- and then concave, while the posterior outline beyond the stalk is latus Tom. 6d. abruptly convex and then almost straight; the dorsal outline beyond Gitochanienand ae emsiclllanis abruptly convex and then moderately so, most raised queen pupal beyond the middle, passing into the rather sudden terminal concavity ; the femur is a little deeper than wide. The tibia, which has a long fairly well-detined stalk, is somewhat longer and wider than the in anterior view. x 24. femur, and is almost 2°2 as long as wide; the anterior outline is first moderately convex and then terminally a trifle concave, while the posterior, beyond the fairly well- marked condylus and elongated, but ill-defined, basal elevation, is very slightly concave and then terminally moderately convex; the ventral outline is moderately convex, while the dorsal is slightly produced ; and the antero-dorsal surface forms a somewhat conical, rounded, and very prominent tubercle. ‘The chela, which is almost three times as long as wide, is scarcely 1-2 as broad as the trochanter is long; the hand, which is about as long as the tibia, but 1-3 as wide, is 1-7 as long as wide, but only 1:3 as long as deep, is 1:3 deeper than wide, and about 1°3 as long as the fingers, which are 1-3 as long as the hand is wide, but only a trifle (1:05) as long as hand is deep, and which gape considerably when closed; the lateral outlines of the hand are only slightly convex and so is the ventral, while the dorsal one is almost semicircular. The 312 MR. C. J. WITH ON SOUTH-AMERICAN CHELIVFERINA number of accessory teeth is very large, viz. anteriorly 10 and 15 in the movable and immovable finger respectively, while both fingers have posteriorly about 30 in addition to the marginal row. The immovable finger has numerous spots, especially posteriorly, while the movable finger has only two posteriorly (cf. fig. 32@ and — text-fig. 79). Coxw.—The coxe are scarcely different from those of Ch. argen™ tinus Thor. (cf. above). Legs.—The hairs are rather short. The legs are long and slender; the tibia of the first pair of legs is almost 1°3 as long as the tarsus, which is 4°5 as long as deep; the femur of the fourth is 9-8 as long as deep, scarcely 1:1 as long as the tibia, and about 18 lower than the tarsus is long. Colowr—The palps and the cephalothorax are dark reddish brown, while the tergal sclerites are pale brown. Measurements.—Cephalothorax 1:38 (1:20); abdomen 3:0 (1:5) mm. Palps: trochanter 0-690 (0-460); femur 1-245 (0°552) ; tibia 1:311 (0°621); hand 1-334 (0-805), depth 1-012; finger 1-058 mm. Leg I.: femur 0°897 (0-253), trochantin 0:140 (0°276); tibia Text-fig. 79. Ch. macrochelatus 0-736 (0:184); tarsus 0°575 (0°125) mm. Tom. ¢. Tibia and Leg IV.: femur 1-058 (0°375) ; tibia 1:00 (0:207); tarsus 0°667 chela in _ posterior (0-156) mm. view. x 24. Female. Palps (text-fig. 80).—The granulation of the palps is perhaps Testeeso somewhat better marked. ‘The dorsal protuberance of the trochanter es is somewhat lower and has a similar, but less characteristic shape, (or as its tip is not marked out from the rest, but regularly rounded, and as the proximal margin is scarcely obtuse-angled; the trochanter is as deep as, not deeper than, the femur, which dorsally does not show any trace of terminal concavity. The ¢bza, the stalk of which appears less slender on account of the almost straight posterior margin, has the antero-dorsal surface very slightly produced. The hand, which is longer, but 1:4 wider, than the tibia, is 1-9 as long as wide, scarcely deeper than wide, and 1°5 as long as the fingers, which are only 1:2 as long as the hand is wide and scarcely gape when closed; the lateral outlines are a little more distinctly convex than in the male, while the dorsal and ventral ones are less distinctly convex. The number of accessory teeth is much smaller, viz. anteriorly about five in both fingers and posteriorly about twelve, while the number and arrangement of spots is practically as in the male. eat ee ae ao iain Nie 5 Ch. macroche- latus Tom., @. Trochanter and fe- mur of right palp in anterior view. x 24. IN THE BRITISH AND COPENHAGEN MUSEUMS. o13 -Legs.—The tarsus of the first pair is 4:8 as long as deep, and the femur of the fourth pair of legs is 2°9 as long as deep, but scarcely longer than the tibia. Measurements.—Cephalothorax 1:380 (1:265); abdomen 2°75 (1°75) mm. Palps: trochanter 0-736 (0-483); femur 1-311 (0:590); tibia 1-334 (0°635); hand 1-495 (0°874), depth 0°874 ; finger 1:012 mm. Leg I.: femur 0:945 (0:276), trochantin 0-161 (0°299); tibia 0°782 (0°184); tarsus 0:598 (0°126) mm. Leg IV.: femur 1:173 (0°405); tibia 1-150 (0-220); tarsus 0°713 (0-150) mm. Variation.—The specimens examined varied rather considerably in size, those of which the measurements are given being the largest at my disposal; the smallest male examined, which was referred to this species (from La Moka), had the tibia .of the palps 1:104 mm. long. A single female from Rio Janeiro, with the tibia 1:15 mm. long, was only with hesitation referred to this species, as the outlines of the tibia were more strongly convex. The characteristic shape of the trochanteral tubercle dorsally, as well as that of the protuberance of the tibia in the male, is in many specimens only slightly pronounced. Material.—Of this species I have examined a number of specimens (22 ? and 12 ¢ ) found beneath the elytra of a specimen of Acrocrinus longimanus from Brazil; three small, almost dark specimens from Tobasis Kapa (Keyserling Coll.); Dr. Meinert collected in Venezuela a single small male in the month of December at Hacienda Elias, Los Trinchéras, as well as two in the month of August at La Moka by sifting ; Reinhardt collected a single female in Rio Janeiro. ‘This species has previously been recorded from the greater part of the sub-continent, viz. Venezuela, Brazil, Ecuador, Paraguay, and Chile. Remarks.—That the species described above is identical with that described by Balzan as well as Ellingsen under this name I regard as certain, in spite of much larger measurements given by Balzan, especially compared to those of a rather small specimen which J had the opportunity of examining by the kindness of Mr. Ellingsen. This species is very nearly related to Ch. argentinus Thor., so nearly that the position of some specimens is not easily ascertained, but I nevertheless regard it as a well-defined species, but one with avery wide range of variation. ‘The female differs generally from that of Ch. argentinus Thor. by its larger size, by the less slender, more depressed abdomen, by the more distinctly granular cephalothorax, by the hand, which is at least not wider than deep, and by the fingers, which are only 1:2 longer than the hand is deep, and by the tarsus of the first pair of legs, which is only 4°8 as long as deep. The male differs besides by the characteristic dorsal tubercle of the trochanter, by the pronounced protuberance of the tibia, by the hand, which is 1°3 as deep as wide, and the fingers, which are only 1:05 longer than hand is deep. According to Ellingsen the main character seems to be the granulation of the head, but this scarcely holds good (¢f. above). Concerning the relation of the described species to that which Témésvary described, I refer to Ellingsen’s discussion (15. p. 154). VOL. XVIII.—ParT 111. No. 13.—October, 1908. 20) 514 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA 47, CHELIFER NODULIMANUS Tém. (Plate XXXI. fig. 33a; text-fig. 81.) 1882. Tomésvary, (4) p. 244. 1884. Témésvary, (5) p. 26, tab. i. fig. 14. 1889. Daday, (8) pp. 173-174, tab. iv. figs. 3 & 9. 1905. Ellingsen, (18) pp. 3-6. 1906. With, (20) p. 171. 1907. Chelifer nodulimanus Tom. pars (?) Tullgren, (23) pp. 46-49, figs. 10 a-f. Male. Cephalothoraz.—Large distinct ocular spots. The cephalothorax, which is somewhat longer than wide, has the almost straight anterior transverse groove fairly prominent in contradistinction to the posterior less prominent one, which is slightly curved forwards in the middle. The integument in front of the median groove is minutely, but distinctly granular ; behind it is almost smooth or rather indistinctly Text-fig. 81. granular (specimen from Tobasis Kapa); the hairs are rather short. In front of the anterior groove a longitudinal depression is found, passing into a deep cavity situated almost in the middle of the head (cf. Ellingsen, 18. p. 4). Abdomen.—The moderately long and depressed abdomen has all the tergites, with the exception of the first and the eleventh, longitudinally divided. The sclerites are indistinctly shagreened, and each tergite has along the hindmost margin about 30 rather short hairs in addition to one or two lateral and a single median hair in front of the row; “tactile” hairs are missing in the specimens examined. Antenne.—The galea, which is very powerful and almost twice as long as the terminal hair, is bifurcate, each branch again being strongly divided. Paips (P). XXXI. fig. 33 a; text-fig. 81).—The palps are almost smooth or minutely granular, most distinctly on the inner surface of the femur and tibia; the fingers are not smooth but fairiy distinctly granular, ‘The hairs are long or moderately so; a number of Ch, nodulimanus long, slender, and completely simple ones are found on the dorsal T%m., d. Femur, tubercle of the tibia in addition to the usual long and slender ones bia, andy oue of the posterior surfaces (cf. 18. p. 4). The trochanter, which is posterionbeses 1:5 as long as wide, is anteriorly very slightly convex and posteriorly x 1355, produced into a prominent and somewhat rounded tubercle ; dorsally it is prolonged into a deep, pointed protuberance (fig. 33 @) somewhat similar to that of the preceding species, but less suddenly attenuated terminally, less distinctly obtuse-angled proximally, and more moderately convex towards the end. it is much of right palp in IN THE BRITISH AND COPENHAGEN MUSEUMS. 315 deeper than wide, but less deep than the femur. The femur has a long and fairly well-defined stalk, beyond which it is scarcely attenuated towards the end. It is 2-3 as long as wide; its anterior outline is almost straight, while the posterior is strongly, but not very abruptly convex and then slightly so; the dorsal outline is first moderately convex and then towards the middle distinctly convex, being most raised just beyond the middle, rather steeply sloping into the terminal concavity; the femur is much deeper than broad. The tibia, which has a very long and fairly well-defined stalk, is distinctly longer and wider than the femur, and is 271 as long as broad; the anterior outline is first moderately convex, and then a trifle concave, while the posterior beyond the not very prominent condylus and elongated basal elevation is moderately concave aud then distinctly convex; the ventral outline is almost semicircular distally and the dorsal one is distinctly produced, as the antero- dorsal surface is prolonged into a deep, conical, and somewhat rounded protuberance. The chela, which is 2°8 as long as deep, is scarcely 1'2 as broad as the trochanter is long; the hand, which is much longer and about 1:4 wider than the tibia, is 1-4 as long as wide, but scarcely 1:2 as long as deep, 1°3 as deep as wide, and about 13 as long as the fingers, which are 1-2 as long as the hand is wide but 1:1 shorter than the hand is deep, and which gape widely when closed; the lateral outlines of the hand are only slightly convex, while the ventral one is distinctly so, and the dorsal almost semicircular. Anteriorly both fingers possess at least ten accessory teeth, and posteriorly about fifteen. The immovable finger has posteriorly about eleven “ spots” behind the basal tactile hairs, and eight in a longitudinal row beyond ; posteriorly only a few are found (as seen in text-fig. 81). Coxe.—The fourth pair are almost exactly like those of Ch. argentinus Thor. Legs.—The hairs are moderately long. The legs are very long and slender; the tibia of the first pair is 1:3 as long as the tarsus, which is 4°9 as long as deep; the femur of the fourth pair of legs is 3:1 as long as deep, scarcely longer than the tibia, and about half as deep as tarsus is long. Colour.—The palps and the head are dark reddish brown, the thorax is yellowish brown, and the abdominal sclerites are dark brown. ‘The legs have the anterior surface yellowish, but the posterior dark brown, a difference which is most marked in the first pair of legs, and which is also indicated in the preceding species. Measurements.—Cephalothorax 2°2 (2:0); abdomen 3°75 (2°25) mm. Palps: trochanter 1-104 (0:736); femur 1:840 (0°805); tibia 1°905 (0°920); hand ‘1-955 (1-265), depth 1-656; finger 1:541 mm. Leg I.: femur 1:265 (0°345), trochantin 0:230 (0°391); tibia 1-150 (0°230); tarsus 0:782 (0°161) mm. Leg 1V.: femur 1°495 (0-475); tibia 1:505 (0°253); tarsus 0-920 (0184) mm. Variation.—One of the two specimens, a badly preserved one from Tobasis Kapa, was almost black, had the granulation of the cephalothorax more pronounced, and showed minor differences in the proportions of the palps and legs. 20 2 316 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERIN © Material.—Of this species I have examined two specimens (6), the one found beneath the elytra of a specimen of Acrocrinus longimanus from Brazil and the other from Tobasis Kapa (Keyserling Coll.), both found in company with specimens of Ch. macrochelatus Tom. Ellingsen’s specimen is from Brazil, found under exactly the same circumstances as the former of the above-mentioned ; all other localities may, as shown by Ellingsen (cf. 18. p. 5), be regarded as doubtful. Rtemarks.—W ell-developed specimens ( ¢ ) of this species can scarcely be confounded with any other form of Chelonethi on account of its powerful palps, with the peculiar shape of the dorsal tubercle of the trochanter and tibia, the long and slender stalks of feraur and tibia, as well as by its exceedingly deep hand with the gaping fingers ; but perhaps smaller animals are less easily distinguished from very large specimens of Ch. macrochelatus Tom., for instance that of which Balzan has given the measurements (11. p. 513)—perhaps Balzan has, like Daday (cf. 8. p. 173, and Ellingsen, 15. p. 154), confounded the two species. As regards the relation of the South-American species to Témésvary’s typical form from Dalmatia, I refer to Hllingsen’s discussion (18. pp. 5-6). 