20 
sent in 200cc. was more than 0'0025cc. of the strong acid, 
it renders the colour deeper than it otherwise would he. 
One cubic centimetre of the standard nitric acid is added 
to each of the cylinders, because (1) It renders the reaction 
much more delicate. (2) Because the colours produced in 
the presence of this amount of free acid are almost always 
of the same tint, being of a pure blue, whilst when no free 
acid is present the colour varies, even Avhen apparently of 
the same depth, from a blue to a bluish green, which ren- 
ders them less easy to compare. (3) Because it destroys 
the effect which the presence of a small quantity of any free 
acid, previously existing in the liquid, might have in alter- 
ing the shade of colour produced, for from a series of experi- 
ments made with reference to this point also, it was found 
that when the amount of free acid present, in addition to 
the Icc. of standard nitric acid added is only small, f.e less 
than 0-05 cc. of the strong acid in 200cc. of water, it has no 
effect on the depth of colour produced. When any free acid 
exists in the water to be examined it must, before being 
oxidised, be made as nearly neutral a,s possible with am^ 
monia, and the iron then determined. 
The following are some of the results obtained on deter- 
termining the iron in solutions of known strength : — 
Iron Found. Iron Calculated. 
Milligrams. Milligrams. 
17-99 19-48 
4-20 4-14 
1-80 1-86 
•61 with salts (C) present *57 
•51 -52 
•40 -41 
•40 ... with KNOg (D) present ... '40 
•33 "30 
•28 -31 
•23 with salts (B) present -22 
•20 -21 
•14 -16 
•12 -10 
•070 -078 
•025 -031 
In order also to test the effect which the presence of dif- 
ferent salts has on this method, four series of experiments 
were made by adding known weights of the following salts 
to 1 litre of the ammonio-ferrous sulphate solution ; 
