7| cool | .... |
| 1788| rainy Amer. | cold |
| 1789| cool spring, hot summer | severe Eur., mild Amer. |
| 1790| .... | .... |
| 1791| very hot Am. | cold |
| 1792| .... | .... |
| 1793| hot, dry Am. | mild Amer. |
| 1794| .... | severe Europe |
| 1795| Amer., hot, rainy | .... |
| 1796| Autumn very Dry Am. | cold Amer. |
| 1797| cool Am. | severe Amer. |
| 1798| very hot } | { long & severe |
| 1799| very dry Am. } | { Amer. & Eur. |
+---------------------------------------------------------+

Still more definite evidence is found in the meteorological tables of Dr.
Holyoke and Dr. Hildreth, and an account, by Dr. Hildreth, of the seasons
when the Ohio River was closed or obstructed by ice, found in Silliman's
Journal, new series, vol. xiii. p. 238.

Thus, we have, from the tables of Dr. Holyoke, the following annual means,
from 1786 to 1825, inclusive. I have arranged them in periods of five
years. It will be seen that there are three peculiarities observable.
First, a marked difference between the first and second periods of the
decade, corresponding, generally, with the presence or absence of the
spots. Second, a difference in the mean of the decades which may well be
supposed to correspond with the difference in the number or size of the
spots since a like difference is observable in number and size, and the
time when they reached their maxima and minima, in the table of Schwabe.
And, third, there are occasional single cold years during the warm period,
and these correspond with what the tables of Dr. Webster show for both the
sixteenth and seventeenth centuries. In relation to this, it should be
remembered that volcanic action is a frequent and powerful disturber of
the regular action of terrestrial magnetism, and that the extremes, for
that reason, are frequently meridional or local and alternating; and to
that cause very great extremes, and marked exceptions, may be due,
notwithstanding the spots upon the sun may exert an influence in producing
hot summers and cold winters toward the close of each decade. Thus, to
select an instance to illustrate this and