266 
NATURE 
1886 
BD: 
“<5 
[Huby 
down against the hot surface and stirred until a uniform glutinous 
mass is obtained. 
added, and the whole thoroughly stirred together. 
This may be left on an air-pump plate for at any rate a couple 
of years without perceptible alteration either in itself or the 
brass. H. G. MApan 
Eton College 
Butterflies’ Wings 
CAN you inform me of any method of relaxing the wings of 
butterflies allowed to stiffen in the closed state ? 
Stretford, Manchester, July 1 JaM. B. 
[If the butterflies are laid on damp sand under cover of a bell- 
glass or other air-tight covering they will soon relax so as to be 
fit for setting-out. A drop or two of carbolic acid on a sponge 
should be placed with them in order to prevent mouldiness. 
—Ep.] 
NOTE ON THE ALSORPTION SPECTRUM OF 
DIDYMIUM 
[x a paper on “ Radiant Matter Spectroscopy ” (Part 2, 
Samarium),! I said that in fractionation of the didy- 
mium earths with ammonia—‘ After a time a balance 
seemed to be established between the affinities at work, 
when the earths would appear in the same proportion in 
the precipitate and in the solution. At this stage they 
were thrown down by ammonia, and the precipitated 
earths set aside to be worked up by the fusion of their 
anhydrous nitrates so as to alter the ratio between them, 
when fractionation by ammonia could be again em- 
ployed.” 
That in most methods of fractionation a rough sort of 
balance of affinities beyond which further separation by 
the same method is difficult, appears to be a general rule, 
I have long noticed this action when fractionating with 
ammonia, with oxalic and nitric acids, and with formic acid. 
The valuable point which renders this fact noteworthy is 
that the balance of affinities revealed by fractionation is 
not the same with each method. It was in consequence 
of the experience gained in these different methods of frac- 
tionation that I wrote in my paper read before the Royal 
Society, June 10 last (Chemécal News, vol. liv. p. 13), after 
saying that I had not been able to separate didymium into 
Dr. Auer’s two earths, “ probably didymium will be found 
to split up in more than one direction according to the 
method adopted.” 
In illustration of this I may mention that, although I 
have not split up didymium into the two earths, or groups 
of earths, which are described by Dr. Auer, other pro- 
cesses of fractionation give me, so to speak, other cleavage 
planes or lines of scission through the compound molecule 
didymium. 
According to Dr. Auer, a line in the well-known yellow 
band, close to the soda line, but less refrangible (w.1. 
about 579), is a component of the absorption-spectrum 
of neodymium, and therefore, under all conditions, its 
intensity should follow the same variations as the other 
bands of neodymium in the blue (wave-lengths 482, 460, 
444). Some of my didymium fractions, however, show 
that the line 579 does not follow the same law as the 
other bands I have named. Thus, ina rather low frac- 
tion (+ 6) of the didymium earths from gadolinite and 
samarskite I found that the neodymium line 579 was of 
the same degree of blackness as the adjacent praseody- 
mium line in the yellow (wave-length about 571), but the 
bands in the blue of neodymium had almost disappeared. 
In the adjacent fractions of didymium I was enabled, by 
appropriate dilution, to keep this set of bands in the yel- 
low as a standard, of exactly the same intensity ; it was 
now seen that in successive fractions the intensities of 
the other more refrangible lines belonging both to neo- 
* Phil. Trans., Part 2, 1885, p. 705. A reprint of this paper is also com- 
menced in No. 1390 of the Chemical News, p. 28. 
Then the proper weight of vaseline should be | 
and praseodymium varied greatly from strong to almost 
obliteration, the bands in the yellow always being kept of 
| the same intensity. 
Didymium prepared from a specimen of fluocerite dif- 
fered somewhat from the other didymiums. Here the 
band 579 (ascribed to neodymium) was very strong, the 
band in the yellow of praseodymium (571) slightly weaker, 
and the bands in the blue of neodymium (482, 469, and 444) 
easily visible. On diluting the solution the bands in the 
blue of neodymium and the one component-of praseody- 
mium in the yellow (571) appeared to follow the same law 
in becoming fainter and fainter with dilution, whilst the 
other component band in the yellow of neodymium (579) 
remained unaffected. 
It seems to me that a possible explanation of this 
variation might be founded on the great strength of the 
bands in the yellow, and that the two fractions of didymium 
then under examination might differ only in the fact that 
one was slightly stronger than the other. To test this 
hypothesis I took the two fractions first experimented on, 
and putting each into a wedge-shaped cell of glass viewed 
them together in the spectroscope. (1) I adjusted the 
wedges so that the group in the yellow appeared to be of 
the same intensity in each spectrum. On examining other 
parts of the spectrum it was seen that in one solution the 
bands in the green were tolerably strong, and the bands in 
the blue scarcely visible, whilst in the other solution the 
bands in the green were very faint, and those in the blue 
quite absent. (2) The position of the wedges was ad- 
justed so that the bands in the green in each case should 
be of equalintensities. It was now seen that the alteration 
had greatly upset the balance of the bands in the yellow, 
the solution in which the bands in the green were faintest 
before, now having much stronger yellow bands than the 
other. The explanation mentioned above therefore falls 
through, and I see no other way of accounting for the facts 
except in the supposition that by the mode of fractionation 
then adopted, didymium had split up in a different manner 
to what it would have done if the method of Dr. Auer had 
been followed. 
The colour of the different fractions of didymium nitrate 
varies from a dark rose-red at the more basic end (+ 17) 
to amber at the less basic end (+ 4). These variations in 
colour do not necessarily accompany a difference in the 
absorption-bands, for in one instance an amber and a rose- 
coloured salt were found to have almost identical spectra. 
It would almost appear from these experiments, coupled 
with the facts I brought forward in last week’s Chemical 
News (p. 14), that the “one band, one element” theory 
I lately advanced in connection with the phosphorescent 
spectrum of yttrium, mav probably hold good in the case 
of the group of elements forming absorption spectra. 
According to this hypothesis, therefore, neodymium and 
praseodymium must not be considered as actual chemical 
elements, but only the names given to two groups of 
molecules into which the complex molecule didymium 
splits up by one particular method of fractionation. 
WILLIAM CROOKES 
HEATING AND COOKING BY GAS 
a FEW years ago the public was led to believe that 
= the use of coal-gas for lighting purposes was on its 
trial, and must shortly give way to the electric light. 
Threatened institutions live long, and even if coal-gas is 
destined to be eventually superseded by electricity for 
lighting purposes, a useful future is now opening out for 
it as a fuel offering many advantages over coal for 
domestic heating and cooking. In these fields it may 
possibly occur in the future that coal-gas—unless the price 
is everywhere considerably reduced—will have to 
encounter rivals such as the petroleum oils on the score 
of their cheapness, but at present, coal-gas, for cooking 
