January 28, 1897] 
question in relation to comets, would indicate that it may be 
equally unobjectionable in the case of the bright line stars. 
But there was more evidence behind with regard to the other 
carbon flutings. 
The spectrum of carbon does not consist of the blue band 
alone, so that some account must be rendered of the other parts 
of the carbon spectrum, more especially of two groups of flut- 
ings in the green, We should not expect the green flutings to 
be so easily visible as the blue in stars, for the reason that they 
fall in the brightest part of the continuous spectrum ; while in 
comets where there is little continuous spectrum, they are the 
most conspicuous bands. 
In the star BD + 36° 3956, for example, Vogel’s observations 
gave indications of the two green bands which are seen in the 
spectra of carbon compounds, and are the chief characteristics 
of the spectra of comets at mean distances from the sun. 
It was in 1890 that Dr. and Mrs. Huggins formulated ob- 
jections, based on some new measures, to my view as to the 
probable carbon origin of the blue band in the bright line stars. 
Thus, in two cases they found, as Vogel had found before 
them, that the brightest band was still more refrangible than 
the brightest part of the modified carbon band; but in each of 
these cases they found, also, a band about the position of 4685. 
As a result of their work, they made the following statement ;! 
“ Our observations appear to us, however, to be conclusive on 
the main object of our inquiry, namely, that the bright blue 
band in the three Wolf-Rayet stars in Cygnus, and in 
DM + 37° 3821, is not coincident with THE BLUE BAND OF 
THE BUNSEN FLAME.” 
The capitals are mine. It will have been seen how carefully 
Vogel in the case of comets, and myself in the case of stars, had 
pointed out that it was not a question of the Bunsen flame ! 
Dr. and Mrs. Huggins do not admit that the observed varia- 
tions of the band in the carbon spectrum are sufficient to 
explain the position and appearance of the band at 4685 in 
the stars, basing their objections on experimental evidence 
afforded by Hasselberg. 
Vogel’s researches, as well as my own, on the carbon spec- 
trum, however, indicate a much greater concentration of 
luminosity of the band about 4685 than appears to have been 
observed by Hasselberg. But this is not to be wondered at, 
since every change in the experimental conditions may have an 
effect on the spectrum. 
Tam not aware of any other objection to my view than the 
above, and it will be remarked that Dr. Huggins is silent 
altogether in regard to the existence of the band in comets. 
Very fortunately for science, a great mass of new work on this 
part of the subject has been brought together since the Meteoritic 
Hypothesis was published, chief among the workers being 
Prof. Campbell. 
Let us turn to this new work, therefore, and see in what 
direction it tends. 
We may take the case of one of the brightest stars of this 
class in Argo, the spectrum of a star which my friend Respighi 
and myself were the first to see on a very hot night in Madras 
in 1871, a beautiful spectrum with many bright lines. Prof. 
Campbell, in 1893, included the study of this bright line star in 
his work at Lick. What is his result with regard to the band 
at 4685? He findsa band at 4688.?— In my first discussion I 
took the position of the brightest band as 4682, depending 
upon measurements made by Ellery, at Melbourne, in 1879. 
Ina more general examination of all the Wolf-Rayet stars in 
1894," he finds a band at 4688 to be the most constant feature, 
and in some stars it appears almost alone! This in itself would 
be almost sufficient to prove carbon. 
Prof. Campbell does not discuss the origins of the lines and 
bands which he has measured, but it will be seen that for such 
a diffuse band as that in question, his wave-length 4688 does 
not differ materially from that already given for the modified 
carbon band. 
There is now, therefore, in the light of the newest and best 
work, ne question about the fact that in the bright line stars 
there is a band at 468, the wave-length of the modified carbon 
‘band ; and this was my original contention. 
The new measures obtained by Dr. and Mrs, Huggins, 
therefore, do not affect my views as to the origin of the blue 
band of the bright line stars recorded at places varying between 
A 468 and A 469. 
