Sadtler.] -08 [Feb. 18, 



sage of the natural gas for 15 or 20 minutes, therefore, should saturate 

 the alcohol with such of these hydrocarbons as might be present. The bro- 

 mine absorption test was made in order to ascertain whether any ethylene 

 or butylene dibromide would be formed by the passage of gas. 



The gases tested in this way were the gas from the Burns Well and gas 

 from the Leechburg Well. The bottles containing the alcohol, having been 

 tightly stoppered and sealed at the time, were afterwards examined in the 

 Laboratory. The Burns Well gas was first examined. The alcohol through 

 which the gas had been passed was transferred to a small flask, and 

 enough additional alcohol was added to completely fill the flask. A cork 

 perforated and carrying a delivery tube was then fitted to it so that the ex- 

 cess of alcohol filled this tube also. 



The flask was now heated, and the gas given oft was collected in a eu- 

 diometer tube over warm water. I obtained thus some 150 cubic centi- 

 meters of gas. The operation was pushed until only alcohol vapor was 

 given oft", which was absorbed by the water. To the gas in the eudiometer 

 I then added rapidly an equal volume of chlorine, and allowed it to stand 

 exposed to diftused sunlight. Oily drops formed at ouce, while the volume 

 of the gas contracted until it occupied only about one-half of the first 

 Tolume. 



This showed the probable formation of ethyl chloride (C2 H, CI) which 

 is gaseous at temperatures over 12°C, and of either propyl chloride 

 (C.,H7 CI) which is liquid at temperatures under 52° C, or of butyl chloride 

 (C4 H,| CI) which is liquid at temperatures under 77.6'^ C. On inverting 

 and applying a flame to the gas in the eudiometer, it was found to be , in- 

 flammable, burning with the greenish flame characteristic of ethyl chloride. 

 The water ujion which the oily drops had collected, was put in a small 

 flask, and an attempt was made to separate them by distillation. This was 

 however, unsatisfactory, as there could only have been small quantities of 

 either propyl chloride or butyl chloride. The water in the eudiometer 

 tube reacted acid from the H CI foi'uied. 



The result of these tests may therefore be summed up as showing the 

 certain presence in the gas of ethyl hydride (C.^ Ilg) and the probable pres- 

 ence of either C3 Hg or C4 H,o. 



The (pialitative examination of the Leechburg gas was made in the same 

 way. I obtained here only about 100 cubic centimeters of gas on heating 

 the alcohol. On mixing with chlorine a contraction of about one-third 

 took place. The inflammability of the residual gas and the formation of 

 oily drops gave the same indications as with the other gas. 



The other qualitative tests were with bromine. This had been tightly 

 bottled and sealed up. The bromine from the test with the Burns gas was 

 then placed in a porcelain dish and water added. Pure sodic hydrate was 

 added to neutralize the excess of bromine. The sodium bromide formed 

 dissolved at once in the water. There remained no trace of oily drops 

 such as would be caused by the bromides of the olefine series of hydro- 

 carbons. The sodium bromide was afterwards crystallized out perfectly 

 pure. The test was therefore entirely negative in its results. 



