19U4.] FROM PAST AFRICA AND ZANZIBAR. 385 



teeth in them varies greatly and apparently increases with the 

 size (that is probably the age) of the animals. 



As regards shape, the teeth present several types : — 



(1) They are simply biiid in ccerulea, semperi, nigrostriata, 

 inarenzelleri, crossei, thalassopora, camoena, lapinigensis. lu 

 these species tlie innermost tooth bears a single accessory denticle 

 on the inner side and hence appears trifid. 



(2) They are bifid with accessoi-y denticles below the bifurcation 

 in gracilis, messinensis, sycilla, hilaris, heiinetti, calijbrnensis, and 

 agassizi. In runcinata the accessory denticles are very small 

 and inconspicuous. 



(3) In SGurra the apex of the tooth is three or five times cleft, 

 and a somewhat similai- arrangement, by which the denticles are 

 all on the upper pai-t of the tooth, appears to occur in pustulans, 

 gloriosa, and riidolplu. 



(4) The commonest form of tooth is hamate, with several 

 (6-10) denticles on the outer side. These are generally fine and 

 minute, but in some forms {e.g., porcata, rosans, mariana, eliza- 

 bethina, jjaitpei-a, tryoni, and godeffroyana) are large and strong. 

 In this class of radula the innermost teeth have generally sevei'al 

 denticles on each side, but sometimes (e. g., mdrchii and gonato- 

 phora) only one on the inside. 



Of the species described below, the first four are closely related, 

 and form a group to which I am almost disposed to accord generic 

 rank. I have not done so out of deference to the high authority 

 of Prof. Bergh, as one of the animals appears to be identical with 

 the Goniohr-cmchus reticulcUus of Pease, examined by him and 

 referred to Chroonodoris. All four forms agree in being oval and 

 flat in shape, with a very ample mantle-margin. The consistency 

 is very soft, and the dorsal pattern composed of spots, not stripes. 

 The innermost teeth of the ro,dula bear a few denticles on both 

 sides; the rest 6-10 denticles on the outer side only. The 

 branchiae are numerous, and the sides of the mouth are connected 

 with the upper lamina of the grooved foot. This latter peculiarity 

 was noted in the living as well as in the alcoholic specimens ; but 

 I am not sure that it is of much morphological significance, for in 

 a very soft animal with ample flaccid integuments such folds may 

 easily be formed at the corners of the mouth without constituting 

 distinct structures. 



Nos. 12-15, described as hilaris, lineata, lineata var., and 

 ? magniUca, are very likely all varieties of one striped species, 

 with bifid teeth bearing accessory denticles, and about 1 branchiae, 

 but a larger number of living animals must be examined before 

 this identity can be established. 



All my species which are not new have been previously found 

 in the Indo- Pacific, which appeais to be the head -quarters of the 

 genus, though not enough is known of the Nudibranchs of the 

 tropical Atlantic to make any comparison as to numbers of 

 species. About half a dozen forms are reported from the West 

 Indies and the Gulf of Mexico. The genus is mainly tropical, and in 



[7] 



