not generally exceed that of a comet, probably; and hence the action takes place invisibly to us. Photographic surveys of the heavens often repeated will doubtless give us more numerous records.
We now return to the regularly condensing swarms. In these the condensation will go on, and the temperature will rise until the loss by radiation equals the increase of temperature due to the fall of meteorites upon the continually condensing center. If we imagine a star to be condensed more and more by the fall of meteoritic material upon it, we shall arrive at a time in which, provided that the supply of material ceases, the increase of temperature of the star from that reason will also cease, and then will arise a condition of things in which the heat radiated from the star will be greater than the heat produced in the body of gas which is ultimately formed in consequence of the tremendous temperature caused by the continual fall of meteoritic matter toward the center.
If it be true that in the nebulæ we begin with meteoritic dust-particles far separate from each other, we must gradually get an increase of temperature so long as they approach nearer the center of the swarm by condensation; and so long as the heat produced by bombardment is in excess of the loss by radiation, the temperature will increase; but when the loss by radiation exceeds the gain by the bombardment we must get a reduction of temperature. A temperature curve like one of the arches of Westminster Bridge flattened at the top will illustrate this idea. We have on the left-hand arm of the curve those bodies in which we get a rise of temperature due to collisions and to condensation; along the top of the curve we have the gradual formation of a globe of gas; the gas begins to cool and gradually condenses, until at the lower end of the right-hand arm of the curve, as a result of the total action, we get the formation of a body like the earth.
Such a temperature curve has been provisionally divided into seven parts, and what has been done so far is to show that there are seven well-defined groups of bodies in space, which may be located, three on the rising part of the curve, one at the top, and three on the descending part; representatives of each of these groups have been classified and their spectra have been carefully studied. There is absolutely no difficulty whatever about placing all the celestial bodies which have been so observed by means of the spectroscope in one group or the other; and further, where the spectroscopic evidence is complete, there is again no difficulty in dividing these groups into species, just in the same way that the biologist deals with organic forms. This has already been done for one group, and in a very few years it will no doubt be done for more, so that here again we are definitely in the region of hard, detailed facts.