440 C. PIAZZI SMYTH ON MICROMETEICAL MEASURES OE 



But my own idea is still, that it may be owing to the electric spark's 

 power of convecting C along its wires ; and then, not merely because such 

 wires are usually coated along their whole length with an easily melted 

 material so rich in C as Gutta Percha, — but because the Induction coil itself 

 is, throughout its chief bulk, little but a huge mass of soft C ; and the rolled 

 up insulated wires inside it, make it a perfect ganglion for accumulating all 

 possible transportable atoms of that element. 



Some small spectroscopic evidence in this direction too, is already in print ; 

 as thus, plate i. of M. Lecocq de Boisbaudran's admirable book Spectres 

 Lnmineux gives two pictures of the electric spark in open air ; one near 

 the positive, and the other near the negative, Pole. They both of them 

 exhibit chiefly the well-known low-temperature Nitrogen bands ; but the 

 latter, or Negative Pole's end, has a glorious distinction from the Positive's, 

 in this, that it has also a very strong violet line, the a, or chief of the whole 

 display, which does not belong to the simple Nitrogen's or to the Air's low- 

 temperature spectrum at all; nor to their high- temperature spectrum either. 



To what then does it belong % 



According to my earlier and perfectly independent "gaseous spectra" 

 paper to this Society in 1880, it is the characteristic line of C} r anogen, or 

 Carbon combined with Nitrogen, the chief constituent of our atmospheric 

 air. Carbon vapour then added to the spark which is producing the Nitrogen 

 spectrum in the open air, can hardly but produce this Cyanogen leading 

 line ; and such Carbon can be obtained by the electric current from nothing 

 so readily as the gutta-percha, vulcanite, and waxed, or resined-paper 

 interstitions, which form so notable a portion of every Induction coil. 



But as I had an opportunity of setting forth something of this view in 

 Nature journal last year, assisted by a woodcut of the spectra, — I here close 

 this part of the present paper on the two compound and opposed gases, CH 

 and CO ; in order to proceed to the next part treating of the three elemental 

 gases H, O, and N. 



PAET IV. 



THE THREE ELEMENTAL GASES, H ; 0, and N. 



Subject 1. — H or Hydrogen. 



There is little trouble in procuring good II tubes ; and they are such 

 excellent illuminators as to get the better of all ordinary impurities, espe- 



in any quantity to speak of; so it could not have come out of the glass. SiH 2 and SiFl 2 were used 

 (very easily prepared pure). CO was the only result he could possibly obtain. That, and 

 always that." 



