20 Memoir upon Coffee. 
General Proportions of the above constituent Parts.—Al- 
though it may be of little service to ascertain the proportions 
of the immediate principles of coffee, since these proportions 
must vary on account of the greater or less maturity or rich- 
ness of the article, yet I considered it my duty to estimate 
these proportions as nearly as possible. After several com- 
parative experiments; I found that eight ounces of coffee 
yielded about one ounce of mucilage, one drachm of resin, 
one drachm of extracto-colouring matter, 33 drachms of 
gallic acid, five ounces 31 drachms of parenchyme, and ten 
grains of vegetable albumen. 
I compared the decoctions and tinctures of the three dried 
coflees of Bcurbon, Moka, and Martinique. Bourbon and 
Martinique coffee seemingly furnished the same principles 
in the same proportions. That of Moka differs essentially 
from the rest. Its decoction was much less saturated, its 
alcoholic tincture was higher coloured than similar tinctures 
of the Bourbon and Martinique coffees; it contains less 
gum, less gallic acid, more resin, and more aroma than the 
others. 
Roasted Coffee. 
In order to ascertain the changes produced upon coffee 
by roasting, I examined the phenomena which took place. 
while it was burnt in the open air. 
At first it augments in volume as it is penetrated by calo- 
ric; it crackles aud becomes fawn-coloured; the pellicle 
which envelops the bean is detached ; as it is very light and 
slender, it flies off with the least breath. |The coffee then 
emits a very agreeable aromatic flavour. This vapour in- 
creases in its intenseness, the bean smokes and becomes 
brown: the smell then changes and becomes slightly em- 
pyreumatic ; the coffee exudes and. becomes oily at its sur- 
face* ; it ceases to smoke; and if the action of the fire is 
continued the coffee is charred. 
# M. Parmentier enveloped roasted coffee, when in a state of exudation, 
in filtering paper, which imbibed the oil, and remained greasy and transpa- 
rent upwards ef a year: this supposes the existence of a fat oil in this grain. 
I could not obtain any such oil, however, in a separate state, either by ex- 
pression, ebullition, or the caustic alkalis: 
The 
