178 TORREY 



and leiotropically, so much so in the last-named annehd that 

 no cross is formed. TJialassema seems to be intermediate ; the 

 posterior arms conforming to the first type, and the anterior to 

 the second. In this division, too, as Tread well ('oi) has pointed 

 out, we have a clear foreshadowing of the bilaterality of the 

 embryo : for the posterior cells divide typically equally, whereas 

 the anterior cells bud off smaller cells toward the rosette. This 

 difference is quite common in annelids, occurring as it does in 

 Nereis, Amphitrite, Clymenella, Arenicola and Podarke. At the 

 same time with the formation of the cross the lower sinistral 

 cells (2<^2> ^tc.) in the second quartet divide unequally (PL I^ 

 Figs. lo and 1 1). In the a, b and c quadrant the smaller cell 

 is budded off upward toward the prototroch (PI. I, Fig. lo), 

 but in the d quadrant downward and lies over the M cell (PI. 

 I, Fig. ii). An exactly similar division occurs in Podarke; 

 and, as there, this cell will be designated x\ o^. Treadwell ('99, 

 '01) in his discussion of cell-homologies, calls attention to the 

 fact that this cell has the same origin in all the forms of an- 

 nelids and molluscs that have been studied. This, perhaps, 

 cannot be said in regard to its fate, for in Amphitrite and Podarke 

 it forms a part of the proctodseal wall, but in TJialassema, in all 

 probability, that of the stomodaeum. The posterior quadrant 

 is now very effectively marked out by the cross and this cell. 

 A cross-section of this stage (Text-Fig. 10,^) shows a peculiar 

 arrangement of the blastomeres. The prototrochal cells are 

 entirely superficial and lie over the second quartet cells, which 

 in turn abut against the cross and intergirdle cells. The lower 

 members of the third quartet, too, extend inwards and upwards 

 for a considerable distance. It will be noticed in the following 

 table that the 64-cell stage in TJialassema does not quite con- 

 form to the ideal, as it does in Lepidonotus and PodarJ^e. This 

 is due to the precocious formation of the cross and the failure 

 of the upper dextral cells in the second quartet to divide. 



Up to this stage in all annelids with equal cleavage the ar- 

 rangement of the blastomeres is practically identical. As a 

 matter of convenience the later divisions of each quartet will be 

 described separately. 



