16G 
PROFESSOR W. N. HARTLEF ON 
Ist Order, obtained by means of an uncondensed spark. That this metal at so high 
a temperature yields such a spectrum is undoubtedly due to the fact that it is 
almost, if not absolutely, impossible to vaporize it witli the oxy-hydrogen flame."" 
LiVEiNG and Dewar have recently obtained a band spectrum by the combustion 
of nickel tetracarbonyl which is also accompanied by lines (‘ Idoy. Soc. Proc.,’ vol.,52, 
p. 117). This spectrum, I expect, will be found to be due to metallic nickel and not 
to the compound substance. 
Yttrium and scandium, in solutions of their chlorides, each yield a line spectrum, 
with a group of bands in the red and orange region, when submitted to the action of 
a condensed spark. From the foregoing facts, and from the descriptions of spectra 
which here follow, it will be seen that several metallic elements emit banded spectra. 
Characteristic Flame Spectra of Elements emitted at High Temperatures. 
I. Line Spectra. —Lithium, thallium, nickel, cobalt. 
IT. Band Spectra. —Antimony, bismuth, gold, tin, sulphur, selenium, 
III. Band. Spectra with Lines. —Copper, iron, manganese, tellurium, lead, and 
silver. 
IV. More or less contimious Spectra with Lines. —Sodium, potassium, magnesium, 
chromium, cadmium. 
V. A continuous Spectrum. —Zinc, carbon, arsenic, aluminium. 
VI. No Spectrum. —Platinum. 
It might be supposed that the band spectra were due to the oxides and not to the 
metallic elements in Group II., but there is evidence against this in the case of 
silvert and gold, since no oxides of these metals can exist at the temperature of the 
flame employed. 
In the case of manganese the evidence is of a different character, and may be 
referred to at somewhat greater length, since Marshall Watts has attributed the 
band spectrum seen in the Bessemer flame to the oxide of manganese, chiefly on the 
ground that it was yielded by manganese chloride (Spectres Lumineux), and in the 
oxy-hydrogen flame by manganic oxide. No evidence was adduced to show that the 
spectrum in either instance was due to the metal. 
* See Appendix 5, p. 211. 
t [Clianiielled emission spectra of silver and tin, produced by the electric arc, have been noticed by 
Liveing and Dewae. 
“ Tin gives llutings in liigbly refi'angilile portions of the spectrum, and silver gives a line fluted 
looking spectrum in the blue.” ‘ Roy. Soc. Proc.,’ vol. 34, p. 122, 1882. 
The same observers haAm described the channelled spectrum of magnesium oxide. A set of seven 
bands in the green beginning X 5006-4 and fading toAvards the violet side of the spectrum are stated to 
be due to the oxide or to the process of oxidation ‘ Roy. Soc. Proc.,’ a’oI. 44, p. 243.—W. N. H. 
September 29, 1893.] 