48. CHELIFER CERVUS Balz. (Plate XX XI. figs. 34 a-e; text-fig. 82.) 1890. Balzan, (10) pp. 412-414, tav. xiii. figs. 2-2 ait, 1891. Balzan, (11) pp. 514 & 548, pl. 9. fig. 4 0. 1902. Ellingsen, (15) pp. 165-167. Male. Cephalothoraz.—Large ocular spots rather indistinct. The cephalothorax, which is distinctly longer than broad, has two broad transverse stripes, of which the more prominent anterior one is curved backwards in the middle, while the posterior almost straight stripe is rather indistinct; in the middle in front of the anterior stripe a rather shallow longitudinal depression is observed. The integument of the head appears polished and is very minutely granular, but is behind the median transverse stripe smooth or almost so; the hairs are rather short. Abdomen.—The fairly long and slender abdomen has all the tergites with the exception of the eleventh longitudinally divided, the first three are wider, but distinctly shorter than the following. The sclerites are almost smooth and the tergites bear along the hindmost margin from 14 to 20 moderately long (?) hairs, and the median tergites possess in addition to these two lateral and a single median hair in front of the row on each side; “tactile” hairs missing in examined specimen. ‘The ventral surface behind the genital area was distinctly keeled, the two sclerites of each sternite making an acute angle with each other. Antenne (Pl. XXXI. fig. 34a@)—The galea, which extends distinctly beyond the terminal hair, is provided with a large number of branches as seen in figure. Palps (Pl. XXXI. fig. 340; text-fig. 82)—The palps are almost smooth except IN THE BRITISH AND COPENHAGEN MUSEUMS. 317 anteriorly, where they are minutely granular. The trochanter and the anterior surface of the femur are beset with a number of moderately long, stiff, and not completely simple hairs; the posterior surface of the femur and the tibia as well as the protuberance of the latter are densely covered—almost shaggy—with long, slender, and completely simple hairs; similar but shorter hairs cover the hand. The ¢rochanter, which is almost 1°5 as long as wide, is anteriorly very slightly convex and posteriorly is produced into a fairly prominent and rounded basal tubercle; dorsally it is prolonged into a very deep, conical, obtusely pointed protuberance; observed from the front this protuberance is most similar to an acute-angled triangle, with the proximal margin slightly convex and the opposite a trifle concave; the trochanter is as deep as long and much deeper than the femur is. The femur has a rather short and well-defined stalk, beyond which it is attenuated towards the end. It is 2:2 as long as wide; the anterior outline beyond the stalk is almost straight, while the posterior is first rather abruptly convex and then slightly so; the dorsal outline beyond the stalk is abruptly convex and then very slightly so, with a fairly pronounced terminal concavity; the femur is distinctly deeper than wide. The ¢tidta, which has a long and slender, not very well-defined stalk, is longer and wider than the femur, and is 2:3 as long as wide; anteriorly it is slightly convex, and posteriorly beyond the well- marked condylus and fairly distinct basal elevation almost straight and then moderately convex; the ventral margin is almost circular, and the dorsal anterior surface is prolonged into a deep, rounded, somewhat conical protuberance. The chela, which is three times as long as deep, is 1:2 as broad as the trochanter is long; the hand, which is somewhat longer than, but 1:4 as wide as, the tibia, is 1:°7 as long as wide but only 1-2 as long as deep, 1:4 as deep as broad, and scarcely 1:3 as long as fingers, which are almost 1-4 as long as hand is wide but a trifle shorter than it is deep, and which gape widely when closed; the lateral outlines are moderately convex, while the Tex t-fig. 82. dorsal is very suddenly raised, being almost perpendicular on the stalk, and then slightly convex; the ventral margin is less suddenly raised and then almost straight. The marginal teeth Gis Py Beene fairly well developed, where the fingers touch each other in halls, of Se walt fia WOME terminal fifth, but more proximally are very low or almost anterior view. x24. obsolete; accessory teeth were not observed posteriorly, but anteriorly the immovable finger has five and the movable seven near to the tip. Anteriorly the immovable finger has about 15 “‘spots” and the movable four (cf. fig. 82), while the immovable posteriorly has seven spots. Coxe.— The second and third pair are more slender than the corresponding in the preceding species. The fourth pair are somewhat triangular, as the inner and hinder 318 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA margins, of which the former is a little longer, gradually merge into each other; the coxe are scarcely widened out towards the extremity. Legs (Pl. XXX1. figs. 34 c-e).—The hairs are moderately long and stiff, not quite simple; long, slender, completely simple hairs are placed dorsally at the tip of the femur, somewhat beyond the middle in the tibia, and one-fourth removed from the base in the tarsus, the latter at least as long as the distance to the end. ‘The legs are moderately long and slender; the tibia of the first pair of legs is 1-3 as long as the tarsus, which is five times as long as deep; the femur of the fourth pair of legs is 2°8 as long as deep, 1:1 as long as tibia, and 1-7 lower than tarsus long. The claws show a remarkable structure; the posterior claw of the first pair is distinctly curved and pointed, but just beyond the middle is widened out to a large, somewhat rounded, and triangular process or tooth (fig. 84d); the anterior claw of the first pair, as well as both claws of the fourth pair, has a similar median, but somewhat posteriorly placed tooth (figs. 834¢ & e). Colour.—Palps and head reddish brown; the thorax is lighter brown and the tergal sclerites are yellowish brown. Measurements.—Cephalothorax 1°38 (1:15); abdomen 2°75 (1:10) mm. Palps: trochanter 0°690 (0°483); femur 1-265 (0°570); tibia 1-380 (0-610); hand 1-449 (0°851), depth 1:173; finger 1-150 mm. Leg I.: femur 0-897 (0°299), trochantin 0°155 (0°322); tibia 0-775 (0-195); tarsus 0:574 (0-116) mm. Leg IV.: femur 1:196 (0425); tibia 1-081 (0-230); tarsus 0:713 (0-161) mm. Material—oOf this species I have examined a single male from the Amazons; it has previously been recorded from Matto Grosso and Surinam. Remarks.—The described form differs from Balzan’s description by smaller size, by slightly different shape of galea (cf. 10. fig. 2 @), by the less produced posterior surface of the trochanter, and by the different proportions of the joints of the palps; the most important of these differences, namely that found in the very wide hand of the palps, which is much wider than the finger is long, is easily explained by the fact that he has measured the hand not in a true dorsal but in a partly lateral view. From Ellingsen’s description it differs in points of even smaller importance; and consequently I regard the identification of the male from the Amazons with those from Matto Grosso and Surinam as beyond doubt. This species shows in the very important character found in the claws of the legs similarity to the following species, but I nevertheless think that it is more nearly related to Ch. nodulimanus Tém. and allied species on account of the large galea, the structure of the head, and striking similarities in all the joints of the palps as well as in the legs. A rather remarkable dissimilarity is found in the structure of the tibia, if Ellingsen is right in the statement that its conical protuberance is equally well developed in the female and the male (cf. 15. p. 167). IN THE BRITISH AND COPENHAGEN MUSEUMS. 319 49. CHELIFER DEPRESSIMANUS, sp.n. (Plate XXXI. figs. 35 a-d.) Female. Cephalothoraz.—Rather indistinct ocular spots. The cephalothorax, which is distinctly longer than wide, has the anterior transverse stripe broad, almost straight, and fairly prominent, whilst the posterior is almost obsolete. The integument appears polished and is completely smooth with moderately long hairs. Abdomen.—The very long slender abdomen has all the tergites with the exception of the first and the eleventh longitudinally divided, but not very distinctly so. The sclerites are smooth, and along the hindmost edge of each tergite is a row of about 14 moderately long hairs, in addition to one or two lateral and a single median hair in front of the row on each side; the “tactile” hairs were missing in the specimen examined. Antenne.—The moderately long, slender galea, which extends somewhat beyond the terminal hair, has six inferiorly placed teeth distally. Palps (Pl. XXXI. figs. 35 a-b).—The palps are almost smooth except on the anterior surface, which is minutely and rather indistinctly granular. The long or moderately long hairs are stiff and not completely simple; long, slender, completely simple hairs are only observed posteriorly just behind the tip of the femur and in the middle of the tibia, ‘The trochanter, which is 1°5 as long as wide, is anteriorly slightly convex and posteriorly produced into a rounded and rather low basal tubercle; dorsally it is prolonged into a not very deep, somewhat conical, and rounded protuberance ; the trochanter is deeper than wide, but scarcely as deep as the femur. ‘The femur has a short well-defined stalk, beyond which it is attenuated towards the end. It is twice as long as broad ; the anterior outline is first slightly convex and then concave, while the posterior beyond the stalk is very abruptly convex, almost perpendicular, and then almost straight; the dorsal outline is almost exactly like the posterior. The tidia, which has a rather short but well-defined stalk, is a little longer and wider than the femur, and is 1‘9 as long as wide; the anterior outline is very strongly convex, almost semicircular in the middle, and then a trifle concave, while the posterior beyond the prominent condylus and slightly marked basal elevation is almost straight and then strongly convex; the dorsal as well as the ventral outlines are moderately convex. The chela, which is 2°7 as long as wide, is almost 1:2 as broad as the trochanter is long; the hand, which is somewhat longer and about 1°3 as wide as the tibia, is 1-6 as long as wide but 1-9 as long as deep, 1:1 as wide as deep, and is 1° as long as the fingers, which are 1-1 as long as hand is wide, but 1-2 as long as it is deep, and do not gape when closed ; the lateral outlines are moderately convex, while the dorsal and ventral are almost straight. Anteriorly each finger has a single accessory tooth, while 320 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA posteriorly each possesses a few in the distal third; the movable finger has only two spots anteriorly near to the base, while the immovable finger has five anteriorly, of which two are very large, two small (fig. 385 6), and one very large placed posteriorly. Coaw.—The coxe are very similar to those of Ch. argentinus Thor. (cf. text-fig. 75) ; the fourth are trapezoidal, scarcely widened out towards the extremity, and with the inner and hinder margins, between which is a distinct obtuse angle, of almost equal length. Legs (P1. XXXI. figs. 35 e-d).—The hairs are rather short, with a long, slender, compleely simple hair placed dorsally, somewhat beyond the middle of tibia IV., as well as one tarsal, one-fourth removed from the base and as long as the distance to the tip; the terminal dorsal one of femur IV. was not observed, but a long, not completely simple one is placed dorsally near the tip of the tibia. ‘The legs are rather short and clumsy; the tibia of the first pair is 1:1 as long as the tarsus, which is four times as long as deep; the femur of the fourth pair of legs is 2°3 as long as deep, almost 1:2 as long as the tibia, and only 1-3 lower than the tarsus is long. ‘The anterior claw of the first pair of legs has a rather small median tooth, and the posterior has a rather large triangular process (fig. 35 ¢), while the anterior claw (cz) of the fourth pair of legs has practically no tooth and the posterior has a rather big, somewhat triangular tooth (cf. fig. 35 d, ce). Colour.—The palps and cephalothorax are reddish brown; and the tergal sclerites are dark brown. Measurements.—Cephalothorax 0690 (0°598) ; abdomen 27185 (0-920) mm. Palps: trochanter 0°335 (0-220); femur 0°552 (0-276); tibia 0°575 (0-310); hand 0:647 (0°391), depth 0°345 ; finger 0:-437 mm. Leg I.: femur 0°410 (0:144), trochantin 0°076 (0°152); tibia 0-296 (0-099); tarsus (0:266 (0-068). Leg IV.: femur 0°524 (0°228); tibia 0-448 (0°132); tarsus 0°296 (0-084) mm. Material.—Of this species I have examined a single badly preserved female from Uruguay (Keyserling Coll.). Remarks.—This species shows in the shape of the palps great similarity to Ch. ovatus Balz. (cf. fig. 6, pl. 9, 11), but as it is much smaller, as the galea is much better developed, and as Ba zan does not mention the presence of teeth in the claws, I do not hesitate in establishing my specimen as a new species. In the shape of the claws this species shows certainly greater similarity to Ch. cervus Balz. than to any other, but, nevertheless, I am not convinced that the two are nearly related to each other (cf. above). IN THE BRITISH AND COPENHAGEN MUSEUMS. 321 IV. Group of Cueiirer Birmanicus Thor. Femur of the first pair of legs with wide oblique articular cavity, with posterior condylus placed near ventral margin. Real eyes always wanting. Fingers of palps without accessory teeth. “Tactile” hair of tarsus IV. basal. Genital area of 3 of birmanicus ¢ype. (Olaws as well as subterminal hair simple. Transverse grooves indistinct or wanting. Hairs pointed and almost simple ; tibia of palps basally with a pair of long, slender, completely simple hairs. Flagellum consists of four hairs, and sexual difference is found in the galea.) I think that the two following species are naturally referred to the same group as Ch. birmanicus Thor. (cf. 20. p. 133), as they entirely agree with that species in structure with the exception of the genital area of the male, which in the badly preserved specimens of Ch. nidificator Balz. at least is less conspicuous than in Ch. birmanicus Thor. (cf. 20. p. 124, tab. iii. fig. 10a), but nevertheless seems to be similar in important features. As seen from the above, only two species of those which Ellingsen refers to Atemnus Can. (cf. 18. p. 2) belong to this group, the other three are more naturally referred to the subgroup of Ch. subrobustus Balz. Synopsis of Species. a. Trochanter not bigibbose ; tibia with short stalk . . . . . . . 