112 



Dr. W. Huggins. 



tion to be very thin," but he prefers to regard this layer as consisting 

 not of gas, but of "a thin smoke-like envelope of precipitated 

 material."* Professor Pickering assumes the existence of an absorb- 

 ing atmosphere about equal in height to the sun's radius, but we 

 shall see further on that there are reasons which make this supposition 

 extremely improbable. 



The light emitted by the corona, whether by the incandescent 

 particles or by the gas mingled with them, which lies outside the low 

 region of absorption, will not have been subjected to the same selective 

 absorption as the photospheric light which is emitted below this 

 region. For this reason the light emitted by the corona will be richer 

 in the more refrangible rays than the sun's light before it enters our 

 atmosphere, and will be in a still larger degree richer in these rays 

 than the solar light which has been scattered by our atmosphere near 

 the sun. These considerations led me to hope that if the corona were 

 observed by this kind of light alone it would be at some advantage 

 relatively to the air-glare which comes in before it. It was of import- 

 ance at the same time to magnify the small advantage the coronal 

 light might have by some method of observation which could bring 

 out strongly minute differences of illumination. Such a power is 

 possessed by a photographic surface. I took some pains to satisfy 

 myself "that under suitable conditions of exposure and development 

 a photographic plate can be made to record (strongly) minute 

 differences of illumination existing in different parts of a bright 

 object, such as a sheet of drawing paper, which are so subtile as to be 

 at the very limit of the power of recognition of a trained eye, and as 

 it appeared to me, those which surpass that limit."f 



* " Constitution of the Sun," " Amer. J. of Science," vol. xxi, p. 33. 

 f " Proc. Koy. Soc," 1882, p. 411. 



Professor Stokes has suggested the following metliod of increasing the intensity 

 of that part of the coronal light which is polarised relatively to the glare from the 

 sky. He says in a letter, which he permits me to add here : — " The light of the 

 corona is known to be strongly polarised, while the atmospheric glare would show 

 no sensible polarisation. Let p be the intensity of the coronal light along any 

 radius vector which is polarised radially, and q the intensity polarised tangentially, 

 and let 2a be the intensity of the glare. Then, without polarising the light, the 

 intensity of the coronal light relatively to the glare would be as p + q to 2a. Suppose 

 now the light falling on the plate to be polarised, say, in a north and south plane. 

 Then north and south the ratio of the coronal light to that of the glare would be 

 increased to p to a, while in east and west directions it would be reduced to q to a. 

 In north-east and south-west as well as in north-west and south-east directions, the 

 ratio would be the same as without polarisation. If in four successive photographs 

 the plane of polarisation were set to north, north-east, east, south-east, we should 

 get a relative increase in coronal light, in one or other of the photographs, all round 

 the sun. It would be least in north-north-east, east-north-east, &c, directions, 

 where it would be p cos 2 22£° + q sin 2 22|° to a, or about - 85jp + 015q to a. 



" The most convenient way of polarising would probably be to use a Nicol of some 

 size not far from the plate." 



