598 EXPERIMENT STATION RECORD. 



amount docs not necessarily apply to other sites, and other observations are needed 

 before general statements respecting loss from this source can be made. 



"The seepage decreases after the lake is first filled, from the effect of silting and 

 from having filled the porous ground underneath and connected with the site. Even 

 in sand there is a limit to the amount of seepage, and the time during which the 

 loss is large. After sand beds connected with the reservoir are saturated the losses 

 from seepage will decrease. The loss increases with the depth, probably nearly as 

 the square. The losses may be lessened, though not entirely prevented, by silting. 

 The silting process is more efficient with small reservoirs, because of the better 

 distribution of the silt. 



"If the loss from seepage is not more than 2 ft. per annum, the sites may be con- 

 sidered as practically water-tight. In the case of canals the losses often average 

 more than that in 24 hours. 



"The losses from evaporation in the cases examined are greater than those from 

 seepage, and the evaporation is not necessarily the same from adjacent bodies of 

 water. The amount of evaporation increases with the temperature of the water, 

 with the wind, and diminishes with increased moisture in the air. From the standard 

 evaporation tank at the experiment station, the average evaporation for 11 years has 

 been 41 in. 



"Evaporation proceeds when the water is frozen, but at a diminished rate, averag- 

 ing about 1 to 11 in. per month. The evaporation at night is the same as during the 

 day, the difference being less with the increase of the size of the bodies of water. 



"The loss by evaporation from several lakes exceeded that from the standard 

 tank, the loss from the lakes being about 60 in. per year. The increase is due to 

 higher temperature of the water and to freer exposure to the wind. In some of the 

 summer mouths the lakes lost twice as much as the standard tank. The lower tem- 

 perature of water at high elevations, and the lower dew points, tend to decrease the 

 evaporation. 



"The diminished barometric pressure tends to increase the evaporation, amounting 

 to 14 per cent at 8,000 ft. and to 18 per cent at 10,000 ft. over the evaporation at 5,000 ft. 



"Since the winter period is longer at the high elevations, it is probable that for 

 the w T hole year the evaporation is considerably less at the high elevations than at 

 the low ones. 



" Every mile of wind movement in 24 hours increases the evapi ration by from 1 to 2 

 per cent over the evaporation if calm. Evaporation is lessened by any influence 

 which diminishes the wind or decreases the temperature of the water. Protection 

 of lakes by wind-breaks is in many cases practicable and in small lakes sometimes 

 desirable. In the large lakes the benefit is by reducing the wind velocity; in small 

 lakes both from effect on wind and by lessening action of sun. 



"The deeper the lake the cooler the water as a whole, the cooler the surface, 

 consequently the less evaporation. 



"Assuming a loss of 5 ft. in depth per annum, an area of 100 acres would require f 

 cu. ft. per second for the whole year to make good the losses from evaporation; one 

 of 500 acres would require 3i cu. ft. per second, considerably more than would be 

 used to irrigate an equal area. 



"The net loss to the reservoir would be the sum of the above losses from seepage 

 and from evaporation, diminished by the rainfall, a combined loss which may be 

 considered as a depth of 6 ft. in one year. As irrigation reservoirs are usually full 

 for a few months only, the loss is much less than this for the high water area." 



Object-lesson roads, R. Stone ( U. S. Depf. Ayr. Yearbook 1897, pp. 878-882, pis. 2, 

 fig. 1). — This is an account of the construction of short sections of roads to serve as 

 object lessons at New Brunswick, l-'lorence, and Englishtown, New Jersey; Geneva 

 and Iliou, New York; Warren, Pennsylvania, and Kingston, Rhode Island, with 

 remarks on State aid to road building in Massachusetts, and suggestions regarding 

 the construction of a coastwise and transcontinental highway to serve as a national 

 object lesson. 