50. Ch. nidificator Balz. 6. Trochanter distinctly bigibbose ; tibia with long and slender stalk . 51. Ch. elongatus Ell. -00. CHELIFER NIDIFICATOR Balz. (Plate XXXI. figs. 36 a—c; text-fig. 83.) 1890. Balzan, (10) pp. 417-418, tav. xiii. figs. 5-5 6. 1891. Atemnus nidificator Balzan, (11) pp. 510-511, pl. 9. fig. 1. 1902. Ellingsen, (15) pp. 146-148. 1905. Ellingsen, (18) p. 2. 1905. Ellingsen, (19) pp. 823-324. 1907. Tullgren, (23) p. 56. Male. Cephalothoraz.—Ocular spots rather indistinct. The cephalothorax, which is scarcely longer than wide, has no transverse grooves, is smooth and provided with fairly long hairs. Abdomen.—The fairly long slender abdomen has generally the tergites indistinctly divided with the exception of the first three and the last one; the division of the tergites is sometimes not complete, consisting of an anterior and a posterior incision separated in the middle; a similar incision may be observed in the second and third segments posteriorly as well. The sclerites are smooth, and the tergites possess along their hindmost margin from 8-12 long or very long, almost simple hairs placed within VOL. XVIIIL—PART 111. No. 14.— October, 1908. 2x 22 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA (5%) white spots; in front of the rowa single hair is observed laterally, but none in the middle. The last two tergites each bear two pairs of very long and slender “ tactile ” hairs and so do the corresponding sternites. The genital area is very inconspicuous. Antenne.—The galea, which extends a little beyond the terminal hairs, is short and slender, with a few terminal teeth. The flagellum consists of four hairs, of which the longer anterior one is serrated, but with smaller teeth than in Balzan’s figure 5 6 (10). Palps (Pl. XXXI. figs. 36 a-b)—The mawilie are smooth and so are the palps, except part of the anterior surface, especially of the femur and tibia. ‘The hairs are long and fairly long, stiff, and not completely slender; the following very long and slender completely simple hairs were observed, viz., three on the femur (two somewhat behind the tip, the one above the other, and the third posterior and terminal), two on the tibia above and below the basal elevation, and three at the base of the hand in a transverse row (one more dorsally placed and two more ventral). The trochanter, which is 1°5 as long as wide, is anteriorly slightly convex, and posteriorly produced into a rounded, not very pronounced tubercle ; dorsally it is prolonged into a low, rounded, and slightly conical protuberance. The femur has a very short, well-defined stalk, beyond which it is suddenly enlarged and then distinctly attenuated towards the end. It is twice as long as broad; the anterior outline beyond the stalk is rather suddenly convex and then slightly concave, while the posterior, beyond the stalk, is very abruptly convex and then slightly so; dorsally the outline is first abruptly convex and then almost straight. The ¢idia, which has a very short and well-defined stalk, is longer aud wider than the femur and is 1-9 as long as broad; anteriorly it is very strongly convex in the middle and a trifle concave terminally, while posteriorly, beyond the not very large condylus and insignificant basal elevation, it is for a short distance almost straight and then distinctly convex; the dorsal and ventral outlines are moderately convex. The chela, which is 2°7 as long as wide, is almost 1°3 broader than the trochanter is long; the hand, which is somewhat longer than, but 1:3 as wide as, the tibia, is 1:6 as long as wide, about 1:1 wider than deep, and 1:4 Text-fig. 83. as long as the fingers, which are 1:1 as long as the hand is wide and do not gape when closed; the lateral outlines are slightly convex, while the dorsal and ventral ones are almost straight. ” No accessory teeth are observed ; anteriorly four “spots” are observed between the three proximal “ tactile” hairs (cf. fig. 36 @) in the immovable finger, but none in the movable finger ; Ch. nidificator Balz., R ee 5 5 5 do. Coxe. x 44. posteriorly a similar number is observed in the immovable as well as in the movable finger (cf. fig. 36 0). Cove (text-fig. 83).—The second and the third pair are somewhat enlarged towards the end; the fourth pair are trapezoidal, scarcely widened out, and with the inner margin somewhat longer than the concave hinder one. IN THE BRITISH AND COPENHAGEN MUSEUMS. 323 Legs—The hairs, which are fairly long or long, are either completely simple or with a few terminal teeth; long and slender, completely simple hairs were missing except the tarsal “tactile” hair of the fourth pair, which is as long as the tarsus at least and one-tenth removed from the base, and the ventral ones of the trochanters and trochantins. The legs are rather short and clumsy; the trochantin of the first pair is much deeper than the femur proper, with a very wide articulate cavity, and its tibia is 1-1 as long as the tarsws, which is 3-5 as long as deep; the femur of the fourth pair of legs is 2°3 as long as deep, 1°3 as long as the tibia, and at least 1:2 lower than tarsus is long. Colour.—Proximal joints of the palps are more yellowish brown and distal more reddish brown ; the cephalothorax and tergal sclerites are brown or yellowish brown. Measurements.—Cephalothorax 0°805 (0°790); abdomen 2°30 (0-920) mm. Palps: trochanter 0°391 (0°253) ; femur 0°690 (0°345) ; tibia 0°713 (0°375) ; hand 0°805 (0°495), depth 0°437 ; finger 0°552 mm. Leg I.: femur 0°494 (0182), trochantin 0-091 (0°205); tibia 0°380 (0°133) ; tarsus 0°325 (0-095) mm. Leg IV.: femur 0°765 (0°334); tibia 0-600 (0°205) ; tarsus 0-410 (0°129) mm. Female. Antennwe.—tThe galea, which extends beyond the terminal hair, is fairly long and slender, with about six terminal teeth, as figured by Balzan (10. fig. 56 a). Palps.—The femur and the tibia are comparatively less slender, being 1:9 and 1:8 respectively as long as wide; the hand is 1‘7 as long as wide, distinctly 1-1 as wide as deep, and 1:5 as long as fingers. Legs.—The legs are less slender, the femur of the fourth pair of legs 2°2 as long as deep, 1:2 as long as the tibia, and scarcely 1:2 lower than tarsus is long. Measurements.—Cephalothorax 0:920 (0-920); abdomen 3°22 (1:495) mm. Palps: trochanter 0-414 (0°276); femur 0:736 (0°391) ; tibia 0-759 (0°414) ; hand 0°897 (0°529), depth 0:460; finger 0°598 mm. Leg I.: femur 0°517 (0:213), trochantin 0:095 (0:228) ; tibia 0-403 (0-144) ; tarsus 0°334 (0-099). Leg IV.: femur 0:798 (0°357) ; tibia 0°646 (0:213) ; tarsus 0-418 (0-137) mm. Material.—Of this species | have examined five males and five females, of which ~ one carried a large bundle of eggs, as well as two young ones, collected by the Danish ‘Galathea’ Expedition in Bahia, a single male sifted by Dr. Meinert in the month of August at La Moka, and one collected by M. Simon in St. Vincent; besides these, a female, collected in Guatemala by Dr. Stoll, and a male from Tobasis Kapa were examined. ‘This species has previously been recorded from Venezuela, Ecuador, Brazil, Paraguay, Argentine, and Haiti. 2x2 324 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA Remarks.—Vhis species, of which the small specimens (var. minor Balz.) examined were exactly like one determined by Mr. Ellingsen, is easily distinguished from all other South-American species by the shape of the palps. 51. CuELIFER ELONGATUS Ell. (Plate XXXI. figs. 37 a-q; text-fig. 84.) 1902. Ellingsen, (15) pp. 149-151. Femate. Cephalothoraz.—Ocular spots indistinct. ‘The cephalothorax, which is distinctly longer than wide, has no transverse grooves, is smooth and provided with fairly long hairs. Abdomen.—The fairly long slender abdomen has all the tergites, with the exception of the first and eleventh, more or less distinctly divided longitudinally. The tergites have the sclerites smooth, and possess along their hindmost margin about ten moderately long hairs in additition to four (six?) in front of the row; tip of the abdomen with tactile hairs. Antenne.—tThe galea, which extends very much beyond the terminal hair, is fairly long and slender and is provided with about six short branches in the distal third. The flagellum consists of four hairs, of which the anterior has about ten longer and shorter teeth along the front margin. Palps (Pl. XXXII. figs. 37 a—f)—The maaille are smooth and so are the palps, with the exception of the posterior surface of the trochanter and the anterior of the three following joints, which are more or less distinctly granular. The fairly long and long hairs are stiff and not completely simple ; long, slender, and completely simple hairs are placed posteriorly in the middle and at the tip of the femur, posteriorly at the base of the tibia, one above and one below the basal elevation, and posteriorly at the base of the hand at least a single one; the hand has ventrally between the base and the insertion of the finger a single hair (fig. 37 d). The trochanter, which is 1°5 as long as wide, is anteriorly slightly convex, almost straight, and posteriorly is produced into a rounded, fairly long tubercle ; dorsally it is prolonged into a somewhat conical and fairly deep protuberance; the trochanter is somewhat deeper than wide, but not as deep as the femur is. The femur has a short, well-defined stalk, beyond which it is slightly attenuated towards the end. It is 2:2 as long as deep; the anterior outline is slightly convex and then concave, while the posterior is rather abruptly convex and then almost straight; dorsally the outline is first abruptly convex and then completely straight ; the femur is about as deep as wide. The tibia, which has a long, slender, fairly well- defined stalk, is longer and wider than the femur, and is twice as long as broad ; anteriorly it is very strongly convex, almost semicircular, and then terminally almost straight, while posteriorly, beyond the prominent condylus and well-marked basal elevation, it is slightly concave and then strongly convex; the dorsal outline is strongly IN THE BRITISH AND COPENHAGEN MUSEUMS. 325 convex, while the ventral is moderately so. ‘The chela, which is 2:7 as long as wide, is about 1:2 as broad as the trochanter is long; the hand, which is longer and 1°3 as wide as the tibia, is 1-7 as long as wide, scarcely as wide as deep, and 1°6 as long as the fingers, which are only a trifle longer than the hand is wide; the lateral outlines are slightly convex, while the dorsal and ventral ones are almost straight. The immovable finger possesses anteriorly seven proximal “ spots” in a longitudinal row, while the movable has none (figs. 37 ef), and posteriorly the immovable finger has two “spots” and the movable a single one. Coxe.—The second and third pair are like those of the preceding species (cf. text- fig. 83), while the fourth are of a more triangular appearance with the inner and hinder margins, of which the former is much the longer, gradually merging into each other. Leqs.—The fairly long or rather short hairs are generally more or less stiff and not quite simple; the femur has dorsally a terminal, long, slender, completely simple hair ; the tarsal ‘‘¢actile” one is one-tenth removed from the base and at least as long as the tarsus. ‘The legs are moderately long and slender; the tibia of the first pair is 1:2 as long as the tarsus, which is 4°3 as long as deep; the femur of the fourth pair is 2:4 as long as deep, 1-2 as long as tibia, and 1:2 lower than the tarsus is long. Colour.—The palps are reddish brown and the cephalothorax as well as the tergal sclerites are pale yellowish brown; the rest of body more or less yellowish. Measurements—Cephalothorax 0°805 (0°690); abdomen 3°56 (1°49) mm. Palps: trochanter 0°368 (0°253); femur 0°640 (0°285); tibia 0°667 (0°335); hand 0°713 (0:425), depth 0°487; finger 0-450 mm. Leg I.: femur 0°456 (0-160), trochantin 0°091 (0°175); tibia 0°365 (0106); tarsus 0-299 (0-072) mm. Leg IV.: femur 0°631 (0°266) ; tibia 0°540 (0°152) ; tarsus 0°329 (0-091) mm. Variation.—Another female mounted with the above mentioned was much paler and had the hairs of the comparatively short abdomen well preserved; along the hindmost margin twelve hairs in addition to six in front of the row in the median tergites; the tenth and last tergites as well as the corresponding sternites with two pairs of tactile hairs each. The palps were somewhat less slender, but in the main as described. Male. Abdomen &c.—The abdomen is exceedingly long and slender, the galea scarcely extends beyond the terminal hair, and the teeth are wanting or obsolete. Palps (Pl. XXXI. fig. 37 g).—The trochanter is much more distinctly bigibbose than in the female, as the posterior tubercle is much longer and more conical and as the dorsal protuberance is very deep and conical. ‘The femur, which is 2°3 as long as wide, is almost straight anteriorly, and posteriorly less abruptly convex. The dédia, which has an exceedingly long slender stalk, is 1:9 as long as wide; the anterior 326 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA outline is much more strongly convex than in the female, and the posterior beyond the basal elevation is very distinctly concave and then more strongly convex than in the female (fig. 37g). The chela is only a trifle broader than the trochanter is long; the hand, which is only 1:2 as wide as the tibia, is 1:5 as long as the fingers, which are 1-1 as long as hand is wide. The immovable finger has anteriorly only six “spots,” less regularly arranged. Colow’.—The palps are dark reddish brown, the cephalothorax is brown, and the abdominal sclerites are almost olive-coloured, and consequently there is no sharp difference between a pale body and dark palps. Measurements. —Cephalothorax 0°690 (0°575) ; abdomen 2:07 (0°69) mm. Palps: trochanter 0°368 (0°253); femur 0°690 (0°299); tibia 0-690 (0°355); hand 0:690 (0:405), depth 0:415; finger 0-455 mm. Material.