Page:Popular Science Monthly Volume 1.djvu/686

670 my researches from the rainfall question to this more general problem; and, with the intention of not being misled by local observations made in the city itself, which, as we have just remarked, are not altogether to be trusted, I have resorted to data of a more general topographical kind, such, for instance, as the times of closing and opening of the Hudson River. Also, with a view of extending the conclusions, whatever they might prove to be, to the Atlantic coast generally, I have used such published records of the meteorology of Philadelphia, Boston, and Charleston, as I could find access to. These reach from 1738, with certain breaks, up to the present time.

The data connected with the Hudson River have been derived from the Annual Reports of the Regents of the University; those of temperature for the locality of New York itself, from the observations taken at Fort Columbus, and by Prof. Morris, for the Smithsonian Institution. The remainder are from the records of this observatory. In the case of other Atlantic cities, the data are chiefly derived from the reports of the United States Army officers to the Secretary of War.

It appears from this that, from 1816 to the present time, we have an unbroken register. Taking 1817 as our starting-point, we have to 1868 five periods of ten years each. The number of days during which the river was closed in each of these five periods is: For the first, 92 days; second, 92; third, 94; fourth, 90; fifth, 91.

The third period gives a greater number of days than any of the others; the general mean is about 91 days.

The conclusion at which we arrive from the evidence thus furnished by the Hudson River is, that during 50 years, that is to say, the whole period of trustworthy records, there has been no important change in the number of days that the river has remained frozen. In this respect the conclusion is the same as that which we have seen in the case of the Baltic rivers for a period of 300 years.

The evidence thus furnished from the closure of a river by ice differs from that of thermometric observations. The latter give merely the intensity of heat at the special moment, and in the special