464 Mr. G. H. Grosvenor. [Nov. 3, 



den Nesselsacken gebildet wird, und ob die anderen von der Hohle der 

 Leberlappen herriihren, von verzehrten Thieren." On the same page 

 he states that " ganz eigenthiimliche, grossere Formen " (of nematocysts) 

 are found in several genera (Glaucus, Coryphella, Flabellina, Pteroeolis). 



Herdman and Clubb (1889), and Herdman (1890), placed beyond 

 doubt, by means of serial sections, the existence of a ciliated canal 

 connecting the cnidosac with the " hepatic " diverticulum. This had, 

 apparently, been doubted by many,* though I think conclusive 

 evidence on the point may be found in several previous papers by 

 Bergh and Trinchese, not to mention Hancock and Embleton. 



In the second paper Herdman states that the cnidosac is " evidently 

 of ectodermal origin." Probably the presence of such typically 

 ectodermal organs as nematocysts was one of the chief reasons of this 

 assertion. 



Davenport proved by tracing the actual development that the 

 cnidosac arises not from the ectoderm, but from the distal extremity 

 of the gastric diverticulum. This has since been confirmed by Hecht 

 and Krembzow. 



The last four authors have all described and figured what they took 

 to be young stages in the development of the nematocysts, and Hecht 

 and Krembzow have attempted to trace this development in detail. 

 The former considers the nucleus to take a prominent part either by 

 " directing " the secretion of nematocysts by the cytoplasm, or, as he 

 thinks, more probably by giving up a portion of its substance to? 

 form the rudiment of the nematocyst. But Krembzow points out that 

 in many cases the cnidoblast is elongated, the nucleus lying at the 

 basal end, while the nematocysts occur from the first at the end next 

 the lumen of the cnidosac. 



Most of the other observations which I have found bearing on this 

 question are scattered through works dealing with the general anatomy 

 and classification of the group, and will be more conveniently mentioned 

 later on. 



Quite recently, however, since a rough draft of the present paper 

 was already written, a note appeared in the Johns Hopkins 

 < University Circular ' by C. 0. Glaser, in which the writer brings 

 forward and discusses the view that the nematocysts of Solids are 

 extraneous and derived from their prey. After pointing out that 

 many ^Eolids are found in association with hydroids or actinians, and 

 that Alder and Hancock observed fragments of a hydroid in the 

 alimentary canal of an ^olid, he shows that the present accounts of the 

 development of nematocysts are unsatisfactory, and that the appearance 

 of nematocysts in endodermal cells, and apparently independent of the 

 nucleus, is at least unexpected. He then records three observations of 

 his own in support of this view 



* See Lankester's " Mollusca," ' Encycl. Brit.,' 9th edit. 



