426 PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 



The carcinceciuDi is very rarely a naked gastropod sliell ; in most of 

 the specimens seen it is either built up by a colony of JEJpizoanthus Amcr- 

 icanus, like the carcinoecinm of Eupagurus Kroyeri, from the same sta- 

 tions, or is made up in a somewhat similar way by the single polyp of a 

 species of Adatnsia, the base secreted by the Adamsia being expanded 

 on either side and united below so as to inclose the crab iu a broadly 

 conical cavity, with only a slight spiral curvature. The nuclei about 

 which these polypean carcinoecia are formed are of various origins; the 

 majority of the Adamsia carcinoecia appear to have been built upon 

 fragments of pteropod shells, in some cases upon bits of worm-tubes, in 

 one case upon the entire shell of a Cadulus, the greater part of the shell 

 being left protruding from the base of the polyp. In the carcinoecia 

 formed by Epboanthus the nucleus seems usually to have been absorbed, 

 so that nothing is left distinguishable from the colony of polyps itself. 

 In some case-s the Adamsia has completely overgrown a small Epizoan- 

 tJius carcinoecium, so that when the Adamsia is removed a jDcrfect Epi- 

 zoatitliiis carcinoecium is found beneath as a nucleus. The carcinoecium 

 of this species, and of H. gracilis as well, does not cover the animal to 

 the same extent as is usual in the species of Eupagurus, the anterior 

 part of the carapax evidently being constantly exposed, its induration 

 fitting the animal for such exposure. The Epizoanthus carcinoecia are, 

 however, very often disproportionally large for the crabs inhabiting 

 them, having grown out either side until they are several times broader 

 than long. In spite of these often enormous carcinoecia, both species of 

 the genus probably swim about by means of the ciliated dactyli of the 

 ambulatory legs, as Spiropagurus spiriger has been observed to do by 

 Stimpson (Proc. Acad. Nat. Sci. Philadelphia, 1858, p. 248 (SG), 1859). 



Stations 865, 870, 871, 872, 873, 874, 876, 877, 878, 880 ; 65 to 252 fatli- 

 oms. At many of these stations it occurred in very great abundance. 



Hemipagurus gracilis, sp. nov. 



This is a smaller and more slender species than the last, and is readily 

 distinguished from it by the smooth carapax, the longer and more slender 

 eye-stalks, the long and acicular ophthalmic scales, and by the narrow 

 dactyli of the ambulatory legs being longer than the correspouding 

 propodi. 



Male. — The carapax in front of the cervical suture is flat, smooth, 

 nearly naked, and scarcely at all areolated. The anterior margin is rather 

 more strongly sinuous than in H. socialis, and the lateral lobes are 

 slightly angular and each is tipped with a minute spine, as iu that spe- 

 cies, but the marginal carina between these spines is much less distinct. 



The eye-stalks are more than half as long as the carapax in front of 

 the cervical suture, flattened and expanded distally, but only about half 

 as broad as long. The eyes themselves are as in 11. socialis. The oph- 

 thalmic scales are more than half as long as the eye-stalks, and are 

 acicular and regularly acute. 



The ultimate segment of the peduncle of the antennula is as long as 



