New YorK AGRICULTURAL EXPERIMENT STATION. 377 
ferent structural parts. If these differences did not exist, the 
substance of the seed reduced to powder would have the same 
specific gravity as the seed itself; but this is by no means the 
case. For instance, the specific gravity of a certain grape seed 
was 1.13, but the specific gravity of the substance of the same 
seed, after having been crushed with a hammer, was about 1.25 
for the endosperm and about 1.385 for the outer seed coat. This 
seeming paradox, that the specific gravity of each part of the 
seed may be higher than the specific gravity of the seed as a 
whole is of course explained by the presence of air spaces within. 
The experiment just reported brings out clearly the fact that 
in studying seeds two kinds of specific gravity must be distin- 
guished — apparent specific gravity, and real specific gravity. 
These two are, or at least may be, very different from each other. 
In speaking of the specific gravity of seeds, apparent specific 
gravity is universally meant, but apparent specific gravity is 
evidently a property of no practical importance unless it can be 
correlated with some cultural property. 
The writer has made few structural analyses with a view to 
determining the specific gravities of the component parts of a 
seed and the relations of these specific gravities to the specific 
gravity of the seed as a whole. Yet such analyses are indispen- 
sable to a correct understanding and interpretation of results 
obtained in the application of the method of seed selection 
according to specific gravity. 
As to the testa, the specific gravity of this part would appear 
to be of little effect on the specific gravity of the seed as a whole 
in the cases of those seeds having very thin covering, such as 
wheat, since the mass of the testa in these cases forms so small 
a part of the mass of the seed as a whole. Nobbe reports obser- 
vations on this point in which it is shown that the thickness of 
the outer seed coat is only about .04 to .05 mm., varying between 
those ranges. Obviously the specific gravity of the seed coat in 
the case of wheat may be disregarded. 
In the case of the grape seed, the writer finds the thickness of 
the testa to vary from .3 to .4 mm. in normal seeds; in hollow 
seeds it is very slightly less, ranging down to about 2.5 mm. In 
this case the mass of the testa is great enough to necessitate a 
consideration of its specific gravity in judging the specific gravity 
