94 Home Produce , Imports , Consumption , a?tcZ 
the better to clean that portion. For the crop of 1891, how- 
ever, the full number of rows were again sown over the whole 
length of each plot. 
In 1889, the produce of the ordinarily or thick-sown portion 
of the land was taken as the produce of the plot ; and this being 
the bottom portion, which is the worst half of the plots, was un- 
doubtedly abnormally low ; the selected plots yielding a mean 
produce of only 27-g- bushels at 61 lb. per bushel, whilst our 
actually adopted estimate was 294 bushels. 
In 1890, the top half of the plots was ordinarily sown, and 
this portion being not only the better half, but having been 
only thin-sown — that is, partially fallowed — in 1889, the produce 
was without doubt abnormally high ; the average produce of the 
selected plots being 374 bushels, whilst we estimated the yield 
of the country at large to be only 314 bushels. Accordingly, 
we have adopted the mean produce of the thick- and the thin- 
sown portions, as more nearly representing the normal produce 
of the plots ; and this gives an average for the selected plots of 
30| bushels, which is not much less than our adopted average 
yield for the United Kingdom. 
Lastly, in 1891, the entire plots were ordinarily sown, but 
after partial fallow of the top half in 1889, and of the bottom 
half in 1890. The produce was, therefore, assumed to be 
abnormally high. The average of the selected plots gave 32 § 
bushels at 6 1 lb. per bushel ; but we estimated the crop of the 
country at only 29| bushels at 61 lb., or 30 bushels at 60 lb. 
per bushel. 
It was, therefore, under abnormal conditions of the land, 
that abnormal results were obtained in the last three years of the 
forty, and hence the same accordance as usual between the actual 
produce of the selected plots and the estimated yield of the 
country in the individual years was not to be expected. It is 
seen, however, that the average yield of the selected plots over 
these three years taken together was, nevertheless, identical with 
the estimated yield of the United Kingdom over the same three 
years — namely, 304 bushels. 
Let us now consider the estimates of yield per acre calcu- 
lated in the three different ways, over the individual periods, 
and over the total period of the first thirty-seven of the forty 
years. Although in the majority of seasons the mean produce 
of the selected plots closely represents the average produce per 
acre of the United Kingdom, as has already been stated it 
generally indicates a somewhat higher yield than the average 
of the country at large in seasons of high productiveness, and a 
somewhat lower one in unfavourable seasons ; and hence, in some 
