of Dr. Holyoke. 111 
observations can be made, seem never to have been extensively 
in use ; and they are too liable to derangement, ever to become so. 
The principal defect, in this point of view in Dr. Holyoke’s 
journal, consists in there being no observation at the coldest 
part of the day. In consequence of this omission, the daily range 
of the thermometer is less than it should be, and the mean tem- 
perature is too high. The observation at sunset is so nearly the 
same with the mean of the whole, that it may be disregarded in 
the result. The observation at noon was evidently designed by him 
to embrace the hottest part of the day, and probably it very 
nearly did so. ight in the morning, or between eight 
and nine, gives a very uncertain result, inasmuch as the tempera- 
ture at that hour, in comparison with that of the whole day, 
varies very much with the different seasons of the year. 
Dr. Holyoke was himself aware of this defect in his observa- 
tions, and proposed a correction for it. But his correction is 
insufficient. The results of his first seven years’ observations, 
as given by himself in the second volume of the Academy’s Me- 
moirs, make the mean temperature for the year, 47,945°, and the 
correction he proposes would reduce it to 47,50°.* 
Dr. Holyoke nowhere, that I can find, explains his reasons 
for fixing upon the number he proposes as a correction, and it is 
probably a mere conjecture. I do, however, find some data 
from which we may substitute a more just correction. 
For the years 1793, 1794, and 1795, the journal contains a 
register of the heat at sunrise, which is nearly complete. It was 
continued partially for several succeeding years, but is in a state 
too incomplete to enable us to draw any deductions from it after- 
* Vol. IL. p. 92. 
