34 
SELECTED ARTICLES. 
at the temperature of 100° c, in the apparatus of M. Liebig, 
until it ceased to lose weight, it was decomposed by heat, and 
the potassa estimated in the state of sulphate, and the iron as 
sesquioxide. For this purpose, a known weight of the salt was 
carefully calcined, the product was mixed with water, and 
super-saturated with sulphuric acid; a small excess of car- 
bonate of ammonia was then added, and the liquor filtered and 
evaporated to obtain the weight of the sulphate of potassa. 
The proportion of the sesquioxide of iron was found by burn- 
ing the filter on which it had been collected, washing the 
ashes with nitric acid, and heating anew to redness. The 
proportion of the tartaric acid was given by the difference be- 
tween the weight of the matter employed, and the weights of 
the potassa and oxide of iron obtained. 
I. 3^,355 of the salt dried at 100° c, gave— 
Sesquioxide of iron, 1.044 or 31.11 p. 100 
Sulphate of potassa 1 .178, or potassa, 0.637 or 18.98 p. 100 
II. 3e r ,017 of the same furnished— 
Sesquioxide of iron, 0.920 or 30.49 p. 100 
Sulphate of potassa 1.018, or potassa, 0.550 or 18.23 p. 100 
These analytic results correspond to the following compo- 
sition: 
1 atom sesquioxide of iron, Fe 2 3 == 978.41 or 30.29 
1 atom potassa, KO, = 589.92 or 18.26 
I atom tartaric acid, C 8 H 6 10 , =1661,42 or 51.45 
3229.75 100.00 
A composition similar to that of tartar emetic deprived of 
its water of crystallization; the oxide of antimony being re- 
placed by an oxide of iron. 
Our experiments to determine the quantity of water which 
the tartrate of iron and potassa would lose at a temperature ex- 
ceeding 100° c, did not give the desired result, but made 
known to us a remarkable property of this salt. When heat- 
ed at a temperature not exceeding 130° c, the oxide of iron 
