26 THE VOYAGE OF H.M.S. CHALLENGER. 



The trichobranchiae exist as bundles, or fasciculi, and are situated at the base 

 of nearly all the branchial plumes, whereas the phyllobranchial plates traverse 

 the stem from the base to the apex, not compressing each other laterally, 

 but implanted with their broad axis in the longitudinal direction of the stem. In 

 some instances the plates slightly overlap one another, but generally they are fixed 

 end to end. 



The trichobranchiate filaments are at the base, just within the margin of the carapace, 

 and exist as a peculiar branchial arrangement corresponding with the lower margin 

 of the branchial chamber, increasing in size and number at the anterior and posterior 

 extremities. 



The phyllobranchiate plumes lie farther within the branchial chamber, and compared 

 with the area, appear to occupy but a small portion of the space. 



Thus we see that the trichobranchiate bundles are arranged mostly where the water 

 plays most actively and freely, that is, along the margin and at the afferent and efferent 

 passages of the branchial chamber, while the phyllobranchiate plumes lie where the water 

 within the chamber is less likely to be disturbed. 



I think there can be little doubt that the phyllobranchial plates and the trichobranchial 

 filaments are derived from one and the same origin, as we see at the base that the one 

 gradually passes into the other. In Cheramus this also appears to be the case, though 

 the close compression of the one against the other is probably an inducing cause ; but 

 probably there are also other conditions brought into play. The foliaceons petals 

 are not implanted one against the other, but are exposed freely in the chamber. 

 Examination of these plates shows that within the several petals the structure is 

 traversed by canals that assume an arborescent appearance, through which the fluid 

 circulates and is brought by the tenuity of the apparatus into closer contact with 

 the aerating agents than it otherwise would be. 



In a respiratory chamber, such as in the genus now before us, the water flows in by 

 the posterior extremity, for which purpose the carapace can be raised or depressed at will 

 within certain limits ; and as we may assume that in a large chamber such as the 

 present, the water flows along the lower margin, passing out at the anterior end only, it is 

 probable that the largest amount of current will correspond with that portion of the 

 chamber where the trichobranchiate filaments are best developed and most abundant, 

 whereas the phyllobranchial plates are present in the centre and deeper recesses 

 of the chamber, where the circulation will be more quiescent, and the power of oxygena- 

 tion less efficient. 



We do not know much of the habits of this animal, but many of the group are 

 burrowers in the deposits beneath the seas in which they live, hence it is more than 

 probable, from the matted condition in which I found the fur that covers many parts of 

 the animal, that it inhabits hollow passages in the mud, and that the circulation of the 



