68 GIBBS. 
During the process of sealing, all tubes were partially immersed in 
ice water and only the protruding constriction heated, in order to avoid 
all-possibility of oxidation. These precautions were found to be especially 
necessary where alcohol of high concentrations was sealed in the tubes. 
These three tubes were then wrapped in a cloth, put in a steam 
bath, and maintained at 100° for four hours. They were then 
cooled, opened, and tested for formaldehyde by the previously 
mentioned three methods. Each tube gave positive results by 
each method; that is, 9 positive tests were obtained. 
Two tubes containing 10-cubic-centimeter portions of a mix- 
ture of 25 cubic centimeters methyl alcohol, 25 cubic centimeters 
of pure water, and 5 cubic centimeters of a 3 per cent solution of 
hydrogen peroxide were sealed in tubes and placed in a steam 
bath at 100°. The first tube was opened after nineteen and one- 
half hours and the second after forty-four and one-half hours. 
The acidity had increased from 0.25 cubic centimeter = at the be- 
ginning to 0.80 cubic centimeter in the first tube and 1.20 cubic 
centimeters = in the second. Both tubes gave positive tests for 
formaldehyde by the three methods and the second tube con- 
tained very much the greater amount. 
The following mixtures were next heated to 100° for varying 
lengths of time. 
TABLE V.—Showing mixtures heated to 100°. 
| | 
| Methyl Hydrogen} Heated 
No. ernst Water. peroxide. = | 
| 13 M. 
5 5 0 | 24hours. 
| 2B. 5 5 0 | 67 hours. | 
=) 
| 2 | 67 hours. | 
Tube number 13, containing the methyl alcohol and hydrogen 
peroxide gave very strong and conclusive tests for formaldehyde 
by the Leach, Hehner, and Rimini reactions. The other two 
gave no positive tests although in one case there was a faint 
color reaction which could not be taken as conclusive. Tubes 
numbered 13 M and 2 B were tested for hydrogen peroxide by 
both the vanadium and titanium reactions, with negative results. 
In order to prove that impurities in the alcohol were not 
responsible for the action with hydrogen peroxide, the following 
samples were employed. 
Two tubes containing 10 cubic centimeters methyl alcohol puri- 
fied by the fourth method, 5 cubic centimeters conductivity water, 
and 1 cubic centimeter hydrcgen peroxide were heated to 100° 
