Mr Newall, Polarization phenomena in the Eclipse 1905. 197 



Polarization phenomena in the Eclipse 1905 August 30. By 

 H. F. Newall, M.A, F.K.S., Trinity College. 



[Read 13 November 1905.] 



Descriptions and illustrations of the visual observations and 

 photographic records obtained at Guelma in Algeiia were given 

 with a view of showing the kind of material collected for the 

 detailed study of the phenomena. It was mentioned that the 

 visual observations combined with the photographic records proved 

 that the plane of polarization of the light diffused by the earth's 

 atmosphere during the eclipse was very nearly horizontal. (This 

 evidence conflicts with the conclusion of M. Salet, who observing 

 at Robertville, about 37 miles from Guelma, had found that the 

 plane of atmospheric polarization was vertical.) Photographic 

 records, obtained with a Savart polariscopic camera for the 

 purpose of quantitative measurements of the relative amounts 

 of polarized and unpolarized light in the corona, showed that the 

 polarized portion of the atmospheric light was equal in intensity 

 to the radially polarized portion of the coronal light at a distance 

 of about 1J diameters from the sun's limb. Thus the phenomena 

 of a radially polarized corona seen through a plane polarized 

 atmosphere are somewhat complicated. The results obtained by 

 photographing the corona through a large Nicol prism, which was 

 set to transmit successively the vertical component and two 

 components which were inclined on each side at 45° to the vertical 

 and consequently perpendicular to one another, not only show the 

 strong radial character of the polarization of the corona, but also 

 seem to suggest that there is a selective action and that the 

 prominent streamers of the corona are markedly polarized. A 

 photograph taken with a new form of polarizing spectograph shows 

 a very marked difference in the intensities of the tangential 

 and radial components; but a curious feature in it is that the 

 Fraunhofer lines are not detected in either spectrum, though the 

 conditions are such as must be regarded as very favourable for 

 their detection. Unfortunately a second photograph taken with 

 an exposure six times as long for the sake of getting further 

 information about more distant parts of the corona was spoilt by 

 the admission of sunlight through the slit of the instrument after 

 the total phase of the eclipse was passed. 



VOL. XIII. PT. IV. 



14 



