1903.] FROM EAST AFHICA AND ZANZIBAR. 367 



stomach is quite free from the hepatic mass. No armature was 

 discernible in the reproductive organs. 



I have some hesitation in classifying this specimen as Peltodoris, 

 as the back is not minutely granulated but covered with small 

 warts. The shape, however, is not that of Archidoris, and both 

 the stiffness and small radula are in favour of the position here 

 assigned to the form. 



9. Thordisa villosa (A. t H.). (Plate XXXII. figs. 1 & 2 ; 

 Plate XXXIII. figs. 1-3.) 



[Alder & Hancock, Trans. Zool. Soc. Lond. vol. iii. (1864) p. 119, 

 pi. xxxiii. fig. 1 ; Bergh, Semper 's Reisen, Heft xii. (1877) p. 540 ; 

 Bergh, Danish Exped. to Siam, Opisthobi^anchiata, 1902, p. 182.] 



One specimen was dredged in Zanzibar Haibovir on a sandy 

 bottom with a little Zostera (PI. XXXII. figs. 1 & 2). The ground- 

 colour of the living animal is a translucent yellow, like a bit of 

 crystallised fruit. On the ample and transpaient mantle-margin 

 were blotches of peaty red and of difl:erent sizes. Smaller spots 

 of the same colour are scattei'ed over the whole bod}^, particulaily 

 above the visceral mass. The under surface is unifoi-m blight 

 yellow with a few brown dots. The whole dorsal surface is 

 covered with colovirless transparent papillae (PI. XXXIII. fig. 2), 

 some simple (especially on the mantle-edge), and some compound 

 with two or more filaments. It is also plentifully supplied with 

 spicules set in a stellate arrangement, but the general consistency 

 is quite soft and not stiff. The rhinophore and branchial 

 openings are slightly raised and tubercv;late, but not stellate. 

 The rhinophores are large and slightly bent back; the stalk is 

 rather longer than the laminated portion. The branchise are six 

 and mostly only bipinnate, though tripinnate branches also occur 

 (PI. XXXIII. fig. 3). They are grey with a brown rhachis. The 

 foot is gi'ooved in front but not notched. The tentacles are thin 

 and digitate. There is no trace of labial armature. The radula 

 consists of about 47 rows of simply hamate teeth, each row con- 

 taining 40-50 on either side of the rhachis. They are all of the 

 same shape and size, except the outermost five or six, which bear 

 from seven to ten long fine hair-like denticles on each side of the 

 much reduced central hook. No armature was discoverable in the 

 reproductive system. 



The alcoholic specimen is quite flat, and is 2'5 centimetres long 

 by 1*6 broad, but the living animal was capable of assuming two 

 shapes — one flat with a broad mantle-edge, and one high with 

 a much narrower edge (PI. XXXIII. fig. 1). 



I think this animal may lie safely identified with the Doris 

 {Thordisa) villosa of A. & H. Bergh seems to think that this 

 species is probably identical with his Thordisa maculigera, and 

 I share this view, though the formation of the outermost teeth 

 is not exactly like either his description or his plate, as the 

 denticles are longer and the central hook, though much reduced, 

 has not vanished, 



[15] 



