MOVKMKNTS ( ) I-' SAI.TS IX SOILS. 73 



oven the second foot is concerned. So, too, if it had transpired 

 that a heavy rain fell, before the rows had been ridged, and es- 

 pecially while they were depressed after applying the fertilizer, 

 the more soluble salts and those less strongly absorbed by the 

 soil might have been carried by percolation into the third and 

 fourth feet, so as to have developed differences at these levels. 

 It appears highly probable, however, that differences due only 

 to such a cause would have been obliterated by lateral diffusion 

 before the date of collecting the samples. 



The more probable explanation of the observed differences is 

 that they had been developed, partly as stated, but also as the 

 result of heavier percolation between the rows after rainfalls 

 and the capillary sweeping which followed. The rainfall rec- 

 ords show that on April 29, 30 and May 2, rain fell to the ex- 

 tent of .35, .20 and .16 inches, respectively, the latter occurring 

 only 3 days prior to taking the samples. With the ridged con- 

 dition of the surface and the generally level nature of the field, 

 a rapid fall of rain does have the effect of sometimes throwing 

 into the furrows the equivalent of 2 or 3 times the amounts of 

 water indicated by the rainfall observed, and in this way may 

 have established such conditions as are associated with furrow 

 irrigation, whose effects upon the movement of nitrates have 

 been cited. 



The table shows that, except in the case of the silica, and 

 perhaps the bicarbonates, the distribution of salts is such as 

 would be expected from furrow irrigation, and it appears more 

 probable that the differences are due to such an effect rather 

 than that the salts have either percolated or diffused directly 

 downward from the furrow where the fertilizers were applied. 



ON GOLDSBORO COMPACT SANDY LOAM AND SELMA SILT LOAM. 



In two other cases similar comparisons were made on sam- 

 ples taken under and between rows of peas, one on the Grolds- 

 boro Compact Sandy Loam and the other upon the Selma Silt 

 Loam. Both crops were planted the last of January, the 24th 

 and 25th. Under the pea rows, on the former soil, were ap- 

 plied 400 Ibs. of guano arid 25 bushels of cotton seed per acre; 



