Prof. Magnus on the Origin of Tar in Olefiant Gas. 423 



When the upper portion of a curved tube or receiver filled 

 with pure gas and shut off by mercury was heated in a bath of 

 Rose's metal, the temperature of which was a little higher than 

 that of boiling mercury, the gas did not assume the smell of tar, 

 but retained its original odour. The formation of tar from 

 olefiant gas begins therefore at a temperature at all events higher 

 than 360° C, and it appears not to occur at a temperature under 

 that of a red heat. 



At a dark red heat, however, the tar is formed in such quan- 

 tities that it collects in drops. It is only partially volatile ; for 

 if we attempt to distil it, a quantity of carbon always remains 

 behind. It is not always of the same character, for it is some- 

 times lighter and sometimes darker. 



When the tar is formed, the volume of the gas from which it 

 comes is diminished. This diminution is different according to 

 the decomposition to which the tar itself is subjected. In dif- 

 ferent determinations the volume of the gas which remained at 

 the temperature and pressure of the original gas amounted to — 



84*4 per cent. 



93-6 



921 



89-4 



88-9 



Mean . . 897 of the latter. 



The volumes of the remaining gas differ from each other, and 

 so do the compositions of the different remaining portions. It 

 always consisted for the most part of marsh-gas, but contained also 

 a quantity of undecomposed olefiant gas, the vapours of tar and 

 hydrogen. The quantities, however, in which these substances 

 were found were different. The remaining gas burnt with a 

 tolerably luminous flame, about the same as that of ordinary 

 coal-gas. 



When the olefiant gas was exposed to a full white heat, its 

 volume was not diminished but augmented. A porcelain tube 

 was closed at one end, and the other end was so connected with a 

 receiver shut off with mercury that the gas might proceed with- 

 out hindrance from the tube to the receiver, and from the latter 

 could return to the tube. After the tube and the receiver had 

 been filled with olefiant gas, the former was introduced into a 

 blast-furnace and heated to whiteness for a distance of eighteen 

 inches. In order that the gas might have full liberty to expand, 

 the receiver which dipped pretty deep into the mercury was 

 slowly lifted out of it, and afterwards during the cooling of the 

 porcelain tube permitted to sink again. When at the conclusion 

 of the experiment the gas was brought under the original pres- 



