Abrit 20, 1882] 
“A crucial test, which can only be applied during an 
eclipse, and with difficulty then, will be to observe what 
happens during the flash to those lines which are 
specially picked out for intensification in spots and 
flames. We might expect to see the lines untouched in 
spots, the lines thickened in spots, the lines brightened 
in prominences, stretching to different heights. 
“They would all appear to rest on the moon’s limb, or 
on the sun’s limb, if the cusps can be observed, because 
we are dealing with the section of a spherical mass, or 
rather, perhaps, of zones of concentric spherical strata. 
““To apply this test under the best conditions, adjacent 
lines should be taken with cross wires, or some equivalent 
arrangement adjusted on the corresponding Fraunhoferic 
lines before totality. 
“The iron lines at 4918-0, 4919°8, and 4923°1 will be 
the best to observe for this purpose, as they are close 
together, and two are always absent from prominences, 
and one is never thickened in spots.” 
When it was decided that an attempt should be made 
to secure observations of the coming eclipse, the next 
thing to do was to try to get over the tremendous diffi- 
culty that we have always experienced, namely, that 
during the eclipse itself, the sun’s light, and therefore its 
spectrum, were absent, so that our familiar scale of refer- 
ence is lost. This is at last got over in a manner so 
simple that the only wonder about it is that it has not 
been thought of before. I allude to the employment of a 
photograph of that part of the solar spectrum which we 
want, instead of micrometer wires or any other more 
elaborate means of determining positions, and this 
method I have already tested, and it works remarkably 
well. 
What is requisite is that instead of a camera replacing the 
eyepiece it should really form part of it. The plate canbe 
taken away and the eyepiece may beused in the ordinary 
manner, or a sensitive plate may be placed in it, anda 
photograph taken. It may then be taken out and deve- 
loped, half of it wiped off before it is exactly replaced in 
its original position, and then we have a field of view, the 
eyepiece never having been separated from the camera 
during the whole of this time, half of which is occupied 
by the photograph, the other half with the spectrum of 
that part of the solar atmosphere which it is desired to 
study. 
The instruments to be used during the eclipse—both 
telescope and spectroscope—will be identically those with 
which Capt. Maclear and myself observed the bright 
lines in 1871, so that instrumentally the chances are 
good. 
I have already pointed out that it is necessary that the 
slit should lie on the narrowest point of the cusps. To 
secure this a 3%’ finder of exquisite definition has been 
‘solidly fitted to the telespectrozcope with adjustments 
easy of application which shall insure this result, and in 
order that the observations may be continuous both in 
the presence and in the absence of the sun, a diagonal 
eye-piece with a prism twice the usual size, is employed. 
This slides easily in two grooves. Half of it is silvered, 
half of it not, and at the instant of totality the silvered 
portion is thrown into use. 
It is hardly necessary to add that the slit of the spectro- 
| scope can be made to lie at any angle from the normal. 
So much, then, for one possible test of the new views, 
There is another—not perhaps quite so direct, but one 
which it will be still of interest to make. Since 1871, 
‘NATURE... - 
577 
when Janssen mate the first observation of this nature, 
those observers who have studied the spectrum of the 
corona under good conditions with small dispersion have 
seen some dark lines as well as the ordinary bright ones, 
and it has been assumed for the most part that these 
dark lines are simply the dark lines of the ordinary sun- 
light reflected to us by particles in the solar atmosphere. 
The possibility of putting this question at rest in the 
most absolute manner by comparing the spectrum of 
the corona with a photograph of the green part—that 
is to say, the most luminous part of the solar spec- 
trum (for too much light must not be expected), renders 
this observation one of importance to make, and, thanks 
to Capt. Abney’s recent researches in the science of 
photography, it is now as easy, however confusing it may 
be to those who believe in chemical rays, to obtain a 
photograph of the green as of the blue, and this will be 
done before the eclipse. 
There is reason to think that if the new views have any 
truth in them the spectrum of the corona may—I do not 
say must—give us the ordinary solar lines changed con- 
siderably in intensity, but it is probable that this observa~ 
tion will be a delicate one at the best. 
But more than our views have changed since 1878. 
The photographic attack now requires seconds only where 
formerly minutes were wanted. Nor is this all: the red 
end of the spectrum awaits a record which it is now easy 
to secure. Indeed, thanks to Capt. Abney’s skill, plates 
have been prepared which it is hoped will grasp the red 
and green and blue light with equal vigour, so that one 
can now more than dream of a permanent record of the 
whole spectrum from the invisible violet at one end to the 
invisible red at the other. 
We got the first photograph of the spectrum during an 
eclipse by means of instruments constructed in 1875 for 
the Siam eclipse in that year.. In these instruments I 
employed a method first used by Fraunhofer, to save as 
much light as possible. The corona was its own slit and 
the prism was placed in front of the object-glass, and the 
dispersion of the prism used was small, because the 
method was new, the plates were slow, and we were 
anxious to secure something. We now know that we 
may safely go ahead, and a prism 3} inches square in the 
side, of 60°, will be placed in front of a lens of 22 inches 
focus. 
The length of the spectrum, if all goes well, will be four 
inches, including the infra-red, which Capt. Abney 
believes will be recorded in one minute’s exposure, and 
this will be available in an eclipse of 72 seconds. 
These extremely rapid plates enable other attempts to 
be made which formerly would have been considered 
hopeless. The ordinary photographs of the corona will 
be taken (by a lens of 5 feet focal length and 43 inches 
in diameter) on plates sixty times more rapid than those 
prepared on the old process. This fact must be insisted 
upon, because it is evident that the shortness of the 
totality during the present eclipse is not such a drawback 
as it once would have been. 
Another attack will be as follows :—An image of the 
sun will be thrown on the slit of a spectroscope by means 
of a heliostat and condensing lens. The size of the solar 
image thus obtained will be about } of an inch. The 
beam of light will be dispersed by a flint prism of 2 
