EGYPTIAN COTTON AS AFFECTED BY SOIL VAEIATIONS. 21 



cotton rows in this field began in good soil where the growth of the 

 plants was normal and where the yield and character of the fiber 

 were as good as could be expected considering the late date of planting 

 (April 12). Toward the ends of the rows where the soil contained 

 gradually increasing quantities of alkali salts the plants were small, 

 produced few bolls, and ripened very late. Finally no plants 

 remained where the quantity of alkali was excessive. 



In two of these rows the soil was sampled to the depth of 3 feet 

 at the base of every tenth plant from the good end of the row until the 

 region was reached where the stand was much interrupted. There 

 samples were taken at more frequent intervals and finally alongside 

 each remaining plant. 



The soil from each 3-foot boring was thoroughly mixed and its 

 electrical resistance when saturated was determined by means of 

 the electrical bridge, the standard container of which has a capacity 

 of about 50 cubic centimeters. 1 From the resistances, the percentage 

 of total salts to dry weight of soil were computed by means of a special 

 correlation curve for the Sacaton type of alkali. This curve was 

 based upon determinations of the electrical resistance and of the 

 total water-soluble salts in 48 samples of soil which had been col- 

 lected two years previously on the experiment farm at Sacaton. 

 Analysis of these samples in the chemical laboratory of the Bureau 

 of Soils showed the average composition of the Sacaton alkali to be as 

 follows : 2 



Per cent. Per cent. 



Ca 1.0 



Mg 3 



K 1.0 



Na 31.9 



CI 24.4 



S0 4 16.4 



HC0 3 22.1 



C0 3 5.0 



Notes were made on the condition of each plant where a soil sample 

 was taken and the seed cotton was collected for examination in the 

 laboratory. Mr. Argyle McLachlan made a series of diagrams for 

 each of these plants, showing graphically the number and position 

 of the fruiting branches and the number of developed and aborted 

 bolls on each branch. Comparison of these diagrams indicated that 



1 The bridge is described and figured and directions are given for its use in determining the salt content of 

 soils in Bureau of Soils Bulletins 15 (by L. J. Briggs) and 61 (by R. O. E. Davis and H. Bryan). In the 

 latter publication tables are given for temperature correction (pp. 22 to 24) and for computing from the 

 resistance the percentage of total salts to dry weight of the soil (pp. 14 to 16). 



2 This composition of the alkali, and especially the presence of carbonates and of large quantities of 

 bicarbonates, probably explains the fact that the Sacaton curve does not agree with that on which is based 

 Table III, p. 14, in Bulletin 61 of the Bureau of Soils. The latter applies to alkali consisting of one-half 

 chlorids and one-half sulphates. For resistances below 40 ohms it indicates higher percentages of total 

 salts and for resistances above 40 ohms it indicates lower percentages than does the Sacaton curve. The 

 disparity increases until a resistance of 170 ohms indicates twice as much total salts on the Sacaton curve 

 as on the chlorid-sulphate curve. On the other hand, the curve for carbonates ('< black alkali") on which 

 is based Table VI, on p. 16, Bulletin 61 of the Bureau of Soils, indicates much higher percentages of total 

 salts for given resistances than on either of the other curves until a resistance of 170 ohms is reached, from 

 which point the Sacaton curve agrees very closely with the carbonate curve. 



[Cir. 112] 



