286 TWENTY-FIFTH ANNIVERSARY REPORT. 
two tests,>* in different years, there was a consistent, though small, 
advantage from applying fertilizer below rather than above the 
seed, and in a later test,°? large quantities of fertilizer applied in 
the row appeared to injure the vegative power of the seed tubers. 
Small quantities applied in the row gave better results than similar 
‘quantities broadcasted, but when larger quantities were used, broad- 
casting gave much better results. 
In a test** on Long Island in 1895, ten brands of commercial 
“potato” fertilizers were applied on eighth-acre plats, each at 
the rate of 1,000 and 2,000 pounds to the acre, and one brand also 
at the rate of 1,500 pounds. (With this brand applications in the 
row and broadcast were contrasted at each rate. The results of 
this last comparison have already been given.) Though, these 
brands were all “ potato” fertilizers they differed considerably in 
every ingredient, least in nitrogen and most in potash; but each 
brand contained in 1,000 pounds more, frequently many times more, 
than enough of each element to supply the demands for a crop 
of 200 bushels of potatoes to the acre. This plentiful supply would 
indicate a liability to great waste, if more than 1,000 pounds to the 
acre was used, for few soils are so poor that they will not grow 
something of a crop, and it should be the purpose of a fertilizer 
to supplement the soil supply, not to furnish all of the ingredients 
necessary. 
The crops in the tests showed that such heavy applications are 
wasteful. In only twelve cases of the twenty-four was the increase 
sufficient to repay the cost of the fertilizer, and only one of these 
occurred where more than 1,000 pounds of fertilizer was used. In 
one case there was a loss from the use of even 1,000 pounds, and 
in almost every case the loss from using a ton to the acre was 
greater than the gain from using half a ton. 
This experiment was continued the next year on the same plats, 
to test the residual effect of the fertilizers. The average yield of 
the plats where 1,000 pounds to the acre had been applied was in- 
creased 48.4 bushels the first year and: 29.6 bushels the second year; 
while the use of an additional 1,000 pounds to the acre increased 
this yield only 4.4 bushels the first year and 14.1 bushels the second 
year. Taking both years into consideration, the use of 1,000 
® Rpts. 7:168-170 (1888) ; 8:253-255 (1889). 
* Bul. 93:277, 278; same, Rpt. 14:35, 36 (1895). 
“ Bul. 93; same, Rpt. 14:25-36 (1895). 
© Bul. 112; same, Rpt. 15:107-118 (1896). 
