STELLAR PHOTOGRAPHY. 203 



graphic plate gives a more precise summing up of all the radiations, since no differ- 

 ence of color appears in the final picture, but the mean wave-length is not far from 

 4000 ten-millionths of a millimetre. Accordingly, blue stars will appear compara- 

 tively much brighter in the photograph, and red stars brighter to the eye. Their 

 relative light can be fully determined only by the comparison of the spectra, which 

 will be considered later. Meanwhile the photograph furnishes an excellent test of 

 the color of a star, since on comparison with the visual brightness the stars which 

 are faint photographically may be assumed to be red, and the bright ones blue. As 

 the difference amounts to several magnitudes, it furnishes a test much more sensitive 

 than that of the eye. Again, this method is applicable to the faintest stars visible, 

 when the difference in color is quite imperceptible b}' any other means. 



The first tests that were made of the photographic lens, and before it was 

 mounted equatorially, consisted in directing it to the pole, and photographing 

 the trails of the polar stars. This is probably the best method of testing the 

 quickness of any given form of lens, plate, or developer. It may be employed by 

 any photographer, as it is only necessary to turn the camera to the Pole-star and 

 leave it exposed for any convenient time, as half an hour. On developing the 

 plate, the faintest stars shown measure the sensitiveness. A^arying either the lens, 

 plate, or developer gives us a means of studying the quality of each. 



An excellent means of secui'ins; an automatic record of the cloudiness durinoc 

 the night consists in exposing a plate in this way. A long focus lens should be 

 used, but it need not be carefully constructed. The slide should be opened in the 

 evenino; as soon as it is dark, and it must be closed before the mornino; twilight. 

 This may be done automatically by an alarm-clock. On developing the plate, the 

 Pole-star will describe a circular arc having a length of 15° for each hour of expos- 

 ure. The time of passage of any clouds will be marked by interruptions of greater 

 or less length. Such an instrument also forms a photographic watch-clock. The 

 watchman must cover the lens at intervals for a minute or so, each of which will 

 be indicated upon the plate when it is developed. 



When the positions of the stars are to be determined photographically, the 

 trails possess some especial advantages. The edges are well defined, and the 

 errors introduced by the irregularities of the clockwork and the shaking of the tel- 

 escope when in motion are avoided. The declinations can be measured with greater 

 accuracy than the right ascensions. For the latter it is best to make a number 

 of breaks at specified times, thus breaking up the lines into a number of dots whose 

 centres can be determined with accuracy. Much care is necessary to avoid touch- 



