1894. 



PROCEEDINGS OF THE NATIONAL MUSEUM. 



545 



larvte belong to two unknown species of that genus. They therefore 

 represent a new larval i^VQ for which I propose the name Odontc rich thus. 

 One of them (No. 9958, U.S.X.M.) was taken by the Albatross Octo- 

 ber 3, 1883, at station 2101, off i^antncket. It is represented in fig, 25. 

 A comparison of this figure with Brooks' fig, 5, pi. xii, which repre- 

 sents a Gonerichthus froin St. Vincent, Cape Verde, will show a strik- 

 ing similarity. They both exhibit the form of body, the shape of the 

 carapace, and telson, that Brooks has shown to be characteristic 

 of the Gonodactylus larvae. An examination of fig. 25 will convey a 

 better idea of this interesting form than pages of description. It will 

 be seen that the specimen before us differs from Brooks' in having a 

 somewhat shorter rostrum with five or six small spines on the ventral 



FiK. : 



Two teeth <iii the 



ODONTKRICHTHUS LAUVA. 

 n are hidden by the .-ye. Uriiwii with a camera Uicida X 14. 



side, and in having a minute additional secondary spine on the ventral 

 edge of the carapace. The dactylus of the raj)torial limb is inuch 

 more developed and shows five lateral teeth beneath the larval skin. 

 The similarities are so much greater than the differences that the latter 

 may be due merely to a difference in age, the one being an older stage 

 of the other. These forms would appear to belong to a species in which 

 the larvie can be distinguished from Gouerichthi only after the teeth 

 begin to form on the raptorial dactylus. 



The other species, however (fig. 2G), is not so similar to the Gone- 

 richthus type, but approaches the Pseuderichthus form, and this is just 

 what we should exi)ect if my view be accepted that this is a larva of 

 Odontodactylvs, because this genus is distinctly intermediate in some 

 Proc. N. M. 94 35 



