240 VEGETABLE PHYSIOLOGY 
the cell in which it lies, and not occupying a definite space 
as do aleurone and starch grains. Whether it is secreted 
from the substance of the protoplasm, or whether the 
materials of which it is made are taken to the latter in a 
state near the condition of the finished fat, is uncertain. 
It is formed by the combination of a fatty acid with gly- 
cerine. Both these bodies can be formed in the plant, but 
how they are finally presented to us in the shape of oil is 
still in need of elucidation. As the oil appears in the cell 
it seems to point to a process of breaking down of the 
protoplasm itself, and not to a direct combination of the 
antecedents mentioned. If we stain cells which are form- 
ing fat with osmic acid, which colours fatty bodies brown 
or black, we see in the protoplasm small specks of fatty 
matter, which, while in the youngest cells mere dots, are 
in older ones larger, and can be recognised as droplets. In 
still older ones the blackness permeates the whole proto- 
plasm, indicating that the latter is saturated with the oil, 
the droplets having run together in consequence of their 
number and dimensions. 
The appearances are, however, not inconsistent with 
the view that the work of the protoplasm is only to effect 
the ultimate changes or interactions of the glycerine and 
the fatty acids which are transported separately to the cells 
or perhaps formed there from some antecedent. 
The deposition of fat in some cases, particularly in 
leaves, has been stated to be effected by the agency of 
certain plastids corresponding to the leucoplasts already 
mentioned in connection with the formation of starch 
grains. These structures, which have been called elaio- 
plasts, are curious bodies of various shapes, sometimes 
round or oval, sometimes irregular in contour, which lie 
near the nucleus of the cell. Like the other plastids they 
consist of a spongy protoplasmic framework, in the meshes 
of which the oil is formed, much as it is in the protoplasm 
of the seeds already described. 
All these bodies, when acted upon by a process ana- 
