juiyjo, I9I7 Evaporation from Water and River-Bed Materials 241 



of 7.4 miles per hour for anemometer i, 14 feet above the ground, and 

 5.3 miles per hour for anemometer 3, 2 feet above the ground. The 

 increase is 39 per cent. An investigation of the variation of wind velocity 

 at different heights in connection with the Salton Sea investigation 

 gave justification for the use of a formula which indicated that the 

 velocity at a point, which from the beginning of the discussion referred 

 to must have been 45 feet above the ground, was 141 per cent of the 

 velocity at the bottom of the tower. This tower was fully exposed and 

 free from interference to wind movement. The variation from the 

 bottom of the tower to the top is shown as a straight line for all veloci- 

 ties (2, p. 30-31). Records giving a comparison between wind move- 

 ments at the top of the Eiffel Tower, 984 feet, and the housetop level of 

 Paris, probably less than 50 feet, indicate that the tower velocity is 

 about lour times that at the housetop level (14, p. 373). 



Each evaporation observation station should have its own anemometer. 

 Weather Bureau records of wind movement taken in connection with 

 future evaporation measurements will probably give figures for the 

 water surface, or approximately so (17). 



RELATIVE HUMIDITY 



An indication of the quantity of moisture mixed with the air is given 

 by the relative humidity percentage. This, determined from the tem- 

 perature of evaporation, should, in connection with other factors, be an 

 index of evaporation depths from water surfaces. The term and its 

 value are somewhat indefinite, being "relative." Taken in connec- 

 tion with the corresponding air temperature, the vapor pressure is 

 obtained, which factor has been used in nearly all of the theoretical 

 evaporation formulas proposed. Relative humidity figures or data form 

 which to get these should accompany water surface evaporation studies. 

 In many cases they do, usually taken, however, but once or twice a day. 

 The use of the wet and dry bulb thermometer is doubtless the best 

 method; however, a continuous record, by far the most valuable, is next 

 to impossible by its use. The two readings a day method, however, 

 gives an approximation to the integrated curve of the hygrograph, if the 

 means are considered for a period of some length. As an illustration of 

 this, a reproduction of the humidity record for one week is shown in 

 figure 7 and the data are given in Table XIV. 



The weekly mean from all 7 o'clock readings from this is 70 per cent. 

 The integrated mean is 65 per cent. If but one reading a day is given, 

 as is sometimes done, the variation is still greater, since the range during 

 the day, particularly in this climate, is great. The hygrograph when 

 properly adjusted and checked frequently with the psychrometer gives 

 the most satisfactory record for evaporation use. 

 98976°— 17 3 



