624 



NATURE 



[October 28, 1897 



periods of national upheaval, when men's minds are deeply 

 stirred— like the rise of Islam, the Protestant Reformation, and 

 the French Revolution —were exceptionally prolific of able men. 

 It does not appear altogether unreasonable to suppose that the 

 stir and movement of modern life may be similarly favourable to 

 the production of " men of light and leading" for the service of 

 the community. The proximate cause of the downfall of states 

 seems always to have been a defective supply of strong and 

 capable men at the head of affairs, and in positions of trust. The 

 doke far nienie is not conducive to the formation of strong 

 characters ; and those who sigh and yearn for social quietism 

 may find comfort in the reflection that the hum and buzz which 

 disturbs them is a sure token of the health and strength of the 

 common hive. 



THE BEHAVIOUR OF ARGON IN X-RAY 

 TUBESy 



] N continuation of some experiments made by Prof. Callendar 

 in the early part of 1896, the authors have studied the 

 behaviour of argon in X-ray tubes of various types. The 

 phenomena presented by a tube filled with carefully dried and 

 purified argon, are in many respects peculiar, as compared with 

 those presented by other gases under similar conditions. 



In the early experiments above mentioned it had been our 

 custom to keep the X-ray tube connected with the pump, which 

 was used as a reservoir of dry air during long exposures. The 

 gas, which was absorbed by the working of the lube at a high 

 vacuum and a long equivalent spark-gap, was restored from time 

 to time, as the vacuum became too high, by letting a little air in 

 from the pump by means of a convenient tap. In this manner it 

 was possible to operate the tube at a very high rate of efficiency 

 for two hours or more at a time. These long exposures were 

 required for some experiments on the velocity of the X-rays, 

 which have been described in a communication to the Canadian 

 Royal Society, May 1896. 



It was noticed on several occasions, after one of these long 

 exposures, that there was considerable blackening and sputtering 

 of the electrodes, and also that the pressure of the air in the tube 

 had increased considerably above the degree of vacuum required 

 for the production of X-rays when the tube was first exhausted. 

 After allowing the tube to rest for a few hours, although there 

 was very little increase in the pressure, it was also observed that 

 no kathode rays were produced until the discharge had been 

 passed for some time. It appeared probable that some of these 

 effects, which are recorded in the paper above mentioned, were 

 due to the accumulation of argon in the tube. The spectral lines 

 of that gas were on some occasions faintly discernible in parts of 

 the tube, but no systematic spectroscopic observations were 

 taken. 



In making further investigations on the behaviour of argon, 

 we hoped to find that, owing to its natural inertness, the vacuum 

 would be of a very permanent type as compared with other gases. 

 We also hoped that its monatomic character would afford features 

 of interest. 



For the preparation and purification of the argon used in 

 these experiments, the Cavendish spark method was adopted, 

 as described by Rayleigh and Ramsay. For this purpose a 

 special transformer was constructed, the primary and secondary 

 of which were wound on different parts of the core. The 

 primary was connected to the loo-volt lighting circuit. The 

 secondary gave 10,000 volts on open circuit, available for 

 starting the arc, but the voltage on the arc when running was 

 only 2,oco. The secondary could be short-circuited, owing to 

 the arrangement of the winding, without materially increasing 

 the current, or running any risk of burning up the coil. The 

 apparatus could thus be left running safely by itself day and 

 night without wasting any power on resistances. After con- 

 centrating the argon to about 60 or 70 per cent, in the flask, it 

 was further purified in a test-tube apparatus, constructed so as 

 to contain the minimum of liquid. The excess of oxygen was 

 sparked off with hydrogen, and the residue removed by absorp- 

 tion with alkaline pyrogallate. The argon thus purified was 

 kept in a bulb containing P.^Og. 



In the first set of trials of this argon in X-ray tubes, a 

 Fleuss mechanical pump was used, which permitted very rapid 



1 By Prof. H. L. Callendar, F.R.S., and Mr. N. N. Evans, Lecturer in 

 Chemistry, McGill University, Montreal. (Read before Section A of the 

 British Association, at Toronto.) 



NO. 1 46 1, VOL. 56] 



exhaustion of the tubes, but had no arrangement for measuring 

 the high vacua. The vacuum was estimated in these cases by 

 the appearance of the tube and the width of the dark space. 



