105 
of flour, while in other years it will yield 52 lbs. ; making 
a difference in every acre, yielding 30 bushels, of 300 lbs. of 
flour, and this making more and better bread than an article 
of inferior quality. 
In the experiment by Mr. Lawes, before referred to, he 
found a difference in the weight of the wheat per bushel, 
made by temperature, in the years 1845 and 1846, (on land 
unmanured and cropped successively for seven years), of 7£ 
lbs; the weight being, in 1845, 56^ lbs., and in 1846 63| 
lbs. And, by an examination of the meteorological tables 
published by Mr. Luke Howard, it is found that the mean 
temperature of these four months varies in different years 
from 57°4' to 63°; nearly 6° of Fahrenheit. But the manner 
in which the mean temperature of the month is calculated by 
meteorologists renders their registers nearly useless in this 
inquiry, since it is the custom in every month to take the 
highest and lowest temperatures that any night and day 
afford, and to call the mean of these extremes the mean 
temperature of the month. Hence, whatever may have been 
the heat of a certain number of days in each month, one cold 
night falsifies the whole, and reduces unnaturally the mean 
temperature. For example, during twenty-three days of the 
present month (August), the lowest temperature of any day 
is 61°; there are several days 66° and 68°, and six days 
above 70° ; but one cold night of 37° reduces the mean to 
only 55°. But, surely, this cold night, or the cold nights 
during May, June, July, and August, do not injure wheat 
at all. A farmer never complains of the frosts in these 
months, while the maximum day temperatures only, and 
*he prevalence of high temperatures, during these months, 
exert a beneficial influence upon the production of wheat, 
and, therefore, a register of day temperature is required. 
For, strictly and physiologically, it is by the action of light 
accompanying the heat that all nutrition in plants is effected ; 
