220 



KNOWLEDGE. 



i October 1, 1894. 



I consider that Expeiimeuts II. and III. prove that 

 oxygen derived from the coal gas ■ used plays no noticeable 

 part in the production of the bright line, for it is evident 

 from the very great width of the line that the light is 

 coming from a considerable thickness of very dense vapour, 



and it is impos- 

 sible to believe 

 that oxygen can 

 be continually 

 diffusing into the 

 central dense 

 regions without 

 being arrested at 

 the outer layers ; 

 and in proof of 

 this, Experiment 

 III. shows the 

 actual effect of 

 allowing oxygen 

 to impinge upon 

 the outer layer, 

 the only result 

 being to obliter- 

 ate the absorp- 

 tion line and 

 produce a central 

 narrow line 

 brighter than 

 the broad line. 



There is, how- 

 ever, another 

 possible source of 

 chemical energy, 

 namely, the por- 

 celain tube itself, 

 whence oxygen 

 may be derived 

 from the silicates 

 of which it is 

 composed ; and 

 some support is 

 given to this 

 supposition by 

 the fact that after the experiments the tube is found to 

 be considerably corroded, black silicon being deposited on 

 the inside. 



To test this I arranged another experiment as follows : 

 Experiment IV. — A short piece of iron tube, two and a half 

 inches long and about three-eighths of an inch bore, was 

 bevelled at the ends and fitted between two pieces of hard 

 glass tube, the joints being ground to fit. This made a 

 practically air-tight joint when the glass became soft by 

 contact with the heated iron and was pressed close up, 

 and it permitted a high temperature to be attained in the 

 iron without the loss of heat by conduction which would 

 ensue if the tube were made wholly of iron. Moreover, 

 the temperature gradient being very sharp at the joints on 

 account of the bad conducting power of the glass, there 

 will only be a thin layer of the cooler absorbing vapour 

 formed, through which one must always observe the 

 bright-line phenomena. 



With the tubes thus arranged the Bunseu was made to 

 play only upon the iron, and when this had reached its 

 maximum temperature the sodium was tipped out from 

 the steel spoon as before. Now, if in the previous trials 

 the luminosity was due to chemical action taking place 

 between the sodium and the oxygen compounds of the 



* Theee experiments liave been repeated with nitrogen in place 

 of coal gas, and with identical results. 



Fig. 2. — n. Continuous spectrum of dense 

 sodium vapour witli D absorption. /;, c, d. 

 Successive appearances after the continuous 

 light has faded; tlu^ glowing sides of tlu' 

 tube give the bright edges to tlu> spectrum. 

 e. Ordinary llamc spectr\im of fodium witli 

 low disptM'sion. 



porcelain, one ought in this experiment to find a striking 

 difference in the intensity of the radiation. No such 

 difference was, however, to be observed, the sequence of 

 phenomena being exactly similar to that represented in 

 Fig. 2. Some more experiments are contemplated with 

 the iron titbe.but I thinli enough has already been done to 

 dispose of the theory that the light is caused by chemical 

 processes. 



lu the foregoing experiments there still remains to be 

 explained the continuous spectrum seen for a short time 

 when the sodium is first volatilized. To investigate this the 

 lamp and reflecting prism are brought into operation, and 

 directly the sodium is pushed in, filling the tube with a 

 reddish glow, the lamp is turned up and the prism 

 adjusted. Then a most beautiful phenomenon is revealed ; 

 the glowing vapour is apparently opaque to white light, 

 but projected upon it is seen a splendid violet light 

 gradually changing to a fine blue as the vapour becomes 

 more transparent, then greenish, and fading until finally 

 no tint is seen — the heated vapour being at this final 

 stage, to all appearance, perfectly transparent. Viewed 

 with the spectroscope one sees, first, a bright band in the 

 violet, the rest of the spectrum being dark except for the 

 feeble continuous emissioti glow visible in the red and 

 green, then a bright green band is transmitted, and later 

 still a red band appears. These three beautiful bands of 

 transmitted light gradually widen, the black absorption 

 spaces between them becoming less and less in width and 

 intensity ; finally, the dark space in the blue fades away* 

 altogether, leaving only the band blotting out the yellow, 

 which soon narrows down and becomes the ordinary D 

 absorption line. 



Now, by comparing the emission with the absorption at 

 every stage, it was found that the continitous glow was 

 briglitest at first, when the absorption was greatest, and as 

 the dark bands diminished the emission also faded, per- 

 sisting, however, as long as the least trace of a dark band 

 could be seen in the blue, vanishing when this dis- 

 appeared. 



The explanation of this result, I have no hesitation in 

 saying, is as follow.^ : — When the sodium vapour is above 

 a certain density it gives the remarkable absorption 

 spectrum described, liut like iodine and other coloured 

 vapours it emits, when in this condition, continuous light. 

 I must here remark, however, that in performing this last 

 experiment great care must be taken to have the apparatus 

 dry and free from oxygen, and the metal must be pure and 

 bright, otherwise the result is complicated by the pro- 

 duction of a spurious glow due to solid particles of oxide 

 forming a smoke-like cloud. 



Here, then, we have found a connecting link between the 

 radiation of the coloured vapours of the metalloids' and 

 the line spectra of the metals. May we not, therefore, 

 assume that all gaseous bodies are subject to the same 

 general laws governing their light radiations, and that a 

 sufficiently high temperature (coupled with not too great a 

 density) is the only condition required to make each 

 element strike its own characteristic light chord upon the 

 ether. The bearing of these experiments on the solar 

 radiation is obvious. Leaving alone the puzzling question 

 as to how the vibrating molecules are enabled to impart 

 their energy to the ether, we find it quite unnecessary 

 to introduce further mysteries, implied by the term 

 "luminescence," into the subject. The glowing calcium 

 and hydrogen of the chromosphere are in all probability 

 merely converting its enormous heat energy directly into 



t Iodine lapour at higher temperatures is said to emit a dis- 

 continuous spectrum, eorresp ndiut; to tlie absorjitioii bands (Sniel. 

 ISpectroscopie, p. 174); thus tlic relation is still closer. 



