294 
POPULAR SCIENCE REVIEW. 
of the heat. He has operated upon oxygen, nitrogen, chlorine, &c. ; and 
in a second letter gives additional details, and states that glass tubes open 
at both ends may be employed instead of the usual ones, which are 
hermetically sealed. 
Mr. J.H. Gladstone, F.R.S.,haspublished in the “Philosophical Magazine” 
a paper “ On the Violet Flame of many Chlorides.” He says it is popularly 
known, that if common table-salt be thrown on red-hot coals, it produces 
violet flames, whilst by the ordinary method of examining bodies by pris- 
matic spectrum analysis, it is well known that that substance gives rise to 
the yellow double line D. 
Many other chlorides, for instance, the chlorides of copper, potassium, 
barium, platinum, gold, mercury, &c., also give the same violet light when 
sufficiently heated, but not the chlorides of calcium, lead, or magnesium, 
and these flames, when analysed, are found to consist not of the yellow 
light, but of three groups of lines ; the first green, the second bluish- 
green and blue, and the third violet. To obtain the violet with chloride 
of barium, the heat must be much more intense. He concludes that the 
violet light of these substances is not due to the chlorine combining with 
the hydrogen or the carbon of the combustible, nor produced by the 
chlorine when it is decomposed by heat ; and that the origin of the violet 
light requires further elucidation. 
Alexander Mitscherlich, in a recent number of PoggendorfFs “ Annalen,” 
has shown that the prismatic spectra of the flames of certain compounds 
of the metals are different from those of the metals themselves. He gives 
drawings of the spectra obtained from four different chlorides and one 
iodide, and shows the bearing of his observations upon solar chemistry and 
other questions. 
Dr. Robinson has published in the “Proceedings of the Royal Society” 
an abstract of a paper “ On the Spectra of Electric Light, as modified by the 
nature of the Electrodes and the Media of Discharge.” He expresses his 
belief that spectrum analysis is not that invariable or absolute test of the 
presence of certain substances as is generally supposed. He made numerous 
experiments, in which the media of discharge were air, nitrogen, oxygen, 
hydrogen, and carbonic oxide, including in some instances the vapours of 
mercury, phosphorus, and bisulphide of carbon ; he employed as elec- 
trodes, graphite, and twenty-three different metals ; and obtained 185 
different spectra, which he examined and measured. His conclusions 
are : — 1st. That many lines are found in all the gases, and in many, per- 
haps all the metals ; the existence of such lines must therefore be inde- 
pendent of the chemical nature of the substance. 2nd. That on varying 
the density of the gases, the spectra -were found to vary, showing that the 
character and even the existence of certain lines are dependent upon the 
mere density of the medium, whilst the chemical conditions remain 
unchanged so far as the presence of the given substance is concerned. 
3rd. That in the case of the presence of more than one substance, the 
spectra of the substances present are not merely superimposed without 
change, as has been generally believed ; and that there is even reason to 
believe, that for certain lines, the action of different bodies upon each other 
may be antagonistic. And 4th. That the brilliancy of the lines is very 
