as one writer has said, "as comparable as stations can be made." These Nancy results showed, for a period of about twenty-five years, and for the best pair of stations, somewhat more rainfall (about one half inch to one inch in the yearly average) in the forest. In the case of the other pair the excess was much greater. This series of comparative observations was unfortunately discontinued a few years since, and although the available data have been widely used, they are, in the opinion of the leading official meteorologist of France, as expressed in private correspondence with the present writer, inadequate to serve as the basis of a serious study.
The Indian Case
The two cases just cited are in the temperate zone. The other two cases are found within the tropics. There is, first, the case of a district in the central provinces of India, where forest protection and reforestation began in 1875, and where the rainfall, as compared with the rainfall of all India, showed an increase of about 12 per cent, in a comparatively few years. This, again, seemed an unanswerable argument in favor of a forest influence upon rainfall. But the complication due to periodic oscillations of climate, various uncertainties and the possibilities of error in the observations, together with the difficulty of "correcting" the catch, acknowledged by the Indian authorities themselves, have led to a feeling that we ought at least to suspend judgment in this case. Nevertheless, because the effect of wind upon the rainfall catch is less in the tropics than in our own latitudes, and therefore the error arising from the increasing protection afforded by the growing forest is greatly lessened, von Hann (1908), the acknowledged authority in climatological matters, is ready to accept the general result of these Indian observations as evidence in favor of an influence of forests in increasing the amount of precipitation at least in the tropics. Dr. G. T. Walker, however, the present director of the Meteorological Service of India, in a recent study of supposed changes of climate in India (1910), does not find evidence of an effect of forests in increasing rainfall.
The Java Case
Finally, we may cite the Java case, which is without question the most striking of all. This case was studied and first discussed a good many years ago by Professor Alexander Woeikof, of St. Petersburg. The facts as given by him are these: There are extensive dense forests in the south of Java, while the north coast has been largely deforested. A station, Tjilatjap, on the south coast, distant from the mountains, has a mean annual rainfall almost twice as large as that of three stations (Batavia, Tegal, Samarang) on the north coast. The difference is, in round numbers, about 150 inches against 75 inches. The north side is the windward side for the northwest monsoon, and during the
