Chemistry and Physics. 373 



>lKrtrnm of the air in the tube should eompletolv correspond with 

 th;it of the aurora. The question then arises— is it permissible to 

 a-sunu' the existence of such low degrees of pressure in the regions 

 .>f our atmosphere in which the aurora is developed? If we sup- 

 tMSr the aurora to be at a height of ten geographical miles, the 

 tiiniierature of the atmosphere at the surface of the earth being 0° 

 and the barometric pressure 760""", the pressure w^ill be 0-078'""' ; 

 at a height of twenty miles, 0-00001'"'". A layer of ignited air 

 1 in thickness would therefore, at a height of ten miles, be 

 • ■'luivalent to a pressure of 78""", and at a height of twenty miles, 

 to a pressure of O'Ol"^'" in a Geissler's tube, in ordir at \\\v same 

 U'!n|H'rature to give a spectrum as bright as that <>f the aurora. 

 I'lit as the thickness of the layer of ignited air must be reckoned 

 i'.v kilometers instead of meters, w^e should have, even at a height 

 '•i ten miles, so high a value for the equivalent pressure in a 

 <n'i>shT's tube— 78'" of mercury for a layer 1 kilonutcr in thiek- 

 iii'vs — that the most powerful induction apparatus could not over- 

 -'Hie the resistance of the air. From this it follows that the 

 luautity of particles of ignited air in a Geissler's tube is i)r()bably 

 fxtioniely small when compared with the actire quantity in the 

 <"i^' <'f the aurora. Since, however, the spectrum of a gas in such 

 a tul^e must have at least the brightness of the auroral spectrum 

 to admit of spectroscopic analysis, it follows that the emissive 

 power of the particles of gas ignited in the tul)e must lie much 

 greater than that of the ignited gaseous particles in the case of the 

 aurora. Such a difference can be produced only by diflerences m 

 temperature. If, therefore, the light of the aurora depen.ls on 

 Jgnited particles of our atmosphere, the temperature at which the 

 Ignition takes place must be much lower than that at which the 

 same gases can be ignited by electricity in Geissler's tubes, 

 from these considerations it appears that all the gas-spectra of 

 the different orders which we can artificially produce belong in 

 'the relatively greater 

 ! quantities of ignited matter renders a greater 

 emissive power necessary, and this can be produced only by a 

 higher temperature. Conversely as in the case of the aurora, 

 ("orona, zodiacal light, &c., a great number of active luminous 

 particles must be as^sumed to exist, it follows that the temperature 

 ot^^the ignited gases must be relatively low.— Fogg. Ann., cxli, 



2. '"Synthesis of coniln.— By the action of an alcoholic solutioa 

 fjrnonia upon butyric aldehyd, at a temperature not exceeding 

 ;"0 ^., ScHiFF obtained two bases, the formation of which may be 

 represented by the following equation : 



Tetrabnt>?aldln. * » t^ 3 



g^8H,^,Ke=2€,H30+NH3-OH, 



