pear as stars in our telescopes, their spectrum closely resembles that of the nebula. Going still further—still increasing the condensation, still increasing the temperature—the region of stars properly so called is reached, until at last we find those which are represented at the top of the curve. These results have been arrived at by spectroscopic work, and the facts recorded have been the chemical changes which take place in these swarms as their temperature increases, from the most sparse condition at the bottom of the curve to the most condensed one at the top.
In the sparsest swarms, in the so-called nebulæ, and those which are so dim as to be with difficulty visible, indications are found of the so far unknown substance or substances to which I have referred at the beginning of this article, together with carbon and hydrogen, and, in all probability, magnesium, one of the most common metals in meteorites, which has a bright spectrum visible at a low temperature; though I should add that the visible presence of magnesium has recently been contested. Its visible presence or absence, however, is not of fundamental importance. As the temperature increases, we find carbon more abundant, and traces of manganese and lead, metals which volatilize at a low temperature.
The next greatest change that supervenes is the addition of more familiar indications of the metals magnesium, manganese, and sodium, while the spaces between the meteorites glow more intensely with the light of hydrogen and carbon, probably brought about by some electrical action. Here the sparseness is still so great that we have little to do with the absorption of light; we simply deal with incandescent vapors due to the high temperature brought about by collisions among the meteorites and to the glow of the gases between the meteorites. But although the particles of meteoritic dust are so far apart that there is no possibility of any obvious absorption of their light occurring at this stage, to any large extent, the story is soon changed, for, when real condensation begins, the light of the meteoritic dust itself is absorbed by the vapors produced at low temperatures which lie between each particle of dust and our eyes. The whole theory of absorption is dependent upon the fact that light must come from the light-source through a vapor which is cooler than the light-source itself.
Thus we get a clear indication that, when this stage is reached, the meteoritic dust is very much closer together, and is on this account capable of forming a background enabling us to see these light-absorption phenomena. Absorption of light by the vapors of substances known to exist in meteorites, such as manganese and lead, is the first to occur, and these absorption phenomena gradually preponderate, and indicate change from low to high
