LIFE HISTORY OF PINUS 1 27 



plate during the ordinary process of cell-wall formation. That 

 the granules are larger and the details of the process more 

 striking here may be accounted for by the fact that, under the 

 influence of each nucleus, three times as much cell-wall must 

 be laid down as is ordinarily formed by the action of a single 

 nucleus. But in any case, we are still far from a satisfactory 

 understanding of the method by which cell-walls arise. 



The eight nuclei are arranged, as usually described, in two 

 tiers of four cells each. The cytoplasm of the upper four cells 

 remains continuous with the cytoplasm of the egg, that is, a 

 dividing wall is not formed along their upper surface (figs. 

 255, «, and 255,3). 



Later Mitoses in the Formation of t/ie Proembryo. — The 

 second set of division figures which occurs at the organic apex 

 of the egg arises in the upper tier of cells, that is, in the four 

 cells which have never been cut off from the general cyto- 

 plasm of the Q^^ (fig. 256). This is contrary to all reports 

 of the development of the proembryo in the Abietinece} The 

 second division occurrmg m the nuclei at the base of the 

 archegonium has not been previously observed, so far as I 

 am aware, and, the third division occurring in the basal tier 

 of cells, the inference seems to have been made that the cells 

 which are not enclosed along their upper sides by definite 

 walls never divide. Coulter and Chamberlain ('oi) make the 

 remark that the upper four free nuclei increase much in size, 

 and they figure them in the spireme stage ; but they do not refer 

 to the fact that they are in the prophase of division, and describe 

 all further mitoses after the eight-celled stage as occurring in 

 the basal tier of cells. Strasburger and Hillhouse ('oo) also 

 describe the further development of the proembryo in Picea 

 after, the establishment of cell-walls, as proceeding from two 

 successive divisions of the four basal cells. 



It seems to me a rather significant fact that the four cells 

 which remain in open communication with the egg should not 

 only divide again, but that their division should be entirely 

 completed before the cells of the lower tier show any signs of 

 dividing. There are thus, in Pimis, four successive mitoses 



' See note at close of Appendix. 



