44 Transactions of the Soutli African PJiilosopJiical Society. 



73 "8 inches during the four summer months, May, June, July, and 

 August, and the evaporation of the ^Yhole year at upwards of 

 140 inches." '''' This is probably the largest estimate for England 

 ever seriously made. 



Symons claimed that the different rates at which the water 

 evaporated from different gauges '' could not fail to convince any 

 one that the key to the whole question was to be found in one item, 

 the temperature of the water." Latham argued, however, that the 

 greatest deviations arise as much, or more, from capillarity, '' the 

 water rising on the sides of the gauge and thus inordinately 

 increasing the amount of evaporation. Consequently a small gauge, 

 having a larger amount of side area, in proportion, than a large 

 gauge, indicates a very much larger amount of evaporation. . . . 

 In a 5-inch evaporating gauge this capillarity adds nearly 40 per 

 cent, to the water area of the gauge, whereas in a gauge 1 foot in 

 diameter the influence of capillarity is less than half that of a 5-inch 

 gauge, whilst in a gauge 6 feet in diameter (or 6 feet square) the 

 influence of capillarity is only one-sixth part of that of a 1-foot 

 gauge." t The same authority showed, moreover, that if any given 

 gauge were coated with some dressing, or pigment, which would 

 lessen the capillary attraction between the walls of the gauge and its 

 water, then the rate of evaporation would become notably decreased. 

 For example, three 5-inch gauges, one plain copper, one enamelled 

 white, and another enamelled black, but otherwise equal and 

 similarly situated, standing in a tank of water up to the level of 

 the evaporating surface, gave the following comparative annual 

 average loss : — 



Inches. 



Copper gauge 27*90 



Black „ 22-97 



White „ 21-74 



while a copper gauge of the same dimensions, freely exposed in air, 

 gave in the same period an average of 36"96 inches, and the 1-foot 

 floating gauge 19'40 inches. Thus a given unprotected and 

 unsheltered gauge lowered its indications by nearly 25 per 

 cent, when relegated to the somewhat cooler and more humid 

 (and presumably less windy) environment of the tank ; whereas 



* Dr. Dobson in some " Observations on the Annual Evaporation at Liverpool 

 in Lancashire." Phil. Tr((ii>f., vol. Ixvii. 



f Many authors agree with Tjiithani in making evaporation proportional to the 

 surface area. But see Preston, who cites Stefan, making evaporation proportional 

 to diameter, 'llicorij of Ilcat. 