—Of this species I have examined two females collected by Mr. Jessen at (2) Essequibo; it has previously been recorded from Ecuador. Remarks.—Vhe male of this species, which I was able to examine by the kindness of Mr. Ellingsen, does not differ from the original description except in a single character ; Ellingsen writes: ‘La main presque deux fois plus longue... . que le femur,” while the animal has the hand about as long as the femur; this mistake was caused perhaps by a slip of the pen or by an error of the printer. That the female described belongs to the same species as the male, is not quite certain. ‘The differences between them in the shape of the body as well as in the structure of the trochanter and tibia, &c., are certainly striking, but not so great that they cannot easily be explained as sexual ones ; consequently I should not think myself justified in establishing a new species. About similarity to Ch. navigator With, cf. 22. p. 63. Variation—In addition to the specimens mentioned I have examined a much larger female, collected in a loaf on board the Danish ship ‘Galathea’ in January 1847, which I feel obliged to refer to the same species, as it only differed in mere trivialities. The trochanter has protuberances somewhat longer (text-fig. 84) and is deeper than the femur; the ¢zdia is less strongly convex anteriorly and more distinctly concave posteriorly beyond the basal elevation. The tarsus of the first pair of legs is only four times as long as deep; the femur of the fourth pair of legs is 9 2°5 as long as deep, 1°3 as long as the tibia, and 1:3 lower than the tarsus Text-fig. 84. Ch. elongatus Ell., var., 2. Trochanter of left palp in anterior view. xX 44. is long. Measurements.—Body somewhat shorter and less slender; the trochanter of the palps 0°506 (0°322); femur 0:875 (0°40); tibia 0°897 (0°437); hand 0°897 (0°552), depth 0°575 ; finger 0°585 mm. IN THE BRITISH AND COPENHAGEN MUSEUMS. 327 The following four South-American species will certainly always remain nomina nuda, if the original specimens do not exist. Not even the papers in which they are described are known to me:— 52. CHELIFER COECUS Gerv. 1849. Chelanops coecus Gervais, (1) p. 18, pl. i. fig. 13. 53. CHELIFER CIMEX Gerv. 1849. Gervais, (1). 04, CHELIFER EXCENTRICUS Holmb. 1874. Holmberg, (2a) p. 299, pl. vi. fig. 6. 55. CHELIFER TIMIDUS Holmb. 1876. Holmberg (2 8). LITERATURE. An asterisk indicates that the book in question was not at my disposal; that I am able to include the pagination &c. of these papers is due to the kind assistance of Mr. A. 8. Hirst ; only the title of the paper numbered 2 6 may be regarded as doubtful (cf. Balzan, ro). r.* P. Gervais in C. Gay—Historia fisica y politica de Chile. Zoologia, iv. (1849). “‘Arachnidos quelifereos,”’ pp. 10-13, pl. i. fig. 18. 2a." BK. Ll. Hotmpure.— Descriptions et notices d’Arachnides de la République Argentine.” Period. Zool. Argent. 1874, i. pp. 283-302, pl. vi. 26.* K. L. Houmpere.—“ Aracnidos Argentinos.” An. de Agricultura Buenos Aires, t. iv. (1876). 3.° T. Tuorrtt.— Sobre algunos Aracnidos de la republica Argentine.” Period. Zool. Argent. 1875, ii. pp. 201-218. 4. O. TémésvAry.— Pseudoscorpiones Faunze Hungarice. A Magyar Fauna Alskorpi6i.”’ Magyar tudomanyos Akadémia Math. és Termeszettud Kézlemények. Budapest, vol. xviii. (1882) pp. 185-256, tab. i.-iv. 5 O. Témésvany.—“Adatok az Alskorpiok ismeretéhez.”” Termés. Fuzetek, vol. viii. (1884) pp. 16-27, tab. 1. 6.* H. Simow.— Arachnides rec. par la Mission du Cap Horn.’ Mission scientifique du Cap Horn, 1882-18838. TT. vi. Zoologie. Paris (1887), 42 pp. 7.* UL. Batzan.—Chernetide nonnulle Sud-Americane. Asuncion, Paraguay, pug. iii. 1887; pug. iii. 1888 (no pagination). 8. E. v. Dapay.—* Ubersicht der Chernetiden des ungarischen Nationalmuseums in Budapest.” Termés. Fiiz. vol. xi. (1889) pp. 165-192, tab. iv. g. E. v. Dapay.—‘ Egy braziliai aj Alskorpidgaj a magyar nemzeti museum 4llattaraban.” Termés. Fiiz. vol. xii. (1889). 328 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA. 10. L. Banzan.— Revisione dei Pseudoscorpioni del Bacino dei Fiumi, Parana e Paraguay nell’ America meridionale.’? Ann. Mus. Civico Stor. natur. Genova, ser. 2, vol. ix. (1890) pp. 401-454, tavv. xili—xvii. ir. L. Batzan.—‘‘ Chernetes (Pseudoscorpiones). Voyage de M. E. Simon au Venezuela.” Ann. Soc. Ent. vol. lx. (1891) pp. 497-552, pls. 9-12. 12. E. Simon.—“ Arachnides rec. & la Terre de Feu par C. Backhausen.” Anales del Museo Nacional de Buenos Aires, t. iv. (1895) pp. 167-172. 13. A. TutteREN.—“ Two new Species of Chelonethi (Pseudoscorpions) from America.” Entomo- logisk Tidskrift, Stockholm (1900), pp. 153-156, figs. 1-2. 14.* E. Srwon.—Hamburger magalhaensische Sammelreise: Arachnoideen. Hamburg (1902), 47 pp. (Pseudoscorpiones, p. 44). 15. KH. Evtinesen.—‘ Sur la Faune de Pseudoscorpions de ’ Equateur.” Mém. Soc. Zool. France, t. xv. (1902) pp. 146-168. 16. K. Evrinasrn.—“On some Pseudoscorpions from Patagonia collected by Dr. F. Silvestri.” Boll. Mus. Zool. ed Anat. comp. di Torino, vol. xix. n. 480 (1904), pp. 1-7. 17. HE, Eximesen.—“ Pseudoscorpiones viaggio dell Dr. Enrico Festa nell’ Ecuador e regioni Vicine. XIX.” Boll. Mus. Zool. ed Anat. comp. di Torino, vol. xx. n. 497 (1905), pp. 1-3. 18. EK, Eriinesen.—“ Pseudoscorpions from South America collected by Dr. A. Borelli, A. Bertoni de Winkelried, and Prof. Goeldi.” Boll. Mus. Zool. ed Anat. comp. di Torino, vol. xx. n. 500 (1905), pp. 1-17. 1g. EH. Exrtinesen.— On some Pseudoscorpions from South America in the Collection of Prof. Dr. F. Silvestri.” Zool. Anz., Bd. xxix. No. 10 (1905), pp. 823-828. 20. C. J. Wirn.—“ TIT. Chelonethi. An Account of the Indian False-Scorpions, together with Studies on the Anatomy and Classification of the Order.” The Danish Expedition to Siam, 1899-1900. D. Kgl. Danske Vidensk. Selsk. Skrifter, 7 Reekke, Naturvidenskab. og Mathem. Afd. i. 1 (1906), pp. 1-214, figs. 1-24, tab. i-iv. 21. C. J. Wira.— On Chelonethi, chiefly from the Australian Region, in the Collection of the British Museum, with Observations on the Coxal Sac and on some Cases of Abnormal Segmentation.” Ann. Mag. Nat. Hist. (7) xv. (1905) pp. 94-148, pls. vi-x. 22. C. J. Wira.—*On some New Species of Cheliferine Hans. and Garypide Hans. in the British Museum.” Journ. Linn. Soc., Zool. vol. xxx. (London, 1907), pp. 49-85, pls. 8-10. 23. A. Tunnernn.—* Sur Kenntnis Aussereuropaischer Chelonethiden des Naturhistorischen Museums in Hamburg.” Mitteilungen aus dem Naturhistorischen Museum, xxiv. (1907) pp. 21-75, taf. 1-v. 24. A. 'furrcren.— Uber einige exotische Chelonethiden.” Entomologisk Tidskrift, Stockholm (1908), pp. 57-64. PLATE XXIX. 330 Fig. la. Fig. 16. lg, 1d. IL @ yp 1g. 1 ib, iliaye Sd NS) 8 bo Ss 2) ¢. 6) ] b ©. bo a cS ow oD oD O92 so asa 4a, 4 6. MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERINA AQ A A Aw Oy we A A Bw PLATE XXIX. 1. Chelifer imperator, sp. n. Left side of abdomen observed from beneath in a some- what oblique position, x 24: 1, first abdominal tergite ; 10, tenth abdominal tergite. Lower surface of anterior portion of body, showing widely gaping genital opening, X 33: 0, entrance-opening to coxal sac; s, coxal spine; sé’, first stigma; 7, ram’s-horn- shaped organ, the left part of which extends somewhat beyond opening. Left palp, x 13:5. Left chela in anterior view, x 24. Left leg I. in anterior view, X 33. Tip of left leg 1. im anterior view, x 200: hi, anterior sub- terminal hair; he, posterior subterminal hair; c¢?, anterior claw ; cex, posterior claw. Left palp, x 17-5. Right chela in posterior view, X 24. Two posterior pairs of coxe, X 33. 2. Chelifer rex, sp. n. Left palp, x 17-5. Right chela of palps in almost anterior view, X 77: pd, pv, and pz, cf. text. Right chela of palps in almost posterior view, X 44. Coxe, X 44: s, rudiment of coxal spine ; ¢o, coxal sac. Left tarsus of first pair in anterior view, X 117: hz, anterior subterminal hair; cex, posterior claw. 3. Chelifer rufus Balz. Seventh to ninth sternites showing spinous areas, X 77: VIL.-IX., corresponding sternites. Right chela in posterior view, X 44. Left leg I. in anterior view, X 57. Femur of left leg I. in posterior view, x 957. Left leg IV. in anterior view, X 57. 4. Chelifer nobilis, sp. n. Left palp, x 24. Right chela, x 44. Oe C.J. With, del. Huth sc et imp. SOU DH JANE RUIIGAN CHE Ele RaNCZzS. Fig. =~] =] a i) Ty ay aj ay =~A “oa S. 5 Ge 8 6. Q ©) G. 2, YG 9 b. 9G. 9 d. 9 e. a ~. I ayes Qs 2 Qo 1 Cy Cy GQ Oh GQ & G G@ IN THE BRITISH AND COPENHAGEN MUSEUMS. 5. Chelifer canestrinii Balz. The seventh and eighth sternites, X 115: VII. and VIIL, corresponding sternites. Left palp, xX 30. Right palp in posterior view, x 44. Coxe and sexual area: @ and p, anterior and posterior plates; IV., fourth sternite. 6. Chelifer longichelifer.Balz. Sixth to ninth sternites, X 77: VI. and IX., corresponding sternites. Left palp, x 20. Right chela in posterior view, x 44. Right chela in posterior view, x 44. 7. Chelifer seqmentidentatus Baiz. The third to the fifth abdominal tergites, x c. 80: III—V., corresponding tergites. Left palp, x 24. Trochanter of left palp from below, x 77. Right chela in anterior view, X c. 50. Right chela in posterior view, X c. 50. Coxee and sexual area, x 44. Left leg I. in anterior view, X 57. Femur of left leg I. in posterior view, x 957. Left leg IV. in anterior view, X 57. Last two pairs of coxe, x 44. 8. Chelifer satanas, sp. n. Left palp, x 24. Trochanter of left palp from below, X 77. Coxee and genital area, x 44. 9. Chelifer insignis, sp. n. Left palp, x 24. Right chela in anterior view, x 44. Right chela in posterior view, X 44. 2? var. Left chela in anterior view, X 39. @ var. Left chela in posterior view, X 39. bo re! he 331 ie AS y is AVC Py (ab ees Bally Ja Toy pee SIE SA rally PMA OOK 334 MR. C. J. WITH ON SOUTH-AMERICAN CHELIFERIN A PLATE XXX. 10. Chelifer rudis Balz. Fig. 10a. ¢. left chela in anterior view, X 24. 106. ¢. Left chela in posterior view. 10 ¢. @. Cuceullus, x 44. NOG nichthpalpy < 2205 See Or a re er ee 21a THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. | Tue scientific publications of the Zoological Society of London are of two kinds—“ BrOvecSREN, 2 s published in an octavo form, and “Transactions,” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices ‘of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published in the “ Proceedings” by the Committee of Publication. A large number — of coloured plates and engravings are issued in the “ Proceedings,” to illustrate the new ‘or otherwise remarkable species of animals described in them. Among such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “Proceedings”? for each year are issued in four parts, in the months of June, August, October, and April, the part published in April completing the volume for the last half of the preceding year. The pagination is now consecutive throughout the issue for the year, so that the year and pa give the complete reference. The “Transactions” contain such of the more important communications made to the cae meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are issued at irregular intervals, Fellows and Corresponding Members, upon payment of a Subscription of £1 Is. before the d the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for year. They are 1ikewise entitled to purchase the Publications of the Society at 25 per cent, le than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1881, if they exceed the value of five pounds. Fellows also have the privilege of subscribing to the Annual Volume of the ‘ Zoological Record? for asum of £1 10s. (which includes delivery), payable on the Ist July in each year; but this privilege is forfeited unless the subscription be paid before the 1st December in each year. Such of these publications as are in stock may be obtained at the Society’s Office(3 Hanover _ Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster noe H.C.), or through any bookseller. P. CHALMERS MITCHELL, October, 1908. Secretary TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Vout. XVITI.—Parr 4. (Puates XX XII-XXXVIL.) LONDON: PRINTED FOR THE SOCIETY, SOLD AT THEIR HOUSE IN REGENT’S PARK; —— fo Msomantnss. > AND BY MESSRS. LONGMANS, GREEN, AND CO,, PATERNOSTER ‘ROW. he ts SE eo 2 February 1911, Ne Price 16s. aa Taylor and Francis, Printers.] [Red Lion Court, Fleet Street 28 191] TRANSACTIONS OF THE ZOOLOGICAL SOCIETY : OF LONDON. To Fellows. To the Public. : cb Se. a, Shed, VOLUME I. (1833-1835, containing 59 Plates). . Price 3138 6 . . . 418 O* VOLUME II. (1835-1841, contamimg 71 Plates). . 5 4 00... 5 6 6% VOLUME III. (1842-1849, containing 63 Plates). . 4, 3 8 3 vote 411 0* VOLUME 1V.(1851-1962, containing 77 Plates) 6 200 ee VOLUME _ V. (1862-1866, containing 67 Plates). . ,, DA de oe oes ene Ore VOLUME VI. (1866-1869, containing 92 Plates). . ,, 11 5 0 ...15 0 0 VOLUME VII. (1869-1872, containing 73 Plates). . ,, 10 4 0... .18 12.0 VOLUME VIII. (1872-1874, containing 82 Plates). . ,, OS Sa eee Leuk nO) VOLUME = IX. (1875-1877, containing 99 Plates). . ,, 12 16. ..16 2 O VOLUME X. (1877-1879, containing 95 Plates) ere degyic ol Ol Oi esos teeitc Oten 7. GENERAL INDEX, Vols. I. to X. (1833-1879) ._,, Os Oe Onda 0 VOLUME XI. (1880-1885, contaming 97 Plates) . ,, Oyler Os tes. oe eo ‘ VOLUME XII. (1886-1890, containing 65 Plates). . ,, Bese Oe raeaheng surh< te.) VOLUME XIII, verre. containing 62 Plates). . ,, 6 8 3... 811 0 VOLUME XIV. (1896-1898, containing 47 Plates) . . ,, beupe Ouse oe i 0) oO) VOLUME XY, (1898-1901, containing 52 Plates). .,, Bype ean re ee 0) * No copies of these volumes remain in stock. Continued on page 3 of Wrapper. fea IV. On New or Rare Crustacea of the Order Cumacea from the Collection of the Copenhagen Museum.—Part IJ. The Families Nannastacide and Diastylide. By W. 7. Cauman, D.Sc., F.Z.8., British Museum (Natural History) *. (Received November 15, 1909; read January 18, 1910.) [Puatrs XXXIT.