the end, the last and the least important of all, modifying slightly, perhaps, the total effect of all the other controls, comes the surface-covering of the earth. This may be snow, or grass, or sand, or lava. Here belongs the forest, a special kind of surface covering.
On the other hand, the forests. What do we mean by forests? Do we mean the vast, dense tropical forests of the Amazon, or a grove of trees on a New England farm? Have we in mind evergreen or deciduous trees, or both? Are the forest trees tall or scrubby? Is their height uniform or varying? Is there undergrowth or is the forest clean? Are we considering the forested slope of a steep and lofty mountain or the trees in a valley bottom; a tropical or an extra-tropical forest; a region of heavy or one of moderate rainfall; of much or of little cloud? Clearly, a complex problem is here before us. No wonder that so much diversity of opinion exists with regard to it. Few of those who discuss the question are at all aware of its extent or complexity. They see only one or two small aspects of it, and upon a very insufficient, and often inaccurate, knowledge they base broad and misleading generalizations.
In a matter of such general interest it is most important to proceed carefully, and to see clearly just what we do, and what we do not know. That is the purpose of the present paper: to set forth, as the writer sees it, the status of the "forest and climate" discussion in the light of the available facts. It may be added, parenthetically, that it is only comparatively recently that a scientific study of the subject has been possible.
The Historical Method of Treatment: its Unreliable Results
The favorite method of attacking the problem of forest influences has been the historical method. Probably the large majority of those who believe in such influences are affected, consciously or unconsciously, by the use of historical arguments. A certain region, we hear, was once forested. There are now few or no traces. "People" say that the climate there has "changed." Hence, the disappearance of the forests must have produced the change of climate. This is not an unfair illustration of the historical argument. Sometimes, of course, simple hearsay, and general impressions, are replaced by actual records of the change in area covered by trees, and by rainfall observations (extending over a relatively short period), or by rough accounts of the depth of water in rivers and streams. But, at best, this method of treatment is very unreliable. All the elements in the discussion are uncertain: the early forest conditions; the supposed "change" of climate; the accuracy of any available meteorological observations. Granted that a "change" of climate has actually taken place, was the so-called "change" the cause, or the effect, of the change in forest cover? And may not the "change" have been the result of the well-known oscillations of the climatic pendulum, which bring periods of wetter and then
