Prof. Magnus on the Origin of Tar in Olefiant Gas. 425 



the tube is longer exposed to the heat and assumes a higher 

 temperature than that formed at another portion of the tube, for 

 example, near the end at which the gas leaves the tube. 



Not always, therefore, was the tar found to possess the same 

 composition; sometimes, as already remarked, it was darker, 

 sometimes lighter ; sometimes it was so volatile that it could not 

 be weighed in an uncovered vessel. After having been some 

 time exposed to the air, a portion was volatilized, and it was pos- 

 sible to weigh the remainder in a small porcelain vessel without 

 a cover. 



When volatile substances are decomposed, it is not necessary 

 that the products of decomposition should be also volatile ; and 

 in the decomposition of olefiant gas, non-volatile products are 

 also formed, for, as has been already remarked, the tar cannot 

 be wholly distilled. 



It would certainly have been desirable to ascertain the com- 

 position of each separate product of decomposition. But as we 

 are not yet in a position to separate these products, I must be 

 content with an analysis of the entire tar. It is, however, evi- 

 dent that it was not possible to obtain it always of the same 

 character for analysis, and hence arise the divergences in the 

 results. 



The analyses were so conducted that the substance was weighed 

 in a small porcelain vessel, and with the latter was introduced 

 into a tube filled with the oxide of copper. During the com- 

 bustion oxygen was conducted over the oxide. Tar possesses 

 the disagreeable property of readily creeping, when warm, over 

 the edge of the vessel which contains it, and then spreading 

 itself over the interior of the hinder portion of the tube. To 

 prevent this a further quantity of the oxide of copper was placed 

 behind the porcelain vessel, which during the combustion was 

 preserved at a red heat, so that the tar was immediately decom- 

 posed as soon as it came into contact with this oxide. 



The analyses gave the following composition for the tar : — 



Carbon. Hydrogen. Loss or Oxygen. Sum. 



I. 94-106 6-066 ... 100-172 



II. 92-461 6-652 0-887 100-000 



III. 92-403 6-808 ... 100211 



This composition agrees pretty well with that of naphthaline, 

 which consists of — 



Carbon . . . 93'75 per cent. =5 C 

 Hydrogen . . 6-25 ... 2H 



10000 5C + 2H. 



The odour of tar is also quite similar to that of naphthaline ; 

 Phil. Mag. S. 4. Vol. 6. No. 41. Dec. 1853. 2 F 