-XXXVII. | ly this, the second and concluding part ‘f of a report on Cumacea from the Zoological Museum of Copenhagen, twenty-nine species are dealt with, all except two of which are regarded as new, and three new genera are established. As in the case of the specimens dealt with in Part I., the authorities of the Copenhagen Museum have courteously allowed a selection of those here described. to be retained for the British Museum. List of the Species dealt with in this Paper. Family Nannastacips£. . Cumella forficula, sp. n. Gulf of Siam. » Clavicauda, sp. n. West Indies. » leptopus, sp. n. op », hispida, sp. n. Gulf of Siam. SS SCENALA. Sp.) 1. West Indies. », levis, sp. D. Gulf of Siam. Nannastacus lepturus, sp. 0. Suez. So zimmer, sp. 0. Ceylon. i gibbosus, sp. 0. - Gulf of Siam. 90 reptans, sp. 0. Hiipdiiss 35 minor, Sp. De 35 i tardus, sp. 0. Ke 99 agnatus, sp. 0. BD Schizotrema depressum, gen. et sp. n. yy = bifrons, sp. 0. India. 3 sordidum, sp. n. Gulf of Siam. Campylaspis platyuropus, sp. n. a i orientalis, sp. 0. Korea. * Published by permission of the Trustees of the British Museum. + For Part I., see Trans. Zool. Soc. xviii. pt. i. No. 1 (1907). (Se) > VOL. XvII.— PART lv. No. 1.—February, 1911. (Su) ne bo DR. W. T. CALMAN ON NEW OR RARE Family Diastyiipa. Gynodiasiylis carinata, gen. et sp.n. New Zealand. 35 levis, sp. n. My os costata, sp. 0. Gulf of Siam. a, bicristata, sp. n. he and Japan. Paradiastylis longipes Calm. 99 Colurostylis pseudocuma, gen. et sp.n. New Zealand. Diastylopsis elongata, sp. u. 2 a crasstor, sp. Ni. BA Diastylis koreana, sp. n. Korea. i fistularis, sp. n. Gulf of Siam. 35 insularum Calm. New Zealand. It is not pretended that this list, with that published in the first part of the paper, accounts for every specimen in the collections submitted to me. A certain number remain over which may represent undescribed species, but they are solitary, immature, or mutilated specimens, undistinguished by any striking features, and descriptions of them, necessarily incomplete, would only add to the perplexities of future students of the group. In addition, there are a considerable number of male specimens of the genera Cumella and Nannastacus which, as explained below, have so far defied my efforts to reduce them to systematic order. In working through this very extensive collection, together with other collections from tropical and southern seas reported on elsewhere, the infrequency with which known species have been recognized has been very striking. We seem to be dealing only with random samples from a very rich and varied fauna, and there can be little doubt that the vast majority of existing species of Cumacea still remain to be discovered. While the group is far more abundantly represented in tropical seas than had been supposed, much more extensive collecting will be necessary before it is possible to compare the number of species with those found in boreal and arctic waters ; certainly the warmer seas have nowhere been found, as yet, to yield the enormous numbers of individuals occurring, for instance, in favourable localities on the Norwegian coast ; and, as in many other groups of animals, the tropical species are far inferior in size to those from the colder regions of the north. The manifest imperfections of our knowledge make it useless to attempt at present any general discussion of the geographical distribution of the Cumacea. Provisionally, however, it may be said that the collections hitherto examined show a remarkable preponderance, in the warmer seas, of the Bodotriide (especially the genus Cyclaspis) and Nannastacide (especially Cumelia and Nannastacus) and a paucity of Diastylide and Leuconide. Further south, in New Zealand, as in the Magellan and other sub- antarctic regions, and in the Antarctic Ocean, the Diastylide and Leuconide reassert themselves. CUMACEA FROM THE COPENHAGEN MUSEUM. 343 From the point of view of the general taxonomy and phylogeny of the Crustacea the results of recent additions to our knowledge of the Cumacea are disappointing. Not one of the new-genera which have been described perceptibly extends the group towards any of the adjacent Orders of the Malacostraca. ‘The separation of the branchial orifices in Zygosiphon and Schizotrema might perhaps be thought to take us a little way back in the direction of the Mysidacea, but it is more likely to be a secondary modification than a primitive character; and although Ceratocuma in its free telson and full series of pleopods unites two characters, presumably primitive, not elsewhere found in combination, it is in all other respects a specialized type. ‘lhe Cumacea still remain a sharply circumscribed group, and although it is as certain as anything of the sort can well be that they have been derived from a Mysidacean-like ancestral form and that their line of descent travelled for some little way along with that of the Tanaidacea and Isopoda (but not, probably, with that of the Amphipoda), none of the intermediate links appear to have survived. It was to be expected that the classification of the Cumacea, established by Sars almost entirely from a study of northern species, would require modification as a result of the discovery of so many new species from other seas. As usually happens in such cases, the diagnoses of existing families have to be greatly modified if they are to admit the new forms, and the limits between the families become harder and harder to recognize. I have elsewhere proposed the union of the Campylaspide with the Nannastacide, and have commented on the increasing difficulty of keeping the Vauntompsoniide apart from the Bodotriide and the Diastylide from the Lampropide. The new genus Colwrostylis described below similarly helps to bridge the gap between the Diastylidee and Pseudocumide. In many cases characters hitherto relied on for the separation of families—such as the number of thoracic exopods or of pleopods— have proved to be of not more than generic value, and in the case of some of the species of Nannastacus described below I have been led to suspect that the number of thoracic exopods may not be even a specific character. Whether it will be possible to group the genera of the order in a smaller number of natural and sharply defined families is a problem which must be left for the future. It is at least as likely that it will be necessary, as it has been in the case of the Gammaridean Amphipods, to establish a much larger number of families and to be content to define these by more trivial characters. At the present moment, however, any attempt at a re-classification could only be of the most provisional nature, and I have therefore preserved, where possible, the existing families, with such extensions as are necessary to admit the new forms. In the descriptions which follow, the measurement of “total length” includes the telson (when distinct) but not the uropods. As it is often impossible to straighten out the abdomen without the risk of damage to the specimen, it is difficult to obtain accurate measurements, and the figures given must be regarded only as approximate. 3A 2 o44 DR. W. T. CALMAN ON NEW OR RARE Family NANNASTACID/. Genus CUMELLA. In addition to the specimens described below there are in the collection a number of males of this genus, collected, for the most part, by Dr. Mortensen in the Gulf of Siam, which I am unable to identify with any described species or to allot to any of the females found with them. As in the case of the genus Nannastacus, I do not attempt to describe these males. Key to the Species of Cumella (excluding C. molossa Zimmer, not yet figured, but described as closely resembling C. gracillima). A. Last somite inflated, longer than the preceding. a. Legs of last pair about as long as carapace, carpus more than twice as long) as) merus);)pseudorostrum! long, alae) Niue i) ae en Canfontiuculanispenmes b. Legs of last pair two-thirds as long as carapace, carpus less than twice as long as merus; pseudorostrum very short . . . . . C. clavicauda, sp. un. B. Last somite not inflated, not,longer than the preceding. a. Peduncle of uropods more than twice as long as last somite; carpus of last leg more than three times as long as merus. A, Carapace one-fourth of total length; pseudorostrum well- developed; eye absent ; antennal flagellum of male not longer (HORNA WOOKERS Gb a) B16 6 oo oo) op Jo) 6 ob co a. Go grncillomn Calman, B. Carapace one-third of total length ; pseudorostrum very short ; eye present; antennal flagellum of male nearly equal to length CHE OWA vo ora aca ae G ka Delco oo Geog Ol 0.) 6 lCy MammyAOS, Bio, b. Peduncle of uropods less than twice as long as last somite. _ A. Last somite about equal in length to the preceding; carpus of last leg three times as long as merus; peduncle of uropods distinctly shorter) than last\somite" “2.7. 3) Ch hispida; sp. ne B. Last somite distinctly shorter than preceding; carpus of last ; leg less than three times as long as merus; peduncle of uropods about equal to or longer than last somite. a. Carpus of last leg more than twice as long as merus. . . . C. serrata, sp. n. 6. Carpus of last leg less than twice as long as merus. a’, Antero-lateral margin of carapace deeply concaye, the lower angle prominent. 1. Third free somite with double dorsal tooth (female) . . C. australis Calman. 2. Third free somite without dorsal tooth. . . . . . . C.pygmea G. O. Sars. b’. Antero-lateral margin of carapace straight, the lower angle with a small tooth. 1p evamikotmnopodsis we quallewan a ache itan nei veermne CallcurS is pants 2. Exopodite of uropods shorter than endopodite . . . . C. limicola G. O. Sars. or CUMACEA FROM THE COPENHAGEN MUSEUM. Jt CUMELLA FORFICULA, sp. n. (Plate XXXII. figs. 1-6.) Description of adult Female.—Total length 1-7 mm. Carapace a little less than one-third of total length, compressed, its depth more than half of its length. The dorsal edge is sinuous, concave in its posterior part, convex anteriorly, and with a small convexity just in front of the hinder margin ; commonly, each of these conyexities bears a curved spine, but sometimes only the anterior spine is present. The ocular lobe is prominent on the dorsal surface and the eye is well pigmented. The pseudorostrum is long, directed slightly upwards, and acutely pointed ; the lateral plates meet in front of the ocular lobe for a distance greater than the diameter of the lobe. The straight antero-lateral margin slopes backwards and the obtuse antero-lateral angle bears a single minute tooth. ‘The surface of the carapace, as of the rest of the body, bears a number of scattered hairs. The abdomen is but little shorter than the cephalothoracic region. ‘The last somite is of peculiar form ; it is longer than the preceding somite and, as seen from the side, it is inflated, with arched upper and under surfaces; seen from above, it increases in width posteriorly and is produced as a rounded lobe for some distance between the bases of the uropods. The antennules are slender, resembling in proportions those of C. limicola, but the unsegmented inner flagellum is still more minute. The mouth-parts appear to differ only in trivial details from those of the type of the genus. The first legs resemble those of C. limicola, but have the basis relatively a little shorter, less than two-thirds as long as the remaining segments. In the second legs the basis is a little less than two-thirds as long as the remaining segments, while that of C. limicola is distinctly more. The remaining legs are much more slender and have the carpus more elongated than in C. limicola. ‘Yhose of the last pair are about as long as the carapace and have the carpus more than twice as long as the merus and four-fifths as long as the basis. The uropods are very short, their greatest length to the tip of the terminal spines being less than that of the last somite. The peduncle is less than half as long as the last somite, tapers a little distally, and is hardly more than twice as long as its greatest width. The rami, especially the endopodites, are curved inwards like the tail-forceps of an earwig. The terminal spines are indistinctly defined from the rami, but to the tip of its spine the endopodite is a little longer than and the exopodite equal to the peduncle. There are two spinules on the inner edge of the endopodite and a few sete on that of the peduncle. Adult Male.—Total length 1°85 mm. The body is a little more slender and elongated than in the female, the depth of the carapace being less than half its length. ‘There are no spines on the carapace and the inequality of the dorsal surface is less marked. ‘The eye is very large, with large and 346 DR. W. T. CALMAN ON NEW OR RARE prominent corneal lenses. ‘The pseudorostrum is shorter than in the female and its upper edge is concave. The antero-lateral margin is very short and the angle has no tooth. The last somite of the abdomen is inflated as in the female. The antennal flagella are nearly or quite as long as the body and the lateral grooves are well-marked on all the abdominal somites except the last. The proportions of most of the appendages are much as in the female, but the uropods are longer ; their length to the tip of the terminal spines is greater than that of the last somite. The peduncle is more than half as long as the last somite and nearly three times as long as its greatest width. ‘There are four spinules on the inner edge of the endopodite and five spinules interspersed with fine sete on that of the peduncle. Occurrence.—Gulf of Siam, “‘ Koh Kam, 5-10 fathoms, 4-6/2/00,” “ Between Koh Mesan and Cape Liant, 5-8 fathoms, 7/2/00,” “‘ North side of Koh Chang, 1 fathom, 3/3/00.” Th. Mortensen Coll., Copenhagen Museum. Co-types in British Museum. CUMELLA CLAVICAUDA, sp. n. (Plate XXXII. figs. 7-10.) Description of adult Male.—TYotal length 1-9 mm. Carapace less than one-third of total length, its depth more than half of its length. Dorsal edge slightly sinuous, without spines. The eye is very large, with about nine large corneal lenses, of which one is in the middle line in front; in all the specimens examined the eye-pigment is aggregated into a pair of small lateral spots, so that under a low magnification the eye appears to be paired and the animal is apt to be mistaken for a species of .Vannastacus. The pseudorostrum is very short, the lateral plates hardly meeting, if at all, in front of the ocular lobe and forming a pair of small tooth-like prominences when the carapace is viewed from above. ‘The antero-lateral margin is short and nearly straight and the angle is rounded off. As in C. forficula, the last abdominal somite is longer than the preceding and inflated as seen from the side; seen from above it increases in width posteriorly, but the hinder edge is broadly convex and hardly produced between the bases of the uropods. ‘The lateral groove is distinct only on the first four abdominal somites and the antennal flagellum is of a corresponding length. The antennules have the peduncle shorter and stouter than in C. forficula, the inner flagellum minute and unsegmented. The basis of the first legs is much more than two-thirds as long as the remaining segments. The posterior legs are relatively short; those of the last pair are little more than two-thirds as long as the carapace and have the carpus about one and a half times as long as the merus and little more than half as long as the basis. The uropods are similar to those of C. forficula. The greatest length to the tip of the terminal spine is about one-third more than that of the last somite. The peduncle CUMACEA FROM THE COPENHAGEN MUSEUM. 