100 AUSTRALASIAN ANTARCTIC EXPEDITION. 



Microscopical study of them, under varyiiig conditions and from different 

 aspects, reveals a new type of bristle ; new not only to this family but, as I think, new 

 to the class. 



Some time previously I had made a drawing of one of the chaeta? from a group 

 separated out and freshly mounted in glycerine; it was symmetrical, finely pointed 

 with a narrow flange on each side, and very similar to that figured by Favivel (1897) 

 for Amfharete grubei (pi. XVIT, fig. 24.) But amongst them T found others in which 

 the bristle is curved and has only one rather broader flange. I supposed therefore 

 that there were two kinds of chaetse in the bundle. 



Some months later, when preparing this account for piil)lication, 1 had occasion 

 to refer to my preparations, one of which was in Canada Balsam. I was surprised to 

 see that all the cheetee are alike, curved, with a single flange. Wishing to ascertain how 

 I could have been deceived in my earlier examination, I cut off a fresh parapod, sep- 

 arated out the cheetse and made a new mount in glycerine. 



Again I saw in most of the cha'tse two narrow flanges. I then pressed the cover- 

 slip so that the chseta^ might be flattened out a little ; now all of them had a single 

 flange. I then lifted the coverslip, turned the cha>tae about and re-examined them. 

 Again I saw several with the two flanges. 



A careful study under a high power informed me that the chteta really has three 

 flanges, two narrow ones lying in one plane, symmetrically arranged, and a third broader 

 one in a plane at right angles to them ; and in this position the chseta is curved. 

 Having made this discovery, it was easy to detect the three flanges in some of the 

 chaetfe, and I have drawn one of them (fig'. 118-120). 



To what extent this observation may vshed light on discrepancies in the accounts 

 of ch»ta? in some families, e.g., the Terebellidte, I cannot say. It is evident that a 

 renewed study of the bristles in certain families is desirable. 



The ventral surface of the thorax is nearly flat, and traversed by a wide shallow 

 median furrow, which increases in depth posteriorly, and after the last gland shield 

 becomes very deep but narrower; the margin of the furrow is formed by the rounded 

 muscular ridge on each side. 



The uncinigerous neuropods commence below the 4th notopod. Those on the 

 anterior segments of the thorax are vertical ridges, limited to the sides of the body, 

 and originating near the hinder boundary of the segments; their edges rise only slightly 

 above the surface. In the hinder segments each neuropod becomes more prominent, 

 thick and fleshy, whfle in the abdomen they are narrower and become flap-like (fig. 

 114). The neuropod is now a quadrangular flap directed backwards and outwards; 

 its free edge carries the uncini. On its upper surface near the body wall is a small 

 rounded papilliform upgro^\'th (which is, perhaps, a dorsal cirrus). By the 12th 

 abdominal the neuropods are already much longer and project still further; the dorsal 



