222 CRUSTACEA MALACOSTRACA. III. 



Length of a male from Diastylis polaris G. O. S. n mm., of a male from Diastylis echinata 

 Bate 0-86 mm. 



Only two males were seen by me, and I found the following differences between them. The 

 male from Diastylis polaris is, as just stated, longer, with only nine processes on the inner margin of 

 first joint of the anteunulse, five pairs of comb-teeth on the last abdominal segment and no rudiment 

 of a sixth pair (fig. 5 i), finally the exopod of the uropods is only a little shorter than the endopod. 

 The male from Diastylis echinata is smaller, with ten processes on the inner margin of first joint of 

 the antennulae (fig. 5 e), five pairs of long comb-teeth and rudiments of a sixth pair on the last ab- 

 dominal segment, while the endopod of the uropods is somewhat longer than the exopod. According 

 to my experience from the second larval stage in Bopyridse and Dajus mysidis I do not suppose that 

 these differences are specific characters, but being unable to decide the question with absolute certainty, 

 especially as the large male is from the cold and the other from the warm area, I will point out the facts. 



First Larval Stage (fig. 5k). The description above may be sufficient, as the number and 

 length of the setae on the antennae and the uropods may be seen on the figure. The body of such a 

 larva is only 0-17 mm. long. 



Occurrence. Found in the marsupium of Diastylis polaris G. O. Sars taken by the "Ingolf" 

 at two stations in the cold area. 



South of Jan Mayen: Stat. 113: Lat. oo^i' N., Long. 7o6' W., 1309 fath., temp, -f- ro; i female, 



i male. 



North-West of the Faeroes: Stat. 138: Lat. 6^26' N., Long. f$f)' W., 471 fath.; temp. + 0-6; 



i female. 



Furthermore in the marsupium of two specimens of Diastylis echinata Bate and of three spe- 

 cimens of Hemilamprops cristata G. O. S., all captured by the "Thor" at the following station in the 

 warm area. 



South-West of the Faeroes: Lat. 6io8' N., Long. g28' W., 436 fath.; 5 females, i male. 



Sub-Order Gnathiidea. 



As already stated on p. 4 I have found it necessary to remove the family Gnathiidse from the 

 sub-order Flabellifera and to establish it as a sub-order, as it differs in important characters from all 

 other families of the order and occupies an isolated position. - - Only one valid genus is represented 

 in the material. 



Gnathia Leach. 



In 1901 A. Dollfns established the new genus Caecognathia for the reception of Gnathia stygia 

 G. O. S. and a new species, but the latter must, however, be cancelled as founded on features due to 

 individual variation. The only difference between Caecognathia Dollf. and Gnathia Leach is the total 

 absence of eyes in the former genus, while well developed eyes are found in the latter genus; Dollfus 



