340 Report or THE HorricuULTURAL DEPARTMENT OF THE 
He remarks further, that previous investigators who have reported 
‘ favorably on the method have disregarded the absolute weight of 
the seed and have thereby been led into error. 
The method of salt solutions as applied to potato tubers and 
to roots is not considered in this article. Much experimental work 
both practical and scientific has, however, been done along these 
lines in Europe. <A critical review of such work is made by 
Wollny. In this country the late Prof. E. 8, Goff* made prac- 
tical application of the method to the selection of potato tubers 
and with good results. 
METHODS OF DETERMINING SPECIFIC GRAVITY AND 
OF SEPARATING SEEDS ACCORDING TO IT. 
Two fundamentally different methods have been used in study- 
ing the specific gravity of seeds. One is the method of separates 
illustrated in the experiments of Grandeau, Lyon and others. 
The other is the method of samples, illustrated in the pycnom- 
eter method. It will be perceived at once that these two general 
methods approach the subject from entirely different standpoints. 
The method of samples is identical in principle with the same 
method as applied in sampling fertilizers, ete. Its object is not 
to separate one constituent from another but only to determine 
the average composition of the whole mass. In the method of 
separates, however, the object is to distinguish between different 
individuals or elements. As applied in seed selection the method 
of separates is historically much the older and has grown up with 
agricultural practice. The method of samples is of comparatively 
recent introduction and belongs to the realm of criticism. 
THE METHOD OF SEPARATES. 
The application of the method of separates has been more or 
less arbitrary. Most frequently it has been the custom to sep- 
arate a lot of seeds, irrespective of their intrinsic quality, into 
two or three lots of generally equal quantities. In so doing the 
fact that seeds vary much in quality under different conditions 
and in different seasons, is not taken into consideration and no 
attempt is made to define with exactness the limits within which 
seeds of different qualities occur. 
* Wis. Agr. Expt. Sta. Rpt., 1895, p. 317, 
