i8 



N.1 TURE 



\Nov.2, 187T 



INSTRUCTIONS FOR OBSERVERS, AT THE 

 ENGLISH GOVERNMENT ECLIPSE EXPE- 

 DITION, 1 87 1 



II. — PoLARiscoi'ic Observations 

 'T'HE chief points to which observers of polarisation should 

 direct their attention appear to be : — 



A. What is the nature of the ouilying corona? 



B. Can the radial polarisation of the circumsolar corona be 

 traced down to the photosphere, or, if not, how low ? 



C. Is secondary atmospheric polarisation traceab'e ? and if so, 

 does the plane change du'ing totality? 



A. We mii^ht suppose this to be due — 

 (i) to circumsolar matter (though at a great distance from the 



sun) reflecting light, 

 {2) to circumsolar matter in the stale of self-luminous gas, 



(3) to circumlunar matter diffracting and, to a certain extent, re 

 fleeting light (most improbable),' 



(4) to lofty atmospheric haze or cloud, of excessive tenuity, 

 diffracting light. 



The light ought to be, for 

 (i) strongly and radially polarised, 

 (2) unpolarised, 

 (3 and 4) insensibly or all but insensibly polarised. 



Hence polarisation observations would only serve to discri- 

 minate between (i) on the one hand, and (2), (3), or (4) on the 

 other. 



From the faintness of the object and its considerable extent, 

 the naked eye, armed with a polariscope, might be best. If a 

 telescope be used, it should be of quite low power, and the 

 aperture as large as the breadth of the pupil multiplied by the 

 magnifying-power. 



Suppose the polariscope be Savart's, the quartz plates being 

 thick enough (if the naked eye be used) to give bands as narrow 

 as, say, 20' diameter. 



<>>. 



Let the observer rotate the polariscope till the bands, if any, 

 seen on the dark moon disappear ; then, untliout rolating the 

 instrument round its axis, let him incline the axis so as to point 

 at the outlying corona in different directions round the sun, and 

 notice whether the bands spring into existence ; and if so, let 

 him sweep round the sun, noticing what lies outside the clearly 

 circumsolar corona of 5' or so height, and let him notice par- 

 ticularly by estimation the direction, relatively to the bands, of 

 the radius vector of the region where they are most vivid, oi", 

 better, the azimuth of both radius and bands. He should also 

 specify, provided he can do so -,oith certainty, whether the bands 

 were black-centred or white-centred. He should also state in 

 his account, and verify the statement by an observation made at 

 leisure before or after totality, wliether his Savart is constructed 

 (or set) so as to have the hands parallel or perpendicular to the 

 principal plane of the Nicol. 



A very useful adjunct to a Savart's polariscope would be a 

 glass reflector, or else a tourmaline, placed so as to cover a small 

 segment of the field of view near the edge. On account of the 

 possible difficulty of illuminating the reflector in the peculiar 

 circumstances of a total eclipse, a tourmaline would seem to be 

 preferable. It should be placed for the naked eye at the least 

 distance of distinct vision — for a telescope, in or in front of the 

 eye-piece, where a real image is formed so as to be seen distinctly 

 — the axis of the tourmaline being parallel to the edge or choril 

 of the segment, and the bands being set perpendicular to this 

 chord, in the event of rotation during the observation, the 

 whole should be rotated together. The question whether the 

 bands are bright-centred or dark-centred, winch, in the case of 

 slight polarisation, is difficult to decide, would thus be replaced 

 by the simpler question, whether the bands in the field were of 

 the same character as in the segment {i.e., bright being a pro- 

 longation of bright, and dark of dark) or of opposite character. 



The observer should previously have practised on the blue 

 sky, rotating his Savart till the bands disappear, and noticing 

 to what degree they are brought hack by small changes of 

 pointing without rotation, so as to be prepared for what he 

 is liable to from secondary atmospheric polarisation during 

 totality. 



Should only very feeble bands be seen in the outer corona, 

 such as might possibly be attributable to atmospheric polarisa- 

 tion operating through small changes of pointing, it would be 

 well for control to rotate the instrument a little till bands are 

 fairly visible on the disc of the moon, and notice whether on 

 jiassing to the outer corona, in lehataer direction, the bands, 

 instead of being reinforced, tend rather to be drowned in white 

 light. Should luminous beams or dark rifts be seen in the outer 

 corona, so as to exhibit contrast of light and shade in close 

 proximity, a good opportunity will be afforded of testing whether 

 the light of the onter corona is polarised or not. If it be 

 polarised, then on rotating the Sivart, so as to make the bands 

 cut at various indications the boundary of light and shade, the 

 bands will in certain azimuths of the Savart be stronger on the 

 luminous than on the dark side of the edge of the beam or rift. 

 If it be unpolarised, tlien, whatever be the azimuth of the 

 Savart, the bands will be rather drowned in white light than _ 

 reinforced on passing from the dark to the luminous side of the 

 edge. 



But Savart's and other colour-polarisccpes. which are ad- 

 mirable for detecting a slight polarisation in liiint which is not 

 particularly feeble, break down when the difficulty arises from 

 the feebleness of the light rather than the slightness of the polari 

 sation. In such cases a simple double-image prism, with a 

 diaphragm-tube, is better. Unless those who have seen total 

 eclipses can decide from trial (suppose on the clear sky after sun- 

 set, or at night when illuminated by the moon), combined with 



their memory of the degree of illumination of the outer corona, 

 it might be well that the observer should be provided with and 

 should try both instruments. 



B. For this a telescope will be required with a magnifying 

 power of, say, 16 or 20. A biquartz seems the best instrument, 

 placed at the common focus of the eye-piece (which should 

 be positive) and objective, and combined with a Nicol's prism, 

 or, if it can be procured, a thoroughly good tourmaline. A tour- 

 maline might be placed over the eye-hole, whereas a Nicol 

 might have to be placed in the body of the eye-piece, which, 

 however, is no particular disadvantage if properly done. 



Let it be ascertained by previous trial how much a Nicol must 

 be turned from the position in which the two halves are purph,' 

 alike to make the tints contrast more vividly. Say it is 30°. 

 Suppose the observer on the line of central shadow, so that the 

 limits of disappearance and reappearance will be on opposite ends 

 of a diimeter. The biquartz and Nicol have been relatively set 

 so that the line of junction is in the plane of polarisa ion of light 

 extinguished by the Nicol, turn them together before totality 30" 

 (or whatever other angle may have been fixed on) to cither side 

 of the diameter of disappearance, and, pointing the telescope to 

 the place of disappearance (Fig. i), await totality without dazzling 

 the eye. The moment the sun is covered, apply the eye to the 

 telescope, and notice whether there is a vivid contrast of colour 

 right and left of the line of junction of the quartz plates all the 

 way dinon to t/ie dark moon (Fig, 2), or only in the higher parts 

 of the circumsolar corona. 



Be ready to repeat the observation before reappearance, with 

 the telescope pointed to the place of reappearance : and mean- 

 while, if time permits, repeat Frazmou-ki's observation by point- 

 ing the telescope, without rotation of the analyser, so that the 

 line of junction bisects the moon, and noticing whether the semi- 



