406 



Prof. H. H. Turner. 



(3) The falling off in intensity is very rapid at first. At one radius 

 from the limb it has fallen to one-twentieth, at two radii to one- 

 hundredth. 



(4) As regards the absolute intensity, it will be more convenient to 

 refer the brightness to the more familiar unit of the moon's brightness. 

 It is a fair assumption to take the total brightness of the moon as 

 0"02 candle at one foot ; or (since an amyl acetate lamp = 0-8 candle) 

 to 0*02 X 81 -^ 0*8 of the units (amyl acetate lamp at 9 feet) adopted 

 in the intensity scale : that is, to 2-025 of such units. Now if the 

 moon were to shine for 50 seconds (the exposure of the photograph) 

 instead of 1, we must multiply this number by 50 ; and if further it 

 shines on the plate through a 4-inch object glass, so that the light 

 falling on this 4-uich circle is collected into an image of 1*5 inches 

 diameter, the brightness of any point of this patch would represent 



2-025 X 50 X (4/1-5)2 

 = 720 units = 2"^'^ units. 



TThus if we subtract 9*5 from all the numbers in Table I, we shall 

 get numbers fairly expressing the coronal brightness in terms of that 

 of the moon, in powers of 2 as before. 



(5) The results from the other photographs need not be given (in 

 this preliminary note) in detail : they confirm those already given 

 remarkably well ; and show that the diminution of light is very 

 gradual indeed after 45 minutes from the limb. They also seem to 

 show that the readings near the edges of the plate in 3 A are too low 

 by about 1-0 ; which is due to the fact that the magnifier is too small 

 to take in the whole object glass at the edges of the field. 



(6) We can now give the comparison of visual and photographic 

 observations. The results from plates 3 and 5 are preferred to those 

 of 3a far from the limb, for the reason stated in the last paragraph, 

 and for measures near the limb, results obtained from special measures 

 with a plain glass reflector. 



In 'Phil. Trans.,' A, vol. 180 (1889), pp. 380—1, Abney and Thorpe 

 give their visual readings for the 1886 eclipse in terms of a Siemens' 

 unit, and remark that the moon would have given an image equal to 

 1-2 candles or 1-4 Siemens' units on the same scale. Hence we must 

 divide their figures by 1*4 to get results comparable with the above; 

 and the same must be done for the 1893 eclipse, the numbers for 

 which are given in ' Phil. Trans.,' 1896 (A), p. 433. We thus get the 

 following table of comparative results, replacing now the powers of 2 

 hy ordii;iary decimals : — 



