90 GIBBS, WILLIAMS, AND PRATT. 
This experiment showed that the methyl salicylate in contact 
with hydrogen peroxide in tube number 26 colored in a few 
minutes and that the presence of formaldehyde in tube number 
29 slowed the reaction since the hydrogen peroxide was partially 
used up by reacting with formaldehyde to produce formic acid. 
The methyl salicylate, in contact with formaldehyde in tube num- 
ber 22, alone remained colorless throughout the investigation. 
Experiments to demonstrate the effect of formaldehyde, methyl 
alcohol, and oxygen upon the coloration in sunlight are deseribed 
in Table VI. 
TABLE VI.—Miaxtures of methyl salicylate, formaldehyde, and alcohol in 
presence and absence of oxygen exposed to the sunlight in sealed tubes 
from March 9, 1911, to August 31, 1911. 
[Quantities are expressed in cubic centimeters.] 
Formal!- |: 
Methyl| dehyde 
Top |salicyl-| 40 per Methyl | Water. | Gas. | Remarks. 
°- | ate. cent cone | 
solution.| | | 
Seer al Noel Sapte ae. ees a = 
4 10 5 0 |. RS RAGE Sic ce = ee eee Very little color. 
5 10 1} 0) 5) || pAir18iceli aera aa oaee as | A little color. 
16 | 10 0! 10 \- ______| Very little air, 0.5 ee __ a Practically no color. 
Co 0} 5 5) Ars 0lcce==eenen eee eeees Light yellow. 
15 2 (a) (0), | PE Se REN ERS ee eee | No color. } 
17 | 2 0 (b) | peme shes ET ese Feu ee eee aa Practically no color. 
"5 ec. 10 per cent solution. »5 ce. of 20 per cent solution. 
From a study of these tubes, the conclusions are drawn that 
oxygen is necessary for the coloration; that formaldehyde delays 
the formation of color and reduces the intensity of the yellow; 
that methyl alcohol, when present in sufficient concentration and 
volume to dissolve the methyl salicylate, materially reduces the 
rate of the color formation, and, when present in smaller quan- 
tities so that the methyl] salicylate is not completely dissolved, the 
amount and rate of color formation are still reduced, but not 
to so great an extent; and that in an atmosphere of hydrogen 
methyl salicylate remains colorless for weeks, the light yellow 
tint finally produced probably being due to impurities dissolved 
from the glass. The colored compound in all cases is found 
dissolved in the methyl! salicylate layer. If formaldehyde forms 
any compound, under the conditions enumerated, it has no color. 