347 is about two-thirds as long as the last somite and not much more than twice as long as its greatest width. The endopodite is not quite so much curved as in C. forficula and the exopodite is nearly straight. ‘There are four spinules on the inner edge of the endopodite, and about six, interspersed with fine sete, on that of the peduncle. Occurrence—* Cruz Bay, St. Jan” (Danish West Indies). Chr. Levinsen Coll., Copenhagen Museum. Co-types in British Museum. CUMELLA LEPTOPUS, sp.n. (Plate XXXII. figs. 11-14.) Description of adult Male.—Total length 2°2 mm. Carapace about one-third of total length, its depth a little more than half of its length. Dorsal edge very slightly sinuous, without spines. The eye is very large, with about seven large and prominent corneal lenses, of which one is in the middle line in front of the ocular lobe; in all the specimens examined the eyes are without conspicuous pigment. ‘The pseudorostrum is practically obsolete, the lateral plates apparently not meeting in front of the ocular lobe. ‘The antero-lateral margin is concave, and the angle, which is slightly prominent, is rounded. The last somite of the abdomen is less than two-thirds as long as the preceding, and is produced and pointed posteriorly. ‘The lateral grooves are distinct on all the abdominal somites except the last and the antennal flagellum is about as long as the body. The antennules have the peduncle rather slender, the last segment being about three times as long as thick, and the inner flagellum minute and unsegmented. The first legs have the basis about equal in length to the remaining segments and the propodus two and a half times as long as the dactylus. The second legs have the distal segments slender, and the carpus more than twice as long as the merus. The posterior legs are very slender. The last pair are a little shorter than the carapace and have the basis about two-thirds as long as the remaining segments; the carpus is about three-fourths as long as the basis and four times as long as the merus. The uropods are long and slender ; the peduncle is more than twice as long as the last somite and has about nine spines on its inner edge; the endopodite is less than half the length of the peduncle, with a long slender apical spine and six spines on the inner edge; the exopodite is about five-sixths as long as the endopodite and its terminal spine is very long and slender. Occurrence.—‘ Cruz Bay, St. Jan” (Danish West Indies). Chr. Levinsen Coll., Copenhagen Museum. Co-type in British Museum. CUMELLA HISPIDA, sp.n. (Plate XXXII. figs. 15-18.) Description of adult Female.—Total length 2°55 mm. Carapace a little less than one-third of total length, its depth about half its length. Dorsal edge slightly arched, with a shallow depression in front of the hinder margin 348 DR. W. T. CALMAN ON NEW OR RARE and two teeth in front of the middle. The ocular lobe is prominent on the dorsal” surface and the eve is pigmented. The pseudorostrum is long, directed obliquely upwards, and acutely pointed; the lateral plates meet in front of the ocular lobe for a distance much greater than the diameter of the lobe. The antero-lateral margin is concave and the lower corner is nearly rectangular. The surface of the carapace and of the rest of the body bears rather numerous scattered hairs. The abdomen is a little shorter than the cephalothorax. ‘The last somite is about equal in length to the preceding, not inflated ; seen from above it increases only slightly in width posteriorly, and is produced as a broad rounded lobe for some distance between the bases of the uropods. The antennules have the peduncle rather slender, the last segment being more than three times as long as thick, and the inner flagellum minute and unsegmented. The first legs have the basis a little less than two-thirds as long as the remaining segments; that of the second legs is about two-thirds as long as the remaining segments. The remaining legs are slender; the last pair are a little shorter than the carapace and have the carpus three times as long as the merus and four-fifths as long as the basis. The uropods have the peduncle about three-fourths as long as the last somite. The straight endopodite, together with the terminal spine which is indistinctly marked oft trom it, measures a little more than the length of the peduncle and is longer by one-fourth than the exopodite. There are two spinules on the inner edge of the endopodite and two on that of the peduncle. Occurrence.—Gulf of Siam, “Koh Kam, 5 fathoms, 6/2/00.” Th. Mortensen Coll., Copenhagen Museum. Co-types in British Museum. CUMELLA SERRATA, sp. n. (Plate XXXIL figs. 19-24.) Description of adult Female.—Total length 2:2 mm. Carapace less than one-third of total length, compressed, its depth nearly two-thirds of its length. The dorsal edge is slightly convex, with about six curved teeth, of which one is set on a prominence just behind a shallow depression and in front of the hind margin. The ocular lobe is prominent on the dorsal surface and is provided with distinct lenses, but without pigment. The pseudorostrum is long, directed obliquely upwards, and acutely pointed; the lateral plates meet in front of the ocular lobe for a distance greater than the diameter of the lobe. The antero-lateral margin is concave and the angle bears an acute tooth. The surface of the carapace and of the rest of the body bears only inconspicuous scattered hairs, The abdomen is equal in length to the cephalothoracic region. ‘The last somite is shorter than the preceding, not inflated; seen from above it increases slightly in width CUMACEA FROM THE COPENHAGEN MUSEUM. 349 posteriorly and projects as a rounded lobe for some distance between the bases of the uropods. The antennules have the last segment of the peduncle about three times as long as thick and a little shorter than the preceding. The first legs have the basis little more than half as long as the remaining segments ; that of the second legs is little shorter than the remaining segments. The posterior legs are rather slender; the last pair are a little shorter than the carapace and have the carpus more than twice as long as the merus and nearly as long as the basis. i The uropods have the peduncle longer by about one-fourth than the last somite. The endopodite without its stout terminal spine is about two-thirds as long as the peduncle and more than half as long again as the exopodite. ‘There are three spinules on the inner edge of the endopodite and four on that of the peduncle. Adult Male.—Total length 2:3 mm. Carapace about one-third of total length, its depth about half its length. The dorsal edge is nearly straight, without teeth. ‘The ocular lobe bears a number of large and prominent lenses, but is without pigment. ‘The pseudorostrum is short, obliquely truncated, with the upper edge slightly concave ; the lateral plates meet in front of the ocular lobe for a distance less than the diameter of the lobe. ‘The antero-lateral margin is oblique and the angle is rounded. ‘The surface of the body is without conspicuous hairs. The abdomen is shorter than the cephalothoracic region. ‘The last somite is shorter than the preceding, not inflated; seen from above it is more than two-thirds as broad as long, increasing slightly in width posteriorly and projecting as a rounded lobe for some distance between the bases of the uropods. ‘The lateral antennal grooves are distinct on all the somites except the last. The antennules have the last segment of the peduncle about three times as long as thick and distinctly shorter than the preceding. The flagella of the antenne are about as long as the body. The proportions of the legs are much as in the female. ‘The last pair are distinctly shorter than the carapace and have the carpus two and a half times as long as the merus and nearly as long as the basis. The uropods have the peduncle longer by one-half than the Jast somite. The endopodite, without its terminal spine, is less than two-thirds as long as the peduncle and distinctly more than half as long again as the exopodite. ‘There are six spinules on the inner edge of the endopodite and the same number on that of the peduncle. Occurrence.—“ Cruz Bay, St. Jan (Danish West Indies), 6/96.” Chr. Levinsen Coll., Copenhagen Museum. Co-types in British Museum. (3s) & VOL. Xvili.—ParT tv. No. 2.-—Mebruary, 1911. 300 DR. W. T. CALMAN ON NEW OR RARE CUMELLA L&VIS, sp.n. (Plate XXXII. figs. 25-27.) Description of adult Female.—Total length 1:15 mm. Carapace a little more than one-third of total length, its depth more than half its length. Dorsal edge horizontal, slightly sinuous, without teeth. The ocular lobe is prominent on the dorsal surface and the eye is pigmented. ‘The pseudorostrum is very short and truncated, the lateral plates meeting in front of the ocular lobe for a distance less than the diameter of the lobe. The straight antero-lateral margin slopes backward and the angle is obtuse and rounded. The surface of the carapace and of the rest of the body is without conspicuous hairs. The abdomen is shorter than the cephalothoracic region. The last somite is less than two-thirds as long as the preceding, not inflated; seen from above it is little longer than broad and its broadly rounded posterior lobe projects only for a short distance between the bases of the uropods, The antennules have the peduncle rather stout, the last segment being about two and a half times as long as wide and shorter than the preceding segment; the inner flagellum is minute and unsegmented. The first legs have the basis about half as long as the remaining segments ; that of the second legs is more than two-thirds as long as the remaining segments. The remaining legs are short and stout; the last pair are little more than two-thirds as long as the carapace and have the carpus about one-half longer than the merus and little more than half as long as the basis. The uropods have the peduncle nearly twice as long as the last somite. The sub- equal rami are more than two-thirds as long as the peduncle and each bears a slender terminal spine sharply marked off from it. There are four spines increasing in length distally on the inner edge of the endopodite. Occurrence.—Gulf of Siam, “Koh Kam, 1-5 fathoms, 1/1/00, 6/2/00, ” < Between Koh Mesan and Cape Liant, 5-8 fathoms, 7/2/00.” Th. Mortensen Coll., Copenhagen Museum. Co-types in British Museum. Genus NANNASTACUS. The discrimination of species in this genus presents difficulties which I have not been able altogether to overcome. The collection of the Copenhagen Museum comprises a very large number of specimens of Nannastacus, most of which were collected in the Gulf of Siam by Dr. Th. Mortensen. The great majority of these are males resembling in general characters the male of NV. zimmeri described below, but differing among themselves in size, in the proportions of the uropods and legs, in the relative size of the eyes, and in other small characters. Most of these characters are subject to individual variation, and after dissecting, drawing, and comparing a number of these males I have reluctantly come to the conclusion that it is impossible, at present, to group them under specific names or to pair them with the different forms of females. CUMACEA FROM THE COPENHAGEN MUSEUM. 301 In the case of WV. zimmeri, the single female specimen was found in a gathering con- taining numerous specimens of the male and no other Cumacea, so that there is every probability that the two sexes do belong to the same species. In the richer gatherings from the Gulf of Siam, however, two or three closely allied types of female frequently occur in the same bottle with numerous males which cannot be satisfactorily dis- tinguished from one another. The new species of which females are described below (NV. zimmeri, N. gibbosus, N. reptans, N. minor, N. tardus, and N. agnatus) differ from the type of the genus (N. wnguiculatus) in having no exopod on the third maxilliped in that sex, and I find that the female type specimen of NV. suhmii G. O. Sars agrees with them in this respect. The males of NV. swhmii and NV. zimmeri, however, have a well-developed exopod on that appendage, and it may fairly be assumed that the males of the other species will be found to have it also. ‘The absence of this exopod in the female sex is already known in several species of Diastylide, and in that family it has been regarded as a distinction of generic value. It might have been so regarded in this case also, were it not for a further peculiarity presented by WV. reptans and N. tardus. ‘The females described under these names differ from all Cumacea hitherto known in having no thoracic exopods at all; they have lost not only that of the third maxilliped, but also those of the first and second legs; and, curiously enough, each of them was accom- panied in nearly all the gatherings in which they occurred by another form hardly differing from the first in any character of importance except in possessing exopods on these legs. Both forms are represented by mature ovigerous females, and I have thought it advisable to describe them under distinct specific names, but it must be admitted as quite possible that NW. reptans may be merely an individual variation or a phase in the life-history of NV. minor, and that NV. tardus may stand in the same relation to WN. agnatus. At all events, the two pairs of species, if they are specifically distinct, can hardly be generically separated ; and since the absence of exopods from the first and second legs seems to be a distinction at least as important as the absence of an exopod from the third maxilliped, there is no good reason for giving generic value to the latter character and not to the former. I formerly gave as one of the distinguishing characters of the genus Nannastacus the absence of a distinct ischium in the second maxillipeds (Fisheries, Ireland, Sci. Invest. 