New York AGRICULTURAL EXPERIMENT STATION. 275 
gains greater than the differences between check plants in the same 
tests; or the tests, when repeated in other seasons, contradicted the 
conclusions of previous trials. 
The failure to secure satisfactory results in so many of these 
tests established the unreliability of the system of plat comparisons 
when the crops under test are sown on square or broadly oblong 
areas. It is practically impossible to secure on soil of glacial origin 
adjacent blocks of any considerable size that are uniform in pro- 
ductive power. This series of tests materially influenced the de- 
velopment of the system of plat testing used by younger Stations 
by which longer, narrower and more numerous plats are used, so 
that errors due to soil inequality are neutralized or eliminated. 
Other factors which prevented satisfactory comparisons were the 
natural strength and moisture-retaining power of the Station soil. 
These are so great that in many cases check plats grew as large 
crops as it was possible for the heat and rainfall of the season to 
produce, so that the influence of minor factors like the addition of 
fertilizer or differences in cultural methods was completely ob- 
scured., 
Two facts were brought out, however: First, that the growth of 
weeds** in corn produces much greater reduction in yield than 
can be accounted for by the fertilizer elements removed by the 
weeds. Beside this, the shading of the soil by the weeds reduces 
the high soil temperature essential to the best growth of corn, the 
weed roots occupy the feeding ground of the plant roots, and 
the weeds draw heavily upon the supply of soil moisture, so that 
the corn often suffers from semi-drought. Second, that root 
pruning” the corn plant is injurious, whether this be done by deep 
spading beside the plants, by cutting off the roots on one side of 
the stalks with a lawn edger, by deep cultivation between the rows 
after the plants were well developed, or even by shallow cultivation 
close to the rows. The more thorough and deeper the preparation”® 
of the soil previous to planting, to give opportunity for deep and 
wide rooting, the better the results, especially in the production of 
forage. 
Variety and race tests The varietal comparisons were usually 
incidental to more careful studies of characteristics, and were on 
too small a scale to justify very emphatic recommendation of par- 
* Rpt. 2:137, 138 (1883); 5:50 (1886). 
= Rpts. 1:53, 54. (1882) 22134 (1883) >. 75173-1778. (1888). 
Rote 7eIgie1 7 salboos). 
