88 
NATURE 
[Dec. 1, 1870 

depending on the angular position of the analyser. He | linesextend outside the hydrogen spectrum and whether 
should notice the form, colour, and general appearance 
of any residual luminosity other than the well-known pro- 
tuberances ; should contrast the appearance, especiaily as 
to colour, with that seen when the light is analysed so as 
to retain light from the same region polarised in the per- 
pendicular direction, and should ask himself whether the 
luminosity is such as could be accounted for by the 
superior brightness near the sun of the unanalysed light, 
even though it were to suffer the same proportionate 
loss by analysis as the Corona at a greater dis- 
tance. Of the instruments named the Nicol’s prism 
is better adapted for a general survey, not  re- 
quiring any limitation of field ; the double-image prism 
is better adapted for a comparison of the oppositely 
polarised images, since the observer can compare them 
directly, not one with is memory of the other. The 
double-image prism will require a diaphragm with a long 
and moderately broad aperture in the focus of the eye- 
piece, rotating with the prism, and in the observation the 
length should be placed in a radial direction. 
Third Observation —The observation of the streamers 
as to polarisation might throw much light on their nature, 
and the observer who undertakes this observation, by 
means of a Savart’s polariscope, or else of a Nicol’s prism 
capped with a plate of calc spar or arragonite, will be in 
a condition to determine with advantage the plan of po- 
larisation, if there be polarisation, of the Corona generally. 
But the streamers must be his first object. 
Fourth Observation.—\t has been supposed by some 
that the polarisation observed in the Corona was really 
due to the secondary illumination of the intervening portion 
of the Earth’s atmosphere, in other words, to the illumi- 
nation produced by reflection from clouds, &c., towards 
the horizon. This effect may be eliminated by using a 
Savart’s polariscope, or, better, a polariscope with quartz 
wedges, and turning the instrument till the bands (if any) 
seen on the moon’s disc disappear. The corona can then 
be scrutinised as to polarisation, and the polarisation 
examined in different azimuths of the Nicol’s prism rela- 
tive to the radius drawn from the sun’s centre, by pointing 
the telescope instead of rotating the analyser. In this 
observation the observer has the choice of two rec- 
tangular azimuths. of the polariscope, for each of which 
the bands (if any) on the moon disappear, and if no bands 
be seen on the moon he is free to scrutinise the polarisa- 
tion of the Corona, by turning the polariscope. 
General Remarks.—The object-glasses of all telescopes 
intended to be used in polariscopic observations should 
be examined before departure as to their freedom from 
defects of annealing. All polariscopes including a Nicol’s 
prism, or tourmaline, should be marked, so that the prin- 
cipal plane may be readily known éy feeding, as sight- 
marks might fail for want of light. Double-image 
prisms should have one side of the aperture in the dia- 
phragm marked so as to distinguish the two images. 
INSTRUCTIONS FOR THE SPECTROSCOPIC OBSERVATION 
OF THE CHROMOSPHERE. 
NOTE: The objects to be obtained are:—1. To de- 
termine the actual height of the chromosphere as seen 
with an eclipsed sun ; that is, when the atmospheric illu- 
mination, the effect of which is doubtless only partially 
got rid of by the Janssen-Lockyer method, is removed. If 
the method were totally effective, the C line, the line of 
high temperature, should hardly increase in height ; but 
there can be little doubt that the method is not totally 
effective, so the increasein height should be carefully noted. 
2. Todetermine if there exists cooler hydrogen above and 
around the vividly incandescent layers and prominences. 
To do this the band of the spectrum just above the stratum 
which gives the hydrogen lines before totality and during 
totality, should be carefully examined, to notice (a) if any 
traces of the hydrogen spectrum exist above the region 
which before totality gave the hydrogen lines, and (4) what 


they also exist with it in the lower strata. 3. To test the 
American observations of last year as to the existence of a 
line at 1474 in the corona spectrum, by seeing if it be 
visible above the region which gives the hydrogen spec- 
trum. 4 Todetermine whether any other gases or vapours 
are ordinarily mixed up with hydrogen, but remain invisible 
with the uneclipsed sun in consequence of the absence of 
saliently brilliant lines in theirspectra. The observations 
should be conducted as follows :—1. Work with a hori- 
zontal slit, ora slit in a parallel of declination, according 
to the instrument used, whether altazimuth or equatoreal. 
2. See that the spectroscope telescope works easily, 
so that sweeping along the spectrum is easy. 3. Find 
before totality an average plain-topped region of the 
chromosphere, where some motion on either side does 
not brighten, or thicken, or lengthen the lines near the 
part covered at the commencement of totality. 4, Observe 
this before and during part of totality, the telescope being 
driven by the clock if an equatoreal is used. 5. Just 
before totality sweep from red to violet ; note the lines 
and their lengths ; mark the 1474 line and the lines be- 
tween D and E carefully. 6. Immediately after the com- 
mencement of totality sweep back ; note new lines—their 
heights ; especially the comparative heights of C, near D, 
F,and near G, with these former heights ; and especially 
whether band over 4, and the thickness of F. Note also 
the lines between D and 1474. 7. Just before the end 
of totality unclamp and bring back the slit to the following 
limb of the Moon ; and note the extinction of the spectrum 
by the reappearance of the Sun, if possible by a rapid 
sweep ; if this is not possible, then watch the behaviour 
of F ; sweep back again to see if there be any variations 
from (5) in the new region now observed. 8. Carefully 
note position angle where slit cuts limb. 9. Record 
the impressions of facts, and facts not already noted, As 
SOON AS POSSIBLE. 
INSTRUCTIONS FOR THE SPECTROSCOPIC OBSERVATION 
OF THE CORONA 
Note.—The word Corona is here used for convenience, to 
include all the light above the prominences. It therefore includes 
that part of the chromosphere which is generally veiled to us when 
observations are made by the Fanssen-Lockyer method. 
The principal object to be obtained is :—To determine 
whether it is possible to differentiate the outer layers of 
irregular outline and streamers from a stratum say some 
5’ or 6’ high round the sun, which may possibly be the 
limit of the gaseous envelopes above the photosphere. 
To attack this question, we require a long slit, a large 
aperture, and long collimator, and small dispersive power. 
The slit must be adjusted for a faint cloud*before totality, 
and on no account is it to be touched before observations 
of a similar cloud can be made after totality, by the heads 
of the party. The most important observation to make 
is, whether there are any dark lines in the spectrum at 
any distance from the sun ; and if so, at what distance? 
Next, whether there are any bright lines; if any, their 
positions must be noted, especially if the lines recorded 
by the American observers are again visible. The obser- 
vations should be conducted as follows :—x. Arrange the 
instrument so that the image of the following limb of the 
moon, at the point of its first contact, will fall on the 
left-hand side of the slit, placed nearly horizontally. 2. 
See if Corona is visible before totality, and note its 
spectrum with utmost care, moving the slit in azimuth, 
so that perhaps, at the instant during totality, while 
possibly with a long slit, the spectrum of the sun or 
prominences on the preceding limb is visible in the same 
field of view. 
SOME PARTICULARS TO BE ESPECIALLY NOTICED BY 
THOSE OBSERVERS WHO MAKE DRAWINGS OF THE 
CORONA 
1. Its extent, and the boundary-line ¢f any, if no 
definite boundary, this should be stated. 2. Whether 
