362 E. W. MACBRIDE. 



As far as regards the liead cavities, I was able by sections 

 to confirm Hatschek's account of the matter, with tlie trifling 

 addition that both head cavities originate from a common 

 evagination of the gut. 



So that, whatever view one may hold of the relationship of 

 Amphioxus and Balanoglossus, the fact remains that in 

 Amphioxus the mesoderm originates from five rudiments — 

 (a) the anterior unpaired pouch of the gut; {b) two dorsal 

 pouches situated far forward, giving rise to the first pair of 

 myotomes; and (c) two dorso-lateral evaginations of the gut 

 wall, which become divided into somites as the animal 

 increases in length. 



Samassa's paper (7) on the development of Amphioxus was 

 published after mine, but the work was carried ou simulta- 

 neously with mine, and in many points I regard his results 

 as a welcome confirmation of my own. It is curious that we 

 should have been both driven to use the celloidin-paraffin 

 method as the only possible way of dealing with such 

 delicate embryos. The points of difference between us which 

 I may notice concern (1) the segmentation, (2) the gastrula- 

 tion, and (3) the formation of the mesoderm. 



With regard to the first point, I asserted that at the close 

 of segmentation no difference is observable between the sizes 

 of the blastomeres at the two poles of the blastula. This con- 

 clusion was based on sections, but as orientation is impossible 

 with spherical objects, it is possible that my sections were 

 horizontal, and I have no doubt that Samassa's views, based 

 on an examination of fresh material, are correct. 



With regard to the gastrulation, Samassa's figures are in 

 many ways similar to mine, but he regards the lip of the 

 blastopore at which there is a well-marked separation of the 

 endoderm and ectoderm as dorsal, whereas I regard it as 

 ventral. 



Now the only method of determining which lip is dorsal 

 and which ventral is by getting hold of a specimen in Avhich 

 the first trace of the dorsal flattening preparatory to the 

 formation of the medullary plate is visible. Such a specimen 



