530 ADAM SEDGWICK. 



this view, e. g. the obviously reticulated nature of the proto- 

 plasm (see particularly fig. 9 in Lankester's paper), the large 

 number of nuclei, and, finally, the germ cells. 



These facts, though not in any sense proofs of the view of 

 the origin of the Metazoa for which I am contending, are at 

 any rate suggestive of it, and, so far as they go, in favour of 

 it. That is to say, they suggest the view that there would be 

 two courses open to a Protozoon after it had reached a size too 

 great for the proper nutrition of its central portion, i. e. a size 

 in which the ratio of surface to mass was unfavorable ; it would 

 either divide, in which case it would remain a Protozoon, or 

 it would develop from its vacuoles a system of connected and 

 specialised channels with a definite communication to the 

 exterior at one or more places. In the latter case it would 

 constitute the first stage in the evolution of the Metazoa. 



There is no reason to suppose that the protoplasm of such a 

 form, even though partially broken up into areas round the 

 various nuclei, would thereby lose the power of taking in to 

 itself foreign substances which were presented as nutriment; 

 in other words, would not prevent it from discharging the 

 functions discharged by the protoplasm of all Protozoa, and 

 by the parenchyma cells or phagocytoblasts of Metschnikofi'. 



To sum up, while fully agreeing with MetschnikofF, that the 

 formation of the endoderm by invagination of the wall of a 

 hollow blastosphere is a secondary process, I cannot accept his 

 position that the hollow blastula is a primitive form, or that the 

 formation of the endoderm by migration inwards of the cells 

 is a primary process. It seems extremely probable that the 

 blastula has arisen to provide for the better nutrition of the 

 growing embryo, and that the inwaudering and invagination 

 are alike secondary processes, the object of which is, when the 

 proper stage is reached, to get the protoplasm back to its 

 central position and ready for the development of the system of 

 channels which render its maintenance in the inside possible. 



(2) The mesoderm in Peripatus arises from certain nuclei in 

 the middle ventral line behind the blastopore. These nuclei 

 may, as I have attempted to show above, fairly be regarded as 



