144 
Agnes Robertson. 
synapsis stage, the spindle of the reduction division was found in 
two cases (Figs 27a and b, and 28); two were also found in material 
collected on June 1st, and one on June 7th, of this year. I have 
not been able to count the chromosomes with any certainty, partly 
because in almost every case the spindle appears in two successive 
sections. At this stage there is a mass of starch at each end of 
the mother cell (Fig. 28). The next step in the development is 
represented in Fig. 29 which shews the two daughter nuclei 
separated by a wall, though still connected by threads traversing 
the partition. Starch is present in both the daughter cells, which 
are of unequal size. Both cells divide again, but the division of 
the lower cell precedes that of the upper one. Fig. 30 shews a 
stage at which the division of the lower cell is completed while the 
upper cell is still in the spindle condition. (In the early division 
of the lower and larger cell and the obliqueness of the upper 
spindle Torreya resembles Lari, r.) 1 In this case abundant starch 
occurred in the lowest cell and in the cell containing the spindle, 
while it was almost or entirely absent from the middle cell. A 
later stage is that drawn in Fig. 31. Four potential megaspores 
have been produced ; the imperfectly preserved wall separating the 
two upper ones is nearly longitudinal, but in the next section to 
that drawn the lower part of it was seen to be more oblique. Here 
starch was again absent in the second cell from the base, though 
abundant in the other three cells. The absence of starch in one 
of the four spores seems curious until we recollect that at the 
period of the reduction spindle it was massed at the two ends of 
the mother-cell, and hence it is now naturally confined to the basal 
spore, and the two sister spores which occupy the upper end of 
the mother-cell. The starchless cell is peculiar in having a tuft 
of fibrils extending laterally from its nucleus to one of the side 
walls. The basal spore—the embryo-sac—now increases greatly 
in size, and the three cells above it are flattened by its growth, 
their nuclei becoming homogeneous and darkly staining. This 
stage is illustrated by Fig. 32 which was sketched from an ovule 
gathered on June 22nd of this year. The three ovules from which 
Figs. 30, 31, 32 were drawn are the only ones in which I have 
actually observed the division into four of the megaspore mother- 
cells. 
In Taxodium, Coker 2 records the division of the megaspore 
1 H. O. Juel. Beitriige zur Kenntniss der Tetradentheilung. 
Prings Jahrb., Bd. 35, 1900, p. 626. 
s loc. cit. 
