132 
Comparison of increase by five-year "periods 
Plat. 
Value of in- 
crease. 
Net Rain over 
cost of fertilizer. 
Fertilizing elements. 
I Mk tsphorus 
Phosphorus and nitrogen 
Phosphorus and potassium 
Phosphorus, potassium, and nitrogen 
First 
period 
§7. 66 
16. 92 
12.62 
21.86 
Second 
period. 
$15. 01 
28. 75 
19.90 
35. 03 
First Second 
period, period. 
$5. 26 
2. 52 
3. 72 
.96 
$12.61 
14.35 
11.00 
14.13 
During the first five years or more of this test phosphorus seemed to be the con- 
trolling element in producing increase of crop, and while the addition of potash 
and nitrogen regularly increased the yield, the increase was not sufficient to justify 
the use of these elements, at least in the quantities employed in this test, quantities 
based upon the chemical composition of the crops, except that nitrogen has been 
used in a much smaller ratio to phosphoric acid and potash than would be indicated 
by the composition of average yields of the crops grown in this experiment. The 
actual quantities used on each five-year rotation are as follows: 
Pounds. 
Phosphoric acid, in acid phosphate, per acre 50 
Potash in the muriate, per acre 130 
Nitrogen, in nitrate of soda, per acre 75 
The fertilizers are distributed over the three cereal crops of the rotation — corn, 
oats, and wheat — the mixed clover and timothy following for two years without any 
further fertilizing. The experiment was planned with the assumption that the 
clover would probably make up the deficit in nitrogen; but latterly the clover has 
been refusing to grow, especially upon the plats treated with acid phosphate, and 
while the hunger of the soil for phosphorus is apparently being gradually satisfied, 
the demand for nitrogen remains constant or increasing, and it is the plats receiving 
large applications of nitrogen which are now not only giving the largest gross yield, 
but also the largest net profit, even though the nitrogen costs 15 cents per pound. 
I am not able to see how this result could have been arrived at by any shorter or 
surer method than the one we have followed. 
One of the great difficulties in field investigation, especially with cultivated crops 
such as corn or potatoes, is to secure a uniform stand. We are sure to have some 
missing hills, and potatoes are often more or less affected by blight or rosette. It 
has been our practice to count the hills on each plat, and, in the case of corn, to 
count the stalks, separating those which bear two ears and those which are barren, 
and to count the ears and nubbins. Unless the stand has been very irregular, how- 
ever, we do not attempt to make any correction, as we have not yet discovered an 
entirely satisfactory basis for such corrections. The additional space given to the 
survivors by the failure of a stalk or hill seems to result in a partial compensation. 
The most satisfactory basis of correction seems to be that of the actual stand, rather 
than that of the full stand. This method gives average results corresponding to the 
yield actually harvested, whereas when we calculate to full stand the yields are 
often exaggerated. 
It is sometimes absolutely necessary to make such corrections or else to lose the 
experiment, or suffer it to tell a misleading story. 
Another difficulty is to avoid mistakes in weighing, and as a check in this direc- 
tion we have adopted the plan of leaving the grain from each plat of wheat or oats 
in labeled sacks until the reported weights can be studied in the office. The first 
operation is to calculate the weight of straw per bushel of grain; if we find a dis- 
crepancy here the sacks are weighed again. By this method we have several times 
secured the correction of such mistakes. In the case of corn, the ears and nubbins 
are counted separately and separately weighed. If a discrepancy appears in the 
weights per ear or per nubbin, a basis of correction is found by comparing the 
counts of ears, nubbins, and stalks. 
It goes without saying that the greatest care should be exercised to secure uni- 
formity in the preparation of the seed bed, in the selection and planting of the seed, 
in the distribution of the fertilizers, and in the culture of the crop. Wherever pos- 
sible machinery should be employed, as properly handled machinery will do more 
uniform work than can be done by hand. 
In conclusion, I would repeat my conviction that conclusions relating to the main- 
tenance or increase of soil fertility must be confirmed in the field before they can be 
