EITTER: OCTACNEMUS. 237 



flange of three or four mm. in height. Instead of being rounded, as in 

 Moseley's animal, it is narrow even at its base, and thins off to an almost 

 sharp edge. The test along the very edge is somewhat hardened, and 

 so changed in structure that it refuses to take stain as do the remaining 

 parts. Even in the preserved animal, this flange with its meandering 

 course is a conspicuous object on the ventral surface of the disc (Plate 1, 

 Fig. 2). Its position is not at the periphery of the disc, in the region 

 of the rays, as Moseley says, but is four or five mm. in toward the centre 

 from the line of the base of the rays. Posteriorly the flange passes on 

 to the atrial part of the body. It reaches back to the region of the 

 adhesive disc, where it gradually disappears, and hence diflers decidedly 

 from its course in 0. bythius, where it is continuous hehiiul the adhesive 

 patch. It would appear that the whole ventral surface rests on the sub- 

 stratum, to which, however, the creature is attached by the adhesive disc 

 alone. The thickenings of the test, or pads on the bases of the rays, 

 mentioned by Herdman, are present here, but extend farther toward the 

 ends of the rays than they did in Herdman's specimen. 



Concerning the microscopic structure of the test, and the circular and 

 radial muscles of the mantle, I have nothing to add to what Moseley and 

 Herdman have recorded. 



The most important diff'erences, both as to observation and interpre- 

 tation, between my results and those reached by these naturalists, relate 

 to the branchial sac and the parts immediately associated therewith. 

 Both Moseley and Herdman sought in vain for branchial stigmata ; and, 

 failing in this, were misled in their conclusions as to the whereabouts of 

 the branchial chamber. Both naturally assumed it to be the great cavity 

 within the oral disc. As it becomes clear from the present study that 

 the branchial stigmata and the branchial chamber are located in quite a 

 different region, the question arises as to the significance of the chamber 

 of the oral disc, supposed by them to be branchial. It will be conven- 

 ient to make the consideration of this the starting point of our account 

 of the internal structure of the animal. We begin by examining the 

 wall of the oral disc external to the chamber. Were the cavity a true 

 branchial cavity, comparable with that of other ascidians, we should have, 

 passing from without inward, the following layers: the test, the ecto- 

 derm, the mantle, and immediately lining the cavity an extension of 

 the respiratory epithelium. I have examined this wall with especial care, 

 both on stained and unstained flat preparations, and on microtome sec- 

 tions, and/ai7 to find anything hut the layer of test. 



Likewise the " horizontal membrane," separating the supposed bran- 



