296 A. F. W. SCHIMPER. 
nothing can be seen of them, or of the corpuscles which have 
produced them. 
The starch-forming-corpuscles of many plants have the 
Same properties and are formed in the same way as those 
which have been described above, but they differ from them, 
more or less, in the mode in which they produce the starch- 
grains. In many instances the starch-grains are produced 
in the manner described above, but the number of them is 
smaller, and they attain a more considerable size. This is 
the case in the rhizome of Amomum cardamomum, which 
supplies excellent material for investigation. 
The starch-grains in the rhizome of 4A. cardamomum} 
(figs. 16—20) are large and club-shaped when mature, and 
they exhibit distinct internal differentiation. The hilum is 
very excentric, and lies towards the thicker rounded end; 
the other end is truncate. Compound grains, consisting of 
two or three smaller ones, are not uncommon. 
The starch-forming-corpuscles resemble those in the epi- 
dermis of Philodendron in respect of size, form, and mode of 
development, but they are paler and less stable. The starch- 
grains are formed in them just under the surface, or even on 
it, and are either solitary or two or three together. They 
are at first hemispherical, attached by the flat surface to the 
starch-forming corpuscles, which are flattened at the points 
of contact ; they subsequently assume a club shape. The 
hilum always lies towards the free end of the grain. When 
two or three starch-grains are formed, compound grains are 
produced. The starch-forming-corpuscles become at first 
larger, and at the same time less dense; at a later stage, 
when the starch-grains have nearly attained their definitive 
size, the corpuscles can no longer be distinguished, or deli- 
cate gelatinous remains of them can be made out by the use 
of iodine solution. When the grains are mature it is im- 
possible to detect any trace of the corpuscles. 
The starch-forming-corpuscles in the rhizome of Colocasia 
antiquorum differ from those which have just been described 
only in that they give rise to starch-grains throughout their 
whole mass. The very numerous grains formed in a cor- 
puscle cohere to form spuriously compound grains. This 
mode of the development of starch-grains is more fully de- 
scribed in the following example, in which I have studied it 
more closely. 
The starch-grains in the endosperm of Beta trigyna 
(figs. 30—32) are large, spherical, or somewhat elongated, 
1 T obtained this from the Botanic Garden at Strassburg. As the plant 
was not in flower at the time I cannot vouch for its identity. 
