376 DISCOVERY REPORTS 



What is perhaps the most important character, namely, the precocious appearance of 

 leg 5, cannot, of course, be used when only isolated specimens of older stages are 

 available, but it may prove that in all cases in which leg 5 has the same predominance 

 over leg 4 it appears before it. Applying these characters, so far as is possible, to the 

 forms here described it seems that the following may with some probability be assigned 

 to Lysmata; Species R.S. II; B.R. I, II, III; A. I, II, III. Species A. I is evidently 

 congeneric with the form referred to Hippolysmata by Kemp. In his form the rostrum 

 had a series of small teeth at the base, and resembled that of some specimens in post- 

 larval stage I which could be determined as H. ensirostris, and were taken at the same 

 time off the Orissa coast. It may be accepted, then, that the larva of Hippolysmata does 

 not differ strikingly from Lysmata in late stages ; but we do not know the early stages, 

 and it may be that there is a difference in the time of appearance of leg 5. It is un- 

 fortunate that I have no late stages of my form R.S. I which so closely resembles 

 Lysmata, but differs remarkably in the late appearance of leg 5. It is not impossible 

 that this may be the larva of Hippolysmata multiscissa, and that the genera may be 

 distinguished by the order of appearance of the legs. There cannot be any doubt that 

 R.S. I is at least closely related to Lysmata, and I refer it provisionally to Hippolysmata. 

 Although the spine on the eyestalk is perhaps the most striking feature of Kemp's 

 larva, and of A. I, I do not think it is such a character as must necessarily be found in 

 all species of the genus. Kemp's figure shows that legs 1-4 are quite normal, and do not 

 have the propod dilated. 



Of the remaining four species from the Atlantic A. IV from Bermuda differs from the 

 Lysmata type only in having the propodus of legs 3 and 4 dilated, and that of leg 5 is of 

 unusually broad form; there is still the same discrepancy in actual length between legs 4 

 and 5. How far such a difference can be regarded as generic it is impossible to say. In 

 the closely allied species Caridio?i gordoni and C. steveni (Lebour, 1930) the larvae differ 

 in the presence or absence of large procurved spines on the carapace, and also in the 

 greater or less expansion of the propod in certain appendages, and we do not know 

 enough to exclude the possibility of large differences between species in Lysmata and 

 Hippolysmata. It seems best to regard this form also provisionally as within one of 

 these two genera. 



The other three larvae differ so much from Lysmata that they cannot possibly be 

 included in it or in Hippolysmata. 



Species A. V has still the same general form, with long eyestalk and, presumably, 

 enlarged leg 5, and its genus must be sought within those allied to Lysmata. The fol- 

 lowing genera are generally regarded as forming the Latreutid group of Hippolytidae: 



Latreutes Stimpson. Paralatreutes Kemp. 



Toseuma Stimpson. Lysmatella Borradaile. 



Trachycaris Caiman. Lysmata Risso. 



Gelastocaris Kemp. Hippolysmata Stimpson. 



Bythocaris Sars. Merguia Kemp. 

 Mimocaris Nobili. 



