[ ASSEMBLY, No. 33. ] 17 
feeding and growth development. -At harvest, it must be observed 
how nearly these conditions have been satisfied, and when discrepan- 
cies occur, these discrepancies must have their effect estimated accord- 
ing to some rule derived from experience, and the resulting figures 
from the investigation either added to or subtracted from the actual 
yield, according ‘to the need. The next process toward our obtaining 
complete results must be the satisfying ourselves of the equivalency 
of the seed used, and of the evenness of the fertility and of the 
physical conditions of the soil, through the study of each plat divided 
into smaller sections, and as well of the conditions of the surrounding 
soil, through the careful use of duplicate series of plats, bearing in 
mind that the figures gained cannot be expected to be identical between 
duplicates, but only to be valued as they fall within wells established 
limits of error. 
That the station has not in all cases followed these a of interpre- 
tation is a misfortune that arises from the newness of its a and 
from the want of sufficient data, and a well-verified series of observa- 
tions which can be regarded as constants for given circumstances. 
It isa good rule, however, for the farm experimenter to consider, 
that when the variation between two plats differently treated is no 
greater than between plats treated alike, then the figures which are 
intended to be for or against any practice are unsatisfactory, and can- 
not be accepted as infallible guides as measuring the influence of the 
changed condition. It is also well to be remembered that ifs it is jus- 
tifiable to interpret gain or loss according to our prepossessions in any 
one case, then it is logical to interpret figures of like meaning and 
like import even when counter to our beliefs. If, for illustration, we 
plant two plats with corn, and to one plat apply fertilizer, and harvest 
from the fertilized plat ten bushels more of corn, then, if we accept 
this ten bushels as the measure of the efficacy of the fertilizer, we 
should estimate weeds also as of ten bushels advantage, when, upon 
two plats, one weeded and the other growing abundance of weeds, we 
harvest ten bushels more of crop from the weedy portion. 
We would, by no means, undervalue our appreciation of the impor- 
tance of the eaten that this station will be of benefit to practical 
agriculture, yet a close acquaintance with the difficulties that beset 
our path leads us to the candid admission, that the most important 
portion of our work is the learning how to experiment, and how 
to interpret our results. We must expect to find much of our labor 
thrown away upon efforts which shall not yield an adequate return, 
and we must expect our path to be marked with failures, saddening at 
the time, but from which lessons may be derived for future progress. 
Indeed, in well-directed experimental trials, failures may be even con- 
sidered at times of Meebo A as teaching lessons which might not 
otherwise be learned. 
The leading aim of an Heparin ental station such as ours must be by 
experimental study of the most careful kind to obtain knowledge of 
the action of the laws which find application in agricuiture, and to 
devise methods for the application of the knowledge gained, in order 
that in practical farming waste of means and effort may be diminished, 
and gain may be increased, and thus profits be forwarded, 
The one discovery in cattle feeding this year opens up a field for re- 
| Assem. Doc. No. 33. ] 3 
