284 



NATURE 



[May 15, 1913 



affords such a surface. This bulb (Fig. 3) can be 

 made to glow like those you have seen before by 

 inserting it into a coil; and if the copper wire is 

 situated in the side tube the glow lasts a long time 

 tor the gas has as yet no access to it. But if I excite 

 the gas again, and turn the bulb round so as to drop the 

 oxidised wire into it, you see that the luminosity is ex- 

 tinguished in a fraction of a second. Combination of 

 the nitrogen atoms occurs much more quickly at the 

 surface, so that the whole quantity of active nitrogen 

 present is almost instantly used up. Incidentally 

 the experiment illustrates the extremely rapid diffusion 

 ot the gaseous residuum in an exhausted vessel, for 

 every particle ot the active nitrogen must evidently 

 find its way to the surface of the wire in the fraction 

 of a second. 



We oass now to consider the effect of nitrogen 

 in this condition on other substances. The yellow- 

 glow we have studied so far is due to the recombina- 

 tion of nitrogen atoms, and accordingly it shows a 

 nitrogen spectrum, though with very curious modifi- 

 cations. 



If we offer to the monatomic nitrogen other sub- 

 stances, it will often unite chemically with them, 

 which, of course, cold ordinary nitrogen will not do' 

 I go back to the apparatus used in the first experi- 

 ment, and admit some acetylene by a stopcock (c, 

 Fig. 1). The jet of active nitrogen now enters an 

 atmosphere of acetylene, and you see that the char- 

 acter of the light is at once changed; it has become 

 lilac. I turn off the acetylene and substitute chloro- 

 form vapour. We now get an orange light. This 

 may appear very different, but the difference is un- 

 essential. The spectrum is in each case that char- 

 acteristic of cyanogen and its compounds, only the 

 violet portion of this spectrum is more intense" with 

 acetylene, the red portion with chloroform. 



Since we get the cyanogen spectrum without having 

 any cyanogen compound originally present, we may 

 suspect that some such compound has been formed. 

 Let us pass from suspicion to proof. Using chloro- 

 form vapour from a bulb containing the liquid (see 

 Fig- 1). we pass the gases through a vessel in which 

 a test-tube is inserted. This test-tube contains liquid 

 air, and any condensable constituent is frozen out on 

 to its external surface (Fig. 4). After a few minutes' 

 run, we take out the test-tube and dip it in a solution 

 of potash. I now add a mixture of ferrous and ferric 

 salts and excess of hydrochloric acid. I pour out the 

 liquid on to this white porcelain dish, and you see 

 that abundance of prussian blue has been 'formed. 

 This proves the presence of some cyanogen compound. 

 We can get the same result w'ith pentane. ether, 

 benzene, or almost any other organic vapour. With 

 these the amount of cyanogen formed is much the 

 same, but the cyanogen spectrum, curiously enough, 

 is far less conspicuous. Benzene, for instance, almost 

 quenches the nitrogen glow, and little can be seen of 

 the cyanogen spectrum either. In most cases it 

 appears that hydrocyanic nc id is formed, but the 

 orange cyanogen glow, only obtained in compounds 

 containing much chlorine, 'is probably due to the 

 formation of chloride of cyanogen in addition. This, 

 when absorbed in potash, forms a cvanate, which has 

 been detected chemically. 



In the case just considered, the spectrum observed, 

 when active nitrogen is mixed with another substance^ 

 is that of the product of the action. In some cases' 

 however, the spectrum developed is that of the sub- 

 stance origfinallv introduced. I admit some of the 

 vapour of perchloride of tin : you see the brilliant 

 blue glow. I introduce a drop of the liquid chloride 

 on a wire loop into the flame of a Bunsen burner, 

 and vou see the same blue colour, though less advan- 

 tageously. The brilliance of the luminous effect does 

 NO. 2272, VOL. qil 



not seem to give any trustworthy indication as to 

 whether much chemical action is going on. If, for 

 instance, we admit bisulphide of carbon vapour to 

 the active nitrogen stream, we do not get very bril- 

 liant effects of luminosity — nothing striking enough 

 to be worth showing you — but none the less interest- 

 ing chemical actions are going on. The tube in 

 which the action occurs gets covered with the dark 

 blue transparent deposit, which I show by projection 

 on the screen. This substance is a known compound 

 of nitrogen and sulphur, originally investigated by 

 Mr. Burt in 19 10. If the gases are condensed farther 

 on in the tube by liquid air, we get a second deposit 

 of brown colour, which can be identified as the brown 

 polymeric carbon monosulphide studied by Sir James 

 Dewar and the late Dr. H. O. Jones. You see, then, 

 that the chemical action is completely traced. Actiye 

 nitrogen takes part of the sulphur from carbon di- 

 sulphide, leaving carbon monosulphide. 



The behaviour of active nitrogen with metallic 

 vapours is of interest, though it has not yet been very 



GAS 



STREAM 



ENTRANCE 



LIQUID AIR 



completely studied. I select the case of mercury to 

 show you. We pass the stream of glowing gas 

 through this tube, which contains a small pool of 

 mercury. While the mercury is cold, the yellow glow 

 passes on unaffected. I apply heat, and green mer- 

 cury light, of the colour familiar in the mercurv- 

 vapour lamps used in electric lighting, is apparent, 

 when active nitrogen mingles with mercury vapour. 

 Soon the tube gets obscured, except when I am 

 actually heating it, by a dirty-looking solid deposit 

 containing much metallic mercury. 



I wish to convince you that an explosive compound 

 of nitrogen and mercury has been form d. For this 

 purpose, to save the trouble of dismounting the tube 

 already used, we will take a similar one prepared 

 beforehand. I heat the mercurial deposit moderately 

 over a Bunsen burner, and, if you will kindly be quite 

 still for a moment, we shall hear a distinct'crackling 

 sound, as the explosive compound decomposes At 

 closequarters it is easy to see flashes of light accom- 

 panying the minute explosions, but these can scarcely 



