38 



Dr. H. H. Turner. On the 



upwards like the curve GH in fig. 2. Now comparing this with CD in 

 fig. 1, it suggests that just as CD could be explained by the addition 

 of a constant to the distance, which made a variable alteration in the 

 log distance, so C4H may be explained by the addition of a constant to 



the brightness, making a variable alteration in the log brightness. And 

 there is a possible physical cause for this constant addition, viz., the 

 general sky illumination or glare which is added to the coronal bright- 

 ness. A value of about 0*012 of the average brightness of the full 

 moon for this illumination seems to satisfy requirements for the 1898 

 photographs. 



8. I proceed to give a brief summary of the measures on the photo- 

 graphs of 1898 so far as they have gone. 



Four photographs have been selected for measurement, three of 

 them taken by me at Sahdol with exposures of 1 sec, 2 sees, and 20 sees.,, 

 and one taken by Capt. Hills at Pulgaon with exposure 8 sees. On 

 these, measures have been made along six radii extending approximately 

 N., S., E., W., N.E., and S.W., the last two being as nearly as possible- 

 in the direction of the main streamers. 



9. The exposures given to the standard squares were all the same.. 

 These squares transmit fractions of the light ranging from to 4 on a 

 scale of powers of 2, a range which might be extended with advantage, 

 seeing that measures on the corona can be profitably made over a range- 

 of to 7 at least. But the smallness of the range is made up for in 

 practice by the measurement of photographs with different exposures.. 

 Thus the longer exposures of 20 sees, and 8 sees, in the above series 

 control the fainter parts of the corona, and the shorter of 1 sec. and 

 2 sees, control the brighter parts near the limb. 



10. In comparing the results from the different plates, it is found 

 that the brightnesses shown by one plate differ from those shown by 

 another in a constant ratio. Since the log (brightness) is tabulated 

 this means a constant difference between similar numbers for the two- 

 plates. Following Sir W. Abney's practice, I have used the base 2 for 

 the logarithms of brightness, and recorded to 0*1, which represents a 



Fig. 2. 



A 



6 



