No. 3.] THE CELL-LINEAGE OF NEREIS. 447 



as if it formed a part of a normal embryo, giving rise to stages 

 of two, four, eight cells, etc., which are practically identical 

 with the corresponding halves of the four-, eight-, and sixteen- 

 celled stages, etc. As in the frog, a perfect half-blastula is pro- 

 duced. The opening then closes, and a blastula is then formed 

 which differs from a normal blastula only in being half the 

 normal size. The gastrulation and later stages follow the nor- 

 mal course. 



This result agrees essentially with Roux's, with the interest- 

 ing difference that the regeneration of the missing half takes 

 place much earlier. It proves conclusively that in the normal 

 development each of the blastomeres is profoundly influenced 

 by the other ; that the cell is not an isolated mechanism whose 

 mode of action is wholly predetermined in its molecular struc- 

 ture. It proves in fact that the form of cell-division is deter- 

 mined by two factors. The first factor is the inherited tendency 

 of the cell to pursue a definite course, a tendency which we may 

 assume exists by virtue of a corresponding molecular or proto- 

 plasmic structure {cf. p. 441). The second factor is the influence 

 upon the cell of other cells in the colony. When the second 

 factor is removed or modified, the first is correspondingly modi- 

 fied, and a complete readjustment takes place. I can see no 

 logical halting-point in the application of this principle between 

 the embryonic and the adult stages. If the experimental results 

 be well founded, — and it is impossible to question the validity 

 of the combined evidence, — then no cell in the embryo or in 

 the adult is isolated, not even the germ-cells. 



In view of these facts there seems to be very strong reason 

 to accept the conclusion that there is a causal connection in the 

 Nereis embryo between the bilateral distribution of material 

 and the ensuing bilateral divisions. I do not mean to assert 

 that this connection is independent of the inherited tendency 

 of the cells. It is probable, on the contrary, that the action 

 has produced an hereditary effect. This result, however, only 

 extends the causal connection over a longer period, and in 

 nowise diminishes its significance. 



We pass now to a discussion of the spiral period itself, which 

 has many interesting and significant features. The principal 

 datum on which the entire discussion rests is the fact that the 

 spiral form of cleavage has no necessary relation to the homology 



