441 



weight of silica. The number 22 being the equivalent for carbonic 

 acid on the hydrogen scale, the equivalent of silicic acid is obtained 



by the proportion g= 22 x wei S ht of silica used - 

 weight ot carbonic acid expelled 



Four experiments are detailed, made with carbonate of potash, 

 which give as a mean result the number 30' 7 for the equivalent of 

 silica. This agrees with the formula SiO 2 , and nearly with the 

 previous results of H. Rose. Then follow seven experiments made 

 in a like manner with carbonate of soda, which give as a mean result 

 the number 21 '3 as the equivalent of silicic acid a number agree- 

 ing nearly with half that represented by the formula SiO 3 , or g 



Some experiments are then related, which go to show that the 

 increased loss resulting with carbonate of soda could not be caused 

 by the action of heat alone. The author had next recourse to car- 

 bonate of lithia, and obtained as the mean result of four experi- 

 ments with this substance, agreeing well together, the number 14-99 

 a number which accords very closely with the formula SiO. 

 These different numbers, obtained with silica, led the author to in- 

 quire whether any other body acting as an acid produces similar re- 

 sults with the fused carbonates of potash and soda. With this view, 

 experiments mere made with dry sulphate of magnesia, as a substi- 

 tute for sulphuric acid, with bi-borate of soda, for boracic acid, with 

 alumina and sesquioxide of iron. Of these bodies only boracic acid 

 gave results similar to those obtained with silicic acid. The other 

 substances all gave the equivalent numbers usually assigned to them, 

 equally with the carbonates of potash and soda. 



Directing his attention then to determine whether the equivalent 

 of silicic acid could be found in other volatile acids than the carbonic, 

 the author relates some experiments made with the hydrates of 

 potash and soda, but he explains that there are circumstances which 

 render it much more difficult to obtain accordant numbers with 

 these bodies than with the carbonates. 



Six experiments made with hydrate of potash gave as mean result 

 the same number as that obtained with the carbonate, viz. 30'8. 

 But with hydrate of soda the mean of three experiments gave the 

 number 17' 2 as the equivalent of silicic acid, a result approaching 

 that previously obtained with carbonate of lithia. 



