364 Scientific Intelligence. 



bands of nitrogen. The bands, whose approximate wave-lengths 

 are given above, extend over the whole length of the arc. In 

 order to show conclusively that these bands are due to nitrogen 

 alone it was necessary to prove, (a) that cyanogen was not pres- 

 ent in the gas used, and (b) that carbon and carbon compounds 

 were not present in the arc proper. Chemical analysis gave 99 -5 

 per cent pare nitrogen and 0*5 per cent for the remaining con- 

 stituents of ordinary air, but no cyanogen. Furthermore, since 

 the nitrogen was obtained from liquid air, and since cyanogen 

 does not ordinarily occur in air, it is extremely improbable that 

 cyanogen was present in the nitrogen as an impurity which had 

 escaped detection by chemical tests. Also, the direct combina- 

 tion of carbon with nitrogen only takes place at very high tem- 

 peratures. 



That carbon was not involved in the radiation of the bands in 

 question is demonstrated by the following experimental evidence. 

 When the gas was driven through a wad or plug the spectrum 

 was not altered in intensity. Hence, the hypothesis that the 

 bands owed their origin, in part, to fine particles of carbon which 

 were torn from the bomb by the rushing gas is not tenable. 

 When small amounts of carbon dioxide were mixed with the 

 nitrogen no appreciable effect on the bands could be observed. 

 When, however, larger quantities of carbon dioxide were intro- 

 duced with the nitrogen, the bands were noticeably weakened. 

 The strongest evidence in favor of the nitrogen origin of the 

 bands is the fact that the spectrum of the center of the arc was 

 precisely the same when the electrodes were made of the follow- 

 ing metals : aluminium, copper, iron, magnesium, and platinum. 

 It would be very extraordinary to find the same per cent of car- 

 bon in all of these elements. Besides, if carbon had been present, 

 even in extremely minute quantities, the carbon bands at A. 4382, 

 473V, 5165, and 5635, and above all the carbon line at A. 2478, 

 would have shown on the negatives. Not the slightest sugges- 

 tion of these radiations was detected. Finally when carbon elec- 

 trodes were used the carbon bands appeared close to the poles 

 while the so-called cyanogen bands persisted throughout the 

 entire length of the arc with the same intensity as when metallic 

 terminals were employed. 



For the sake of completeness, the authors show how the 

 hypothesis of the nitrogen origin of the " cyanogen " bands 

 accounts for all the phenomena observed by earlier investigators 

 who erroneously ascribed these bands to a nitro-carbon compound. 

 — Phys. Zeitschr., June 1, 1914, p. 545. h. s. it. 



6. The Influence of Magnetic Fields on Explosions. — It has 

 been suggested by Sir J. J. Thomson and other physicists that 

 combustion is concerned not only with atoms and molecules but 

 also with rapidly moving electrons. In particular, it was thought 

 that the great rapidity of combustion in an explosion-wave might 

 be due to the molecules at the instant of combustion sending out 

 electrons having very high speeds. These may precede the 



