134 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



the loiigitnde of the observiiifj station* ; (2) observations with a prismatic spectroscope during the 

 total phase ; (3) meteorological observations during the occupation of the station ^d especially at 

 the time of the eclipse ; t (4) observations of the first and fourth contacts. 



SPECTROSCOPIC OBSERVATIONS DURING THE ECLIPSE. 



The plan of spectroscopic work designed by Dr. Hastings included observations with pris- 

 matic spectroscope, which were assigned to me. The instrument used was a telescope made by 

 GrIjnow, belonging to Mr. C. H. Eockwell. It has an aperture of 2J inches, a focal length of 30 

 inches, and was for these observations furnished with an eye-piece magnifying 29 diameters. 

 Before the objective of this instrument was placed a prism of flint glass, having a refracting angle 

 of 30°. The refracting edge of the prism was placed parallel to the polar axis of the telescope 

 in order that the spectrum might extend in the field of view in the direction of the diurnal move- 

 ment of the heavens. By this means a slight movement of the telescope upon the polar axis 

 only was sufficient to keep the spectrum in the field during the total i)hase of the eclipse. The 

 apparatus was mounted equatorially upon a simple tripod, and was adjusted with its polar axis 

 approximately in the meridiah and elevated to an angle of 10°. At the time of observation, the 

 telescope pointed east of the sun. 



With this instrument I was requested to make a special study of the coronal rings, to note 

 what changes occurred during the progress of totality and to estimate the relative brightness and 

 height of the several rings at mid-eclipse. In addition, it was designed to estimate the duration 

 of the reversal of the Fraunhofer lines if this reversal was seen. 



In order to form an estimate of the appearance of the spectrum in the instrument, I attached 

 to it, the day preceding the eclipse, a slit and collimator and examined the spectrum thus formed. 

 The field of view was sufticient to take in the portion of the spectrum included between B and F, 

 and the principal lines between these limits were plainly seen. 



During the partial phase preceding totality, the encroachment of the moon ui)on the spectrum 

 enabled me to properly focus the instrument. The solar spectrum, while comi>lete on its violet 

 end, was cut off at the red end by the advancing moon. The obscuration affected the apparent 

 lower portion of the red end, but not the whole of it, since the moon was moving from a point 

 several degrees south of west, and the lunar cusp gave a narrow horizontal spectral line, which 

 was capable of a sharp focus. Seventeen minutes before the total phase began, indications were 

 noted of the formation of the Fraunhofer lines on the lower edge of the spectrum. Eight minutes 

 before the total phase, the principal lines were sharply defined. 



Five seconds after the beginning of totality four curved lines were seen, in the expected 

 positions of the lines C, Da, 1474, and F. A light cloud immediately afterward obscured the sun 

 and obliterated the lines; on its passage and the return of the spectrum, a small line, of about 

 one-third the height of the others, was seen between the third and fourth. One hundred seconds 

 after the beginning of totality, the coronal rings supplanted the lines above described, in color 

 red, yellowish green, and green, supposed to be C, D3, and 1474. They overlapped each other, and 

 the exact outlines of the overlapping portions could not be made out. At mid-eclipse a careful 

 estimate of their width was made ; the first (at the red end) was of the same width as the second, 

 but the third was over one-half this width. The estimates above given refer to the width on the 

 eastern or western sides. The width could not be estimated on the upper and lower portions on 



* This is given in IV, §. t This is given in IV, J 4. 



