208 STELLAE PHOTOGRAPHY. 



the instrument once an hour. If desired, the intervening time could be employed 

 in other observations. The average length of a night, after allowing for twilight, 

 is about ten hours. It would not be difficult to find a location where four nights 

 in eveiy week would be clear. This would give for the maximum capacity of 

 a single photographic telescope nearly two thousand plates annually. The area 

 covered by each plate is twenty-five degrees square. The total area of the sky is 

 about forty thousand degrees square. Sixteen hundred plates would therefore be 

 required to map the entire sky. Two stations must be employed to reach both 

 northern and southern stars, and it therefore follows that it would be possible to 

 prepare in this way a map of the whole sky in a single year. The final charts 

 would not show the faintest stars that could be obtained by photography with 

 larger instruments, but would give about as many stars in a given area as are 

 contained in the charts of Peters and Chacornac. The charts should be carefully 

 compared with the original negatives, to remove defects which might be mistaken 

 for stars. To avoid the need of this comparison, the polar axis of the instrument 

 may be moved slightly in azimuth. As shown on page 195, each star will then 

 leave a short vertical trail. These can be distinguished with certainty from defects 

 in the plate, and will give a more accurate indication of the brightness of the stars 

 than can be derived from circular images. 



Stellar Spectra. 



An investigation of the photographic sjDcctra of the stars 'was conducted on an 

 entirely different method from that employed by previous investigators, which has 

 been described on page 182. A large prism was constructed, and placed in front 

 of the object-glass, as was first suggested and tried by Father Secchi in his eye 

 observations of stellar spectra. 



The great advantages of this method are, first, that the loss of light is extremely 

 small, and, secondly, that the stars over the entire field of the instrument will 

 impress their spectra upon the plate. As a result, while previous observers have 

 succeeded in photographing the spectrum of but one star at a time, and have not 

 obtained satisfactory results from stars fiiinter than the second or third magnitude, 

 we have often obtained more than a hundred spectra on a single plate, many of 

 them relating to stars no brighter than the seventh or eighth magnitude. 



The first experiments were made in May, 1885, placing a 30' jDrisni in front 

 of the object-glass of the lens described on page 182. No clockwork was used, 



