HABITS, ETC., OF CEEEBRATULUS LACTEUS. 151 



6. The blastomeres may be entirely separated during early 

 segmentation, with the result that each one produces a per- 

 fect undersized pilidiura. 



Conclusions. — Recent investigations (3 and 5) have dis- 

 closed in living eggs of Echinoderms, Annelids, and Molluscs 

 activities similar to those just described, and filaments which 

 were probably of the same nature were found in preserved 

 eggs of Amphioxus and the Amphibia. The first observer of 

 these activities (5) interpreted them as establishing the physio- 

 logical continuity of the blastomeres during cleavage, and all 

 subsequent investigations have only emphasised this inference. 



Our first conclusion, therefore, reiterates the same truth : 



1. The spin-threads connect the blastomeres and 

 prevent any physiological separation during seg- 

 mentation. 



Few eggs undergo more complete separation of the blasto- 

 meres than those of Cerebratulus, but when most distinctly 

 separated they retain their physiological unity by means of 

 numerous cytoplasmic commissures. Evidently these spin- 

 threads are portions of the internal protoplasm which have 

 flowed out to the exterior, and are formed in the same way 

 as the filose pseudopodia of Radiolarians and the larger 

 pseudopodia of amoebae. 



We should expect to find them, therefore, as suggested 

 (3), " high roads for the passage of living substance " be- 

 tween the blastomeres, and between them and the polar 

 bodies. And under specially favourable circumstances the 

 actual movement of granules along certain threads may be 

 seen. The current runs now outward and again inward, as 

 in a Radiolarian's pseudopod. 



Hence in those filaments which reach from one blastomere 

 to another, and remain for any length of time, there would 

 be a greater or less transference of living material. Seg- 

 mentation in such eggs cannot be regarded as a separation 

 of the egg substance into independent " cells," but becomes 

 rather a differentiation of different portions of the same unit 

 inass for different purposes, 



