26 



Messrs. Morton, Mayer, and Thomas on 



II. Determination of the average light of the lamp in all azi- 

 muths. — It was noticed at once that there was a vast difference 

 between the amount of light given out by the lamp in a direc- 

 tion transverse to the plane of the loop and in the direction 

 of that plane, the former quantity being about three times as 

 great as the latter : and while it would of course be possible, 

 on certain assumptions, to estimate what should be the average, 

 it was also perceived that a direct determination by experiment 

 would be far more reliable and important than any amount of 

 theoretical discussion. 



The lamp was therefore mounted on a divided circle with 

 the axis of the lamp passing through its centre. A fixed index 

 measured the angle of rotation of the circle and lamp, and was 

 so placed as to read 0° when the plane of the loop was in the 

 axis of the photometer. This was indicated by a well-defined 

 line of shadow of the nearer half thrown by the further half of 

 the loop on the photometer-disk. 



The lamp was rotated 10° at a time : and several readings 

 were made in each position, the averages of which are given 

 in the accompanying Table. 



Candle-power of Loop in various Azimuths. 



lane oi loop 

 photometer. 



Candle-power. 







6*7 



10 



6-9 



20 



8*4 



30 



12-8 



40 



14-3 



50 



16-3 



60 



17-7 



70 



19-1 



80 



19-8 



90 



20-6 





10)142-6 





Average = 14*26 = 69 percent 





of maximum 



This shows the results for one quadrant. Similar experi- 

 ments were made for the three other quadrants, with like 

 results. Fig. 3 exhibits the results of the Table plotted in a 

 curve. 



