11 
gives the number of c. c., to which 10 c. c. of potassium cyanide solu- 
tion are equivalent. 
In the titration of a dilute formaldehyde solution a definite volume 
or weight (known not to be in excess) is added to 10 c. c., or more if 
necessary, of the standard potassium cyanide solution. The latter, 
after stirring, is added to an acidified excess of decinormal silver nitrate 
solution, the whole made up to 100 a. c. or 200 c. c., and 50 or 100 c. c. 
of the clear filtrate titrated with sulphocyanate as above. From the 
amount of sulphocyanate required the volume of silver nitrate precipi- 
tated by potassium cyanide is calculated as above, and the difference 
between this and the volume of silver nitrate equivalent to the amount 
of cyanide solution originally taken is the number of c. c. of silver solu- 
tion that represents the formaldehyde present; this number of c. c. 
multiplied by 0.003 (more exactly 0.002979) gives the weight of abso- 
lute formaldehyde, HCHO. 
The silver cyanide precipitate must be removed before titrating an 
excess of silver by sulphocyanate, because it interferes with the sharp- 
ness of the end point and causes error. 
In order now to determine the quantity of formaldehyde in the 
atmosphere of the room during any particular experiment a given vol- 
ume of air, usually 10 liters, was drawn through a series of tubes con- 
taining a known amount of the standard solution of potassium cyanide 
or water by means of the apparatus shown in the cut facing this page. 
The quantity of formaldehyde absorbed under these conditions was 
determined in the manner already described. It was soon discovered 
that the air bubbling through the solution carried away a little hydro- 
cyanic acid. To avoid loss of the latter the last tube containing the 
cyanide solution was followed by one containing silver nitrate solution, 
which effectually precipitated all the hydrocyanic acid carried over 
into it. The silver nitrate tube was followed by one containing dis- 
rocyanic acid vhen 
tested by the ferric ferrocyanide reaction, and no formaldehyde when 
tested by Schiff’s fuchsin-bisulphite reagent, ° or ammonia silver 
nitrate solution. 
In preliminary experiments the contents of the silver nitrate tubes, 
after having passed formaldehyde-laden air through the series, were 
tested for formaldehyde to determine if any had escaped absorption in 
the cyanide tubes. To some of the liquid, about 2 c. c. of Schiff’s 
reagent were added; a precipitate of silver chloride was produced, but 
no pink or purple color, indicating the absence of formaldehyde. For 
comparison the reagent was also added to a few c. c. of silver nitrate 
a Schiff’s reagent: This is prepared by adding 20 c. c. of a solution of sodium bisulphite 
(sp. gr. 1.27) to 1,000 c. c. of aqueous fuchsin solution (1:1,000) and, after one hour, adding 
10 c. e. of pure concentrated hydrochloric acid. It should be preserved in a well-stoppered 
bottle. 
tilled water. This tube showed no trace of hyd 
