168 
NATURE 
was approached, until they were thoroughly understood ; 
and recollect that we owe the discovery of them—by 
which we are enabled now to determine the pressures 
acting in the atmospheres of the most distant stars— 
simply to the fact that Dr. Wollaston, instead of drilling 
a round hole, used a slit ; and to the other additional fact, 
that Mr. Simms, instead of using that slit with a mere 
prism, used a lens and made the beam parallel, and then 
allowed that parallel beam, after it had passed through 
the prism, to pass into another telescope,and form an 
image of the slit. You see how closely connected are the 
grandest discoveries with the skill and suggestiveness of 
those who supply different instruments for our use. 
Now I must ask you to come back again to the prism. 
I have already told you that dispersion is the measure of 
the difference of the refrangibilities. Ifwe take a prism 
which appears like an ordinary one, but really is com- 
posed of several layers of different kinds of glass, and 
pass an ordinary beam of light through it, it will be dif- 
ferently acted upon by the various layers, and we shall get 
a difference in the spectra. We have here, in fact, three 
distinct spectra, showing that there is something in the 
different layers of which this prism is composed which 
turns the light out of its path, and which disperses it 
more in some cases than it does in others. The cause of 
this is the density of the glass composing each layer: 
some kinds of glass are nearly twice as heavy as others, 
and fortunately we are not limited to glass, for if we were 
we should not be able to go so far in these inquiries as 
we do. The prism in reality consists of three separate 
pieces of glass of different density, and it may be seen 
that the three spectra obtained are differently refracted. 
It is a very natural conclusion that the heavier and 
denser glass should have a stronger action on the light 
than the lighter glass has. So that, in these inquiries, if 
we want to get great dispersion, not only must we use 
heavy glass, but we leave glass behind altogether, as 
amongst the liquids we find some which give even a 
greater dispersion than the densest glass. If a beam 
is passed through a hollow prism of glass filled with 
bisulphide of carbon, the spectrum obtained is much 
longer than that produced by the densest flint glass we 
can get. But there is another consideration to be borne 
in mind, The dispersive power and refractive power 
not only depend upon the density of the glass, but on the 
angles of the prism. Ifa beam of light is sent through 
two prisms of unequal angles the effect is extremely dis- 
tinct. Thus, if we take one prism with an angle of 20°, 
and another with an angle of 60°, the larger angle gives 
us a much greater deviation and dispersion ; therefore, 
we not only have density to help us, but we have also the 
angle of the prism. 
And now let us go on to a third important point in the 
matter. We are not limited to one prism if we wish to 
get a great amount of dispersion ; if you will think the 
matter over, you will see that there is no good reason 
why we should not employ two, and then you will find 
that the dispersion will be considerable. So you see, 
first, we have a single prism of a dense substance; by 
increasing the angle we get increased dispersion, and 
then we get it still further increased by adding another 
prism, and so we might go on, adding prism after prism, 
until we get to any number of prisms arranged in the best 
possible manner for the light to be successively dispersed 
by each of them. First of all, you have the dispersive 
power of glass, then you have the angle of the prism, and 
then you have a number of prisms, all of them capable of 
being so arranged that we can make them all useful in 
these inquiries, until at last we get a dispersion of such 
an enormous amount that the spectrum of the sun, as 
mapped by Kirchhoff and Biinsen, is several yards in 
length, although it is nothing but a succession of images 
of one of the finest slits which our best opticians are able 
to make, 
You see, therefore, that our spectroscope depends first 
of all on Newton’s work with the prism in 1675, and on 
the fact which Newton found out incidentally, that it is 
important that the prism should be used at the angle of — 
minimum deviation. We then get the slit added by 
Wollaston in 1812; then the collimating lens, added by 
Simms, in 1830. 
spectroscope improved and modified as an instrument, 
until at last we get spectroscopes so arranged that the 
glass is of the finest possible material, the angle being 
the largest possible, the glass the densest possible, and 
the number of prisms as great as possible, 
There are some other considerations connected with 
the manufacture of spectroscopes which it is hardly 
necessary I should bring before you, as they are rather 
more in the nature of detail than of general principles ; 
but I must point out that where liquids are employed, it 
is absolutely essential that the temperature should be 
as equable as we can get it. A current of warm air ina 
room is quite sufficient to render any spectrum obtained 
by these liquid prisms perfectly useless ; hence, although 
their great dispersive power is of great value in some 
cases, where we want dispersion more than anything 
else, still, as a rule, we are limited for nearly all our 
researches to these dense glass prisms of great angle, 
to which I have already alluded. But there is another 
consideration of great importance which comes in here. 
If the angle of a prism be large, a ray of light travelling 
from one prism to another, enters the second at an 
extremely small angle, under which circumstances a large 
amount of light is reflected, but still it is not better to use 
a greater number of prisms of a smaller angle than a 
smaller number of a larger one. Again in spectroscopes 
of many prisms it is essential that there should be some 
arrangement by which each part of the spectrum should 
be observed with each prism at the angle of minimum 
deviation for that ray. This may be done in many ways, 
and the beam may be made to pass back again through 
the prisms, thus doubling the amount of dispersion, On 
these points I shall have more to say presently, 
Another important consideration, besides the purity of 
the material, is the perfect figure of the slit, You might 
imagine that the slit of a spectroscope was perfectly easy 
to make ; but, judging by the results of the manufacture, 
it is extremely difficult, for a perfect slit is still very rare, 
the best being made by Steinheil of Munich, Mr. 
Browning has suggested making the slit of a compound 
of gold, which will not rust, or be acted upon much by 
temperature, and which also will take a good figure with- 
out any very great difficulty. 
NORMAN LOCKYER, 
(To be continued.) 
NOTES 
TuE Académie Royale de Belgique has elected Dr. Hooker 
‘‘Membre associé.” This has been done as their contribution 
to the Kew controversies. Prof. Monen, writing from Liege, is 
glad that Dr. Hooker has received ‘‘le plus haute distinction 
scientifique que notre pays peutconferer . . . dans une moment 
on vous (Dr. H.) soutenez une lutte vive et penible pour 
Vhonneur de Ja botanique.” 
IN their desire to uphold the standard .of medical teaching, 
the authorities of Charing Cross Hospital have committed an 
act which presents the appearance of an injustice. Not only 
have they not elected anyone to the vacant post of Demonstra- 
tor of Anatomy at their school, but they have sent the following 
announcement to each of the candidates :—‘‘ Resolved: That in 
the opinion of this Committee, the gentlemen who have offered 
themselves as candidates for the Demonstratorship of Anatomy 
have not had sufficient practical experience in teaching anatomy 
to justify the Committee in selecting any one of them.? The 
Committee would, of course, have been perfectly warranted in 
[fon 2, 3893 4 
In this way we have arrived at the 
= % 
7, 
