275 COTTER : SODA INDUSTRY OF SIND. 



From the above analyses it will be seen that while the dhands of 

 the Nara region are typical "triple" waters containing mixtures 

 of sulphate, carbonate, and chloride, those of the Kot Jubo area 

 are more of the carbonate-chloride type with subordinate amounts 

 of sulphate. 



Laniwari, Nau Rait, and Rait Pario are exceptions to this, 

 being " triple " waters, while Leyara is a sulphato-chloride dhand, 

 as is also Lambro. 



Laniwari, Nau Rait, and Rait Pario lie in the more open country 

 north of the great sand-plateaux or dra-ins of Pur Chandar and 

 Sanoi ; it may then be said that those dhands which are in close 

 association with the dra-ins are of the carbonate-chloride type 

 or lend to that type, while those of the open country composed 

 of bhit and tali tend to be of the "triple" type. 



The proportion of carbonate to chloride in the dhands is much 

 lower than it is in the efflorescence'. This means that the sim or 

 percolating water carries greater quaafciiie. of chloride than the 

 efflorescence which it causes. We would naturally expect this to 

 be the case, since the predominence of carbonate in the efflorescence 

 is due to fractional crystallisation, whereby trona is deposited 

 (or sodium bicarbonate), while a large part of the chloride stays in 

 solution. 



Analyses of ckamho. The J*^ **> *he analyses of the speci- 



mens of chamho collected. 



The last column shows the quotients obtained by dividing the 

 quantity of sodium carbonate present by the amount of sodium 

 bicarbonate. This may be called the " carbonate-bicarbonate index." 



The theoretical carbonate-bicarbonate index in trona is 106/84 

 or 1'262. In practice, in natural trona the carbonate-bicarbonate 

 index appears never to be quite so low as the theoretical value. 

 The index for the trona of the Lonar Lake (dulla), calculated from 

 F. J. Plymen's analysis given in La Touche and Christie's paper 1 

 is 1*413. 



In the above analyses, the indices in the cases of fourteen 

 dhands are below 1*600. The chaniho of these fourteen dhands 

 may be regarded as trona, not altogether pure perhaps, but still 

 trona. 



La Touche and Christie, Rec. GeeL Sur. Ind., XJLI, p. 260. 



