THE DENSITY AND VOLUME-WEIGHT OF SOILS. 113 



In the arid region it will frequently happen that such soils 

 when not tilled to a sufficient depth, will during the later part 

 of the summer so shrink and crack beneath the shallow-tilled 

 surface layer that the latter will bodily fall into the cracks, ex- 

 posing the roots to all the deleterious influences of mechanical 

 lesion and drying-out. It is thus obvious that the cultivation 

 of such soils should not be undertaken at all by those not nat- 

 urally able and willing to bestow upon them, to the fullest ex- 

 tent, the deep and thorough tillage which is absolutely essential 

 in the utilization of their usually high productive power. 



Extent of Shrinkage. The extent of this shrinkage in drying, and 

 subsequent expansion in wetting, have been measured by the writer by 

 the use of the sieve cylinder described below (chapt. n, p. 209), a^ 

 serving for the determination of the water capacity of soils. When a 

 soil of the kind above referred to is placed in the sieve cylinder in the 

 tilled (flocculated) condition, then allowed to absorb its maximum of 

 water and then dried at 100 degrees C., the contraction in drying can 

 be very strikingly seen, and its amount measured by filling up the 

 empty space with mercury ; then measuring the latter after expelling 

 the surplus by means of a ground glass plate laid on top. The con- 

 traction of several heavy clay soils, thus measured, has been found by 

 the writer to range from 28 to as much as 40 per cent, of the original 

 bulk. 1 The soil thus contracted, when again wetted, does not return 

 altogether to its original bulk, but remains in a more or less compacted 

 condition, like that of a soil which has been rained upon. 



The expansion and contraction of a heavy clay soil on wet- 

 ting and drying are well illustrated in the figure below, in which 

 the soils are shown in the shallow cylinder which serves for the 

 determination o'f water-holding power (see chapt. i i, p. jo<)). 

 The middle figure shows in profile the expansion of a dry, 

 pulverixed " black adobe." struck level, when allowed to absorb 

 its maximum of water; it rises above the rim of the sieve-box 

 to nearly the half height of the latter. The outside figure to 

 the right shows the same soil after drying; that to the left, a 

 red clay soil similarly treated. It is easily seen that the<r 

 variations in volume may bring about very marked results in 



1 Wollny ' Korsch. Vol. 20, p. 13 ff, 1897) records similarly high shrinkages in 

 his experiments. 



8 



