VI.] PRODUCTION OF NITRE. 55 



appears to exercise a similar order of preference. If a 

 double silicate of alumina and soda exists in the soil, 

 and lime should be brought in contact with it, the 

 silicate of alumina gives up the soda and takes up the 

 lime instead, and thus we get silicate of alumina and 

 lime. The presence of soda will not enable it to dis- 

 place the lime, as the silicate of alumina prefers the 

 lime to the soda. If, however, some potash be added, 

 the lime is given up and the potash is taken into com- 

 bination, because the silicate of alumina prefers the 

 potash, and thus we obtain silicate of alumina and 

 potash. But if ammonia comes within the influence 

 of this compound, there is so much preference for the 

 ammonia, that even the potash loses its position, and 

 then we get silicate of alumina and ammonia formed. 



105. The chief difficulty, as far as we know, appears 

 to be in getting the silicate of alumina to commence 

 taking in one of these substances. If by any means 

 you can produce even the lowest form, viz., the double 

 silicate of alumina and soda, there is no difficulty in 

 advancing through the higher stages. The real diffi- 

 culty is in the commencement of the series. In this 

 respect the energy of the caustic lime appears to be 

 very valuable, and thereby the double silicate of 

 alumina and lime is probably produced. The special 

 action that caustic lime, which has been slaked with 

 water containing salt, has upon silicate of alumina, 

 favours this view. We may, therefore, regard the 

 action of caustic lime upon clay which latter sub- 

 stance, you will remember, consists very largely of 

 silicate of alumina as contributing to the production 

 of that most valuable class of bodies which have been 

 called the double silicates. 



1 06. Caustic lime also favours the production of 

 nitrate of potash in the soil. This action when it 

 takes place in the compost heaps of the farm admits of 

 more careful observation, than when the change is ac- 

 complished in the soil. There is, however, no reason 



