32 DEVELOPMENT OF STARCH. . 
sight it is almost impossible to help believing that the granule 
must have been built up in the same manner as a crystal, 
namely, by the deposition of fresh matter over the older, or, 
in other words, that the outer rings of the starch granule have 
been deposited over those which are more internal, and that 
therefore they are the youngest portion of the granule. But 
the observations of Nageli have proved this not te be the case, 
for he has shown that the appearance of stratification in the 
starch granule is really due to the difference in the quantity of 
water which exists in the different parts of the granule, and he 
has also proved that the outermost layer, instead of containing 
the greatest amount of water, as it ought to do if it was the 
youngest part of the granule, contains the least, while the 
nucleus on the other hand is the most watery of all. Nageli 
concluded from these observations that the growth of the starch 
granule was precisely the same as that of the cell-wall (see page 
42); namely, by intussusception of fresh particles of the starch- 
compound between those of an older date ; and hence that the 
regular alternation of dense layers with more watery ones around a 
nucleus or hilum produces the peculiar 
appearances of starch granules. That 
the different layers vary in density 
may be at once proved by the action 
of polarised light, when each granule 
usually exhibits a black cross. Seein 
then that the growth of the arch 
granule is by intussusception, it will 
be readily understood why it is that 
this growth cannot be carried on ex- 
cept so long as the granule is imbedded 
in the substance of the living proto- 
plasm, and that as soon as the proto- 
plasm of the cells in which the starch 
is being formed is used up or killed, 
all further development of starch be- 
comes impossible. 
In some cases, as for instance in 
the Huphorbiacee, starch granules are 
found floating in the contents (latex) 
of the laticiferous vessels (fig. 53), and 
Ree eorousrcerel — this would seem to be in contradiction 
liste opiiiusstarch granules to the above-mentioned law that 
of a peculiar dumb-bell and starch granules can only be formed 
eee while enveloped in protoplasm, but 
the mode of formation of these gra- 
nules has not been observed. 
The starch granules of different plants vary very much in the 
character of their hilum and in the distinctness and general 
arrangement of their concentric lines, in the same way, as we 
