120 [ ASSEMBLY 
Hence we are to conclude that as the variation from the average of 
the plats are no greater than the variations within the plats, that these 
discordant figures are as near duplicates as the circumstances of the 
case admitted of our obtaining, and if we assume a uniform soil upon 
our plats it would be logical to figure that unless between opponent 
plats there was a greater difference than fourteen or fifteen per cent, 
the figures of the crop could not be interpreted either for or against 
individual treatment. As a matter of fact, however, the soil of our 
plats is not of uniform fertility, as appears evident from the study of 
the crops yielded, as also is certain from the varying quantities of fer- 
tilizer which had been previously applied. 
The whole interpretation of plat experiments depends upon recog- 
nizing the limits of error for each case within which differences of fig- 
ures can have no exact meaning. The curse of plat experimentation 
is that this limit of error is liable to be excessive, as in our present case. 
In the case of our series I. B., one to five, we have a variation of crop 
between the five yields from single eyes used as seed of one hundred 
and thirteen pounds, or forty per cent from the average; between the 
five yields from ordinary cuts one hundred and seventy-two pounds, or 
forty-eight per cent from the average; and between the five yields 
from whole potatoes of seventy-five pounds, or ninteen per cent from 
the average. Thevariation from the average of the yields in anh three 
duplicate : rows of each plat was for the good potatoes. 
Single eyes. Ordinary cuts. Whole potatoes. 
TSI ot hee Rea Cle ats 10 per cent. 8 per cent, 5 per cent: 
ELUTE ed eds tacts OY a tank Pg Ue BUR “F 
Tes Bo 3 aa scale n . es wee brains TOG 17 ed 7 i 
(Bde ag rae eh Ps rch ed Ae Lo cuaiee egy eae 
i Wigs Sein se eh LAY peg Sa 2 ag Brae! OS 
These two calculations both show that the variation in the yields 
from the whole potatoes used as seedswas less than when single eyes 
were used, and that ordinary cuts produced the most uniform crop. 
We must, therefore, conclude that under the methods of planting, and 
in the soil of this year’s trials, single eyes gave less yield than the other 
kinds of seed used. The order of the figures of crop have, however, a 
greater significance than the actual amounts. In the five duplicate 
trials it will be observed that in but one case was the yield from the 
single eyes in excess over that from the whole potatoes. 
Although these plats received no fertilizer this year, yet last year 
two hundred pounds was applied to I. B. two, four hundred pounds to 
I. B. three, eight hundred pounds to I. B. four, and sixteen hundred 
pounds to I. B. five. Let us calculate the gain or loss over the unfer- 
tilized plat I. B. one, in bushels: 
Single eyes. Ordinary cuts. Whole potatoes. 
TBS se haa Be Bis. vcs —18 bush. +33 bush. -+23 bush. 
DB Baio hye ae eas le Ad Bale coh ra 0 Fe +23. ° §6 +18 §§ 
| Degas 2 Se Nadie ay eke eS ce A ag +29 +50 +37 <“ 
Te Bad) sc. Sa ete +62 << +86. << Guia 
We find here an influence indicated in the crop for the fertilizer left 
over, as well marked as might be expected, but yielding no certain in- 
formation, but only the general value of the fertility present. 