The first tube tried had two aluminium electrodes, and had 

 been lying open to the air for some time previously. It was 

 exhausted and wa.>5hed out two or three times with dry argon, 

 and then sealed off at a good X-ray vacuum. Each operation 

 occupied only two or three minutes, and the vacuum has since 

 that date deteriorated slightly, probably owing to insufficient 

 removal of residual gas from the electrodes, but it still gives 

 sufficient light to see "the bones of the hand. The tube during 

 exhaustion presented exactly the same appearances, except in 

 colour and spectrum, as if it had been filled with air. 



The second tube had been worked up to a sparkless vacuum 

 some weeks previously, and had been frequently renovated by 

 heating. It had an aluminium kathode and a platinum anode. 

 It was connected to the pump and exhausted as soon as possible 

 after opening. It was then filled with dry argon up to a pressure 

 of one-fifth millimetre, and exhausted to an X-ray vacuum five 

 times in succession. The glow on the kathode inside the dark 

 space showed the F line of hydrogen, and also the C line more 

 faintly. These lines probably indicated the elimination r)f 

 hydrogen from the electrodes, especially the kathode, as they 

 became fainter with each repetition of the process of washing 

 out. 



At the sixth filling of the tube, the pump was worked for ten 

 strokes only. The kathode then began to sputter and blacken 

 the tube, and the argon was apparently absorbed, as the dis- 

 charge refused to pass in three minutes. Fresh argon was again 

 admitted, the coil was left running, but the pump zvas not 

 worked at all. The spectroscope this time showed only blue 

 argon without any trace of hydrogen. The concave aluminiunk 

 kathode sputtered violently and partly mehed down. In less than 

 two minutes the discharge refused to pass through the tube, 

 which was then sealed off. 



The coil used in these experiments was a very small one, 

 which gave a two-inch spark with difficulty when running on a 

 large 8-volt battery. 



The next tube upon which we experimented was a double 

 focus tube, containing two aluminium kathodes and a platinum 

 antikathode. This was washed out with argon and exhausted 

 eight times with the two-inch spark coil running all the time. The 

 direction of the discharge was frequently reversed, but no trace 

 of absorption could be observed. The argon lines always 

 disappeared, and the hydrogen lines, especially F, became 

 faintly visible inside the kathode, as the tube approached an 

 X-ray vacuum. The tube at each exhaustion gave fairly bright 

 X-rays, and showed no blackening or sputtering. The hydrogen 

 lines showed more brightly close to the kathode than in the 

 body of the tube, where the argon lines were most conspicuous. 

 The hydrogen appeared in fact to be coming out of the metal. 

 The glass walls of the tube were in a very dry state, as it had 

 been previously heated and exhausted. 



Finding that we could not get rid of the residual hydrogen 

 with the coil, we had resort to the alternai.ing current, which 

 we had previously found very effective in tubes with double 

 electrodes. It appears that the elimination of hydrogen takes 

 place chiefly, if not entirely, at the kathode. With the first 

 application of the alternating current, the hydrogen lines showed 

 extremely bright. The tube was then exhausted. In fifty 

 strokes, the discharge refused to pass. On refilling with argon 

 to a pressure of one-tenth of a millimetre, the blue glow 

 inside the dark space .showed only argon and no hydrogen. 

 The pump on this occasion was not worked at all, but the gas 

 apparently was absorbed, and the discharge refused to pass in 

 about three minutes. There was some sputtering of the elec- 

 trodes and blackening of the tube, but the aluminium, though 

 blistered, was not melted. The experiment was repeated twice 

 with the same results. On reconnecting the tube to the two- 

 inch spark coil, the same absorption was observable but less 

 rapid. The electrodes were larger, and were less heated than 

 in the case of the first tube. 



We concluded from these and similar observations, of which 

 the above may be taken as a sample : ( i ) that the hydrogen occluded 

 in the kathode played the part of carrier of the discharge from 

 the metal to the gas. (2) That if there were sufficient occluded 

 hydrogen, there would be little or no sputtering of the aluminium. 

 (3) That when no hydrogen was present, the discharge was 

 conveyed from the kathode by particles of the metal itself, which 

 were capable of exciting fluorescence of the glass, and of gener- 



i 



