188 EXPEKIMENT STATION EECORD. 



river at the point in question. The percolation into the caual per square foot 

 of percohiting surface was found to be 0.01 to 0.02 gal. per minute per foot of 

 head. These figures are said to be low, " due to the vegetation and slime that 

 obstructs the sand through which water must pass, but it is mainly due to the 

 fact that stratified gravels do not readily transmit water if the water must 

 travel across or perpendicuhir to the direction of bedding." 



The cost of construction and subsequent expenses for maintenance are found 

 to be so large that the author expresses the opinion that a pumping plant con- 

 sisting of a Corliss condensing engine and centrifugal pump woiild be more 

 satisfactory. 



Use of water in agriculture — a contribution to the study of irrigated 

 meadows in the Vosges, M. R. Olry (A tin. Inst. Xut. Ayron., 2. *er., J {1906), 

 No. 1, pp. 283-382). — A thorough study of a system of agriculture in the depart- 

 ment of "Vosges, France, corresponding more or less closely with that found on 

 certain elevated plateaus of the western United States. 



The systems of water distribution described are similar to those existing in 

 many western States, and the same confusion and difficulty is experienced in 

 the strict enforcement of priorities in times of drought, to remedy which the 

 writer recommends irrigation associations between whom agreements shall be 

 made to use water- in turn for a period of 5 days each. 



With regard to the quantity of water used the author finds that " In the 

 great majority of cases the quantity of water used in the districts studied varies 

 between 150,000 and 300,000 caibic meters per hectare per year (sufficient to 

 cover soil to a depth of from 50 to 100 ft.). The amount diverted by the canals 

 amounts usually to 100 liters per second per hectare " (1.4.3 cu. ft. per second per 

 acre). — In commenting upon this apparently exorbitant use of water, the 

 author says : 



" Nobody has been able up to the present to stop this harmful practice, for 

 the farmer does not consider it prejudicial to the maintenance of soil fertility, 

 the general idea being that ' too much water can not be used,' yet by following 

 these methods the soluble constituents of the soil disappear, the humus becomes 

 less and less capable of nitrification, the free phosphoric acid is unable to com- 

 bine with bases and is slowly drained away, lime is dissolved in the presence of 

 the excess of carbonic acid and is likewise lost through drainage." 



A sand trap for irrigating ditches, II. O. Crafts {Eiujin. Rcc, ali (1906), 

 No. 6. pp. l.~)0, l')l). — Some considerations bearing on the necessity of such de- 

 vices in the case of canals diverting water from streams carrying large amounts 

 of foreign matter, and a description of a typical sand trap. 



The underflow in Arkansas Valley in western Kansas, C. S. Slighter 

 (U. 8. Gcol. Hurroj. Watcr-iSiippJij and Irrig. Paper \o. l')3. pp. 90, pis. 3. flgs. 

 2//). — Investigations on the underflow of the Arkansas River were made during 

 the summer of 1904. The following principal conclusions were reached: 



"(1) The underflow of Arkansas River moves at an average rate of 8 ft. 

 per 24 hours, in the general direction of the valley. 



"(2) The water plane slopes to the east at the rate of ~l ft. per mile and 

 toward the river at the rate of 2 to 3 ft. per mile. 



"(5) The underflow has its origin in the rainfall on the sand hills south of the 

 river and on the bottom lands and plains north of the river. 



"(7) The influence of the floods in the river upon the ground-water level 

 does not extend * mile north or south of channel. 



"(8) A heavy rain contributes more to the underflow than a flood. 



"(9) On the sandy bottom (JO per cent of an ordinary rain reaches the water 

 plane as a permanent contribution. 



