MR. J. S. TOWNSEND ON MAGNETIZATION OF LIQUIDS. 
545 
Ferric Nitrate. 
An experiment was tried with ferric nitrate and estimated gravimetrically, like the 
chloride and the sulphate, and it also gave the formula 
10 7 & = 2660 W — 7*7. 
thus showing* that the coefficient k depends only on the quantity of iron present, and 
the nature of the solvent. 
Ferrous Chloride. 
A solution of FeCl 3 , which contained '0409 gram of iron per cub. centim., had 
a coefficient of magnetization k given by the formula 
10 7 & = 84*5 — 7*7 at 12° centigrade. 
The solution had been boiled for several hours with iron wire, so as to reduce 
any ferric salt to the ferrous state and then filtered. So that for ferrous chloride we 
have the formula 
10 7 k = 2060 W - 7*7. 
Ferrous Sulphate. 
A more complete examination was made of ferrous sulphate, the values for 10 7 & for 
two solutions, one containing twice the amount of iron per cub. centim. as the other, 
were found to be 
127*3, and 60*5, 
so that when they are corrected for the water in which they were dissolved they 
have the ratio 2:1. 
A solution was examined similarly to the ferrous chloride and gave a formula 
10 7 & =i 2090 W — 7*7, the quantity of iron having been estimated gravimetrically 
by boiling a specimen with nitric acid and proceeding in the ordinary way. 
As this formula did not coincide with that got for the ferrous chloride a more 
complete examination was made of a solution which contained a small amount of 
ferric sulphate. Having found the value of k, 25 cub. centims. of the solution were 
extracted and the quantity of ferrous salt found by a standard solution of per¬ 
manganate of potash, another specimen having been treated with nitric acid gave 
the total quantity of iron present, thus providing a means of finding the amount of 
ferric iron in the solution. 
The numbers thus obtained were :— 
MDCCCXCVI.—A. 
4 A 
