STELLAR PnOTOGRAPHT. 223 



time of exposure, we may hence obtain a method of determining relative intensities 

 of light. If two stars give equally distinct trails, their intensities must be inversely 

 as their polar distances. The last column in Table VII. shows that this condition 

 is nearly fulfilled. The difference in the photographic magnitude of 8G' 269 and c 

 is 9.8 magnitudes, and the photometric difference is 9.7 magnitudes. The image of 

 the first star in this case moves nearly seventy times as fast as the image of the 

 second star. Individual stars show a greater discordance, but the entire series fails 

 to indicate any appreciable systematic variation. The evidence is not conclusive, on 

 account of the small number of stars and the color of some of them. A much better 

 test is afforded b}' Plate 245. This photograph was taken on January 5, 1886. 

 Two exposures were made of the vicinity of the north pole. The first was made 

 with the clock attached, but the axis was so far out of adjustment that the results 

 ing elongation of the images caused them to appear as trails of a considerable length. 

 The second exposure was made without clockwork. All the trails during the first 

 exposure had the same length, whatever the position of the stars. The trails formed 

 during the second exposure had a length proportional to their polar distance. Ac- 

 cordingly, for stars near the pole the second trail was much the most intense, while 

 the opposite effect was produced for the more distant stars. A discussion of the 

 relative intensity of the trails of all these stars affords an additional determination 

 of the light ratio corresponding to one magnitude. Plate 238, taken on December 

 30, 1885, was made in a similar way. Another determination might be made by 

 varjing the velocity of the images. This could be done by changing the rate of 

 the clock, or the position of the polar axis. 



The above methods depend on observations of trails. Similar processes may be 

 applied to the images formed with clockwork, when the instrument and the clock 

 are accurately adjusted. A scale for comparison may be made by taking a series 

 of photographs with different apertures, or varying the time of exposure, moving 

 the telescope a little in right ascension or declination between successive exposures. 

 Two such exposures may be made on the plate to be measured, and the images com- 

 pared by the method eraplojed by Argelander for observing variable stars. Each 

 image will thus be compared with others of nearlj' equal brightness, and the final 

 values in grades may be reduced to ratios of light by the relative times or aper- 

 tures employed in the two exposures. Plate 209, described on page 210, may be 

 reduced in this way. Two exposures were made, of one and five minutes respect- 

 ively, so that the brighter image of each star represents five times the intensity 

 of the fainter image. Plate 375, taken on February 22, 1886, can be used in the 



