208 MR. J. T. CtTKNINGHAM OX THE [Mar. 15, 



In reckoning 9 teeth on the antero-lateral margin, Stebbing 

 follow 8 Thomas Bell (' British Crustacea,' 1853), and, I believe, all 

 other writers who have described the species. Whether the 10th 

 tooth should be counted or not may be considered a matter of 

 opinion if the description of the species is considered apart from 

 its relations to other species ; but m)' own observation has convinced 

 me that the teeth correspond exactly to the broad teeth or lobes 

 of Cancer pagurus, and in both species there is a 10th tooth at 

 the posterior end of the series. Thomas Bell noticed this 10th 

 tooth in Cancer j)agim(s, giving as one of the specific characters 

 " latero-anterior margin ten-lobed ;" and when, as in his work, only 

 9 teeth are attributed to Atelecyclus the homology of the teeth in 

 the two species is obscured. 



It is stated as a peculiarity of the legion Corystinea, containing 

 the single family Corystidse, that the third pair of maxillipeds do 

 not usually make a complete closure of the mouth-cavity, and are 

 extended over the anterior margin of its frame. But it is mentioned 

 as an exception that in Atelecyclus the third maxillipeds do make 

 a complete closure of the mouth-cavity. The real significance of 

 this exception is that in Atelecyclus as in Cancer the inner edges 

 of the third maxillipeds meet in the middle line. 



The form and proportional size of the chelipeds, and of the 

 other legs, are very similar in Atelecyclus and in Cancer. In 

 Atelecyclus there are 7 longitudinal rows of tubercles on the 

 propodus of the chelipeds. In the adult Cancer there are no 



{)rominent tubercles, but it is easy to verify the fact that the five 

 ower rows are represented by granulated ridges, while the upper 

 two are obsolete. On the preceding segment, or carpus, in 

 Atelecyclus there are 4 rows of tubercles, of which the uppermost 

 bifurcates anteriorly. These are likewise represented in Cancer 

 by ridges, but the bifurcation of the uppermost is not visible. In 

 the young stages of Cancer pagurus which I have figured and 

 described in this paper the rows of tubercles on both the segments 

 mentioned are very similar to those in the adult Atelecyclus. 



In all the points mentioned in which Atelecyclus heterodon 

 approaches to Cancer joagurus it differs from Corystes cassivelaunus. 

 The form of the carapace in the latter is quite different ; it is 

 much longer than broad, and its sides are almost straight and 

 parallel to the antero-posterior axis. There is no median tooth to 

 the rostrum, the extremity of which forms two divergent teeth, 

 and the sides of which slope outward to the orbits without any 

 projecting tooth, but with only a slight rounded prominence at the 

 inner boundary of each orbit. Instead of 10 teeth on the antero- 

 lateral margin, there are 4 widely-separated lateral teeth, with a 

 blunt projection between the 2nd and 3rd. The maxillipeds of 

 the 3rd pair are long and narrow, and their inner edges do not 

 meet in the middle line. 



The chelipeds of Corystes, instead of being short and robust as in 

 Atelecyclus and Cancer, are, especially in the male, long, and slender, 

 with long segments, and the rows of tubercles mentioned above 



