552 Proceedings of Royal Society of Edinburgh. [sess. 
very beginning of the catastrophe. I admit that my exposition 
contains no direct allusion to this point, which, however, seemed to 
me too obvious to require a special explanation. We must grant, 
I think without hesitation, that the appearance of the nebular 
spectrum presupposes not only the presence of nebulous matter, 
but also those special conditions of temperature under which 
alone this matter can emit the peculiar lines of gaseous nebulae. 
Nobody denies now that the materials of which the stars are 
composed once formed nebular clouds, and that under such condi- 
tions they emitted the typical nebular spectrum, of which at 
present, with a few exceptions, we see no traces in their atmo- 
spheres. It is one of the great achievements of modem spectroscopy 
to have shown that the same substance emits essentially different 
spectra under different conditions ( e.g . the spectrum of hydrogen 
at low and high temperature). Hence we are clearly not permitted 
to think that nebular matter — an infinitely more complex structure 
than the simple hydrogen atom — will betray its existence by one 
and the same typical spectrum under all circumstances. The 
spectrum of nebular matter at a high temperature will most likely 
be essentially different from that at a low temperature. If our 
ideas of cosmic evolution be correct, the former must resemble 
that of incandescent cosmic matter in the star atmospheres, i.e. 
it must be chromospheric, while the latter is typical of the condi- 
tions in nebulae which our modern views suppose to be at very 
low temperatures, and luminous rather than incandescent. Doubt- 
less the nebular matter round a temporary star is under the former 
conditions immediately after the outburst. It is only after the 
subsidence of impacts that the star and the nebulous matter round 
it gradually cool down and approach those conditions of low 
temperature which finally lead to the appearance of the typical 
nebular spectrum. In the ordinary process of evolution, therefore, 
cosmic matter begins its spectroscopic existence by showing the 
low temperature nebular spectrum, and thence develops its high 
temperature or chromospheric character ; in temporary stars we 
notice the inverse process — so to speak, a negative evolution. 
These remarks will suffice to explain why the nebular spectrum 
should be absent at first, and should gradually develop with the 
cooling of the star. 
{Issued separately April 15 , 1905 .) 
