ELIOT— NUDIBRAXCHS. 423 



brownish rods. The radula consists of 56 rows, each containing from 90 to 100 teeth 

 on either side of the rhachis. The teeth are of the ordinary hamate shape : the 

 innermost have low shafts and long bases which are prolonged a little in front of 

 the point where the shaft rises ; the outermost are erect and spike-like. 



The stomach lies outside the liver and its inner walls bear numerous moss-like folds 

 near the exit of the intestine. This latter, after issuing from the stomach, makes a 

 bend under the liver-mass and comes up again on the same side. The blood-gland is 

 double and olive-coloured. The ampulla of the hermaphrodite gland is thick and 

 coiled. The prostate is large and distinct, partly greenish and partly pinkish. The 

 spermatotheca is large and green. The spermatocyst is small. 



Alder and Hancock's D. fragilis has been neglected and the name has not been used 

 in describing recent specimens. If I had only the evidence of the present specimen, 

 which lias suffered from self- mutilation and is otherwise poorly preserved, I should 

 hesitate to revive the species. But I have also examined fresh specimens from Ceylon 

 and the type specimens preserved in the collections of the Challenger and of Semper 

 under the name D. morphcea. I have no doubt that the D. morphea of the Challenger 

 Collection is D. fragilis, for it agrees in anatomy with the specimens received from 

 Ceylon and also in external appearance, except that it has lost its colour. These 

 Ceylon specimens agree with Alder and Hancock's descriptions in all points except that 

 the upper lamina of the foot-groove is notched in the middle *. 



Bergh himself has pointed out certain differences between the Challenger specimen 

 of I). morphcBa and that in Semper's Collection, and I feel less certain that this latter is 

 identical with D. fragilis. But, on the whole, I think Bergh is right in regarding the 

 two forms as varieties of one species. 



The remarkable powers of self -amputation possessed by D. fragilis are also exhibited 

 by a species of Ge)ia (probably O. levis) which is common in the Indo-Pacific. The 

 present collection contains some of the tails of this animal which were doubtless 

 mistaken for nudibranchs. The mistake is very natural, as I can testify, for the tails 

 continue to adhere and possess some power of movement after they have been thrown 

 off. I once spent some time in examining them in the living state (if such an expression 

 can be used) at Zanzibar and had difficulty in persuading myself that they were not 

 complete nudibranchs. It would seem that Oena throws off its tail not only when it is 

 attacked by some enemy, but whenever it feels uncomfortable. I have kept numerous 

 specimens in confinement and have invariably found that they threw off their tails in 

 a few hours. 



11. Discodoris modesta, Bergh. 

 Bergh, Mai. Unters. in Semper's Reisen, Heft xii. pp. 534-6. 



One specimen from Coetivy. 



Length 35 mm., breadth across back 20, width of mantle-brim about 7. Dorsal 



• The ends of the diyided lamina overlap, so that the incision is not conspicuous. 



