1922. j S. Kemp : Notes on Crustacea Decapoda. 121 



E. Dactylus of last three legs with basal pro- 

 tuberance double [rostrum toothless, concave 

 above; associated with echinoidsj ... Stegopontonia, p. 



E'. Dactylus of last three legs with basal pro- 268. 



tubcrancc single. 



F, Rostrum laterally compressed, frequently 

 with teeth ; inner lacinia of maxillula nar- 

 row ; dactylus of last three legs with a 

 single claw and a hoof-shaped basal pro- 

 tuberance ; living on coials ... ... Coralliocaris,^ p. 



F' . Rostrum depressed, toothless ; inner 268. 

 lacinia of maxillula very broad ; dactylus 

 of last three legs with two claws and flat 

 basal protuberance; living in lamelli- 

 branchs ... ... ... Conchudytes, p. 279. 



B'. Antenna! scale rudimentary. 



C. Rostrum present, with or without teeth ; distal 

 lacinia of maxilla well developed ; all maxillipeds 

 with rxopods ; dactylus of last three legs biun- 

 guiiulnte [associated (? always) with sponges] ... Typfon, p. 286, 

 C. Rostrum absent ; distal lacinia of maxilla rudi- 

 mentary ; second and third maxillipeds without 

 exopods ; dactylus of last three legs simple .. Parafypton, p. 286. 



In this key Nobili's Onycocaris, originally proposed as a sub- 

 genus of Coralliocaris, is not included (see p. 278). I am not con- 

 vinced that the two species for which it was founded are related to 

 Coralliocayis, and as I have not seen either I prefer to leave their 

 position undetermined for the present. The generic position of a 

 number of other species is doubtful 2 ; when they are better 

 known it is probable that some modification will be necessary in 

 the generic arrangement here adopted. 



Balss' Bathypalacmonella* evidently does not belong to the 

 subfamily, as it possesses a series of arthrobranchs in addition to 

 five pleurobranchs. 



Of the seventeen genera which I recognise Pcriclimenes com- 

 prises by far the largest number of species. No less than eight 

 genera are monotypic and the majority of these are known from 

 single specimens only. 



In the keys to the species I have followed Borradaile's ex- 

 ample and have in each instance inserted the rostral formula. An 

 expression such as R. n-14 : 2-3 indicates that the teeth on 

 the upper border of the rostrum vary from 11 to 14 and that 

 there are 2 or 3 teeth on the lower border. The length of a 

 specimen, as given in the descriptive parts, represents the distance 

 between the tip of the rostrum and the tip of the telson with the 

 animal extended as nearly as possible in a straight line. The 

 figures in the text, even when forming part of a single text-block, 

 are not necessarily drawn to the same scale. 



1 Not including Onycocaris Nobili. 



3 In my attempts to readjust the generic classification of the subfamily I 

 have found myself greatly handicapped by our inadequate knowledge of a number 

 of species. It is important thr.t we should have fuller knowledge of Onycocaris, 

 of the two species from Japan which Balss referred to Periclimenes (see p. 138) 

 «nd of the three forms attributed to Coralliocaris by Miss Rathbun (see p. 268). 



» Balss, Zool. Am. XLIV, p. 598 (1914). 



