128 
pied by living organisms, of whose functions we as yet know but little, although that 
little is sufficient to show that they are of the utmost importance. These conditions 
never recur in orderlj succession, but each season has its peculiar influences which 
work for the welfare or the detriment of the crop. Some of these conditions may 
be more or less completely reproduced in the lal (oratory ; others are altogether beyond 
control. All that the laboratory can do, therefore, is to point out the way in which 
progress is indicated; the final and crucial test must be made in the field itself, for 
it matters not how promising a method may appear under the conditions of the 
laboratory, it it be not also applicable to those of the field it is of no practical value. 
No greater mistake can be made than to assume that the scientific investigation of 
this problem ends with the laboratory; that field research is merely a simple affair, 
which may be conducted by the ordinary farmer, and that it is his business to take 
up the work at the point where the laboratory leaves it. The fact is that it is the 
laboratory investigation which is the relatively simple matter, while the making of 
such a field investigation as shall add to the sum of human knowledge involves an 
expense in preparation and equipment, a patience and exactitude in execution, and a 
discernment in interpretation which makes this form of research one of the most 
costly and difficult known to science. 
While, therefore, the student of the soil will continue to use the chemist's, the 
physicist's, and the bacteriologist's laboratories, his chief laboratory must be the 
field itself; and I believe that as the magnitude and vast importance of this work 
are more fully comprehended, and as the scope and province of the field experiment, 
its limitations, and its possibilities, are more fully understood, the employment of this 
form of investigation will steadily increase. 
A field experiment in fertility maintenance, if it is to accomplish any useful pur- 
pose, must be continued, not merely for a single summer, but for many years and on 
the same soil, in order to meet the varied conditions of changing seasons and to 
adequately study the peculiarities of the soil. As we look back over ten years of 
such work at the Ohio Station we feel that we have only made a beginning," and in 
the final endowment of the Rothamsted investigations, after fifty years of continuous 
work, the greatest field experimenter the world has ever known has expressed his 
conviction of the necessity for long-continued, stationary effort. 
For a work involving such an enormous outlay in time and labor as does a field 
experiment conducted along the lines I have indicated, it is of the highest impor- 
tance that the preparation be as thorough as possible. As a matter of course the 
soil upon which a comparative test is instituted should be as uniform in character as 
possible, but this uniformity is extremely difficult to secure. If the land lies level 
there will be shallow depressions from which the surplus rainfall escapes more slowly 
than elsewhere, thus increasing the water supply to an injurious extent in wet sear 
sons and giving to these depressions an undue advantage in dry seasons, when water 
may be the controlling factor in producing increase of crop. On the other hand, 
slopes from which the rainfall rushes rapidly must be avoided, as on such slopes 
more or less fertilizing material will be carried from plat to plat, or the land will be 
washed into gullies, thus interfering with cultivation. The ideal topography for this 
work is a broad, even slope of about 1 or 2 per cent — or just enough to give easy 
though not rapid drainage. 
Not only the surface but the subsoil must be considered in locating a field experi- 
ment. Where beds of drift gravel or loosely stratified rocks lie near the surface they 
will materially modify the drainage conditions, and uniformity of drainage is a mat- 
ter of prime importance. 
Most of the agricultural colleges have been located with reference to other matters 
than field experiment; or, if this question was considered at all, it was with the idea 
that an experiment farm should have a variety of soils, an idea which seems to be 
practically universal in the minds of men not trained in the actual work of field 
experiment. The result is that very few of the farms attached to these colleges pos- 
sess any qualifications for this work. 
The Ohio station had the good fortune to be permitted to select a farm for its 
work, after some ten years' experience on an agricultural college farm. We found a 
soil possessing in an unusual degree the points which we had found to be necessary 
for successful field investigation, it being a sheet of drift, lying in gentle slopes upon 
and largely modified by a shaly sandstone of the Waverly series, the drift sheet being 
sufficiently thick to have formed a subsoil comparatively impervious to water. AVe 
were not able to secure all the desirable points, as some of our slopes are too steep 
for the best results, and in some places it has been impossible to avoid cross drain- 
age, while in others differences in previous treatment, which 1 will refer to again, 
have brought about permanent inequalities in the natural fertility of the soil; but, as 
we have gone forward with the work and observed other soils throughout the State, 
we feel t hat we have probably secured as many of the important features as are likely 
