PHILOSOPHICAL SOCIETY OF WASHINGTON. 117 



vations of the total eclipse of 1878, which were made at Creston, 

 Wyoming. 



One special object ou that occasion was to search for lines in the 

 ultra-violet part of the spectrum. The eye-piece of the spectro- 

 scope-telescope was removed and a fluorescent eye-piece substituted. 

 The spectrum used Avas that of the second order, and upon examin- 

 ing the ultra-violet region prior to totality, it was found remarkably 

 distinct. During totality, however, neither bright lines nor contin- 

 uous spectrum were seen. It may be that the corona has no ultra- 

 violet spectrum, but these observations are not conclusive on that 

 point. It may be so faint as to have been invisible because of the 

 great waste of light by the use of a diffraction grating. Probably 

 it would have been better to have used the first order spectrum 

 instead of the second, but the knowledge we now possess points to 

 the conclusion that a crown glass prism would be superior to a 

 diffraction grating. 



Prof. Harkness also described in detail the operations of pho- 

 tographing the corona during totality. A point which it was desira- 

 ble to investigate with care was the determination of coefficients to be 

 used in formulae expressing the time of exposure of a photographic 

 dry-plate under such circumstances. If we denote the focal dis- 

 tance of the camera objective by F ; its working aperture by d ; 

 the exposure coefiicient by C ; and the length of the exposure ia 

 seconds by t, then 



^ = ^(^) 



d 



In the present instance the negative which had an exposure co- 

 efficient of 1.505 seems to exhibit nearly or quite all of the corona 

 which is visible to the naked eye, excepting only the streamers, yet 

 as the whole series of negatives taken on this occasion, shows that 

 every increase of exposure produced a corresponding increase in 

 the size of the image of the corona, there is no certainty that the 

 final limit was attained. It seems desirable during the next eclipse 

 to push the coefiicient to a value of 5, and greater if possible. 

 This can only be done with the most rapid portrait lenses, and 

 Prof. Harkness believed that the objectives employed should have 

 a ratio of aperture to focal distance of not less than one-fourth to 

 one-half. But in passing outward from the moon's limit, the light 

 of the corona diminishes very rapidly, and hence to depict all its 

 47 



