The experiments of the Rev. Dr. Robinson on a variety of gases and vapours, inclosed in glass tubes, show that a greater change is produced by pressure than by heat. At the ordinary atmospheric pressure, the spectra show a number of bright lines on a coloured ground, the light of which is, in general, stronger towards the red than the violet end, and strongest in the green. In some the ground is so bright as to efface all but the most luminous lines. This is especially the case with hydrogen. On gradually exhausting the tube in which the vapour is contained, the spectra rather suddenly fade away, leaving only a suspicion of one or two lines, but upon exhausting the tube still more, these transition spectra become bright again, fresh lines appear, and they are changed into new spectra which are never so bright as those at ordinary pressure. Fewer lines are visible in the rarefied spectra, and of these four-tenths are not found in the spectra of atmospheric pressure. The difference between the common pressure spectra, the transition, and the rarefied spectra shows, that the character and even the existence of certain lines depend upon the mere density of the media, the chemical circumstances remaining unchanged. Dr. Robinson also observed that spectra are not superposed without a change; the spectrum of atmospheric air does not always exhibit all the lines of oxygen and nitrogen, and occasionally there are some lines not visible in either of them. It appears also that for certain lines the actions of bodies may be antagonistic. Metals do not always give the same spectrum, whatever may be the combinations in which they are found. Among various instances M. Mitscherlich mentions that the spectra of copper and the chloride and iodide of copper present essential differences, and Mr. Roscoe has found that a similar difference prevails in the spectra of carbon compounds when in a state of incandescent gas, which have hitherto been supposed to yield the same spectrum. ‘The spectrum obtained from the flame of olefiant gas is different from that obtained by the electric discharge through a vacuum of the same gas; while the spark passing through a cyanogen vacuum produces a spectrum identical with that of the olefiant gas flame, and through the carbonic oxide vacuum a spectrum coincident with that of the spark through olefiant gas vacuum.’ The chlorides, bromides, and iodides are the most easily vaporized of all the metallic salts, and give the most brilliant flames and the most intense spectra, especially the chlorides. A small piece of the chloride of barium volatilized by a colourless gas flame tinges the flame green, and the red and green lines on the spectrum stand out with extreme brilliancy. The scattered yellow light on the spectrum of the chloride of sodium is comparatively dark by contrast with the bright lines, and upon shading off the more luminous part of it, traces of lines are visible in the more refrangible portion.