The general results of these researches, which are believed to be sufficiently definite for general questions, are not only prized by the scientist, but they are of value to the engineer and the builder. Indeed, to all interested in costly structures or extended works liable to harm from wind pressures, the factor of safety is {49} a matter of no small pecuniary importance. These experiments show that, as was formerly believed to be the case, the wind pressure varies as the square of the velocity of the wind, expressed in miles per hour; but a most important fact has developed, namely, that the pressure in pounds per square foot is equal to the miles of hourly velocity multiplied by 0.004 instead of 0.005, as was formerly assumed. Professor Marvin was not content with one system of experiments, but he further attacked the problem in a direct manner by a method which checked and verified his experiments with the whirling machine. On the summit of Mount Washington, at an elevation of 6,300 feet, he obtained simultaneously and under the same conditions, by automatic and electrical apparatus, continuous registration of the pressure of the wind in pounds per square foot and of the velocity in miles per hour. The results thus verified can be considered as conclusive from a general standpoint. The corrections for the Robinson anemometer thus determined from these experiments are comparatively unimportant at low velocities, say from 10 to 15 miles per hour, being only a fraction of a mile per hour. The uncorrected velocities, however, are in all cases too large, and by greater and greater amounts the higher the velocity. At 60 miles per hour the observed velocities are about 12 miles per hour too high, and for an indicated velocity of 90 miles the experiments show that the actual velocity is but a fraction over 69 miles per hour. The anemometer formula found to satisfy most closely the entire range of experiments has the following form for velocities in miles per hour: Log. _V_ = 0.509 + 0.9012 log. _v_. This difference indicated by the formula may seem small and insignificant, as it is in the case of light winds, but at very high velocities the differences are very great. For instance, an actual velocity of 60 miles per hour may occur at some time in almost any locality of the United States for a few minutes, and even greater velocities are occasionally reported, apart from severe tornadoes. Under the old coëfficients for the Robinson anemometer an actual velocity of 60 miles per hour would have been reported as 77 miles per hour, which under the old factor of 0.005 would mean a pressure of 29.6 pounds per square foot; but when considered with reference to the true velocity of 60 miles, under {50} the new factor of 0.004, the pressure would only be 14.4 pounds per square foot--a reduction of over 50 per cent. from the pressure-values formerly accepted.