367 



on the main question under consideration, I have been led to recognize 

 some other points in the development that heretofore have been either 

 iücompletely or incorrectly described. That my results are fuller and 

 more exact than have been those obtained by other workers on this 

 subject I suppose to be chiefly due to the large quantity of variously 

 prepared material that has been at my disposal. Owing to the great 

 abundance of Perophora annectens Ritter on the Californian coast, 

 and owing also to the very numerous and closely placed ascidiozooids 

 in the cormi, it is extremely easy to collect this species in almost 

 unlimited quantities, and consequently correspondingly easy to get 

 great numbers of buds in all stages of development. 



I find, on the other hand, that here in the Bay of Naples, P. 

 Listeri occurs not only much less abundantly, but also that the 

 stolons are so delicate and cling so closely to the substratum on 

 which the animal lives, that the collection of the buds in any con- 

 siderable quantity is a very laborious process. 



I give the results of my work on this genus in the same cate- 

 gorical way as for Goodsiria, reserving all details of description and 

 general discussion for my final paper. But I mention only those 

 points in which my observations either extend or correct those made 

 by other investigators: 



1) When the differentiation of the "endoderm" into the branchial 

 and two peribranchial sacs takes place, it does so in such a way that 

 the developing blastozooid is connected with the double walled par- 

 tition of the stolon, not by the branchial sac, as has been hitherto 

 supposed, but by the left peribranchial sac. 



2) This communication of the peribranchial sac with the stolonic 

 partition is entirely severed at an early stage in the development of 

 the bud; viz. at a time when the two peribranchial pouches have 

 merely begun to envelop the branchial sac. From these two last facts 

 which I have established in the clearest possible manner, for both 

 P. annectens and P. Listeri, it is obvious that there can not be in 

 Perophora an epicardium corresponding to the structure called by 

 this name in Clavelina. 



3) The common anläge of the central nerve ganglion and the 

 dorsal tube appears nearly contemporaneously with the beginning of 

 the differentiation of the "endoderm" into branchial and peribranchial 

 sacs; therefore it is one of the very first organs to appear. 



Every step in the differentiation of this anläge 

 into nerve ganglion and dorsal, or hypophyseal duct 

 can be followed with great ease and clearness. 



