10 Stopes and Fujii, Tlie nutritive relations of tlie surrounding tissues etc. 
The migration of the food also follows much the same course 
as in the Cycads. We had however a more complete series of 
materials of different stages of development after the f'ormation of 
the starch grains in the endosperm, than we had for the Cycads, 
and will now summarise the results of observations on material 
collected at intervals of every 2—3 days. TV^ith Ginkgo there is 
great uniformity in the development of the different ovnles on the 
same tree or even on different trees growing in the same place, so 
that more value attaches to dated observations lor Ginkgo than for 
the Cycads where there is mnch irregularity even in one and the 
same cone. 1 ) 
For this pnrpose we nsed principally microtome series, passed 
into water or alcohol as the case demanded, and examined in 
lodine. In the conrse of one month (Ang. 28^ to Sept. 30 th ) the 
changes were as follows. Beginning with a stage in which the 
endosperm cells in general contain-ed many starch grains, we fonnd 
the grains smaller in the cells nearer the archegonia and a little 
langer in the cells of the jacket layer, while the egg cell was free 
from starch. In the next stage the grains in the jafcket cells were 
smaller and stained a brownish rather than the true blue violet of 
the storage starch of the rest of the endosperm: very small 
brownish violet grains also began to appear in the egg cell. This 
difference in the natnre of the' starch in the sheath cells was also 
observed in permanent preparations of microtome series which had 
heen stained with Flemming ? s triple stain from which the Gentian 
violet was almost entirely washed out. In these the large storage 
starch grains of the usual endosperm cells were stained pale llesh 
colour, while the grains in the jacket cells were blue. These 
reactions certainly show that the starch grains in the two regions 
were in somewhat different conditions, probably indicating that the 
starch in the jacket cells was just being transformed into soluble 
carbohydrate by diastase secreted in the jacket cells. In later 
stages the starch steadily decreased in the jacket cells, in which 
the grains were sometimes grouped together to one side of the 
cells in a curious manner (like the arrangement in the **statolith v 
starch grains) for which we have as yet no explanation. In these 
stages the number of starch grains in the egg inereased. The 
jacket cells then emptied themselves of starch, beginning at the 
base of the Archegonium tili they were finally completely emptied 
in ab out 3 weeks from the first mentioned date. The emptying of 
the cells of starch spread in an outward direction in the endosperm 
tili the zone immediately round the archegonia for 10—15 cell 
layers was free from it, just as was the case in the Cycads 
(cf. figs 7 and 8). We judged in both cases from the various facts 
observed, that this starch, temporarily stored in the endosperm, 
was being transformed into soluble carbohydrate and passed into 
the egg cell. 
fr Dr. Iliyake informed us of great irregularity in tlie development of 
tlie ovules of Zamia floridana. 