1 Roy. Soc. Proc., vol. xlix. p. 46. 
2 Astronomy and Astrophysics, 1893, p- 555: 
NO. 1422, VOL. 55 | 
3 [bid., 1894, p. 448. 
NATURE 
395 
With regard to this band, then, the first conclusion is now 
ares than ever, strengthened as it is by Campbell's new 
WOrk, 
Dr. and Mrs. Huggins object to another point. 
It has already been remarked that Vogel’s observations sug- 
gested the presence of the green flutings in one of the stars, 
On this point Dr. and Mrs. Huggins remark that, when 
observing the bright line stars with small dispersion, ** it might 
easily be supposed that the spectrum is brighter at the position 
of the green carbon band” ; with high dispersion, however, they 
can see *‘no sensible brightening” in that part of the spectrum. 
In the case of another line distinctly seen with small dispersion, 
they remark that with high dispersion it was so indistinct that 
they could not determine whether it was D or D,;! so that their 
observations do not demonstrate the absence of the green fluting 
of carbon. 
Observations made at Kensington strengthen the idea that the 
green flutings are present in the spectra of bright line stars. 
When high dispersion is employed, flutings are weakened in 
much greater proportion than lines; so that comparatively small 
dispersion must alone be employed in observations of this kind. 
Campbell shows that while the average position of the brightest 
blue band in one group of stars is 4688, in another group it is 
4652. These two bands are frequently associated in the same 
spectrum, but occasionally each occurs by itself, 
As to those stars in which a band appears about A 465, it is 
quite possible that we may still have to deal with carbon, At 
present I am not aware of any experimental evidence ; but the 
possibility of a band at this wave-length, under a certain still 
untried condition is suggested by the fact that a band about 
this position was observed in Brorsen’s comet in 1868 and 1879. 
But the existence of a band at 465 surely does not negative 
the existence of a band at 4685 ! 
The present position of the question of carbon, then, is this. 
The new work of Campbell justifies us in associating, not only 
in comets, as first suggested by Vogel, but in bright line stars, 
as suggested by myself, the blue bands at 468 with carbon ; and 
a study of the spectra of comets suggests, but does not demon- 
strate, that the other band at 465 has the same origin. The 
feeble appearance of the green bands is no doubt due to their 
superposition upon the brightest part of the continuous spectrum. 
Hence the idea of the chemical and physical kinship of comets, 
nebula, and bright-line stars is strengthened. 
J. NorMAN Lockyer. 
THE SAVING OF VANISHING KNOWLEDGE. 
T is well from time to time to take stock of our knowledge 
and of our methods of inquiry, to see whether we are 
working on sound lines. As the business man finds it necessary 
to periodically go over his stock and balance his books, so, too, 
the scientific man, especially the biologist, should perform an 
analogous operation, lest perchance he find out too late that he 
has been entering on a comparatively unprofitable line of work, 
or has been neglecting valuable opportunities. While it is im- 
possible to say what scientific work is ultimately unprofitable, it 
may not be difficult to suggest that particular subjects for 
investigation are of more immediate importance than others. 
Let us for the moment divest ourselves of all preconceived 
ideas and pet fancies, so as to discover what is at the present 
time the most urgent need of science. In order not to compli- 
cate the question, we will dismiss the practical applications of 
science by admitting that they are of immediate importance, 
This leaves the field clear for scientific subjects which are 
studied solely for their own sakes. ‘ 
We can, perhaps, gain a clearer view of the question by look- 
ing at it from the standpoint of our successors—What will be the 
opinion of the naturalist of a hundred, or of a thousand, years 
hence? What is the scientific work that he would wish us to 
have undertaken? This question is an easy one to answer. 
He would not consider it very necessary for us to elucidate 
the structure, development or physiology of every commonanimal ; 
these matters can be done at any time. The investigation of 
the life in the oceans—whether on the surface, in shallow water, 
or in abysmal depths—can be done by him as well as by us. 
We may safely leave for the present the problems of the Ant- 
1 Ray. Soc. Proc., vol. xlix. p. 44+ 
