CHAPTER XXIV. 173 



gauges record all that falls actually within them, but, except in very 

 still weather and with gentle rains, they do not intercept all that they 

 should. The wind sweeping across them carries away a portion o£ 

 the precipitation, varying from 14 to 7 per cent, for protected gauges, 

 and from 7 to 40 per cent, for unprotected gauges. The decrease in 

 catch of rain gauges raised above the ground, formerly believed to have 

 been due to height, has now been shown to be due to the increased force 

 of the wind. Some very interesting data on this subject has been pub- 

 lished by the United States Department of Agriculture. It also ap- 

 pears that the deficit during gentie, fine rains is very much greater, 

 Eccording to the strength of the winds prevailing at the time, than 

 during heavy rains. There was from 18 to 52 per cent, shortage during 

 fine rains, and 6 tp 17 per cent, during heavy rains. The standard for 

 comparison being what is known as the "pit gauge" at ground level. That 

 the loss should vary with the strength of the wind is readily explained. 

 Draw a diagram of forces, the perpendicular representing the down- 

 ward force of the rain drops, the horizontal the force of the wind in Its 

 sweep across the rim of the gauge. It is obvious that the resultant 

 will deviate more and more from the vertical as the rain movement Is 

 light and the wind strong. 



OTHER ELEMENTS OF UNCERTAINTY, 

 There is still a greater element of uncertainty, viz:, the smallness 

 of the actual collecting area of a gauge and the comparative immensity 

 of area of country to which its readings are applied. If to every four 

 square miles there was a 10-inch gauge, and this is much closer than 

 we find them, the ratio of areas would be as about 200 million to 1. At 

 Rothamsted, in England, they have a rain gauge measuring 7.26 feet 

 by 6 feet, giving an area of 43.56 square feet on exactly one thousandth 

 C.I an acre. The catch on this gauge from 1853 to 1880, twenty-eight 

 years, was about 9.8 per cent, more than the catch in an adjacent 

 5-inch gauge. The ratio of their respective catchment areas is as 

 320 to 1. Now, if this ratio of area gave a variation of nearly 10 per 

 cent, in the catch, we must not rely too implicitly on the results shown 

 by applying the readings of a gauge to an area of country several 

 hundred million times the area of the gauge. 



All the various difficulties considered we are not warranted in 

 hoping for any decisive direct quantative comparison between the 

 rainfall over forests and open ground. However, while lacking in this 

 direct proof we do know from the records of the various forest stations 



