366 BOTANICAL GAZETTE [may 



groups (fig. 1,3). Sometimes the middle third of the proembryo is 

 entirely free from nuclei; sometimes the nuclei are distributed 

 throughout, but are smaller and more numerous above; sometimes 

 the nuclei are dividing above but resting below, and later they 

 divide below while the nuclei above are resting; there may be 

 evanescent segmentation above, but no trace of it below; later, 

 there is a migration of nuclei toward the base, followed by seg- 

 mentation below, with none above; the embryo is formed below, 

 while the upper region furnishes nutrition. In fact, there is a con- 

 stant manifestation of polarity from the appearance of the arche- 

 gonium initial throughout the life of the plant. The source of 

 nutrition is probably the cause of a part of these phenomena, but 

 much is still to be explained. 



Cause of free nuclear period. — Why the cycad embryogeny 

 should begin with a period of free nuclear division is not hard to 

 imagine. It would seem a physical impossibility for the small 

 mitotic figure of the first division to segment the comparatively 

 immense mass of the egg. Further, the large amount of nutritive 

 material doubtless causes the nuclear divisions to follow in rapid 

 succession. That they do follow in rapid succession is shown by 

 the fact that the 2-, 4-, and 8-nucleate stages are very rarely found; 

 and also by the fact that nuclei which have reached the resting 

 condition are not likely to be found before the 16-nucleate stage. 

 Further, a glance at the illustrations accompanying this paper will 

 show that the mitotic figures, nearly 100, are all in the equatorial 

 plate stage, a fact which indicates that other phases are compara- 

 tively rapid. 



The small size of the mitotic figure, the great mass of the egg, 

 and the rapid succession of mitoses doubtless cause the free nuclear 

 period which characterizes not only cycads, but nearly all living 

 gymnosperms. As the divisions proceed and the mass of the 

 nucleus approaches that of the surrounding cytoplasm, segmenta- 

 tion, which may be more or less evanescent, appears; and later 

 permanent segmentation with the formation of cellulose walls 

 takes place. In the Bennettitales and earlier cycadophytes the 

 seeds were much smaller than in any of the living cycads yet studied. 

 It would be interesting to know the embryogeny of these forms. 



