298 Mr. Porrett on Ferro-chyazate of Potash, [Oct. 



certainly detected by weighing the metal reduced by its means 

 than by weighing the water which it produces : the only circum- 

 stance which it is necessary to guard against, when this proces* 

 is resorted to, is the oxidation of any portion of the azote. 



To return to the analysis. — It now only remained that I should 

 ascertain the quantity of water of crystallization in the salt ; but 

 renewed experiments to determine this added nothing to what 

 I had before stated, and which had been confirmed by Dr. 

 Thomson ; namely, that only 13 per cent, could be separated 

 from it at a temperature below that which would decompose its 

 a«id. In taking 13 then as the quantity of water in 100 of the 

 salt, we cannot be certain that it indicates the whole quantity, 

 but must rest satisfied with it as the nearest approximation to that 

 quantity which experiment is capable, of giving us. 



Collecting now, from the preceding experiments, the propor- 

 tions of all the constituents of 100 gr. of ferro-chyazate of 

 potash, they appear as follows : 



CirainB. 



Potash 41-68 



rlron 12-60 



^ , • , \ Carbon 22-64 



Ferro-chyazic acid J Azote 1MB 



(. Hvdrogen 00-80 



Water 13-00 



104-04 



being a surplus of four grains, arising from the unavoidable 

 inaccuracies in determining experimentally on small portions of 

 the salt the proportions of so many constituents. 



These inaccuracies are easily removed by the application of 

 the atomic theory ; for by taking as our guide the weights of 

 the atoms of each of the elements, we obtain the following 

 numbers : 



1 atom potash = 60-00 40-34 



r{ atom iron = 17-50 .... 11-76 



1 atom ferro-chyazic ) 4 atoms carbon. . . = 30*00 .... 20*17 



acid = 66-25 J 1 atom azote = 17*50 .... 11*76 



' 1 atom hydrogen . . = 1*25 .... 00*84 



2 atoms water = 22*50 .... 15*13 



1 ferro-chyazate of potash , = 148*75 .... 100-00 



which doubtless gives the true proportions of the several ele- 

 ments of this salt. 



Thus it appears that in attempting to reconcile the analysis of 

 ferro-chyazate of potash with the numbers generally received as 

 representing the relative weights of the atoms of its several 

 elements, we are reduced to the necessity of considering one of 