1904, 1. (1905) p. 25). In some of the species described below I find indications of a separate ischium in this appendage, butas the character is one which it is difficult to verify without undue sacrifice of material I should prefer to omit it from the generic definition. As many of the species are known only in one sex it is not possible to give a satisfactory key to their arrangement. I have attempted, however, in the remarks appended to the description of each new species to indicate as far as possible its systematic relations. 3B2 302 DR. W. T. CALMAN ON NEW OR RARE NANNASTACUS LEPTURUS, sp. n. (Plate XX XIII. figs. 1-3.) Description of adult Male.—Total length 2:0 mm. Carapace a little more than one-third of total length, slightly compressed, dorsal margin slightly arched. ‘The pseudorostrum, seen from the side, is short and truncate. Antero-lateral margin concave, forming a widely open antennal notch ; antero-lateral angle rounded, hardly produced. Eyes prominent, darkly pigmented, set rather close together, the distance between them not more than one-eighth of the length of the e carapace. The first leg-bearing somite is distinct. The pleural plates of the succeeding thoracic somites are hardly expanded laterally. The abdomen is nearly equal to the cephalothoracic region. The somites are cylindrical, each with a well-marked lateral groove; the fifth somite is one and a half times as long as the fourth and twice as long as the last, which is truncate posteriorly. The peduncle of the antennule has the three segments (as seen without dissection) subequal, the second with a short distal process on the upper side, not reaching beyond the end of the segment ; both flagella have two segments, and the inner flagellum is a little shorter than the first segment of the outer. The first legs have a slender dactylus, about two-thirds as long as the propodus. in the last pair of legs the basis is little more than one-third of the length of the limb, the carpus is less than twice as long as the merus, and one-third longer than the propodus, which again is about equal to the dactylus with its claw. The peduncle of the uropods is more than twice as long as the last somite, slender, with five spinules on the finely serrated inner edge. ‘The endopod is about half as long as the peduncle, with a terminal spine of less than half its length and seven spines, increasing in length distally, on the inner edge. The exopod is about four-fifths of ne jength of the endopod and bears a slender terminal spine of its own length. Remarks.—N. longirostris G. O. Sars, N. brachydactylus Calman, and the species described above form a group defined from the other species of the genus by having the peduncle of the uropods at least twice as long as the last somite. WV. hirsutus Hansen, which has the peduncle only about equal to the last somite, agrees with the first two of these species in the length of the pseudorostrum, of which the lateral plates meet above for at least one-sixth of the total length of the carapace. From all these species WV. lepturus is distinguished by the very short pseudorostrum and by having the abdomen nearly equal to the cephalothoracic region. Occurrence.—* Suez, 3/1/98, H. Mortensen.” Copenhagen Museum. NANNASTACUS ZIMMERI, sp. n. (Plate XX XIII. figs. 4-15.) Description of adult Female.—Total length 1-5 mm. Carapace a little more than one-third of total length, inflated behind, where the CUMACEA FROM THE COPENHAGEN MUSEUM. 309 width is greater than the depth, much narrowed in front. Seen from the side the dorsal edge is convex, with a slight elevation at the posterior margin bearing a small tubercle ; the pseudorostrum is short, directed obliquely upwards, narrowly truncate ; antero-lateral margin deeply concave, antero-lateral angle produced into a blunt spine with a few sete at the apex. Seen from above, the lateral plates of the pseudo- rostrum meet below but not above the respiratory channel, so that the opening of the latter extends back to the interocular margin; only the tips of the antero-lateral angles project laterally. The eyes are small and darkly pigmented, and the distance between them is about one-fourth of the length of the carapace. ‘The surface of the carapace is smooth, with very minute scattered sete, and there is a slight depression on each side behind the eye. The first leg-bearing somite is reduced to a very narrow strip behind the carapace, expanding below into the pleural plates, which, like those of the following somites, are spread out laterally. The abdomen is about equal in length to the cephalothoracic region, slender and cylindrical, without teeth or spines; the fifth somite is about one-third longer than the preceding somite. The third maxillipeds have no exopods; the basis is a little more than one-third of the total length. ‘The first and second legs carry exopods. The remaining legs are slender and do not differ greatly in length. ‘The proportions of the last pair are much as in the male. The uropods have the peduncle about half as long as the last somite. The endopod is about three times as long as the peduncle, with a stout terminal spine of less than half its length and two spinules on the inner edge, which is finely serrated throughout. The exopod is less than one-eighth of the length of the endopod, with a slender terminal spine which reaches beyond the middle of the latter. There is an irregular patch of dark pigment on each side of the anterior part of the carapace and a band encircling the penultimate somite of the abdomen. Adult Male.—Yotal length 1:6 mm. Carapace a little more than one-third of total length, not so deep as in the female, not much narrowed in front, its width greater than its depth. Seen from the side the dorsal edge is nearly straight, with a median depression between the inflated branchial regions; the pseudorostrum is very short and blunt; antero-lateral margin deeply concave, antero-lateral angle rounded and armed with a curved spine, behind which are a few irregular serrations. Seen from above, the lateral plates of the pseudorostrum meet for a short distance above the respiratory channel and the antero-lateral angles project laterally. The eyes are large and darkly pigmented, and the distance between them is not more than one-sixth of the length of the carapace. ‘The surface of the carapace is very minutely granular. The first leg-bearing somite is hardly visible and its pleural plates are very small. 304 DR. W. T. CALMAN ON NEW OR RARE ‘The pleural plates of the succeeding somites are not so much expanded laterally as in the female. The abdomen is about equal in length to the cephalothoracic region, without teeth or spines on the dorsal surface, but with a lateral groove overhung by a serrated crest on each of the somites; the fifth somite is at least one-half longer than the fourth. The antennule has a rather slender peduncle, the first segment about as long as the other two together; the second segment is produced distally into a narrow process bearing a group of plumose setze ; the inner flagellum is unsegmented and equal in length to the first of the two segments of the outer. The flagellum of the antenna extends, in the natural position, at least as far as the penultimate somite of the abdomen. The mouth-parts closely resemble those of NV. wnguiculatus, but the palp of the maxillula is relatively a little longer. The third maxilliped has a well-developed exopod and the basis is about half of the total length. The first legs have the basis little more than half the combined length of the remaining segments; there are no teeth on the outer margin of the ischium and merus. The legs of the last pair are less than one-third of the length of the body; the basis is about one-fourth of the length of the limb, the carpus nearly four times as long as the merus and nearly one and a half times as long as the propodus; the dactylus with its claw is a little longer than the carpus. The uropods have the peduncle about two-thirds as long as the last somite. The endopod is about three times as long as the peduncle, with a terminal spine of one- third of its length and five spinules on the distal part of the finely serrated inner edge. The exopod is about one-tenth of the length of the endopod and its terminal spine reaches beyond the middle of the latter. As in the female, there are some irregular patches of pigment on the side of the carapace and a band encircling the penultimate somite and there are also less distinct and less constant bands on the third and fourth abdominal somites. Remarks.—Vhis species is closely allied to WV. suhmii G. O. Sars (‘ Challenger’ Rep. Cumacea (1887), p. 63, pl. x. figs. 4-5). As I have elsewhere mentioned (Herdman’s Rep. Ceylon Pearl Fisheries, Royal Soc. pt. ii. (1904) p. 177), the type specimens of that species are now more accessible for examination than when they were described by Prof. Sars, although they are by no means in a good state of preservation. In the female specimen I find no trace of an exopod on the third maxilliped, and I believe that the species agrees with WV. zimmeri in this character, Sars’s mention of a small exopod on ‘“¢ the second pair of gnathopoda ” being based on some misinterpretation of the mounted specimen. ‘The female WV. swhmii, however, differs from the species described above in the form of the carapace, which, as shown in Sars’s figure, has the branchial regions very strongly inflated and the posterior part of the dorsal surface much more prominent ; CUMACEA FROM THE COPENHAGEN MUSEUM. 305 there seems to have been a median spine or tooth a little in front of the middle of the dorsal surface ; the anterior abdominal somites have strong paired teeth on the dorsal side; and the exopod of the uropods, with its spine, is not more than one-third of the length of the endopod. The males of the two species are much more alike, but in N. suhmii the exopod of the uropods is much shorter than in the species described above, being even shorter, relatively, than in the female (the proportion of one-fourth of the exopod, which I formerly mentioned for the males, is, however, slightly exceeded in some specimens). Occurrence.— Trincomali, Ceylon, surface, 3/1889, K. Fristedt Coll.” Copenhagen Museum. Co-types in British Museum. NANNASTACUS GIBBOSUS, sp. n. (Plate XXXIII. figs. 16-21.) Description of adult Female.—Total length 1-8 mm. Carapace two-fifths of total length, little broader than deep. The pseudorostrum is very short and truncate and the two parts meet together in the middle line below the respiratory orifice. The antero-lateral margin is deeply concave, the antero-lateral corner produced and triangular, sometimes with one or two teeth at the tip. The eyes are prominent and darkly pigmented. The surface of the carapace is very uneven, but the inequalities are developed in varying degrees in different specimens. The branchial regions are inflated to form rounded bosses, between which the dorsal surface is depressed in the middle line. Posteriorly, the dorsal surface is elevated more or less and there is a small boss on each side just behind the eyes. One or two spiniform teeth are commonly set on the posterior elevation of the dorsal surface and another in the middle a little way behind the eyes, but all of these may be absent. A number of long sete are scattered over the surface of the carapace. The first leg-bearing somite is represented only by the pleural plates. The remaining thoracic somites have the pleural plates expanded and rounded. The second and the last somites may have a spiniform tooth in the middle of the dorsal surface, as may also the first abdominal. The abdomen is a little shorter than the cephalothoracic region, the fifth somite is about twice as long as deep and rather more than one-third longer than the preceding somite. The antennule has the third segment of the peduncle longer than the second and, together with it, longer than the first; the outer flagellum has a minute terminal segment. The third maxillipeds have no exopods; the basis is about one-third of the total length of the limb. The first legs have well-developed exopods; the basis is about two-sevenths of the total length of the limb. The second legs have well-developed exopods; the basis is inflated and is about 306 DR. W. T. CALMAN ON NEW OR RARE three-fourths as long as the distal segments together; the distal segments are broad, the carpus is distinctly longer than the merus, and the dactylus is twice as long as the propodus. The remaining legs do not differ greatly in length; the last pair are about one-third of the total length of the body. The basis is a little over one-fourth of the length of the limb, the carpus is a little more than twice as long as the merus and about equal to the propodus. The uropods have the peduncle a little over three-fourths as long as the last somite, with two small teeth on the inner edge. ‘The endopod is nearly three times as long as the peduncle, with a stout terminal spine of about one-third of its length and with five small spines on its inner edge. ‘The exopod is less than one-eighth as long as the endopod, with a slender terminal spine which does not reach to the end of the latter. Remarks.—TVhe specimens described above resemble the female of Sars’s VV. suhmit even more closely than does that of NW. zimmeri, especially in the inflated branchial regions and the posterior gibbosity of the carapace. They differ from NV. swhmii in the shorter and more obtuse pseudorostrum, in the less prominent dorso-lateral teeth of the first abdominal somite and their absence from the following somites, in the shorter carpus of the posterior legs, and in the much longer exopod of the uropods. Occurrence.—Gulf of Siam, “Koh Chang, about 1 fathom, coral,” “ Koh Kam, 5 fathoms, mud.” Th. Mortensen Coll., Copenhagen Museum. Co-types in British Museum. NANNASTACUS REPTANS, sp. n. (Plate XX XIII. figs. 22-28.) Description of adult Female.