140 
On Modes of Comparing the 
tables wliich afford to animals substantial nourishment. Now It 
happens that all these nutritive jirinciples In roots are specifically 
heavier than water ; for with water at 1000 as the standard, sugar 
and starch vary in weight from 1500 to 1600. The specific 
gravity of gluten and albumen I have not been able to find in 
])ublished works, but they sink in water, and are of course speci- 
fically heavier.* 
It may perhaps be supposed that the soluble salts, and Insoluble 
earths and oxides, are sufficient to disturb the accuracy of any in- 
ference from the relative specific weights of the water, and of the 
other constituents of the plant. Their amount, however, seems 
so trifling (the soluble salt In the potato giving only a little more 
than 8, and the insoluble earth, &c., less than 5 parts in the 
1000) that the difference of their specific weights would do little 
more than turn the scale ; and It should be observed that, as, for 
the most part, the soluble salts, If not in excess, perform a part In 
animal nutrition, they may he considered as partially nutritive, 
while they are also specifically heavier than water ; and of the in- 
soluble earths and oxides, which are not taken up in the process 
of digestion, some are lighter and others heavier than water. 
From the details of Eynhoff^ given in Thomson's ' Vegetable 
Chemistry,' their proportions seem to be such that the sum of 
their specific gravities (in the potato 1093) varies only a trifle, as 
compared with water — a difference which, as applied to only 5 
parts In 1000, could but very slightly disturb the relative weights 
of the 700 of water, and of the '274 of the nutritive matter, which 
are to one another as 1000 to 1,500 or 1600. It is true that, from 
Sprengel's tables, published by Dr. Daubeny in the ' Royal 
Agricultural Journal,' the fixed ingredients in beet and swede 
turnip are considerably greater than In the potato or white turnip ; 
but it seems probable, as with the potato, that the several specific 
weights of these Ingredients, in each root, either neutralise one 
another, or tell in positive aid of nom-ishment. Admitting this 
mode of estimating nourishment to be — as, in fact, it only can be — 
an approximation, it Is certain at least that, in as much as so large 
a proportion of these roots consists of water, the absolute weight of 
a crop alone must be a still more deceptive estimate of Its value. 
The weight of the nutritive ingredients, separate from the water. 
Is what we want, and this the specific gravity of the root seems to 
give nearly enough ; for it is, In fact, a pretty close measure of 
their weight, minus the weight of water in the root, which Is lost, 
* The proportion of nourishment due exclusively to gluten, and the other 
azotized in>^redients of food, is of course not to be detected by the mechanical 
process here proposed. The sum of the solid ingredients, as compared with 
the water of the plant, is all that can be learnt from it. — W. H. H. 
