are other irregularities which deserve to be noticed, in this
connection, although the analogical evidence they afford is far from being
decisive.

I have already said that it was within my own observation, that
alternating lines of heat and cold, as well as rain and drought, existed
frequently, without regard to latitude, following, to some extent, the
course of the counter-trade. Such lines have been observed by others.

Thus, Mr. Espy, after describing a snow-storm, which was followed by a
very cold N. W. wind, of several days' continuance, says:

"This cold air covered the whole country, from Michigan to the
eastern coast of the United States, till the beginning of the great
storm of the 26th January; and, what is worthy of particular notice
is, that _the temperature began to increase first in the north and
north-west_. On the morning of the 25th, in the north-western parts
of Pennsylvania, and northern parts of New York, the _thermometer_
had already _risen in some places 30 deg._, and, in others, _above 40 deg._.
While in the S. E. corner of Pennsylvania, and in the S. E. corner of
New York it had not _begun to rise_. The _wind_ also began to change
from the _north-west_ to _south_ and _south-east_, _first_ in the
north-west parts of Pennsylvania and New York, some time before it
commenced in the south-east of those States; and, during the whole of
the 25th, the thermometer, in the north of New York, continued to
rise, though the wind was blowing from the southward, where the
thermometer was many degrees lower."

Thus, too, Mr. Redfield (American Journal of Science, November, 1846, p.
329):

"On the contrary, in times of the greatest depression of the
thermometer, in numerous instances, the cold period has been found to
have first taken effect in, or near, the tropical latitudes, and the
Gulf of Mexico, and has thence been propagated toward the eastern
portions of the United States, in a manner corresponding to the
observed progression of storms."

This was because the cold N. W. wind which _followed_ storms began to
follow them as the storms curved and passed to the N. E.

They occur in Europe also. Says Kaemtz:

"Such contrasts are not uncommon in Europe, and, in this respect, the
Alps form a remarkable limit; for they separate the climates of the
north of Europe from the Mediterranean