Imagini ale paginilor
PDF
ePub
[graphic][merged small][merged small]
[blocks in formation]

SPECTROSCOPICALLY krypton bears a close analogy to argon. Like argon it emits two different line spectra, one with Leyden jar and spark gap in the secondary circuit of an induction coil, the other without Leyden jar and spark gap. As in the case of argon this latter spectrum consists on the whole of less refrangible lines, and I believe, if it were possible to fill a vacuum tube with pure krypton, the color of the tube would change from yellow to blue, when the Leyden jar and the spark gap are interposed. But as far as I know there are no means as yet of getting rid of the admixture of argon. I prepared the gas after the prescriptions of W. Ramsay and Morris Travers. Professor E. Warburg kindly let me have about 3/4 liter of liquid air condensed in the Linde machine of the Physical Institute of Berlin. It is well known that liquid air may be kept for a considerable time in an open silvered vacuum vessel,' by which the influx of heat and consequently the evaporation of the liquid are greatly reduced. In an open vessel of this kind I brought the liquid

[ocr errors]

I W. RAMSAY and MORRIS TRAVERS, Proc. Roy. Soc., 63, 405, 1898.

2J. DEWAR, Proc. Chemical Society, Jan. 14, 1896.

air in my hand from Berlin to Hannover. It lasted four days before all was evaporated. The last fifteen liters were caught in inverted flasks over water. They contained about eighty-five per cent. of oxygen. A few liters of this supply were set apart and from the rest the oxygen was removed partly by metallic copper and partly by pyrogallic acid. I have to thank Professor Eschweiler for helping me in some of these operations. After removing the oxygen, the nitrogen had to be got rid of. Before proceeding to do this I wanted to see what the spark spectrum looked like. I took the very last that had evaporated, removed the oxygen and through the remaining sixty cubic centimeters of gas I passed a spark. The gas exploded violently. I repeated the experiment with the next but last that had evaporated and the gas again exploded, shattering the flask as in the first case. I repeated it a third time with a stronger flask. This time the flask withstood the explosion, but the large beaker glass containing a weak solution of potash into which the neck of the inverted flask dipped, was broken by the pressure propagated through the solution, although there was a considerable amount of free surface all around the neck of the flask. I then tried to explode the gas again over mercury at a reduced pressure, in order to measure the amount of contraction after explosion. But I did not succeed in making the gas explode again. The explosive ingredient, therefore, must have a smaller tension then either krypton or argon, as its percentage is increased relatively to the percentage of krypton and argon by the process of evaporation. But I am at a loss to explain of what it consists. There is an observation of Theodore de Saussure (Ann. de Chim. et Phys., 44, 52 and 53), who found that 2000 parts of air, from which all carbonic acid has been removed, when exploded with pure hydrogen contain one part of carbonic acid. There is another observation by Boussingault (Ann. de Chim. et Phys., 57, 148) that the air contains hydrogen, and lately Armand Gautier (C. R., 127, 693, 1898) has confirmed these observations as regards the air near human habitations, while he found that pure sea air contains free hydrogen to the extent of about 15 cubic

cm in 100 liters. So far as I can see, however, these facts alone do not explain the explosion.

The remaining gas was now mixed with the supply rich in oxygen that had been set apart and sparked for several days over a weak solution of potash. The last sparking was done with a surplus of oxygen until no further appreciable contraction took place and no traces of nitrogen lines were to be seen in the spectrum of the spark. After removing the oxygen by means of pyrogallic acid about 30 cubic centimeters of the gas remained. At atmospheric pressure the condensed spark between platinum electrodes showed besides some platinum lines a great number of argon lines. In the less refrangible part the principal lines of the red spectrum of argon were to be seen besides the principal lines of the blue spectrum of argon and traces of the green and yellow krypton lines. In the more refrangible part photographs of the spectrum show the "white" spectrum of argon' and the stronger lines of the second of the two krypton spectra described below. The argon lines are mostly widened and rather diffuse. In the vacuum tube without Leyden jar and with not too low pressure the krypton lines come out bright. At the same time the carbon bands are very conspicuous. I believe their origin is the same gas that Ramsay and Morris Travers have called metargon. A. Schuster has already called attention to the fact, that the spectrum of metargon as described by Ramsay and Morris Travers seems to be identical with the spectrum of carbon. It is indeed remarkable, as Ramsay and Morris Travers have pointed out, that if metargon is a compound of carbon, it should not be absorbed by sparking it with oxygen over a solution of potash. I have convinced myself, however, that the bands in my vacuum tube coincided accurately with the carbon bands, although the gas was also sparked with oxygen over a solution of potash. In the following list I give in the first column the wave-lengths of a number of edges determined from the neighboring argon lines 'See J. M. EDER and E. VALENTA, Denkschriften der Wiener Akad., 1896.

[blocks in formation]

and in the second column the determinations of H. Kayser and myself of the edges of the carbon bands in the spectrum of the electric arc.

[blocks in formation]

The differences may well be due to errors of observation, as it is more difficult to determine the wave-length of an edge than the center of a symmetrical line. I think there can be no doubt that these bands are the same as those in the electric arc. As regards the intensities of the lines composing the bands I have, however, observed a considerable difference between the bands in the vacuum tube and in the electric arc. In the green band 5165 of the electric arc there are a series of weak triplets between strong close doublets (see the photograph given in Kayser and Runge's article, Abhandlungen der Berliner Akademie, 1889). In the vacuum tube the triplets are much stronger and the doublets much weaker than in the electric arc. The dispersion of the short focus concave grating with which I have been working is not great enough to study these differences satisfactorily and I have therefore not followed out the subject at present. I do not think it possible that the carbon bands are due to impurities introduced after sparking the gas. For impurities caused in the mercury pump never, as far as I know, produce the carbon bands, but invariably produce the so-called bands of carbon monoxide, which were not to be seen in my vacuum tube. I think it most likely that there is a combination of argon and carbon that is able to resist the sparking with oxygen. I noticed that the carbon bands are to be seen on photographs

« ÎnapoiContinuă »