1906] CHAMBERLAIN—OVULE OF DIOON 347 
still undivided; but such ovules were from cones picked about the 
middle of March, and the ovules finally decayed without any division 
of this nucleus. The top of the typical archegonium in such material 
is shown in fig. 28, which is from a cone picked March 18, the material 
being fixed May 13, nearly two months later. The two neck cells 
are in perfect condition, but the central cell has begun to degenerate. 
In normal material the spirem is broad but very tenuous in consis- 
tency. In the only cases noted it had evidently suffered from reagents 
(fig. 29). The shortening, condensation, and segmentation of the 
spirem were not observed, the next available stage being shown in 
fig. 30, where the. segmentation into chromosomes has taken place. 
_ The splitting of the chromosome in the equatorial plate is shown in 
fig. 31. In this preparation the number of chromosomes was deter- 
mined with reasonable though not absolute certainty to be twelve. 
In a late anaphase the chromosomes, while still retaining the U-shape, 
are becoming irregularly moniliform, looking as if they might break 
up into small pieces (fig. 32). Only a few stages in the formation of 
the spindle were observed. The first indication of it is a granular and 
fibrillar appearance which is more marked at the lower pole of the 
nucleus (fig. 29). There are no centrosomes, and in fig. 30 the poles 
are rather blunt. Fibers like the spindle fibers are abundant in the 
cytoplasm of the papillate projection in which the figure lies. In 
fig. 31 the spindle is more sharply bipolar, and long mantle fibers are 
more conspicuous, although many of the spindle fibers are still con- 
tinuous from pole to pole. Fig. 32 shows no trace of the granules 
which mark the beginning of a cell plate, and later stages make it 
certain that no wall is formed between the daughter nuclei. The 
ventral canal nucleus remains free in the cytoplasm of the egg, as 
shown in jigs. 33-35. 
A ventral canal cell in cycads was first described by STRASBURGER 
(18) in 1876 for Cycas sphaerica. The next year WARMING (23) 
described one in Ceratozamia robusta, but soon concluded that he had 
been mistaken. Treus (22) in 1884 failed to find any ventral canal 
cell in Cycas circinalis, and from that time it was generally believed that 
the cycads have no ventral cell. However, in 1898 IkENo (10) made an 
unmistakable demonstration of the critical mitosis in Cycas revoluta, 
At that time no ventral canal nucleus not separated from the egg 
