346 Thos. IT. Montgomery. 



of each protozonite appear to move caudad, and others at its posterior 

 border to migrate cephalad, whereby transverse clefts come to sub- 

 divide the mesoblast. I could not positively ascertain whether this 

 itself is due to change in position of cleavage spindles at certain 

 points or to a creeping of cells. Immediately afterward each pro- 

 tozonite divides into a right and left somite, between which is the 

 ventral nucleus lined only by ectoblast. This movement also is 

 limited to the mesoblast, it marks the separation of a right and left 

 half of the body, and seems to be directly due to automatic movement 

 of mesoblast masses away from the midline. One of its immediate 

 effects is the outgrowth of the appendages with their enlarging 

 coelomic cavities. There is no likelihood that the ectoblast pulls out 

 the mesoblast in the formation of the appendages, rather mesoblast 

 must push out ectoblast. But whether the appendages are originated 

 by outpushing of the somatic mesoblast itself or by movement of yolk 

 fluid into the coelomic cavities is an open question. 



After the first appearance of the appendages other changes occur, 

 of which the first three are movements of the ectoblast that had up 

 to this time remained rather passive. First, the ectoblast thickens 

 right and left of the midline, forming the central nervous system. 

 Unlike the mesoblast segments, the ganglia are paired from the start ; 

 this is due to vertical mitoses at segmentally separated paired points. 

 Second, there is invagination of the stomodaeum and of other parts 

 of the cephalic lobe (cerebral ganglia, anterior and jDOsterior lateral 

 vesicles) ; this is in part due to vertical position of the mitotic spin- 

 dles, but ecpially to invagination of layers of cells engendered probably 

 by mechanical pressure of other parts. Third, from the extraem- 

 bryonic ectoblast arise by delamination the blood cells, and these 

 gradually wander into the embryonic body probably by amoeboid 

 movement. Lastly, there is gradual extension dorsad of the somites 

 by increased cell division at their dorsal borders. 



Summarizing the various movements just detailed, we find they 

 all can be reduced to two groups of phenomena: (1) cytoplasmic 

 movements, (2) interaction of cytoplasmic movements with external 

 pressure and strains. Of the first group we find : (a) movements of 

 the cytoplasm as a whole before cell membranes become established ; 



