T. Deighton -l-n 



The principal objection to this is the fact that no stage corresponding 

 to the curd stage in soaps has been observed in sihca or other similar 

 gels hke gelatine; but if the curd fibres in soap had been transparent 

 instead of white it is by no means certain that they would have been 

 observed, and in this case since silica crystals are transparent while 

 soap crystals, of which the curd is thought to consist, are white, it seems 

 not impossible that silica curds may exist in the gel under certain con 

 ditious and yet have remained unnoticed uj:) to the present ; only further 

 experiment can decide the matter. Laing and McBain found the specific 

 conductivity of the curd to increase with concentration, but as the highest 

 concentration employed was -QN no deduction germane to the present 

 case can be made from this. 



It is worth mentioning that the first kink in the curve seems to be 

 related in some way with the "unfree" water of Bouyoucos(iO). The 

 mean of his values for various sands is 1-6 per cent, which is in good 

 agreement with the moisture at the first kink in the sands dealt with 

 here, which varies from 1 to 2 per cent. In the soils we have greensand 

 4-4 per cent., glacial loam 3-2 per cent, and boulder clay 8-4 per cent, 

 at the first kink. Bouyoucos' averages for unfree water in sandy loams 

 and clays come out 4-4 per cent, and 12-2 per cent. 



As regards the effect of the resi.stance of the distilled water used 

 with artificial mixtures this is quite what one would expect. In natural 

 sands and soils, salts are present in sufficient quantity completely to 

 mask any eft'ect due to this cause. 



It appears then, that for any pair of electrodes it should be possible 

 to construct an empirical curve from which, given the resistance between 

 them in any definite place the moisture content can at once be deduced. 

 The degree of accuracy attainable in practice would probably not be as 

 great as the lie of the observations about the curves would lead one to 

 expect as, apart from the error in sampling a large volume of soil for 

 standardisation purposes, the hysteresis eft'ects on wetting and drying 

 the colloids would undoubtedly have some effect. Moreover, the curve 

 obtained for the greensand seems to indicate another possible source of 

 error which might prove very serious. The portion drawn with a dotted 

 line^ was obtained quite by accident, after the last observation on this, 

 an abnormal fall in resistance was noted — the author therefore went 

 backwards and obtained the full fine curve. On adding water afterwards 

 this curve was consistently followed and it was found impos,sible to pro- 

 ceed again along the dotted portion. Thus there seems to be a possibility 



^ Corresponding observations marked with an asterisk in Tal>le VII. 



