294 A. #, W. SCHIMPER. 
treatment, for days or weeks, with alcohol, causes these cor- 
puscles to become smaller and more resistant. This effect 
is immediately produced when they are treated with a 
watery iodine solution, and this is the best means for 
examining them ; they then become stained of a darker or 
lighter yellow colour, according to the concentration of the 
iodine solution. Millon’s reagent colours them, when coa- 
gulated, a brick red, and nitric acid colours them yellow. 
These reactions indicate that the bodies in question consist 
of albuminous substances. 
The investigation of the earliest stages shows that these 
corpuscles are present before the starch-grains, and that the 
starch-grains which are subsequently produced bear the 
same relation to these corpuscles as regards their develop- 
ment as the starch-grains which are developed in assimilat- 
ing cells do to the chlorophyll-corpuscles which are present 
in those cells. The starch-grains may be produced at any 
‘points within these corpuscles, or their formation may be 
confined to the peripheral portion. 
The starch-grains which are developed in the peripheral 
portion of these albuminous corpuscles, and which therefore 
have one side free at an early period, have an excentric struc- 
ture, and the hilum lies near to the free end (figs. 19, 23, 
36—40, 48), just as in the excentric starch-grains which 
are produced by chlorophyll-corpuscles. If the starch- 
grains come into contact with other of the corpuscles 
prominences of various forms are developed upon them. In 
the cases in which I have as yet investigated, the starch- 
grains which had been developed in the interior of one 
of these corpuscles were compound; the minute grains 
composing them rarely exhibited any perceptible differen- 
tiation, but when they did it was always concentric (figs. 
24-29). These corpuscles, like the chlorophyll-corpuscles, 
become larger at first after the formation of the starch- 
grains, and this is usually accompanied by a diminished 
refrangibility ; they then become smaller and gelatinous, 
and finally they altogether disappear. 
The behaviour of these bodies indicates that they are the 
starch-forming organs in cells which do not assimilate, that 
is, that the conversion into starch of the assimilated sub- 
stances which have been conveyed from other parts of the 
plant is effected by means of them. 
In the cases in which a starch-grain is formed in the 
interior of one of these corpuscles its function is obvious. 
In those cases in which the starch-grain is formed in the 
peripheral portion of the corpuscle, and soon projects freely 
