378 PROCEEDINGS OF SOCIETIES.
valuable information bearing on tlie point. These low temperatures de- pend to a great extent, no doubt, on the latitude, but they more par- ticularly are determined by the state of the sky and by local situation. If the climate is characterized by a dry calm atmosphere, the loss of heat by radiation during the nights of winter is great, and the temperature falls correspondingly low ; and owing to the greater density and conse- quent weight of cold air, it flows down the slopes of a country which is being cooled down, much in the same way as water, and settles in the low-lying situations. It is well-known that it is in such situations where the destructive effects of fro?ts are greatest on vegetable forms. Now what we wish to draw possible atlention to is this:— of the two con- ditions, heat and moisttu-e, we have in the distribution of rain a body of facts which will lead to a knowledge of the laws which regulate the dis- tribution of plants more quickly and more certainly than the facts of the distribution of temperature. It is not merely the daily and annual fluc- tuations of the temperature of a climate which may be inferred from a knowledge of its humidity ; it is not merely the greater immediate results which accrue to plants arising from a change in the humidity than rises from a change of temperature ; but what especially concerns the question is this vital distinction between the two, viz. as regards temperature, climates range into each .other by comparatively nice gradations, whereas on the contrary, the most diverse climates, as respects moisture or dryness, are frequently placed sharply side by side. I have been long impressed with the importance of a knowledge of the rainfall viewed in these aspects, and have recently constructed thirteen maps, showing for each of the months and for the year the rainfall over the greater part of the land of the northern hemisphere. The result of the whole discussion in its rela- tion to climate is vtry striking. Everywhere the rainfall is dependent on the prevailing winds and on the configuration of the earth's surface ; and since the prevailing winds depend on the distribution of land and water over the globe with respect to the heat of the sun, it follows that the pre- sent climates of the earth are determined by the relative distribution of land and water, and that with a different distribution we should have different climates. Perhaps nowhere on the globe does there exist such diversity of climate as in America, west of the Rocky Mountains ; and there is also very great diversity in India, the region of the Caucasus, and Spain and Portugal. The United States of America, both the prairie and Alleghany regions, British America, the Mediterranean regions, Sweden and Norway, and the south and east of Asia, have peculiar and well- marked climates. Not only so, but even in Russia, there are great dif- ferences arising from the different way in which the rainfall is partitioned through the months, especially from May to August. Now it is in those regions which present climates the most diverse from each other, and many of them sharply defined, that the geographical distribution of plants may be best studied, because these climates afibrd the conditions best suited for tracing the influence of climate in this distribution, and in the changes it eflects, or does not effect, on the habits and different organs of the plants themselves. But to develope this part of the subject so as to arrive at some really scientific knowledge of the laws regulating the dis- tribution of plants, it is absolutely necessary that the confusion in our present catalogues of plants be rectified, and that the areas of the distri- bution of species be stated with a precision much greater than has hitherto