=106. The rain gauge=, as the illustration shows, is a cylindrical vessel with a funnel-shaped receiver at the top, which is 8 inches in diameter. The receiver fits closely upon a narrower brass vessel or measuring tube in which the rain collects. The ratio of surface between receiver and tube is 10 to 1. For readings covering a general area, the rain-gauge is placed in the open, away from buildings or other obstructions, and is sunken in the ground sufficiently to keep it upright. In localities where winds are strong, it is usually braced at the sides also or supported by a wooden frame. In measuring the amount of rain in the measuring tube, the depth is divided by ten in order to ascertain the actual rainfall. The depth is measured by inserting the measuring-rod through the hole in the funnel until it touches the bottom. It is left for a second or so, quickly withdrawn, and the limit of the wetted portion noted. In the case of standard rods, the actual rainfall is read directly in hundredths, so that the division by ten is unnecessary. After each reading, the measuring tube is carefully drained, replaced, and the receiver put in position. No regular time for making readings is necessary. During a rainy period, it is customary to make a measurement each day, but it has been found more satisfactory for ecological purposes to measure each shower, and to record its duration. These two facts furnish a ready clue to the relative amount of run-off in each fall of rain. The measurement of snowfall is often made merely by determining its depth. For comparison with rainfall, the rain gauge with receiver and tube withdrawn is used. The snow which falls is melted, poured into the measuring tube, and measured in the ordinary way. The U. S. Weather Bureau standard rain gauge, with measuring stick, may be obtained of H. J. Green, or of J. P. Friez for $5.25. =107. Precipitation records.= From the periodic character of precipitation, rainfall sums, means, and curves have little importance in the careful study of the habitat. The rainfall curve for the growing season is an aid in explaining the curve of water-content, and the mean rainfall of a region gives some idea of its vegetation, though even here the matter of its distribution is of primary importance. The rain and snow charts published by the U. S. Weather Bureau furnish data of some importance for the general study of vegetation, but it is evident that they can play little part in a system which is founded upon the habitat. Precipitation records, for reasons of brevity and convenience, are united with wind records, and the form will be found under the discussion of this factor. WIND [Illustration: Fig. 23. Simple anemometer.]