CELL DIVISION IN EGGS OF CREPIDULA. 545 



egg and sperm centrosomes, since the latter had disappeared as such, before the 

 first appearance of the cleavage centrosomes. 



In the eggs represented in figs. 159-163 various stages in the division of the 

 egg centrosome and in the formation of the egg amphiaster are shown. That this 

 centrosome and amphiaster actually belong to the egg and not to the sperm 

 is beyond doubt, the sperm nucleus being so far removed from the egg nucleus 

 and so isolated from it by intervening yolk that there is no possibility of confusing 

 the two. In the figures mentioned the sperm centrosome and aster are not visible, 

 if present. In fig. 164 two spindles are present in the egg, one of which is certainly 

 the egg spindle and the other the sperm spindle; the same is true of fig. 168, in 

 which the two spindles are entirely distinct. In all other figures on plate LV, 

 viz., figs. 165-167, 169-170, the two spindles have not remained distinct, but 

 their poles have interfered, thus giving rise to tetrasters. 



I have discussed these figures elsewhere (Conklin, 1904) and need not com- 

 ment on them here further than to call attention to the fact that the origin of 

 amphiasters in connection with each of the germ nuclei, when the latter are kept 

 apart, indicates that the egg centrosome does in this case persist, and that it 

 may be stimulated to divide and form an amphiaster by methods which have 

 elsewhere been employed to produce artificial parthenogenesis. In all proba- 

 bility this is just what occurs in every case of normal or artificial parthenogenesis; 

 the cleavage centrosomes being derived from the egg centrosome, and not being 

 formed de novo. 



Kostanecki (1906, pp. 386-394) has commented at some length on my con- 

 clusions as to the origin of the cleavage centrosomes in Crepidula under normal 

 and artificial conditions. So far as the origin of these centrosomes under normal 

 conditions is concerned I again admit, as I have done in previous papers (1901, 

 1902, 1904), that the evidence is not conclusive that they come, one from the 

 egg centrosome and the other from the sperm centrosome, since both of the 

 centrosomes have disappeared as such before the cleavage centrosomes appear. 

 On the other hand the persistence of the egg and sperm spheres until the union 

 of the germ nuclei, the fusion of the spheres and the appearance of a cleavage 

 centrosome in connection with each of the germ nuclei, the absence at all stages 

 of the normal process of an amphiaster in connection with either the egg or the 

 sperm nucleus— these incontrovertible facts do not afford evidence for the view 

 that the cleavage centrosomes always come from the sperm centrosome. I freely 

 admit that in many instances such an origin may be clearly traced. In certain 

 ascidians which I have studied (Conklin, 1905) there are no centrosomes in either 

 the first or second maturation divisions, and none at any stage in connection 

 with the egg nucleus; while the origin and division of the sperm centrosome and 

 its transformation into the sperm amphiaster and into the first cleavage spindle 

 can be determined with the greatest certainty. There is here a great difference 

 between Ciona or Cynthia and Crepidula or any other gasteropod. 



These differences may be summarized in the following tabular statement: 



36 JOURN. ACAD. NAT. SCI. PHILA- VOL. XV. 



