712 
Proceedings of the Royal Society 
by small induction sparks, but seen very brightly by the end-on 
method of viewing. 
A comparison of the different spectra thus obtained follows, and 
some curious results are elicited as to the prevalence of certain 
impurities among gases, as well as alterations and even transforma- 
tions of some of them with time and use. 
These facts are contained chiefly in Appendix 1 and Appendix 2 ; 
while a third appendix, kindly contributed by Professor Alexander 
Herschel, contains some further observations of his with the same 
apparatus but higher dispersion introduced. See his account of the 
same. 
Two plates of spectra accompany the paper. 
4. On the Diffusion of an Impalpable Powder into a Solid 
Body. By R. Sydney Marsden, D.Sc., F.R.S.E., &c. 
In a note on “ The Effect of Heat on an Infusible, Impalpable 
Powder,” by Professor P. G. Tait, in the Proceedings of the Royal 
Society of Edinburgh, vol. ix. p. 298, for the year 1876-77, Pro- 
fessor Tait points out that such a powder becomes very fluid under 
the action of heat, and behaves in many respects in the same way as 
a liquid would do — viz., convection currents are distinctly to be 
observed, and small particles of the powder are thrown up from the 
surface, in the same manner as we perceive little drops of water 
thrown up from the surface of a glass of soda-water. And Professor 
Tait then asks the question — If, supposing we had two such 
infusible, impalpable powders, would they diffuse into one another 
as do gases and liquids ? This is a question which as yet has not 
been answered. Professor Tait and I have been engaged in some 
experiments on the subject for some time, but the difficulties 
(chemical and physical) to be overcome are much greater than at 
first sight appear, and at present we are unable to say definitely 
whether they do so or not. But I think an answer may be obtained 
from another source. In some recent experiments I had occasion to 
have a number of Berlin porcelain crucibles and amorphous carbon 
in an impalpable powder kept in contact with each other at very 
high temperatures for from ten to twelve hours, with the following 
effect, that, although the crucibles did not become fused, but 
