364 BOTANICAL GAZETTE [may 



equatorial plate stage. In all of these the number of chromosomes 

 is 12, but the double character of the chromosomes is evident; 

 and since 12 and 24 are the haploid and diploid numbers, I believe 

 that there is a pairing of chromosomes at fertilization, as described 

 by Hutchinson (14) for Abies. 



Embryogeny 



First mitosis. — At the first division in the fertilized egg, the 

 mitotic figure is surrounded by a remarkable display of achromatic 

 structures, arranged in irregular nets or grouped into cones resem- 

 bling half-spindles (figs. 10, n, 25). This achromatic area easily 

 identifies the first mitosis, even when a second sperm has entered 

 and divided, since the achromatic area surrounding the dividing 

 sperm is much smaller (fig. 11). The division of an extra sperm 

 was observed in two cases, and in both the mitotic figure showed 

 12 chromosomes. After the first mitosis the achromatic display 

 gradually diminishes, but is still quite conspicuous as late as the 

 fourth or fifth division. After the first mitosis, however, the dis- 

 play consists almost entirely of slender threads like the spindle 

 fibers (fig. 26). 



First free nuclear period. — According to previous accounts 

 of the embryogeny of cycads and Ginkgo, there is a regular series 

 of free nuclear divisions, so that the nuclei at the successive divisions 

 number 2, 4, 8, 16, 32, 64, 128, 256, 512, and in some cases 1024, 

 with only such variations as might be expected from the occasional 

 failure of a nucleus to divide. In Dioon edule (12) it was noted 

 that the 8th, 9th, and 10th mitoses were irregular, especially in 

 the upper part of the proembryo, so that the number of nu'clei 

 beyond the 256-nucleate stage was likely to vary widely from the 

 theoretical estimate. 



In Stangeria, irregularities begin to appear earlier and are more 

 pronounced. The first 4 mitoses, giving rise to 2, 4, 8, and 16 

 nuclei, are likely to be strictly simultaneous and regular, and 

 the 32-nucleate stage was observed in two cases; but after the 

 16-nucleate stage (fig. 12), the number of nuclei not only varies 

 from the anticipated 32, 64, 128, etc.,, but the discrepancies are 

 great. An examination of a large number of cases showed that the 



