360 E. W. MACBRIDE. 



adjacent portion of tbe atrial cavity has given the impression 

 that the atrial cavity grew upwards and displaced other 

 structures, which is not the case. 



(6) Whilst Kowalevsky's main idea as to the manner in 

 which the atrium is formed is therefore correct, the descrip- 

 tions given by Lankester and Willey of the structures seen 

 by them are quite correct, but the consideration of additional 

 facts renders it impossible to accept their interpretation of 

 the processes of growth involved. 



Before closing this account of the development of Amphi- 

 oxus I should like to refer to another criticism of my former 

 paper, and also to some papers on the same subject which 

 appeared subsequently to its publication. The criticism 

 referred to is that of Klaatsch (1). In discussing my paper 

 he remarks (1) that my figures do not awaken much confidence 

 in the state of preservation of the specimens ; (2) that my 

 position as regards the formation of the mesoderm is governed 

 by a rash and exaggerated comparison with Balanoglossus. 



In reference to Klaatsch's first remark, I should like to say 

 that the figures do not satisfy me, and are very far from doing 

 justice to the preparations. I regret to say that my powers 

 of draughtsmanship are not first-class, and that the figures 

 were somewhat hurriedly executed owing to my having to 

 leave Cambridge for Montreal, and not being able on that 

 occasion to take my preparations with me. Nevertheless, 

 imperfect as they are, they give a far better idea of what is 

 actually seen in well-preserved preparations than the highly 

 schematic figures of many authors, who represent an epithe- 

 lium of clearly defined pillar-like cells with a nucleus showing 

 in each. It is astonishing how often a first glance at a pre- 

 paration will give the impression of such a structure, when 

 more careful examination will only reveal a row of elongated 

 nuclei, with here and there a cell limit seen. In particular, 

 in Amphioxus the whole protoplasm both of ectodei^m and 

 endoderm is so loaded with yolk that cell limits are excessively 

 difficult to make out once the blastula stage has been passed. 

 I was careful to base my work only on preparations which 



