594 
ASTRONOMY: IV. H. WRIGHT 
The nebula N. G. C. 40, the spectrum of which is shown in figure 2, 
is a large dim ring with a comparatively bright nucleus. The spectrum 
was photographed with a very short camera, so the scale is small. The 
nebular spectrum is shown to consist principally of the hydrogen lines 
and the ultra-violet line at 3727A. As is usual in such cases the lines 
show knots at their extremities corresponding to the bright edges of 
the ring. Ni and N2 are present, though they are exceedingly faint, 
and appear to be short. The spectrum of the nucleus is over-exposed 
on this plate and is shown to better advantage, with reduced exposure, 
in figure 3. The original negatives of this spectrum are very small and 
the necessary enlargement has been so great that the definition is not 
of the best. It will be seen, however, that the spectrum of the nucleus 
is made up of numerous hazy bright band' and measurement proves 
these to be well-known Wolf-Rayet radiations. There are seventeen 
of them in addition to the hydrogen bands. The brightest is 4650A^ 
and the lesser one beside it is 4686A. If it is true that the nebulae 
^727A Hy 3889A 465OA 
FIG. 3. SPFXTRUM OF NUCLEUS 
FIG. 2. SPECTRUM OF N.G.C. 40. N.G.C 40. 
are condensing into stars there are reasons for supposing that this 
one is further along in the process than any of the others we have studied. 
In presenting these spectra an endeavor has been made to arrange 
them according to the degree of concentration of 4686A and some of 
the neighboring lines. It will be seen that the successive objects stand 
in very close relation to one another, yet at one end of the scale is a purely 
gaseous nebula, and at the other a banded star with only the shreds of 
nebulosity clinging to it. 
Quite aside from any bearing which these observations may have 
on the subject of nebular evolution it appears to me that they offer 
the basis for a satisfactory classification of the nebulae. The obser- 
vation of many of the nebulae for the purpose of investigating the de- 
gree of concentration of the lines which have been discussed would be 
impracticable but this, I believe, would hardly be necessary, for it 
will be observed that, having arranged the spectra in the order shown, 
there are certain progressions in the relative intensities of some of the 
comparatively bright lines which would serve to fix the positions of any 
of the spectra in the adopted scale. For instance, consider the doubtlet 
