210 
PROFESSOR W. N. HARTLEY OY 
Appendix. 
[1. Reference has been made to the fact that Mitscherlich (‘Fogg. Ann.,’ vol. 121, 
p. 459, 1804) compares the band spectra of metalloid elements with those of compound 
substances. He used both the oxy-hydrogen and oxy-coal-gas flames. He attributes 
only line specti’a to copper, bismuth, lead, gold, iron, manganese, chromium, tin, 
potassium, sodium, lithium, zinc, cadmium, mercury, silver, barium, strontium, and 
calcium. He figures banded spectra of the following elements, magnesium, lines and 
bands ; sulphur, selenium, tellurium, phosphorus, boron, iodine (bromine and chlorine, 
by absorption), and carbon. 
Cyanogen and ammonia are also figured as giving channelled spectra, as well as the 
following metallic chlorides and oxides :— 
PbO, PbClo, AuClg, Fe^Og or FeO, MnO or MiigOg, CuClj, CuBrg, CuTj, CuFo, and 
CuO or CujO, BiClg, BiBi’g, Bilo, Bi^Og, BaO, SrO, CaO, BaFj. 
The following salts gave lines, or lines and bands together :— 
BaCl^, BaBi’o, BaL, CaFo, CaCb, CaBi’j, CaTg, SrFg, SrCb, SrBi’o, SiTg. 
It will thus be seen that several metals enumerated on pp. 174 and 179 yield 
channelled emission spectra, and that these are not credited by Mitscherlich with 
other than line spectra, except in the case of magnesium, to which he assigns lines 
and bands. The most refrangible rays observed by Mitscherlich were about X 4,000, 
and, though wave-lengths were not determined, the positions of lines and bands were 
measured and the spectra very carefully drawn. 
2. Liveing and Dewar, in their “Investigations on the Spectrum of Magnesium,” 
‘ Boy. Soc. Proc.,’ vol. 44, p. 243, give the following description of a spectrum ascribed 
to the oxide or to the process of oxidation :— 
The component parts of the spectrum are the following—(1) The h group, X 5183- 
5172-516G. (2) The MgH series, close to it, 5210, &c., and 5186, &c. (3) Bands in 
the green. (4) The triplet near L, X 3838-3831-3829. (5) Triplet near M of the 
flame of burning magnesium, X 3730-3724-3720, with the group of bands in that 
region. (6) The line, X 2852. 
The spectrum wliicli I have described differs from the above inasmuch as the least 
refrangible ray photographed was X 3929, which is at the edge of a strong band 
degraded towards the less refrangible side. Next, there is a strong line and a well- 
marked band, very strong from 3834 to 3805. Liveing and Dewar place the triplet 
near L, in or about this region. The triplet near M, and group of bands mentioned 
above, occupy the place of a band with lines upon it, extending on my photographs 
from X 3805 to 3682. 
Lines belonging to triplets near L and M were not recognized, though by vaiying 
the exposure and using sulphate, nitrate, and carbonate of magnesia, the conditions 
under which the specti'a were obtained were modified. It is possible to obtain an 
intense continuous spectrum by prolonging the exposure to one hour and using the 
