Behaviour of Chemical Compounds. 281) 



portion of it from the remaining quantity of free base and acid, 

 nearly the same percentage would again combine to form a 

 salt as before. This portion would also be precipitated ; and 

 the process would go on in the same manner, until all the base 

 and acid had combined and formed an insoluble salt. 



(3) On employing increasing quantities of acid to 1 equiva- 

 lent of ferric oxide, the amount of free acid in the solution 

 first diminishes until the total quantity of the (free and com- 

 bined) acid in the solution amounts to somewhat more than 

 one equivalent. With a further increasing total quantity of 

 acid in the solution the amount of the free acid increases. 



(4) If we divide the quantity of ferric sulphate formed when 

 one equivalent of ferric oxide is employed, by the number of 

 equivalents of acid used, we obtain the remarkable result, that 

 the amount of ferric oxide combined ivifh the same quantity 

 (1 eq.) of the acid is the greatest ichen the acid in the solution 

 amounts to somewhat more than 1 equivalent to 1 equivalent of 

 ferric oxide. When the acid is in larger quantity the amount 

 of ferric oxide combined with 1 equivalent of acid is, as was 

 to be expected, smaller. (With indefinitely much acid it 

 would be indefinitely little.) The same relation, however, 

 enters with smaller quantities of acid, where, on the contrary, 

 one w^ould a priori suppose that, an excess of ferric oxide 

 being present, the acid would in greater measure be saturated 

 with it. 



Similar results are obtained with solutions of ferric nitrate 

 containing different proportions of nitric acid. Thus, accord- 

 ing to the preceding experiments: — 



Iron in 100 ^ i<r r\ 



■1 , . le : IN ., Oj. X. y, 



cub. centinis. ^ ^ 



0-479 100 : 73 31-9 0*93 



0-266 103 18-8 0*79 



0-185 132 8-2 0-69 



0-139 164 6-7 0-57 



0-111 197 5-6 0-48 



In a solution containing about equal numbers of equivalents 

 of ferric oxide andniti'ic acid, therefore, about 19 per cent, of 

 the total amount of iron in the solution is present in the un- 

 combined state, together with free acid ; and the percentage 

 falls, on the addition of 1 eq. more of nitric acid, to from 5 to 

 6 per cent. With these solutions a diminution of the value y 

 with less content of acid does not appear to enter ; to decide 

 this point, however, further observations shall be instituted. 



It is evident that, if the alteration of the magnetism with 

 the temperature be determined for a solution of colloid ferric 



Fhil. Mag. S. 5. Yol. 4. No. 25. Oct. 1877. U 



