(A) Calcium Bulpbate, magnesium sulphate^ ammonium 
chloride^ sodium chloride, and potassium carbonate, in all 
lb grin. 
(B) Ditto, in all 0’9 grm. 
(C) Magnesium sulphate, ammonium chloride, potassium 
carbonate, sodium chloride, and calcium chloride, in all 
0‘8 grm. 
(D) Potassium nitrate 0'4 grm. 
The solutions thus obtained were oxidised with perman- 
ganate and sulphuric acid diluted to one litre, and the iron 
estimated as previously described. The results obtained 
are given below, the letters attached denoting to which of 
the preceding series they severally belong, and from them it 
will be seen that the presence of these salts has little or no 
effect. 
It was also found that neutral organic matter is not 
detrimental to the method. 
With reference to the delicacy of the method it was found as 
a mean of seven experiments that 0'0055 mgrms. of iron give 
a very distinct colour on the surface, and that 0’015 mgrms. 
give a blue colour on being stirred with 200cc. of water, and 
therefore that 1 part of iron produces a blue coloration in 
13,000,000 parts of water containing ferrocyanide of potas- 
sium and nitric acid. According to Hartig,* however, 
1 part of iron (in the form of sulphate) only produces a 
colour in 600,000 parts of water containing ferrocyanide. 
The difference of these results is due to the effect which the 
presence of the small quantity of free nitric acid, added in 
the new method, has in increasing the delicacy of the 
reaction. 
As to the smallest differences of reading which can be 
detected, it was found that when any quantity of iron solu- 
tion below Icc. had been added, a difference of 0‘05cc. can 
be discriminated ; above 1 and below 2cc. a difference of 
OTcc. ; above 2 and below 4cc. a difference of 0*2cc. ; and 
above 4 and below 5cc. a difference of O'Scc. 
Jour. Pr. Cliem, 22. 51. 
