HABITS, ETC., OF CRREBRATULUS LACTEUS. 143 



be either by the withdrawal of the filose threads of which it 

 was composed and the protrusion of others from a new point, 

 or by a gradual slipping along the surface. The latter hypo- 

 thesis is untenable from the very nature of the spin-threads, 

 and the most persistent watching failed to reveal any tempo- 

 rary withdrawal of the connecting protoplasm. We are left, 

 therefore, to the conclusion that the change is only apparent, 

 and that the actual point of attachment remains the same. 

 This affords a means of identification whereby we can follow 

 one particular portion of the surface during segmentation. In 

 this way we find, first, that there is no investing membrane in 

 contact with the surface of the egg, but that the protoplasm 

 of the latter forms an ectosarc similar to that of the amoeba. 

 This is lighter in colour and more transparent than the 

 remainder of the egg, partly because it contains no yolk 

 granules. 



Again, we find that the change of position is due to the fact 

 that the point of attachment remains stationary, while the 

 groove advances; in other words, the old ectosarc travels 

 inward, during the formation of the groove, only the limited 

 distance necessary to accommodate the more rounded outline 

 of the blastomeres, and the ectosarc which covers the bulk of 

 the groove is a new structure. This is easily formed, since it 

 consists essentially of egg protoplasm minus its yolk granules. 



Anticipating a little, we find that after the two blastomeres 

 have come together again, and the furrow has been filled in 

 by appression, the point of attachment does not return to the 

 bottom of the furrow, but, approaching it closely, remains at 

 a distance which varies within narrow limits for different eggs 

 (fig. 50). 



The two edges of the old ectosarc do not quite meet, but 

 a portion of the new ectosarc remains exposed along either 

 side of the furrow. This portion increases with subsequent 

 segmentations until the point of attachment comes to lie 

 near the centre of the cell in the 128-cell stage (fig. 45). 



This is the natural result, since the egg increases slightly 

 ill size with each segmentation. The protruding surface of 



