water do you suppose evaporates from that reservoir per day, on an 

 average, from May to October? Most people guess in gallons, and when 

 we suggest barrels they look incredulous ; yet it is a fact that on an 

 average during that whole period 20 barrels a day were lost by evapora- 

 tion, a depth of one-fifth of an inch. The greatest loss on any one day 

 was 50 barrels, which occurred between 6 o'clock on the evening of 

 August 24th and 6 o'clock on the evening of August 25th. The three 

 days preceding had been excessively warm, but about 4 o'clock on the 

 24th the temperature dropped suddenly, and a very strong wind rose 

 which continued throughout the night and the following day. 



In measuring the evaporation we use a graduated glass standpipe 

 of water which feeds automatically into an evaporating cup so arranged 

 that the wind cannot blow the water out although the evaporating surface 

 is level with the top of the cup. The amount that has passed out of the 

 standpipe gives the depth of water evaporated since last observation, and 

 from this we can calculate the amount in barrels. When we have another 

 season's work on this and on evaporation from soils, we hope to publish 

 a detailed report of our methods and our results. 



An evaporation of 20 barrels a day from an area 100 feet by 60 feet 

 is equal to about 140 barrels per acre. The amount will, of course, vary 

 with the situation, exposure, temperature, etc. What the exact loss 

 from soils would be during that period, we are not yet in a position to 

 say, that problem lies all before us; but, from preliminary tests, we have 

 reason to believe that so long as the soil is bare and looks moist on the 

 surface, evaporation is robbing it of its moisture about as fast as it takes 

 water from the reservoir. But as soon as the soil looks dry, or is hidden 

 by a crop, the rate of evaporation falls off very rapidly. 



These latter conditions are best brought about by cultivating and 

 seeding as soon as the land is dry enough. If there are two plots of soil 

 side by side, and one is cultivated and the other is not, the evaporation 

 from the cultivated one is much greater for a day or so than from the 

 other, but this evaporation takes place chiefly from the loosened portion, 

 and hence in a very short time, provided no rain falls, this layer becomes 

 dry and acts as a blanket to protect the soil below, diminishing the evap- 

 oration in one test we made by 62.5 per cent. Hence it is a matter of 

 vital importance that the soil should be cultivated at the earliest possible 

 moment. A delay of one week in this operation after the soil is fit will 

 rob the soil of from one to two inches of water, an amount sufficient to 

 tide the crop over the critical period of a drouth. Deep cultivation is not 

 advisable, for all of the loosened layer dries out in time of dry weather, 

 and since the deep blanket is little, if any, more effective than the thinner 

 one, the extra loss from the thicker blanket itself is not atoned for by 

 greater saving of water in lower layers, and is therefore a net loss to the 



plant. 



With cereals the conservation of moisture by cultivation may be 

 continued until the grain is nicely up. If a rain has come, packing down 

 the soil and destroying the loose blanket and thus setting^ up rapid 

 evaporation again, it is good practice to run over the crop with a light 

 harrow and restore the blanket. The saving in moisture will more than 



