230 



type. If Professor Roule had studied Tornarla^ by optical sections 

 only, I have no doubt he (or any other investigator under similar con- 

 ditions) would have claimed a mesenchymatous origin for the meso- 

 derm, and, who knows, might have been led to deny the existence of 

 body-cavities in Balanoglossus. The only real distinction between the 

 two methods of mesodermic origin is that of localized ingrowths and 

 diffused ingrowths. If the evidence of microtome-sections is not to be 

 preferred, in demonstrating cellchanges in an epithelium inside an 

 organism, to the method of optical observation of the entire animal, then 

 I have nothing more to say in the matter. 



2) The structure of the larva. 



Here Professor Roule disagrees with me in many essential points. 

 Actinotrocha is to him a trochophore, hence it cannot have the organs 

 described by me. Professor Roule's Actinotrocha has, according to 

 him, no organs of the kind; hence mine has them not. Is Professor 

 Roule going to deny the existence of the ascidian notochord because 

 some ascidians do not have a tailed larva? Surely this is one of the 

 most extraordinary lines of criticism to adopt. The Actinotrocha with 

 which I have been working has mesenteries separating the parts of the 

 coelom which with a little practice can be recognized with the naked 

 eye. The presence of the lower mesentery has been described by others 

 before me, and the front mesentery was figured in optical sections by 

 Wagener (in 1847), though he took it for a nerve-cord. 



In connexion with these mesenteries we may mention one point. 

 The Professor seeks refuge in "bad preservation" to account for my 

 finding a nerve-ganglion, and in shrinkage for producing the Epider- 

 mistasche and subneural glands; these are plausible although not true, 

 but I do not think that either of these could in any circumstances ac- 

 count for my mesenteries. It is worth while to record, however, that 

 I have very successfully reduced the mesoderm of some Actinotrochae 

 to Professor Roule's condition. The most efficacious method appears 

 to be a sudden change from sea-water either to fresh water or to ab- 

 solute alcohol. Rapid rotation and contortion of the larva results and 

 as a rule, the sections shew a remarkable approximation to the figures 

 of Professor Roule. I do not wish to imply that he has pursued any 

 so crude method, but I desire to emphasize the fact that mesenteries 

 may disappear under treatment, but that they can hardly be created 

 where none such existed. 



The four years that have elapsed since my paper on Actinotrocha 

 have led me to modify some of my theoretical conclusions, and an 

 anatomical detail in Ceplialodiscus which I have corrected in the An- 