—Votal length 15 mm. Carapace more than one-third of total length, little broader than deep. The pseudorostrum is very short and truncate, and the two parts meet together in the middle line below the respiratory orifice. ‘The antero-lateral margin is deeply concave, the antero-lateral corner produced and triangular. The eyes are prominent and darkly pigmented. The surface of the carapace is very uneven, being depressed in the middle line between the inflated branchial regions and with a median convexity anteriorly. A few long sete are set on the branchial regions and on other parts of the surface. The first leg-bearing somite appears to be represented only by the pleural plates. The remaining thoracic somites have the pleural plates expanded and rounded, without spines. The last thoracic and the first abdominal somites have each, on the dorsal surface, a pair of small spiniform teeth. The abdomen is nearly equal to the cephalothoracic region; the fifth somite is a little less than twice as long as deep and about one-third longer than the preceding somite. The antennule has the second and third segments of the peduncle subequal and CUMACEA FROM THE COPENHAGEN MUSEUM. 307 together about equal to the first segment. The outer flagellum has a minute terminal segment. The mandibles, maxillule, mavxille, and first maxillipeds appear to differ in no conspicuous characters from those of WV. wnguiculatus. The second maxillipeds have the basis only about one and a half times as long as broad; the basal plate carries seven sete. The third maxillipeds have no exopods; the basis is less than one-third of the total length of the limb; the distal segments are relatively very stout. The first legs also are without exopods. The basis is not more than one-quarter of the total length of the limb. The second legs have no exopods. The basis is not inflated and is little shorter than the remaining segments together. The distal segments are broad, the carpus is hardly longer than the merus, and the broad dactylus is not twice as long as the propodus. The remaining legs increase slightly in length to the last pair, which are about two- fifths of the total length of the body; the basis is a little over one-fourth of the length of the limb, the carpus is not three times (in the third pair it is less than twice) as long as the merus and a little shorter than the propodus. The uropods have the peduncle about two-thirds as long as the last somite, with two small teeth on the inner edge. The endopod is about twice as long as the peduncle, with a stout terminal spine of about half its length and with three minute spines and a small seta on the inner edge. ‘The exopod is less than one-fifth as long as the endopod, with a slender terminal spine which reaches nearly to the end of the latter. Remarks.—This species differs from that last-described in the form of the carapace (which has the branchial regions less inflated and the anterior part of the dorsal surface more strongly arched), in the greater length of the posterior legs, and in the different proportions of the uropods. It differs more remarkably in the absence of exopods from the legs of the first and second pairs, in which respect it resembles the species described below as WV. tardus. Occurrence ——Gulf of Siam, ‘“‘ Koh Kahdat, about 1 fathom, coral, i—1./00,” ‘‘ Koh Chang, 1 fathom, coral, i—iii./00.” Th. Mortensen Coll., Copenhagen Museum. Co-types in British Museum. NANNASTACUS MINOR, sp. n. (Plate XXXIV. figs. 1-3.) Description of adult Female ——Total length 1-3 mm. Carapace not quite two-fifths of total length, little broader than deep. Pseudo- rostrum very short and truncate, the two parts meeting in the middle line below the respiratory orifice. Antero-lateral margin deeply concave, the antero-lateral corner produced and triangular. Eyes prominent and darkly pigmented. Surface of carapace not very uneven, the branchial regions only slightly inflated, with a depression between VOL. XVIII.—PaART Iv. No. 3.—Lebruary, 1911. 3 ¢ 358 DR. W. T. CALMAN ON NEW OR RARE them in the mid-dorsal line. There are a few scattered sete on the surface of the carapace. The first leg-bearing somite is represented only by the pleural plates. The remaining thoracic somites have the pleural plates expanded and rounded, The abdomen is shorter than the cephalothoracic region ; the fifth abdominal somite is barely twice as long as deep, and less than one-third longer than the preceding somite. The antennule has the third segment of the peduncle shorter than the second and, together with it, a little shorter than the first. The third maxillipeds have no exopods ; the basis is less than one-third of the total length of the limb ; the distal segments, especially the carpus and propodus, are very stout and only slightly flattened. The first legs have well-developed exopods; the basis is about two-sevenths of the total length of the limb ; the distal segments, especially the carpus and propodus, are expanded and flattened. The second legs have well-developed exopods; the basis is inflated and is about equal to the remaining segments together ; the distal segments are broad, the carpus is distinctly longer than the merus, and the dactylus is only a little longer than the propodus. The remaining legs do not differ greatly in length. The last pair are about two- fifths of the total length of the body ; the basis is more than one-fourth of the length of the limb, the carpus is about twice as long as the merus and a little shorter than the propodus. The uropods have the peduncle nearly two-thirds as long as the last somite. The endopod is about twice as long as the peduncle, with a stout terminal spine of more than half its length, and with three small spines on its inner edge. The exopod is about one-fifth as long as the endopod, with a slender terminal spine which reaches nearly to the end of the latter. Remarks.—Apart from the possession of exopods on the first and second pairs of legs this species resembles the preceding somewhat closely, especially in the relative length and proportions of the posterior legs and of the uropods. ‘There are, however, small but constant differences in the form of the carapace, which has the branchial regions less inflated and the dorsal surface less arched; in the stoutness of the distal segments of the third maxillipeds and first pair of legs; and some small differences of proportion noted in the description above. It has already been mentioned as a possibility that this species may represent only a variation or a phase of development of VV. reptans, but I do not think that this is probable. Occurrence—Gulf of Siam, “Koh Kahdat, about 1 fathom, coral, 1/2/00,” “Koh Chang, 1 fathom, coral, 1/3/00.” Th. Mortensen Coll., Copenhagen Museum. Co-types in British Museum. CUMACEA FROM THE COPENHAGEN MUSEUM. 359 NANNASTACUS TARDUS, sp. n. (Plate XXXIV. figs. 4-11.) Description of adult Female.—Yotal length 1-6 mm. Carapace about three-sevenths of total length, considerably broader than deep. The pseudorostrum is very short and truncate, and the two parts meet together in the middle line below (and apparently also above) the respiratory orifice. The antero- lateral margin is deeply concave, the antero-lateral corner produced and triangular, extending in front of the pseudorostrum, the lower edge serrated anteriorly. ‘The surface of the carapace is not very uneven and the branchial regions are only slightly inflated. The first leg-bearing somite is represented only by the pleural plates, which, like those of the following somites, are greatly expanded laterally and bear a marginal row of laminar spines (partially detached in the specimen figured). The second leg- bearing somite has a small median tooth dorsally. The abdomen is shorter than the cephalothoracic region: the first somite has a pair of small dorso-lateral teeth, but these are often absent, or perhaps readily become detached ; the fifth somite is at least twice as long as deep and nearly one-half longer than the preceding somite. ‘The antennule has the second and third segments of the peduncle subequal and together about equal to the much stouter first segment. The third maxillipeds have no exopods. ‘The basis is less than one-third of the total length of the limb, and its distal process is very long and narrow. The first legs also have no exopods. ‘The basis is about two-sevenths of the total length of the limb, and has a prominent rounded lobe on its inner edge. The second legs are without exopods. ‘The basis is about three-fourths as long as the remaining segments together. ‘The carpus is much longer than the merus and the dactylus is not twice as long as the propodus. The remaining legs are subequal in length and less than one-third as long as the body. ‘The basis is about one-fourth of the length of the limb, the carpus is less than three times as long as the merus and about equal to the propodus. The uropods have the peduncle little more than half as long as the last somite, The endopod is a little more than twice as long as the peduncle, with a stout terminal spine of less than half its length and three small spines on the distal part of its inner edge. The exopod is about one-sixth as long as the endopod, and its terminal spine does not reach to the end of the latter. Remarks.—In the form of the carapace, with its strongly produced antero-lateral angles and slightly inflated branchial regions, this species approaches more closely to N. wnguiculatus than to any other species hitherto described. It differs from that species, however, in having no exopod on the third maxilliped, in the absence of spines from the carapace and from the dorsal surface of the posterior abdominal 8) 0) 7A 360 DR. W. T. CALMAN ON NEW OR RARE somites, the much feebler serrations on the lower margin of the carapace, the shorter posterior legs and exopod of the uropods, and in many other details. It also differs from all other Cumacea, with the exception of JV. reptans described above, in having no exopods on the first and second pairs of legs. Occurrence.—Gulf of Siam, “Koh Kam, 5-10 fathoms, 2/00,” “Koh Kahdat, 4—5 fathoms, 16/2/00,” “ Between Koh Mesan and Cape Liant, 5—8 fathoms, 8/2/00,” “North of Koh Mesan, 10-15 fathoms, 15/2/00.” Th. Mortensen Coll., Copenhagen Museum. Co-types in British Museum. NANNASTACUS AGNATUS, sp.n. (Plate XXXIV. figs. 12 & 13.) Description of adult Female.—Total length 1:3 mm. Resembling the specimens of WV. tardus very closely, but having well-developed exopods on the legs of the first and second pairs. ‘The basis of the first legs is a little shorter, but is shaped as in that species; the basis of the second legs is a little longer; the posterior legs appear to be slightly stouter and shorter. The last somite is distinctly shorter and the peduncle of the uropods is more than half its length. Remarks.—Apart from the presence of exopods on the first two pairs of legs, I have not been successful in finding any characters other than those mentioned above to differentiate this species from WV. tardus. The possibility suggests itself, therefore, even more strongly than in the analogous case of WV. reptans and N. minor, that we have to do here with a single species in which the exopods may be lost (1) by accident, (2) as a matter of individual variation, or (3) normally in the course of the breeding- period. The analogy of other Cumacea makes the first of these very improbable ; as regards the third it is to be noted that both forms are represented by ovigerous females, and that the specimens of J. tardus are on the whole, a little longer than those of WV. agnatus. I do not think that any importance can be attached to the fact that, except in one case, the two forms occurred together in the same gatherings, and in the absence of conclusive evidence that they belong to a single species, it seems convenient to assign specific names to both. Occurrence.—Gulf of Siam, “Koh Kam, 5-10 fathoms, 2/00,” “Koh Kahdat, 4-5 fathoms, 16/2/00,” ‘‘Between Koh Mesan and Cape Liant, 5-8 fathoms, 8/2/00,” “North of Koh Mesan, 10-15 fathoms, 2/00.” Th. Mortensen Coll., Copenhagen Museum. Co-types in British Museum. SCHIZOTREMA, gen. nov. Carapace more or less depressed, not overhanging the anterior free somites ; lateral plates of pseudorostrum not meeting in front of head, exhalent respiratory orifices paired and widely separated from each other; third maxillipeds and first two pairs of legs with exopods in female: other characters as in Nannastacus. CUMACEA FROM THE COPENHAGEN MUSEUM. 361 Type-species S. depressum, sp. n. This genus bears the same relation to Wannastacus that Zygosiphon bears to Cyclaspis. No males have yet been identified, but the analogy of Zygosiphon makes it probable that the division of the respiratory openings will be found in that sex also. In the type-species the second maxilliped has a distinct ischium, but for reasons already mentioned (p. 351) I prefer not to regard this character as of generic value. (ey to the Species of Schizotrema. A. Carapace very broad and depressed ; a conspicuous lateral series of spines on carapace, posterior thoracic somites, and abdomen; peduncle of uo podslonserthanwiastssOmi teen rit sila sei een ae cen eR Ce depressum, sp. D. B. Carapace less depressed; no conspicuous lateral spines on carapace or abdomen ; peduncle of uropod not more than half as long as last somite. a. Surface of body with numerous small spines and tubercles; antero- lateral angle of carapace produced as a stout cylindrical process . . S. bifrons, sp. n. 6. Surface of body for the most part smooth; antero-lateral angle produced asa slender tooth . . . ..... 2.2.4. =. =. . S. sordidum, Sp. 0. SCHIZOTREMA DEPRESSUM, sp. n. (Plate XXXIV. figs. 14-17.) Description of adult Female.—Total length 1-5 mm. Carapace a little less than one-third of total length, very broad and depressed. The pseudorostrum, seen from the side, is very short and truncate, directed obliquely upwards. Seen from above, the two parts of the pseudorostrum are widely separated, each completely surrounding the respiratory orifice, from which a long branchial siphon is protruded. ‘The antennal notch is obsolete, the antero-lateral margin descending nearly vertically to the rounded antero-lateral corner, which carries a stout curved spine.