The Location of the Rain Gauge.—In establishing any sort of a station where the measurement of rainfall is to be recorded, at least two rain gauges are desirable. One of these should be a standard Weather Bureau gauge or one of similar pattern; the second may be any vessel with an 8-inch circular opening in the cover.
In cities which are solidly built the flat roof of a building offers about the only suitable place for a rain gauge. If the edge of the roof is a parapet, so much the better, for the drive of the wind is less apt to blow aside the rain that should fall into the receiver. In a position of this sort the catchment of the two gauges is not likely to differ materially.
In suburban localities and in communities where buildings are 100 feet apart the gauges are better placed in such positions as have the full sweep of the rain-bearing winds. If two places 100 feet or more apart show no material difference in the catch, either location is probably suitable. With gentle rain and still air the two gauges should be in close agreement; if the wind blows in strong gusts there may be a material difference.
The wind is the chief obstacle to accuracy of rainfall measurement and shielding the gauge from the full strength of the wind is the best means to insure an accurate catch. The Nipher shield is a trumpet-shaped metal device about 20 inches across which surrounds the mouth of the gauge. It is surmounted by a rim of copper mesh which prevents insplashing. J. O. Alter, Observer at Salt Lake City, constructed a much simpler shield by fastening a strip of canvas about 9 inches wide, to a metal ring about 30 inches diameter. The screen thus constructed is suspended about the gauge by metal struts. The edge is about 2 inches higher than the mouth of the gauge. The Weather Bureau regards this shield with favor. The author has found a similar shield made of copper mesh, such as is used in window screens, a most excellent device. In the long run, a shielded gauge will catch from 6 per cent to 10 per cent more rain than one unshielded, according to the experience of the Weather Bureau. P. R. Jameson, with measurements covering many years, finds a gain of about 9 per cent in the case of shielded gauges.
The pit-gauge is favorably considered by C. F. Marvin, an authority on precipitation. The pit-gauge is merely a