THE BRITISH ECLIPSE EXPEDITION 



595 



disentangled from one another on the plates. A little reflec- 

 tion will show that, under these circumstances, it is only pos- 

 sible to measure the diiterence of the displacements between 

 the stars nearest the sun and those on the plates farthest from 

 it. The resulting value of the gravitational displacement has 

 therefore only about half the weight which it would possess 

 if the scale could be determined independently. But now 

 suppose that another region of the sky is photographed on 

 both the comparison and the eclipse plates, the chosen region 

 being at about the same altitude as the ecHpse field. Then 

 this region enables the scale to be determined so that the full 

 amount of the deflection can be deduced from the eclipse field 

 itself. This was the method that was tried. It possesses the 

 disadvantage that the position of the telescope has to be 

 changed during totality, and should an error occur in the 

 setting, valuable photographs would be lost. The method 

 was well adapted for use with our telescope, which possessed 

 an accurately divided setting circle. By marking the circle 

 with small strips of white gummed paper the telescope could 

 easily be moved from the one position to the other in twelve 

 seconds of time to an accuracy of one second in right ascension. 

 It was not necessary to look at any faint graduations, and the 

 setting could be made as well in the dim light of a total eclipse 

 as during broad daylight. 



The duration of totality was estimated as nearly five minutes, 

 and the programme of observations which I decided upon was 

 as follows, using 7-inch aperture : 



If the duration of totality were to prove shorter than antici- 

 pated, the last exposure could be shortened. If the sky were 

 somewhat cloudy, the first four plates to be secured "w^ould 

 enable a reasonably good determination of the deflection to 

 be made without losing much time in changing the position 

 of the instrument. With good conditions ample material 

 would be available for an excellent determination. 



