236 



Journal of Agricultural Research 



Vol. X, No. 5 



nights. Tank 3 has the low mean temperature; tank 7 comes between 

 No. 3 and the lake, which is high. Figure 5 shows the influence of heat 

 upon evaporation pans 2 feet in diameter under exactly the same con- 

 ditions otherwise. There is no reason to believe that this relation would 

 not extend to surfaces of larger water bodies of the same type. The 

 curve would indicate that an increase of one degree mean temperature, 

 all other conditions being the same, would increase the evaporation 

 approximately 5 per cent at Denver for the season under consideration. 

 The mean temperature of the lake is approximately i degree higher than 



1 ^f 



111 



Fig. 6.— Relation of water and air temperatures. 



that of tank 7. From that cause the lake evaporation would be then 

 approximately 5 per cent more than that of tank 7. However, there is 

 the ratio of perimeter to area, which v/ould tend to make the lake 

 evaporation less. The vapor-blanket efifect mentioned previously 

 would also lessen the lake evaporation in comparison with that from a 

 land pan. The wind movement was practically the same at the lake as at 

 the laboratory. A comparison between the 3-foot square tank at the lab- 

 oratory and the floating tank would tend to strengthen the vapor-blanket 

 theory. The mean temperature of the floating tank is higher, its wetted 

 metal perimeter the same, yet its evaporation is less. 



