286 CEYLON PEART- OYSTER REPORT. 



are subequal and without a special terminal hamulus distinct from the dentate portion 



(Plate V., fig. 108). The jaw-pieces of the second pair are the most difficult to 



make out satisfactorily ; one side, so far as I could ascertain, showed three large 



teeth, and a long edentulous shaft reaching back to the radices maxillarum (Plate V., 



fig. 110) ; on the other (probably the left) side the second jaw-piece, which was 



fractured near the apex, showed a strong apical tooth and a long posterior dentate 



portion (Plate V., fig. 109). There are, as usual, five pairs in all, the fifth being a 



simple hamulus. As stated above, the diagnostic character is the absence of a 



pronounced diastema between the apical tooth and the succeeding teeth of the first 



pair of jaws. 



Family: GLYCERID^E. 



Glycera lancadivae, Schmarda Plate V., figs. 113 to 116. 



Schmarda, op. cit., 1861, p. 95 ; Michaelsen, ' Jahrb. Hamburg. Wiss. Anst.,' ix., p. 102. 



Locality: Galle, 14th February, 1902. Two specimens. 



This species is a representative in these waters of the Atlantic Glycera capitata, 

 and is characterised by the absence of branchiae and by the biligulate anterior lip 

 of the parapodium, the posterior lip being rounded and slightly emarginate. It 

 appears to present a dimorphism analogous to that existing between G. capitata 

 and the variety G. setosa. In the typical form the parapodia are longer than in 

 the second form ; we may refer to these as A and B respectively. In both, the body 

 segments are biannulate, and there are two kinds of compound setae differing in the 

 structure of the articular end of the shaft ; in one kind the transparent guard 

 projects beyond the edges of the cup, in the other the guard is level with the cup 

 and shows fine sulcations (Plate V., figs. 113 and 114). In form A, the ligules are 

 fusiform and approximately ecpual in bulk, the ventral ligule rather longer (Plate V., 

 fig. 115). In form B, the ligules are obtuse and the ventral ligule, besides being 

 longer, has about twice the bulk of the dorsal ligule (Plate V., fig. 116). If these 

 specimens had been taken at different times and places they would probably have 

 been described as separate species. The extruded proboscis (form B) is covered with 

 minute papillae arranged irregularly, and of two kinds, acuminate and rounded, the 

 former greatly predominating. Length of specimen of form A (incomplete behind), 

 upwards of 60 millims. ; form B, complete with subulate anal cirri, 45 millims. long, 

 including the proboscis (7 millims.). 



Family : SPIONID^l. [See also p. 325.] 

 Polydora hornelli, n. sp. Plate V., fig. 117. 



From pearl oysters, Gulf of Manaar. 



It is hard to say whether this species is really distinct from Polydora ciliata ; the 

 differences, so far as can be judged from preserved material and published records, are 

 very slight, but such as they are, they seem to afford ground for separation when 

 taken in conjunction with the geographical distribution and the character of the host. 



