136 THE VOYAGE OF H.M.S. CHALLENGEK. 



The branchlets and pinnules in certain portions of the corallum are modified to 

 form a long cylindrical reticulum which serves as the habitation of an Annelid (PI. 

 III. fig. 2). In such portions the lateral pinnules arch over so as to meet above, and 

 are connected together by numerous bridge-like processes, the whole forming a close 

 tubular reticulum. The pinnules of adjoining branchlets are often united together at 

 various points, but there is no regular reticulum formed, excepting in the case of the 

 worm-tubes already referred to. The type of corallum is intermediate between that of 

 certain paniculate species of A-phcwijxtthes and the fan-like reticulate species, such as 

 Tijlopathesf flabellum (Pallas). 



The spines are short and subtriangular, with a sharp apex. The upper margin may 

 be concave or almost straight. A spiral arrangement is not well marked, but an 

 irregular steep dextrorse spiral may be sometimes observed. The spines are arranged 

 in regular longitudinal rows, six of which may be counted from one aspect of a pinnule. 

 The members of a row are from two to two and a half lengths apart (PI. III. fig. 4). 



The coenenchyma has a dirty yellow colour washed with light brown. The polyps 

 are of a deep rich brown colour, excepting a circular area, including the mouth, which 

 is white. The polyps are elongated in the transverse axis, being usually about twice as 

 broad as they are long ; there appears to be no marked tendency for the polyps to 

 assume a rounded outline on the stronger portions of the corallum. On the pinnules 

 the polyps usually form a single row, wdiich, however, is not always confined to one 

 aspect. On the branchlets the polyps rarely form a single series, they are sometimes 

 arranged irregularly, but often two rows are formed, in which case the polyps are 

 arranged in pairs on opposite aspects of a branchlet. The tentacles in living specimens 

 are evidently of moderate length. In spirit preparations most of them become reduced 

 to small rounded elevations of the peristome, which recall the appearance in Aphani- 

 pathes. Some, however, are better preserved, and appear as conical processes having a 

 length about equal to that of the polyp (PL III. fig. 3). Often four tentacles are all that 

 appear to remain in old individuals, the sagittal tentacles having atrophied. A com- 

 parison of a number of specimens has led me to suppose that after a time the sagittal 

 tentacles leave their normal position at the margin of the peristome and sink to a lower 

 level, after which they become reduced in size and ultimately disapj>ear. The lateral 

 pairs of tentacles next become reduced, apparently without changing their position. 

 Specimens which possess only two tentacles are not rare. Ultimately all disappear, 

 and in the older portions of the corallum the position of a polyp is often only to be 

 recognised by the white circular area on which the mouth opens. 



Habitat.— -Station 310 ; January 10, 1876; lat. 51° 27' 30" S., long. 74° 3' 0" W., 

 Sarmiento Channel ; depth, 400 fathoms ; bottom, blue mud. 



