Fan. 12, 1882] 
NATURE 
255 
They generally end somewhere about 650 wave lengths. Of course 
there is one well known exception, which is the potassium line, 
which has the wave-length of 770 (in fact below the A line), and is 
visible with a certain amount of difficulty. But as a rule all the 
bright metallic lines end above 700, and when they get towards 
700 they are always thin lines and poor lines. When you come 
to think of it, it seems highly improbable that you should obtain 
lines very low down except in the case of metals of low fusing 
point. If you have a metal of low fusing point, of course it is 
much more likely that you get lines of low refrangibility than 
you would if the metals have a high fusing point, and as a fact, 
taking a metal which has got a high fusing point such as iron, 
you find no line in the ultra red part of the spectrum, whereas in 
the case of sodium, which has a low fusing point, we do find a 
pair of solitary lines about W L 840. 
When you heat platinum wire by a current of electricity, at 
first though it may be hot to touch, it remains dark. Then, as 
you increase the current, it gets red and hotter still, and new waves 
—green—put in an appearance; and finally we get white 
light. At 550 degrees centigrade the body shows redness. 
At a white heat we have the whole of the visible spectrum 
present. Whether all the waves exist at ordinary temperatures in 
the platinum wire is a matter for future consideration. It is, I 
think, possible that such may be the case, provided the amplitude 
be very small indeed. At all events, the molecules of the body 
on which the source of radiation falls must be in a state ready to 
vibrate with the higher wave-length. Each wave as it puts in 
its appearance has a certain amount of energy, and a comparison 
between the energy of the two waves may be shown by photo- 
graphy as well as by their heating effect. But the heating effect 
is the true comparative measure of energy if the body on which 
it falls completely absorbs the radiations. Lamp black, perhaps, 
is the most perfect absorbent of all radiations, and the energy is 
shown by the heating effect on it. This heating effect in its 
turn is converted iato an electric current by the use of a thermo- 
pile. Here is a thermo-pile whichis capable of movement by a 
screw in any required direction, and we will take a very brilliant 
spectrum and cast it upon its face. It is connected with the 
galvanometor, and the galvanometer reflects a spot of light on 
this scale, and when a current passes that spot of light is 
deflected. You will see whereabouts it is. Now if I move the 
face of the pile gradually into the yellow I think you will find 
that it will move slightly up the scale. We will bring it more 
into the red, and if the galvanometer is in order you will see that 
it ought to be deflected still more. We have now got it in the 
infra-red part of the spectrum. ‘The deflection is still greater. 
It was by noting the deflections in somewhat this way that Dr. 
Tyndall was able to construct his spectrum monogram of the 
electric light, using as his mateial prism rock salt. The 
limit of the red is shown by an arrow, onthe left is the thermo- 
gram of the visible part of the spectrum, and on the r.ght of the 
invisible part of the spectrum (Fig. 20). Now this thermogram 
Fic. 20.—Tyndall’s thermogram of the spectrum of the electric light. 
is a puzzle, or rather was a puzzle, in a great many ways. For 
instance, if you know the wave-length of any two points you 
may find the theoretical limit of the spectrum by the method 
which I showed you in my last lecture. Unfortunately it lies 
well within the heat curve, or rather the effect of heat which 
is shown in the diagram. 
When working it out in this way it will be found that what I 
may call the thin tail of the thermogram lies beyond the 
theoretical limit of the spectrum. But before I go any farther 
I want to show you a possible cause for this, This rock 
salt prism has been very finely polished by an optician, and 1 
will mount it in the place of the bi-sulphide of carbon prism 
which we have so far employed. Now rock salt is supposed 
to allow the low radiations to pass through much more readily 
than a glass prism. We will try to get the spectrum tolerably 
pure. You willsee that there is a fairly bright spectrum upon a 
piece of card, 
Now in making observations with rock salt prisms, Col. 
Festing and myself were rather mortified to find that we got 
beyond the limits of the spectrum when employing photography. 
Our curiosity was therefore raised, and we endeavoured to find 
out why this was the case. We therefore placed some bichromate 
of sclution of potash in front of the slit of the spectroscope and ob- 
served the spectrum. I will repeat the experiment,and you see that 
all round the spectrum we have a wide spreading yellow halo which 
is due to the imperfection in the rock salt. The rock salt sur- 
faces are as perfect as grinding can make them, but still there is 
a certain amount of diffused light which passes irregularly through 
the prism, and gives us that yellow halo. It is totally different 
as you will notice when you replace the bi-sulphide prism in 
position. You get a pure spectrum, You will see that each side 
the green and the red is tolerably sharp, and when you use a 
properly adjusted spectroscope the perfections, and for that 
matter the imperfections, are much more apparent than they are 
when making a lecture-table experiment. I have shown you 
this experiment that you may see with what caution measure- 
ments taken with rock salt should be received. I may say that 
we tried not only one prism but three or four, made out of 
different samples of rock salt, and all gave a like result. The 
only way we can use a rock salt prism when it is well ground is to 
allow an excessively narrow beam of light to pass through it. 
Directly any large surface of the prism (as is the case when a 
Jantern, or condensing lens for condensing the beam upon the 
slit), is used, the action of diffused light at once renders the 
results liable to suspicion, 
Now I will show you other figures obtained from a thermo- 
pile when using very delicate apparatus. I wish to show 
you how the thermo-pile and photography can work hand in 
hand. We havea thermogram taken with a glass prism, and 
you will see that it presents some features of similarity—not 
quite like Tyndall’s thermogram. ‘There is a reason for this 
difference, which is that the one is a thermogram of the positive 
pole of a powerful electric light, whereas I believe the other was 
Fic. 21.—Thermogram of the spectrum, the positive pole of the electric light. 
taken with the whole of the radiation coming from both of the 
poles and from a less powerful electric light. The negative 
pole of the light we used has been calculated to have approxi- 
mately a temperature of 3,000°, whilst the positive pole approxi- 
mately a temperature of about 4,000°. When using a source of 
one temperature, and that temperature of about 4,000°, you will 
see that the curve forms a cusp, that is at the place of maximum 
heating effect it comes very nearly to a point, and I believe that 
if we obtained a spectrum of a source of heat at a perfectly 
= 
# D A 
Fic. 22.—Energy curve of the same spectrum as Fig. 21, obtained by means 
of photography. 
even temperature we should get that thermogram with an abso- 
lutely sharp point. By taking photographs we are able to check 
the results of the thermo-pile, . 
Fig. 22 is an energy curve as depicted by a photograph, You 
