No. 2.] CLEAVAGE OF ARENICOLA CRISTATA. 8 1 



more generations than do the paratrochal cells in Amphitrite. 

 A cell in the segmenting egg of Arenicola arises in exactly the 

 same manner as the head-kidney cell of Nereis, but remains as 

 a part of the ectoderm. 



These facts seem to bear more or less directly on the ques- 

 tion of the mosaic theory, for, if we have a true mosaic in 

 these cases with so great uniformity of cleavage up to a stage 

 of a hundred cells or so, may we not expect to find a somewhat 

 closely corresponding uniformity in the fate of the blastomeres? 

 Of course the view is possible that each egg is, so to speak, laid 

 out in a different mosaic, even though the forms of cleavage 

 may correspond, but is it probable ? 



There is, however, a third argument against the mosaic 

 theory as related to the oblique cleavage, which seems more 

 conclusive than either of the preceding. 



About a year ago Crampton's ('96a) experimental work on 

 Illyonassa appeared. It was followed by an appendix by Wil- 

 son ('96b). In the latter paper Wilson asserts that the results 

 of Crampton's work confirm his view that the cleavage of 

 Nereis is "a visible mosaic-work." Crampton succeeded in 

 separating the segmenting egg of Illyonassa into halves, quar- 

 ters, and eighths, which were capable of further development. 

 He asserts that the parts, with the exception of a few details, 

 segment exactly as they would if the rest of the egg were pres- 

 ent, and that therefore Illyonassa possesses no " postgenerative " 

 or "regenerative" power, and that the cleavage is a mosaic. 

 He noticed, it is true, a few changes in direction of cleavage, 

 or in size of the products in the partial embryos, but explains 

 them as the result of lack of pressure, owing to the absence of 

 the other blastomeres. If differences of pressure are able to 

 change the form of cleavage so easily here, it is strange that 

 changes do not occur in the normal cleavage where in many 

 cases, as far as can be seen, the spindles lie nearly or quite in 

 the direction of greatest pressure, and the division pushes 

 whole groups of cells from their positions temporarily. 

 Furthermore, in no case was Crampton able to bring the 

 larvae up to a stage late enough to determine that "regenera- 

 tion " could not take place. He did, it is true, obtain partial 



