62 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 5 



Ambulatory legs diminishing regularly in length, cylindrical, pu- 

 bescent; dactyli almost as long as propodi, tips curved. 



Sternal trough extending past tip of abdomen, which incompletely 

 fills it. 



Remarks: The proposed new species is the Pacific analogue of the 

 Atlantic A^. brasiliensis Miers (1886). Since A^. lamellatus Stimpson 

 (1871) occurs on both sides of the continent, careful comparison was 

 necessary to ascertain that A^. brasiliensis does not likewise. Examination 

 of the types of Hyastenus caribhaeus Rathbun (1893) (syn. "N . brasili- 

 ensis Miers) Rathbun (1925) shows the following differences: 



1. 2V. brasiliensis possesses a long rostrum. This is undoubtedly an 

 age character and might be shown by more mature specimens of 

 the new species. 



2. The hepatic region is not tumid on H. caribbaeus, although Miers' 

 figure of the type of "N . brasiliensis shows a tumid region as in the 

 new species. 



3. The basal antennal article is thick and bears a prominent lobe on 

 the outer margin. In the new species the edge is thin with but 

 the shallowest of lobes indicated. (See pi. 14, fig. 3.) 



4. The merus of the outer maxilliped, squarish in brasiliensis , is in 

 mexicanus almost triangular. (See pi. 14, fig. 4.) 



5. The cardiac region is slightly less protuberant in brasiliensis. 



In addition to the above, the spongy papillae of the rostrum and 

 second antennal segment appear to be unique in the genus. They may 

 disappear with age. 



It should be borne in mind, in connection with the above, that the 

 material available for comparison consisted of the unique holotype of A^. 

 mexicanus, a young specimen, and the mature types of H. caribbaeus. 

 Such great disparity exists between the young and old of this genus, as 

 shown by a long series of A^. lamellatus in the Hancock collections, that 

 one unfamiliar with the genus might easily mistake the juveniles for a 

 different species, or even refer them to another genus. 



The one character constant in individuals of any age should be the 

 shape of the maxilliped. In the case in question, the dissimilarity to 

 brasiliensis is sufficient, in the opinion of the writer, to establish the 

 new species on this point alone. 



