26 



REPORT — 1869. 



ratio of luS, 1-27, and 1-00. Fig. 1 therefore lost 78 per cent, more 

 water by evaporation than fig'. 3. 



HOOK 



GAUGE 



E^ 



» 



r/c. t. 



K 



Fia.3, 



(2) During- the daytime the sunshine heats figs. 1 and 2 to such an extent 



that the ratios" of evaporation become about 2-150, 1-50, and I'OO. 

 (.3) Durins? the night there are indications of a sliglit addition to fig. 3 from 



condensed vapour. 

 (4) The evaporation, as computed from the hygrometer, bears no reguhnr 

 relation to any of the others, being sometimes greater than any of them, 

 and sometimes less. The total computed evaporation is 3-39 inches. 

 As already stated, the authors consider that the accuracy of an evaporator is 

 largely dependent on its capability of retaining the temperature of the Avater at 

 a? nearly as possible that of large volumes of veater, such as reservoirs. In the 

 few coniparisons they have been able to make, they have found that the tempera- 

 ture of the water iii fig. 3 has l3een nearly identical with that of a rather shallow 

 reservoir one acre in extent. The average temperature of the water about 2 p.m. 

 was in fio-. 1, 80°-7 ; in fig. 2, 75°-8; and in fig. 3, 73°-8, showing an average excess 

 of 7° in the temperature of fig. 1 over that of fig. 3. In sunshine there is an excess 

 of twice that amount ; in fact, at times the metal becomes so hot as to scorch the 

 hand. 



The authors conclude with a strong plea for fm-ther investigation, by quoting 

 the words of M. Yalles, the French "engineer, who first called attention to the 

 great inconsistency of existing experiments. " "We do not understand how, in a 

 country like o\irs, and with reference to one of the most important hydraulic 

 data, we can rest content with only knowing that the numerical value to be attri- 

 buted to this datum Ijes between two limits,' one of which is double the other."' 



