EESEARCHES UPON" THE CHEMICAL PROPERTIES OF GASES. 179 



Coal gas which had been stored under pressure in an iron cylinder, and which 

 had lost all its original free hydrogen, did not reduce palladium chloride at 100°. 



Pure methane (prepared by Gladstone and Tribe's method) exerted no action 

 until heated neai'ly to 250°. 



Palladium chloride heated in an inert gas, such as carbon dioxide, was found to 

 yield chlorine at about 250°. In presence of oxygen the case is very different. Pal- 

 ladium chloride heated in dry air loses chlorine readily at 160°, being apparently 

 converted into an oxychloride. After long-continued heating to 100° in air, chlorine 

 in minute traces is set free and recognizable by silver nitrate containing some ferrous 

 sulphate. 



If air containing any hydrocarbon (paraffin, olefene or acetylene) be led over 

 gently heated palladium chloride a decomposition occurs at once. The palladium 

 salt is reduced and hj'^drogen chloride is promptly set free. Alcohol, ether and ben- 

 zol vapor cause similar results. Repeated trials have shown that less than 0.1 per 

 cent, of hydrogen in air may be recognized by the reaction above described if the 

 temperature of the palladium chloride is not increased above 50° C. If the tempera- 

 ture rises to 100°, chlorine will be evolved from palladium chloride b}'^ the action of 

 air alone, as may be easily shown by causing the air to bubble through silver 

 nitrate solution containing a little ferrous sulphate (free chlorine is not easily 

 detected by silver nitrate alone, and ma}'" bubble through it unabsorbed and unrecog- 

 nized). 



The reduction of anhydrous ruthenium chloride by hydrogen is curiously influ- 

 enced by the presence of oxygen. Ruthenium chloride was reduced by pure hydro- 

 gen at 190°. In another experiment, using a mixture of hydrogen 4 volumes and 

 air 6 volumes, no hydrochloric acid was produced, even on heating to 320°.. The 

 following is the most convenient method of applying the test : 



The gas, previously dried (1) by calcium chloride and (2) by phosphoric anhy- 

 dride, is conducted through a narrow tube to the bottom of a dry test-tube contain- 

 ing about 0.2 gm. of palladium chloride. The test-tube has a rubber cork with two 

 holes. Through a second tube the gas escapes and passes into a solution of nitrate 

 of silver. The test-tube may remain cold, but, in the absence of oxygen or air, it 

 is better to immerse in water at 40° or 50°, provided no hydrocarbons likely to reduce 

 palladium chloride are suspected. The palladium chloride may be placed in a glass 

 tube of one-eighth-inch bore, with asbestos plugs to prevent its becoming displaced 

 by the gas stream, and then the tube connected with the vessel containing silver 

 nitrate solution. If traces only of hydrogen are to be tested for, oxygen must be 

 completely removed by prolonged contact with pyrogallol and soda, or better with 



