268 ON THE COMPOSITION OF ALCOHOL. 



zoic acid : but as the weight of all three tof>ether amounted 

 to about -^^-^ only of the fluid that held them in solution, it 

 may be considered as pure water, without any risk of errour, 

 in an analysis like the present. 

 Oxkarburetted 4. The oxicarburetted hidrogen gas, the barometer being 

 hidro^en gas. ^^ ^^ inches, and the thermometer at 17° R. [70j° F.], 

 occupied the space of 7199 cubic inches ; and weighed, the 

 day after it w,is collected, taking a mean between the weight 

 of the gas that came over at the beginning, middle, and 

 end of the process, 17S6-6l grs*. Though the heat of the 

 tube 'lid not perceptibly vary, the gas obtaiticd at the be- 

 ginning of the experiment was lighter, and contained less 

 C5ri*Dn, than at the end. This was owing to the charcoal 

 deposited by the alcohol accumulating gradually in the 

 tube, and reacting on the fluid that was decomposed in 

 proportion to this accumulation. However slowly I con- 

 ducted the distillation, I could not prevent the gas from 

 carrying over with it pretty copious white fumes, the weight 

 of which I could not directly calculate, jmd the loss of 

 which occasioned a deficiency in the result^ of the analysis. 

 These fumes smelled of benzoin; and appeared to me to af- 

 , ford on condensation similar products with those collected 

 in the receiver, namely, a great deal of water, and a very 

 small quantity of oil. The latter could only be in very 

 small proportion ; for, on detonating the gas immediately 

 after its developement, and while these fumes were sus- 

 pended in it, I did not obtain more carbonic acid gas front 

 the combustion, than when it was detonated after the fumes 

 bad been condensed in the water under the jars. The un- 



Composition of * At 28 inches of the barometer therefore, and 10° [54|*'] of the 

 the gas affected thermometer, 1000 cubic inches of this gas weigh 266 grs. This result 

 by the manner ^,j^^^^ ^ j-^,,^ ^^^^ ^^^^^ ^f ^^ Cruikshank, who makes it 237 in the 

 ot conductrnj; . . , ,. . , - • , . 



the process. ^^^^ circumstances. I have performed this experiment three times, 



changing the diameter of the tube a Httle, and likewise its inclinatioa 

 in the furnace, and each time 1 found a perceptible difference in the 

 \veight of the gas and its composition. But the sum of all the pro- 

 ducts, in each of the experiment?, afforded siniilar results for the 

 composition of alcohol. Thus it appears, Ihat we should be liable to, 

 considerable errour, if we did not compare together all the products of 

 each expeiimeat. 



certainty 



