which were thought to have resulted therefrom. However, the aurora has not yet been satisfactorily explained. With the exception of the aurora, there is no known relation between terrestrial magnetism and atmospheric phenomena.
The question as to whether or not forests affect weather and climate has been much debated. Recent investigations have brought out the following facts: Whatever influence forests have upon meteorological conditions is purely local, and even that influence is not marked. In one case it was found that the mean annual temperature within a forest was only a few tenths of a degree cooler than at a point a half mile or a mile outside the forest border, the greatest difference amounting to 2° F. The relative humidity was at times 7 per cent. greater within the forest than in the open country. In the United States the wholesale destruction of forests, which has been going on since colonial times, has not been accompanied by any marked increase or decrease in rainfall. On the other hand, the reforestation of large tracts in central Europe and in northern Africa during the past century has not resulted in an appreciable effect upon the precipitation observed during that period. Forests are the effect rather than the cause. There is still considerable confusion in the public mind concerning rainfall and flowoff, when the supposed influence of forests is considered. Deforestation has undoubtedly increased the frequency and the intensity of floods in small constricted districts, notably in certain mountain valleys, but where the removal of the forest cover over large areas has been followed by cultivation of the soil the rate of flowoff has remained unchanged. From hydrographs of the principal rivers of the United States it is apparent that high waters are neither higher nor low waters lower than they were fifty years ago, and they are neither more frequent nor of longer duration now than they were then. Notable floods like that of Paris, France, during the spring of 1910, and that of the Ohio Valley in the spring of 1913, are the result of a number of causes, in which the excessive rainfall was in no way related to the presence or absence of forests, and in which the rapid flowoff was more dependent upon the frozen soil than upon the recent removal of the forest cover. That the flowoff is more rapid when the ground is frozen explains the greater frequency of floods during spring than during any other season of the year. Moreover, forests tend to preserve the snows of winter, as well as to retain the fertile elements of the soil from washing away. While forests are thus of importance to the agriculturalist and the engineer, they are of little concern to the student of the weather.
The deep-seated notion, held by many individuals, that the climate is changing is often referred to in expressions like "old-fashioned winter," "the storms we used to have," and "the deep snows when I was a boy," etc. Subjective phenomena like these are of inter-
