tems, a classification of the zones by winds also emphasises the conditions of rainfall. In such a scheme, the torrid, or tropical zone, with its regularity of weather through the year, and the comparative simplicity of its climatic features, is bounded on the north and south by the margins of the trade wind belts, and is therefore larger than the classic torrid zone. This trade wind zone is somewhat wider on the eastern side of the oceans, and properly includes within its limits. the equable marine climates of the eastern margins of the ocean basins, even as far north as latitude 30° or 35°.

Most of the eastern coasts of China and of the United States are thus left in the more rigorous and more variable conditions of the north temperate zone. Through the middle of the trade wind zone extends the sub-equatorial belt, with its migrating calms, rains, and monsoons. On the polar margins of the trade wind zone lie the sub-tropical belts, of alternating trades and westerlies. The temperate zones, with the great irregularity of their weather phenomena and their marked seasonal changes, embrace the latitudes of the stormy westerly winds, having on the equatorward margins the sub-tropical belts, and being somewhat narrower than the classic temperate zones. Towards the poles, there is no obvious limit to the temperate zones, for the prevailing westerlies extend beyond the polar circles. These circles may, however, serve fairly well as boundaries, because of their importance from the point of view of insolation. The

polar zones in the wind classification, therefore, remain just as in the older five-zone scheme.

A compromise between the rigid division by latitude circles and the isothermal and wind classifications has been suggested by Woeikof, who objects to limiting the torrid zone by the tropics on the ground that the high temperatures of that zone, as well as its characteristic winds, extend beyond these parallels. Latitude 30° would be a more natural boundary; but as the westerlies, which are characteristic of the temperate zones, prevail there in winter, latitude 25° is chosen as a compromise between 2312° and 30°. The polar zones are bounded by latitude 65°. When bounded by these several limits, the areas of the different zones are as follows:

Summary and Conclusions. Reviewing what has been said regarding the climatic zones, it would seem that, all things considered, a simple division by isotherms, such as that suggested by Supan (1896), is the best for general use. The early division by latitude circles, while it has the merits of great simplicity, and emphasises the all-important element of sunshine, is too arbitrary, and hence does not accord sufficiently well with the facts of actual climate. Nevertheless, we should not discard the classic zones without recog

nising that they have a real meaning in relation to solar climate. The grouping of the climatic zones according to wind systems has much to recommend it from a meteorological standpoint, but is not quite simple enough for general use. Its adoption involves an understanding of the great wind and calm belts of the world, and of the migration of these belts. The shifting of the boundaries of the torrid zone also brings in an element of uncertainty which is somewhat confusing, although, as a place in the sub-tropical belt really changes its climate with the seasonal change from westerlies to trades, and vice versa, it may reasonably be expected to change its zone. other words, actual climatic conditions are recognised; and in any case, this is a more reasonable plan than to limit the torrid zone by means of the tropics, which arbitrarily cut across the trade wind belts and separate areas which are climatically the same. The temperature zones proposed by Köppen, while useful in special studies of plant distribution, are too detailed for general adoption.

In

Whatever climatic zones we adopt, we should certainly abandon the word temperate altogether as the designation of the middle zone in each hemisphere, and substitute some such adjective as intermediate for it. The words torrid and frigid should likewise disappear, and be replaced by tropical or equatorial, and polar.

Necessary Subdivisions of the Zones. However we may classify them, the climatic zones are far from

being uniform in character throughout their whole extent. Hence, no brief, simple description of the climate of a zone can be given. For this reason, suggestions have been made regarding subdivisions of the different zones. Thus, in the case of the classic north temperate zone, it has been proposed to subdivide it into sub-tropical, temperate, and sub-arctic, but the question how to limit these subdivisions is difficult to settle. A more rational scheme is that which, in view of the great differences in the climatic relations of land and water, recognises a first large subdivision of each zone into land and water areas. Then, as continental interiors differ from coasts, and as windward coasts have climates unlike those of leeward coasts, a further natural subdivision would separate these different areas. Finally, the control of altitude over climate is so marked that plateaus and mountains may well be set apart by themselves as separate climatic districts. If each of the zones, whether bounded by latitude circles, or by isotherms, or by wind systems, be considered under these general subdivisions, as close an approach to actual conditions of climate will be made as is possible in general description. Obviously, however, when the larger zones are subdivided to such an extent as is here suggested, we are dealing with a classification of climates rather than with climatic zones.

Need of a Classification of Climates-Relation of Continental and Ocean Areas to Temperature: Reasons for the Slow Change in the Temperature of Ocean Waters-Marine or Oceanic Climate-Continental Climate-Desert Climate-Coast or Littoral Climate-Monsoon Climate-Mountain and Plateau Climate-Mountains as Climatic Divides.

Need of a Classification of Climates. A broad division of the earth's surface into zones is necessary as a first step in any systematic study of climate, but it is not satisfactory when a more detailed discussion is undertaken. The reaction of the physical features of the earth's surface upon the atmosphere complicates the climatic conditions found in each of the zones, and makes further subdivision desirable. Under the control of these different physical conditions, the climatic elements unite to produce certain fairly distinct types of climate, and these may be classified in various ways. The usual method is to separate the continental (near sea-level) and the marine. An extreme variety of the continental is the desert; a modified form, the littoral; while altitude is so important a control that mountain and plateau climates are further grouped by themselves.