The Development of Theridinm. 347 



and {h) niuvemonts of particular cells or particular groups of cells 

 after cell membranes have appeared. That is, the formation of rela- 

 tively firm cell membranes inhibits or restricts movements of the 

 cytoplasm as a whole, it reduces such movements to intracellular 

 processes. Therefore, cell membranes are not only of infportance in 

 metabolism^ acting as selective osmotic membranes, as the physiolo- 

 gists have proven, but their presence is likewise a factor in embryonic 

 differentiation, for they divide the embryo into somewhat indepen- 

 dent territories, each of Avhich then pursues its own dilferentiation 

 process more or less apart from the others. Conklin (1902) has most 

 ably and minutely analyzed cell movements during development and 

 resolved them into cytoplasmic flowings, and has shown how impor- 

 tant a factor they are in differentiation; his observations were made 

 upon the holoblastic egg of a Gasteropod. But I believe the modify- 

 ing influence of the presence of c^ll membranes upon these move- 

 ments has not been pointed out. Then, in the Spider's egg, the cyto- 

 plasmic movements after establishment of cell membranes are of two 

 kinds: (1) intracellular movements leading to change in position of 

 axes of mitotic spindles; and (2) cytoplasmic flowings in the rest 

 stage that lead to amoeboid locomotion of single cells (as vitellocytes 

 and blood cells), and to dehiscence, probably also by amoeboid move- 

 ment, of masses of cells (as in the formation of the somites). Cells 

 that exhibit such locomotion are in a sense the most autonomous and 

 independent of all ; and their locomotion may be referable to par- 

 ticular chemotactic influences, though it must be borne in mind that 

 the idea ''chemotaxis" is not explanatory, but is often evasive of the 

 question at issue; it simply puts substance in the place of motion. 



Finally, I would discuss one complex movement of the embryo as 

 a whole, the reversion process, and we may first mention the explana- 

 tions that have been given before. Claparede (1862) explained the 

 process of reversion as accomplished by widening of the ventral 

 sulcus bringing the bases of the legs to the dorsal side, while the 

 movement of the tail lobe would be a consequence of this process.' 

 Barrois (1878) ascribed reversion to displacement of the anal seg- 

 ment producing modification in the angle of divergence of the 

 abdominal "germ bands" ; but he did not explain the cause of the 



