234 
SELECTED ARTICLES. 
the 59.30 parts of oil must correspond to some integer, or at 
least simple number of atoms. And, reciprocally, if we find 
such to be the case, we shall be fortified in the conclusion 
which we have drawn. 
With a view to this method of verification, let the numbers 
which represent the iodide of potassium and iodine, and that 
which is supposed to represent the oil, be divided by their 
respective atomic weights, and let the quotients be reduced to 
others in the same ratio, and so that the iodide of potassium 
may be represented by unity. When these arithmetical ope- 
rations are performed, we obtain the numbers in the second 
and third columns of the following table, the former being the 
quotients themselves, and the latter other numbers bearing to 
each other the same proportion. 
(1.) (2.) (3.) 
Iodide of potassium, . 12.55 0.075 1.000 
Iodine, . . . 28.14 0.223 2.973 
Oil of cinnamon, . . 59.30 0.442 5.893 
The numbers, it will be seen, in the last column approxi- 
mate so closely to the integers 1, 3, and 6, as to leave little 
doubt that the true empirical formula is 
IK + I 3 + Cin 6 ,* 
a conclusion which is strikingly confirmed by the following 
statement of the composition of our substance in 100 parts 
calculated upon this hypothesis: 
Iodide of potassium, . . 12.26 
Iodine, .... 28.0S 
Oil of cinnamon, . . . 59.66 
100.00 
To apply, however, to this conclusion, the most decisive 
test, it remained to burn the substance with oxide of copper, 
* Cin is assumed as the symbol for the oil of cinnamon. 
