OCTOBEE 14, 1898.] 



SCIENCE. 



489 



about the presence of a new gas in the 

 pulverized glass. I subsequently exposed 

 this lot of glass several days to the atmos- 

 phere, spread out in a thin layer. "When 

 tested again after this exposure it yielded 

 the new gas, both before and after heating, 

 as freely as at first. This rejuvenation of 

 pulverized glass by exposure to the air was 

 fully confirmed with another lot made from 

 common window glass, and in connection 

 with the results obtained by the diffusion 

 of air, hereafter described, leaves no room 

 for doubt that the new gas is a constituent 

 of the atmosphere. 



Pulverized glass appears to begin evolu- 

 tion of the new gas as soon as the atmos- 

 pheric pressure is reduced. In one case, 

 while making a slow preliminary exhaus- 

 tion, I tested the air which remained when 

 the pressure was still 132 mm., and, to my 

 great surprise, found it contained not only 

 a trace, but a considerable quantity of new 

 gas. The amount increased rapidly as the 

 pressure was further reduced. This leads 

 to the belief that finely pulverized glass 

 gives up the greater part of its new gas on 

 simple reduction of pressure to a low point, 

 and without heating; somewhat as pal- 

 ladium gives up occluded hydrogen. 



Several substances other than glass were 

 examined, and nearly all found to contain 

 the new gas. A specimen of old charcoal 

 made from pine-wood sawdust by long ex- 

 posure to a bright red heat yielded as was 

 expected, comparatively large quantities of 

 several gases when heated in vacuum. At 

 some stages of the evolution these were rich 

 in the new gas. 



It would seem easy in such cases to re- 

 move the diluent gases by oxidation and 

 absorption ; but it is not. After making 

 the necessary additions to my apparatus, I 

 made many attempts of this kind. The 

 reagents used were in large excess on ac- 

 count of the very small quantities of gas 

 treated; so that some of the observed re- 



sults may have been due to impurities. 

 Cupric oxide and lead chromate evolved 

 gases of their own almost indefinitely on 

 high heating in vacuum, and then freely 

 absorbed carbon flioxide and moisture at a 

 lower temperature. Soda-lime dried in 

 vacuum in presence of phosphorus pentox- 

 ide was almost indifferent to carbon dioxide. 

 All the reagents named, especially phos- 

 phorus pentoxide, absorbed the new gas; 

 and all but the latter give it up again on 

 heating. 



A specimen of very fine white silicious 

 sand, when heated in vacuum, gave a large 

 quantity of gas consisting priacipally of 

 hydrogen and hydrocarbons, with a con- 

 siderable amount of the new gas. Some of 

 this mixture was exposed successively to 

 the action of red-hot cupric oxide, soda- 

 lime and phosphorus pentoxide. By this 

 treatment the gases were reduced to less 

 than three per cent, of their former volume. 

 The residue was not very rich in the new 

 gas, because of the absorption of the latter 

 by the soda-lime and phosphorus pentoxide, 

 as I afterwards learned ; but I have shown 

 the curve o^ of its heat conductivity, be- 

 cause the pressure was carried to a lower 

 point than in any other case. At the last 

 station in the curve, representing a pressure 

 of .12 of a millionth, the conductivity was 

 equal to that of 5.1 millionths, or forty-two 

 times as much hydrogen. From this it 

 seems reasonably certain that the curve ef, 

 if carried to as low a pressure, would have 

 shown a conductivity at least a hundred 

 times greater than hydrogen. And yet the 

 new gas in that experiment must have been 

 very far from pure on account of the con- 

 tinuous evolution of ordinary gases, as 

 shown by the rise of pressure and loss of 

 conductivity during the next few hours. 



It will be observed that when the new 

 gas was present the form of the conduc- 

 tivity curves was very different from those 

 of the known gases, the effect of the new 



