THE BRITISH ECLIPSE EXPEDITION 589 



being reflected into the telescope tube by a coelostat mirror. 

 Such instrumental equipment is very much easier to set up 

 in a temporary fashion than an equatorially mounted telescope. 

 The discordances in the 19 19 results were traced to a slight 

 astigmatism of the coelostat mirror due to its surface having 

 been unequally heated by the sun during the eclipse. In the 

 case of one of the instruments this was shown by the diffuse- 

 ness of the star images. The images given by the other instru- 

 ment were quite sharp. It was the former instrument which 

 gave the results which did not agree with the Einstein theory. 



On account of the importance for scientific thought gener- 

 ally of a decisive verdict on the question, it was decided by the 

 Eclipse Committee that a further attempt must be made to 

 determine the gravitational deflection at the eclipse of 

 1922, September 21. The circumstances of this eclipse were 

 not so favourable as were those of 1919 as the stars in the 

 neighbourhood of the sun at the time of eclipse were much 

 fainter. Preliminary examination of the field at Greenwich 

 showed, however, that the stars were sufficiently bright to 

 obtain good results under fairly favourable conditions. In 

 order to avoid the distortion which had been produced by 

 the heating of the coelostat mirror it was decided to use a 

 telescope mounted equatorially. For this purpose, the astro- 

 graphic telescope of the Royal Observatory, Greenwich, seemed 

 most suitable. This instrument has an aperture of 13 inches 

 and a focal length of 1 1 feet. It was therefore of sufficient 

 focal length for the purpose in view, and at the same time not 

 so large as to make it impossible to erect it in an out-of-the- 

 way part of the world. 



The path of totality of the eclipse passed across the Maldive 

 Islands, Christmas Island, and Australia from the north-west 

 to the south-east. There was nothing to commend the Mal- 

 dives, w^hich are not easily accessible, are very unhealthy, 

 and where there might have been trouble from vibration due 

 to the surf. Christmas Island was easily accessible and had 

 excellent facilities for the landing and erection of heavy 

 instruments. It possessed the further advantage that the 

 eclipse occurred at local noon, so that the sun was very high 

 in the sky at the time of eclipse. This is of some importance 

 when faint stars have to be photographed, as atmospheric 

 absorption is then a minimum. The north-west coast of 

 Australia is an inhospitable region, difficult of access, and as 

 landing is only possible in surf-boats, the use of heavy instru- 

 ments was out of the question. Incidentally, it may be 

 mentioned that the Australian Government subsequently 

 offered many facilities to expeditions which decided to go to 

 the north-west coast, but at that time our preparations were 



