228 On Fuchs’s method for the determination of Iron. 
the reaction being Fe2 Cle + 2Cu = 2FeCl + Cu2Cl. The 
quantity of iron is then calculated from the loss of weight of 
copper, for according to the reaction, as the atomic weight of 
copper is to that of iron, so is the loss of weight of copper to the 
quantity of iron sought. 
As the idea contained in the method is capable of many ana- 
lytical applications, it became a matter of interest to determine 
first the accuracy of the method itself. The iron used in the 
analyses was fine piano forte wire, and as a preliminary experi 
ment the copper, which was of the purest Lake Superior variety, 
was boiled for one half hour in concentrated chlorhydric acid, 
with a loss of 0-69 per cent., and in acid of about one quarter the 
strength, 0-11 per cent.; showing that the very dilute free acid, 
in the solution of the perchlorid, can have no sensible effect on 
the result. 
Tron taken. Copper dissolved, Tron found. 
I 20074 2°2010 1:9441 = 96-84 per cent. 
It 20591 2:3065 20372 =" 96-94" 
Ill. 1:9262 2°1811 19265 = 100-015 
IV. 1-6682 1:8875 1-6671 = 99:93 * 
7 22574 2°5045 22121 = S799 
VI. 2-0084 22855 20187 = 100-52 “ 
VIf. 1-9807 22015 1:9445= 98:17 “ 
VILf. 20671 2:3074 2:0380 = 9859 “ 
2:0618 2:3380 2:0651= 10016 “ 
X. 1:0637 11907 1:0517 = 9887 “ 
The analyses show that although the iron is reduced to proto- 
chlorid, the change in color of the solution does not indicate with 
sufficient sharpness the exact time when such reduction is cour 
plete, thus rendering the method inaccurate and unreliable. The 
method can be made to give accurate results, as soon as some UD 
objectionable process is given whereby the reduction is rendered 
manifest independent of mere change in color. , 
ince the above investigation was completed, a paper on this 
same subject has been published (Erdmann, Ixi, 127) by Dr. Jur 
lins Lowe, in which he states that the method, in point of acc’ 
racy, leaves nothing to be desired, and gives as proof the follow- 
ing examples: 
I. Il. IIt. Iv. Vv. vi. ped 
Taken, 0-185 0-161 0-126 0-212 0-084 0-282 oy 
Found, 0-179 0-158 0-122 0212 0081 0278 020, 
Difference, 0:006 0-003 0-004 0-003 0-004 000 
Although in these analyses the absolute difference 18 very 
small, unfortunately the error in parts per cent. is large; 
pressed in this manner we have: 

