68 THE INDIAN ECLIPSE, 1898. 



On the day of the eclipse the actual procedure was as follows : 

 About ten minutes before totality the heliostat was started 

 going, and was carefully adjusted to get the spectrum central 

 in the field of the prismatic camera. Then the exposing cap 

 was put on, and the first plate drawn up into position by the 

 rack and pinion arrangement. Then the exposing shutter of the 

 slit spectroscope was closed, and the dark slide drawn out ready. 



About two minutes before totality was due the large spectro- 

 graph was moved in R.A. until the image of the cusps touched 

 a certain mark made on the slit plate exactly ^ in. from the 

 centre of the widely opened slit in the direction of the diurnal 

 motion.* This adjustment was easily made without assistance 

 by looking at the slit image with a miniature telescope placed 

 for the time being in front of the 6-in. object-glass, but facing 

 the opposite way. The slit being in the focus of the large O.G., 

 the latter formed a collimator for the small telescope, in which 

 the slit with the cusp image upon it could be clearly seen when 

 focussed for parallel rays. The eyepiece of the small telescope 

 contained a total-reflection prism, so that I could look in at 

 the side and not obstruct the incident light on the large lens. 



The image of the crescent sun was kept on the mark by 

 following in R.A. until the chronometer I was using indicated 

 88 seconds before totality, then it was allowed to drift ; I then 

 quickly removed the small telescope, covered up the 6-in. O.G. 

 with a plate of blackened aluminium, and drew out the camera 

 dark slide. 



During the last half-minute before totality was due I began 

 exposures with the prismatic camera, taking two instantaneous 

 photographs of the cusp spectrum and then drawing another 

 plate into position ready for the flash spectrum. 



Now, all being ready, only a few seconds remained before the 

 bright lines might be expected to appear. The gloom of the 

 approaching shadow was already increasing at an alarming rate. 

 I turned to the tele-spectroscope, took off the slit head, and 

 watched the spectrum of the last remaining thread of sunlight 

 without any slit, observing in the green region near b. All the 

 curved dark lines of the ordinary spectrum were seen at first 

 j ust as though the semicircular slit ordinarily used for observing 

 prominences had not been removed. But almost immediately 

 the rapidly narrowing band of continuous spectrum broke into 

 a number of strips, and the dark lines disappeared ; at the same 

 moment the bright lines flashed out in hundreds between and 

 across the streaks of continuous spectrum. They were not 

 faint, short lines, but long and brilliant arcs very sharply defined 

 and extending in many cases over 30 of the limb. 



The most astonishing part of this beautiful display was the 



* The large spectrograph was fitted with an ordinary slit, but, as the jaws 

 were opened very widely, it was practically a " slitless " spectrograph. 



