A leur one-grams in the Lupin, 165 
having however been formed. In the ripe seed the grains, 
which are roundish or somewhat angular through mutual 
compression, are still separated by a protoplasmic network 
in which oil-drops occur, while starch is wanting. Hanstein’s 
solution brings out the network and nucleus very well, 
staining these a deep violet, while the grains scarcely stain 
at all (Fig. 7). 
Solid inorganic constituents were repeatedly sought for, 
but without success. Sections of the ripe seed, from which 
the oil had been removed by ether, were treated on a slide 
with 1 per cent, of potash, which was allowed to diffuse in so as 
not to wash away any small globoids which might be present ; 
individual cells or grains were carefully watched meanwhile, 
sometimes under Zeiss’ F objective, at others under the D, 
but in all cases an empty space was left in the protoplasmic 
network after solution. Some granules scattered over the 
section, but especially, and almost exclusively, near the few 
cell-layers with very granular contents beneath the epidermis, 
and with no definite relation to the grains, proved to be small 
starch-grains washed out from these cells. No crystals could 
be detected by double refraction when such a section was 
examined under a polarising microscope. Hence we may 
conclude that the aleurone-grains of Lupinus digitatus have 
no solid mineral contents. From the foregoing facts it ap- 
pears that the presence of mineral matter is of very secondary 
importance in the development of the grains, whereas in the 
process as described by Pfeffer the mineral matter was es- 
sential, forming the point of attraction for the aggregation 
of the proteid. But Pfefifer’s suggestion is too mechanical, 
and moreover gives no reason whatever for the fact that the 
grains in the ripe seed are always embedded in a protoplasmic 
matrix ; they should rather be lying loose in the vacuole. 
The earliest stage, namely, secretion in the protoplasm of 
matter soluble only in dilute potash, has also been observed 
to occur in precisely the same way as above described in 
another species of lupin ( ?L . ravins). 
It is most interesting to note that the development of 
