254 McMURRICH. [Vol. II. 



The Hertwigs showed that they are solid structures and that 

 two portions are recognizable (three in the Sagartiadae which 

 possess acontia). The upper portion consists of two wing-like 

 lamellae attached to the edge of the mesentery, which, in con- 

 sequence, presented a somewhat trilobed condition in transverse 

 section. The two lateral lobes are characterized by the epithe- 

 lium of their outer surfaces being composed of long and very 

 narrow cells, each of which bears a single cilium. These cells 

 vary somewhat in length at regular intervals, so that longitudi- 

 nal sections show the lobes to have a wavy contour. No glan- 

 dular or nematocyst cells occur in this portion of the filament, 

 which the Hertwigs termed the Flinimerstrcif. 



The lower portion consists of a more or less coiled cylindrical 

 cord attached throughout its entire length to the edge of the 

 mesentery. Its epithelium consists of gland cells, nematoblasts, 

 supporting cells (Stiitzzellen), and sensory cells, the nerve fibers 

 from these last forming a plexus between the bases of the other 

 cells. This portion was termed the NesseldriisensU'eif. In its 

 upper part it is an almost straight cord, and in perfect mesen- 

 teries is continuous with the central lobe of the ciliated bands, 

 and through this with the stomatodaeal ectoderm ; in imperfect 

 mesenteries, however, there is no such continuity with the 

 ectoderm, the glandular streaks gradually fading out above, and, 

 the central lobe of the ciliated bands being wanting, the lateral 

 lobes are separated by a depression lined by ordinary endoderm 

 cells. It would seem from this that the central lobe of the 

 ciliated bands is in reality the upper part of the glandular streak. 



The arrangement just described may be regarded as the 

 typical one for the hexactinian filaments, though certain 

 departures from it have been observed. Thus in the Madre- 

 poraria the ciliated bands have not yet been observed, and they 

 are also absent in certain Actiniaria, such as Protanthea and 

 Gonactinia (Carlgren, '93) and Corynactis, Ricordea and Rho- 

 dactis (Duerden, '98).^ 



^ I may state that I can confirm Duerden's observation as to the absence of the 

 ciliated bands in the last two forms. 



So far, however, as the majority of the Hexactiniae are concerned both parts 

 occur. 



