288 Ishikawa—Studies on the Embryo Sac 
and a generative one (PL VII, Fig. i>g) in the peripheral portion. The 
former is presented as a spherical chromatin mass showing a sign of disinte¬ 
gration, and embedded in a plasma mass, while the latter is smaller and 
ellipsoidal, and is surrounded by a thin plasma sheet. The rest of the 
contents is impregnated with immense numbers of hyaline, fusiform 
corpuscles (Pl. VII, P'ig. i), which scarcely stain with the usual chromatin 
dyes, so that they are apt to be overlooked, and the whole contents of the 
pollen grain are taken for an alveolar plasma mass. These corpuscles are 
typical fusiform or oblong ellipsoidal in shape, with acute or often obtuse 
ends, and show the reaction characteristic to starch by treating with iodine 
solution. Starch grains were often found in pollen grains of many plants; 
they were spherical, ellipsoidal, or rod-like, but not fusiform as in the case 
of Oenothera , although Guignard ( 24 ) reported that there were fusiform 
starch grains in the pollen and pollen-tube of Najas major . Thus, it was 
found necessary to test whether they were really starch grains or not. But, 
having no fresh material at hand, Oenothera odorata was used for this 
purpose, whose pollen grains are quite identical with those of Oe. pycno - 
carpa and of Oe. nutans in every minute respect. The fresh pollen grains 
were crushed under the cover-glass, and several reagents were applied upon 
them. The result was as follows : By adding iodine solution the corpuscles 
in question turned blue, while by heating they became decolorized and 
swollen ; they swelled also when treated with chloral hydrate; they 
disappeared after standing 6-20 hours in 1-2 per cent, solution of ‘ Taka- 
diastase ’, keeping them in 6o° C. These reactions leave no room for doubt 
that they are no other than starch grains. 
The pollen-tube reaches the embryo sac at least in 48 hours after 
pollination. In an experiment, some pollen-tubes have arrived in the ovules 
in 53 hours and 30 minutes; in this case some eggs were already fertilized 
or just being done, but some embryo sacs were laid free from invasion of 
tubes. The plasm of the pollen-tube contains a large number of starch 
grains which are fusiform, ellipsoidal, or spherical, but not so uniform as 
observed in the pollen grain ; this fact suggests that starch grains undergo 
some chemical changes, owing probably to the use as nutrition of the tube. 
PL VII, Fig. 2, illustrates the lower portion of the tube travelling 
through the stylar tissues, and carrying a disintegrating tube nucleus (v) 
and two male nuclei ( m.n .) which are in the last telophase. In the upper 
portion of the tube, though not shown in the illustration, a series of 
stoppages or cross septa are laid down, which showed the reaction charac¬ 
teristic to callose, being stained red with corallin-soda. Nodal swellings 
and branching of the tube were often observed ; the latter case was always 
found in the micropylar portion, and made it difficult to distinguish whether 
two different tubes or a pair of branches were present. The wall of the 
tube is never found of a uniform thickness, particularly in the apical portion. 
