46 BOAED OF AGRICULTURE. [Jan. 



oration. The first is usually of little consequence for our 

 present reriew, except in its absence, for cultivated soils 

 should be well drained, either naturally or artificially. The 

 second is of great import, as the loss through e*^^aporation is 

 greatly in excess of the loss through percolation (if percola- 

 tion can be considered a loss to the crop), and this our 

 figures heretofore given are sufficient to prove. We may 

 state, however, that the water which flows from drains dur- 

 ing the growing season in our climate is rarely the water of 

 immediate rainfall, but that of displacement, or more usually 

 that of the water-table, to which drains may be considered 

 as standing in the relation of an overflow. At Geneva, N. 

 y., the water-table in 1887, as measured by the depth of 

 well water from the curb (the well, happily for our purpose, 

 being situated on a swell of land which carried the soil sur- 

 face below the depth of the well bottom at a distance of a 

 few hundred feet in each direction), was 4 feet 2| inches 

 below the surface on May 1 , 6 feet 10 inches on June 1, 

 9 feet 2 inches on July 1, 10 feet 5| inches on Aug. 1, 11 

 feet 3 inches on Sept. 1, 13 feet 4^ inches on Oct. 1, and 

 15 feet 4 1 inches on Nov. 1, soon after which date the dis- 

 tance gradually decreased. This water-table was unaffected 

 by the rainfall during the growing season. Capillarity, as 

 measured by soil in tubes (a far less satisfiictory way than 

 in natural soil, but far more certain for experimental pur- 

 poses), was traced vertically 44 inches, and probably would 

 have extended farther had the experiment been designed for 

 this purpose, but slight breaks in the column are apt to 

 retard or terminate capillary progress. In natural soil, or 

 blufis along a shore, the base in contact with water, I have 

 noted the effect of capillarity in raising water apparently a 

 dozen feet or more. Horizontally the action of capillarity 

 carries water more rapidly and farther than it does verti- 

 cally. In nature capillarity brings an immense body of water 

 to the surface, to be removed as aqueous vapor by evapora- 

 tion, and in this manner, during some months, an amount of 

 water exceeding the rainfall of the month thus disappears 

 into the atmosphere. Evaporation is practically a surface 

 phenomenon, and may be checked in great degree by any 

 device that shall prevent capillary water from reaching the 



