As usual, fakers step in where science fears to tread. With goose-bones (not their own) and hickory nuts they prophesy with all their might. And if their prophecies come true, as sometimes they must, there is wide rejoicing in the newspapers and the cause of science is set back by just so much. But science cannot be thwarted in the end and every year new discoveries are made, new speculations proved true or forever false, and some time, doubtless, the weather will be predicted from year to year with the same 85% accuracy with which the 36 hour forecast is now made. Experimenting is worth the little that it costs, too, for to know when the summer is to be dry or wet, hot or cold will be a boon to everybody and to the farmer most of all. One conclusion has already been reached by officials in the Weather Bureau and scientists generally. It has been decided by long search through creditable records, painstaking comparisons of averages coupled with the most accurate investigations for half a century, that, on the basis of ten years, our seasons do not change. That is, counting the decade as a unit, our weather keeps to the same level of efficiency through the centuries. This statement comes always as a blow. It always provokes argument and citations of grandmother's blizzards. There is a great and universal hesitation in believing that our weather is as good to-day as it used to be. The good old times when there was a general debauch of snow and you could skate all winter on anything but the Atlantic Ocean certainly appear no more. As a matter of fact there has been a change, but it has been in our memories. In grandmother's youth the trains,--if they had trains then,--doubtless were stalled by a big snow for then they did not have rotary plows. In father's day they may have had an unbroken winter of sleighing. We couldn't now; sleighs are extinct. But in our time, in fact every year, some record is being broken and the records go back a respectable length of time. For example in Philadelphia the most accurate records made by standard instruments have been kept for 43 years. During this time the highest wind velocity was recorded in 1878 (75 miles an hour). The greatest rainfall in 24 hours occurred in 1898 (5.89 inches). The lowest temperature was registered in 1899 (6 degrees below zero); the highest in 1901 (103 degrees). The greatest number of thunderstorms for any one year took place in 1905 when we had 51. As late as 1909 the heaviest snowfall ever recorded at this station, amounting to 21 inches, occurred. And just a few weeks ago (April 3rd, 1915) it snowed 19 inches in half as many hours. All these items do not indicate a climate decreasing in virility very swiftly.