400 Mr^ix. \V. H. M. Christie and F. W. Dyson. 



instruments erected on them. On one of these a large clock case, 

 5 feet long, 2 feet broad, and 1 feet high the heliostat was placed, 

 a light wooden frame holding the two objectives. This was covered 

 when not in use by Willesden canvas on a light wooden framework 

 which could be readily lifted on and off. The spectroscopes stood 

 on a mahogany table, 5 feet by 4 feet, which rested on the topmost of 

 two boxes. The boxes were of such a height that the middle of the 

 slits of the two spectroscopes were respectively half an inch below and 

 half an inch above the centre of the mirror. This had to be arranged 

 somewhat carefully, as a 5-inch and a 4^-inch lens had to be supplied 

 with light by a 12-inch mirror whose normal at the time of the eclipse 

 was inclined more than 40 to the incident and reflected rays. The 

 spectroscopes were in a hut, 8 feet square, made of Willesden canvas, 

 facing north and south, and with the north side open when the instru- 

 ments were in use. 



The adjustment of the polar axis of the heliostat was made by 

 means of an attached theodolite, the altitude of the axis being first set 

 to the latitude, and the azimuth then adjusted by observing the sun's 

 declination at different hour-angles from 9 h to 16 h . In this way 

 the instrument was readily adjusted till the observed declinations of 

 the sun agreed to within 2' with those of the Nautical Almanac 

 throughout the above range of hour-angle. The stability of the 

 mounting of the instrument on the sand was quite satisfactory, only 

 very small changes in level occurring, and no perceptible changes in 

 azimuth. The only difficulty experienced with the heliostat was in the 

 driving, which is not very satisfactory at large azimuths. 



Programme of Exposures. The two spectroscopes were adjusted to be 

 as nearly as possible on the sun's limb simultaneously, and the pro- 

 gramme of exposures was the same for both. The cameras of the 

 two spectroscopes were provided with rack movements, so that a 

 number of exposures could be made on the same plate. A flap was 

 arranged so that the exposures for both spectroscopes were made at 

 the same time. The programme, as arranged with an expected dura- 

 tion of totality of 90 sees., was as follows : 



Ten exposures, were made of 1 sec. duration beginning 10 sees, 

 before totality at about 1 sec. apart for the spectrum of the " flash " at 

 the beginning of totality. The plates were then changed, and at 

 20 sees, from the beginning of totality the plates were exposed for 

 50 sees., i.e., till 70 sees, from the beginning of totality for the 

 spectrum of the corona. The plates were again changed and the 

 image on the slit moved by Mr. Berry by means of the slow motion of 

 the heliostat, and ten exposures of 1 sec. duration, l>eginning at 80 sees, 

 after first contact, were given for the spectnim of the " flash " at 

 second contact. The conditions under which the " flash " was photo- 

 graphed, as determined by the circumstances of the eclipse at Ovar and 



