406 



Mr. W. J. Dibdin. 



[Apr. 28, 



between the two perforated plates 2'4 inches apart. These can be 

 conveniently brought into position for modifying the colour and 

 intensity of the comparison light. The coefficient of absorption of 

 each of the coloured glasses was determined photometrically, and 

 tables prepared by reference to which corrections for their use can 

 be made. 



As this method was found to present an objectionable feature in 

 regard to the nncertainty attending the use of multiple glasses for 

 reducing intensity, the third method was employed. The lamp was 

 placed on a graduated bar in such a manner that its distance from 

 the ground blue glass could be conveniently altered, and the illumi- 

 nation of that glass reduced on strictly photometrical principles. 

 This method was found to answer so well that the experimental 

 apparatus first employed is being altered and improved. When it is 

 completed, a further and extended series of observations will be 

 made. In the meantime, the results already obtained may be dis- 

 cussed. 



Details of the observations are given in the tables presented with 

 the paper. By plotting the average results on a diagram a mean 

 curve is drawn. From the value thus found, the relative intensity 

 of stars of all other magnitudes can be calculated. 



It then only remains to convert the comparative into actual values 

 by the correction for the distance of those stars whose positions are 

 known, when their actual intensities will be ascertained. 



The following series is given of the average results of the determi- 

 nations of the intensity of a sequence of stars in descending order of 

 brightness, together with their respective magnitudes, and a com- 

 parison of their theoretical intensities on the assumption that a 

 second magnitude star equals 0'00075 candle placed at a distance of 

 109 feet, which factor is deduced from the mean curve of all the 

 determinations made : — 



