229 



inability to do so may be due to one of two causes: it may be due to 

 disability on tbe part of Professor Roule, or it may be due to disabi- 

 lity on the part of the larva. As a matter of fact it is probably a 

 little of both, for I do not think the Professor has used the best me- 

 thods (see ultra), and there is also no question that the larva of P. Sa- 

 hatieri is an abbreviated larva compared with mine. 



However, there are three special points upon which I wish to 

 touch : 



1) The development of the mesoderm, 



2) The structure of the adult larva, 



3) Professor Roule's phylogenetic conclusions. 



1) The development of the mesoderm. 



All the observers before Caldwell employed the method of opti- 

 cal sections and they mostly agreed in the so-called 'mesenchymatous' 

 origin of the mesoderm, i. e. that the first indication was the appea- 

 rance in the blastocoele of isolated masses of mesoderm. Caldwell 

 applied the method of serial sections, and although he apparently made 

 some error regarding the posterior diverticulum, he saw and figured 

 paired localized cell-proliferations in the entoderm, which grew into the 

 blastocoele to form mesoderm. This I corroborated, and also gave some 

 tentative evidence for a similar origin to a posterior pair, and an an- 

 terior unpaired part. 



Professor Roule reverts to the old 'mesenchyme' notion, and we 

 might at least have expected some fresh evidence, but none is given. 

 He does not figure a single true section until a stage long past the 

 origin of the mesoderm. Nothing is given except a few optical sections, 

 which as such must be more or less diagrammatic. They merely show 

 a few loose cells, which may have come from anywhere, lying in the 

 blastocoele. He has not seen fit to furnish us with a single microtome- 

 section showing the origin of these cells, though he argues in the text 

 that they originate from the mesoderm. If he has sections illustrating 

 the origin of these cells, why does he waste two or three plates with 

 optical sections which shew nothing more, but rather less, than his 

 predecessors of twenty years ago. And it is for evidence such as this 

 that my researches are not to have 'la moindre créance'! It is almost 

 incredible that Professor Roule should attempt in 1900 to prove the 

 origin of the mesoderm in a much controverted case by a series of op- 

 tical sections which do not exhibit a single approach to a cell-division. 

 In my paper 1 drew a close comparison between the method of meso- 

 derm formation in Tornarla (according to Morgan) and that of Acti- 

 notrocha. Neither has a true enterocoelic formation, but a modified 



