56 



takes place from the surface of a considerable body of water. A polished 

 slate tank 3ft. by 2ft., and ij^ft. deep, was fixed in one of the filter-beds 

 of the Cockermouth Water Works, where it was partly immersed in water 

 which was being constantly renewed. The filter-beds are supplied by a 

 self-acting valve from a storage reservoir, and after percolating through 

 the gravel the water passes into another reservoir. This appeared so 

 admirably adapted for the end in view, that I was very sorry to be obliged 

 to relinquish the experiment before obtaining any good results. 



The observations were extended over part of three years, with 

 occasional interruptions from various causes, so that I cannot give 

 accurate figures ; but I believe the annual amount of evaporation under 

 the conditions named to be about 21 or 22 inches, i. e. about one half 

 of the rain-fall. 





