106 GENERAL CONSIDERATIONS. 



seems fairly generally to happen that at the moments of 

 activity the segments round themselves off, and in the intervals 

 of rest flatten out against each other. These facts seem to me 

 to indicate that it may be possible to find a purely mechanical 

 explanation of complete cleavage. However this may be, it 

 seems pretty clear that the holoblastic segmentation of small 

 ova has not the phylogenetic significance usually ascribed to it. 



To sum up, the ancestral Metazoon has generally been 

 assumed to be a colonial Protozoon, and when we examine the 

 evidence for this view we find that the holoblastic segmenta- 

 tion, which really suggested it, is totally opposed to it; and 

 further, that the facts of incomplete cleavage which were 

 thought to be opposed to it are somewhat in its favour, though 

 much more suggestive of another view, which I will now 

 consider. 



In Chapter II I suggested that the ancestral Metazoon was 

 not a colonial Protozoon, but a multinucleated Infusorian-like 

 animal with possibly a mouth leading into a central vacuolated 

 mass of protoplasm ; and that the evolution of the higher 

 forms has consisted mainly in a definite arrangement of the 

 nuclei and of the specialisation of certain of the vacuoles in 

 the internal protoplasm into the cavities of the organs, and 

 of the protoplasmic strands between into the walls of the latter, 

 and into nerves, muscles, &c. 



This is not a new view ; it is the old view of the origin of 

 the Metazoa, and has been held recently by Saville Kent 

 (No. 45, vol. ii, p. 480) and others. It is entirely in 

 accordance with the facts of the development of Peripatus 

 capensis. 



With regard to this development, we have to observe that 

 we cannot speak of cells till a comparatively late period 

 (Stage b), and that the intimate structural change underlying 

 the processes of growth of the young embryo is not an increase 

 of cells, but a multiplication of nuclei. First of all a cortical 

 layer of nuclei, lying in the peripheral layer and entirely sur- 

 rounding, excepting at one point, the vacuolated spherical mass 

 of protoplasm of which the embryo consists, is differentiated. 



