314 The Shrinkage of Soils 



of studying in detail departures from the prevalent soil type over 

 limited areas. 



Recently the subject has been further investigated in relation to 

 certain West Indian soil types in the writer's laboratory, and the results 

 of these investigations are given below. 



These investigations fall naturally into two parts ; in the first the 

 relationship existing between shrinkage and water loss is considered; 

 in the second considerations affecting the proportionality existing 

 between the observed shrinkage and the content of colloidal clay are 

 dealt with. 



PART I. 



THE RELATIONSHIP BETWEEN CONTRACTION AND WATER LOSS. 



The method of attacking the problem consisted in studying in 

 detail the relationship existing between water losses and the corre- 

 sponding linear contraction on blocks of soil from 20 to 25 c.c. in volume, 

 measuring in each case incremental water losses and the corresponding 

 increments of linear contraction. 



Experimental. • 



Description of the apparatus employed^. 



The apparatus adopted took the form described below : 

 It comprised a small rectangular cage of brass wire gauze 1-6 mm. 

 mesh, and of the following dimensions, 8 cms. x 2-5 cms. x 2-5 cms. ; 

 this cage was provided with a slit 6 mm. in width running along the 

 centre of the upper surface parallel with the length. One end of the 

 cage was capable of being opened and closed at will so as to permit 

 the insertion therein of the block of soil under observation. The 

 dimensions of the cage and the blocks of soil were such that the blocks 

 fitted snugly inside the cage. 



The cage was supported in a frame of stout copper wire, provided 

 with a loop for suspending it. To the horizontal bar of the frame a 

 scale graduated in millimetres was fixed, the arrangement being such 

 that when the cage was placed in the frame the scale lay immediately 

 above the slit in the upper surface of the cage previously referred to. 



1 For certain features in the apparatus the writer is indebted to the paper by B. A. Keen 

 on "The Evaporation of Water from Soil," This Journal, 6, 456. 



