e whole number of days in the year.

+------------------------------------------------+
|Course.| 1804. | 1811. | 1812. | 1813. | Total. |
|------------------------------------------------|
| N. | 143 | 105 | 90 | 111 | 449 |
| N. E. | 99 | 207 | 138 | 138 | 582 |
| E. | 33 | 18 | 22 | 23 | 96 |
| S. E. | 131 | 108 | 135 | 110 | 484 |
| S. | 58 | 69 | 113 | 80 | 320 |
| S. W. | 224 | 255 | 153 | 261 | 893 |
| W. | 81 | 69 | 102 | 57 | 309 |
| N. W. | 329 | 264 | 345 | 315 | 1253 |
+------------------------------------------------+

This work of Mr. Coffin has been brought to my notice since the foregoing
pages were written. The facts embodied in it will be found to comport with
what I have observed and stated. In relation to the proportionate number
of days in the year during which the wind blows from the different points
of the compass at the several stations it is very full and able.

But it has cardinal defects. It does not show the _main currents_ of the
atmosphere. It treats the surface-winds, which are incidental, as
principals. The direction of the main currents is indeed shown frequently
by the mean course of the surface winds, but not uniformly or
intelligibly. Nor does it distinguish between the fair weather and storm
winds; nor always between the trade winds during their northern transit,
and the variable winds north of the trade-wind region. Hence, the
deductions derived from it disclose no general system, and sustain no
theory, although many very important facts appear. Some of these,
Professor Coffin found it difficult to reconcile with received theories,
or satisfactorily explain. For instance, he found the prevailing winds of
the United States, in Louisiana and Texas, S. and S. E.; in western
Arkansas, and Missouri, southerly, and in Iowa and Wisconsin, S. W.,
forming a curve, and evidently connected together.

Thus, alluding to the winds west of the Mississippi, and between the
parallels of 36 deg. and 60 deg., he says:

"On the American continent, west of the Mississippi, there appears to
be more diversity in the mean direction of the wind, yet here it is
westerly at sixteen stations out of twenty, from which observations
have been obtained. The most peculiar feature in this region, is the
_line_ of southerly winds on the western borders