THE BBOADBALK WHEAT BOILS. 
21 
lssi. After a careful study at t he time it was considered by Sir John 
Lawes and Sir Henry (Gilbert that the lighter weights of 18»>5 were 
more probably correctly representative of the first 9 inches, and that 
the 1881 samples had included, in the actual 9 inches taken, more 
than the first 9 normal inches of soil ; in fact, that the quant ity taken 
on the original weight represented more nearly 10 inches of real depth 
than 9 inches, or that it included about 1 inch too much of subsoil. 
Accordingly, after analysis of the soils, the estimates per acre of the 
nitrogen, for example, in the 1 SSI samples, were calculated on the 
assumption that the weights should have been approximately one- 
tenth Less, a deduct ion of one-tenth being made from the originally 
indicated acreage weight of the soil, lint more recently, when the 
weights of the 1 S9:j samples were available for comparison, it was found 
that the average weights of the 1SS1 and IS!*:} samples of the first 9 
inches agreed, both being very sensibly above those of 1865. A care- 
ful study of the climatic features of the three individual seasons, 
both before and after t he removal of their respective crops, and of 
the probable condition of the land as affected thereby, together with 
the notes that had been kept relating to the mechanical operations to 
which the land had been subjected, led to the conclusion that in 1865 
the surface soil was in a lighter condition than in either 1SS1 or 1S93, 
and that probably the weights were more normal in the two later 
years. After much tabulation and much thought and calculation as 
to the effects of assimilating the weights of 1S65 with those of 1881 
and lsiii), it was found that no inconsistency was introduced and that 
the truth was more probably arrived at by correcting, on this assump- 
tion, the est imates derived from the weights of ISO') t lian by altering, 
as had previously been done, the estimates for 1881. As regarded the 
subsoils, there seemed to be litt le reason for supposing t hat the weights 
would materially alter from period to period, and. after careful corre- 
lation of all the details, it was decided to take a fixed weight of fine, 
dry soil for each depth at -each of the three periods, this being found 
by taking the mean of all the determinations in the three or the two 
years, as the case might be. 
There were considerable variations in the weights of fine, dry soil 
in the samples from the different plats in one and the same year, 
which were largely attributable to the varying amounts of stones; 
but comparing the averages for all the plats of one year with those of 
the others the differences were comparatively immaterial. 
Except, therefore, in the case of the plats receiving large quantities 
of organic matter, such as farmyard manure or rape cake, there has 
DOW been adopted, for the three years, a uniform average weight of 
fine dry soil for the first 9 inches, and also a uniform weight per acre 
of fine dry soil for each depth below the first 9 inches, including, in 
1 he case of the subsoils, those of the dung and rape-cake plats. As the 
estimates of weight of soil per acre now adopted will continue to be 
