5i8 POPULAR SCIENCE MONTHLY. 



binaries have been discovered. As the objective prisms employed do 

 not permit the use of a comparison spectrum, the binary character is 

 apparent only when both the components are bright. In such cases 

 the lines of the spectrum are alternately single and double. A study 

 of the spectra of the southern stars photographed with this telescope 

 has been made by Miss Cannon as a part of the Henry Draper Memorial. 



The focal length of this telescope is about sixteen feet, so that an 

 arc a degree in length in the sky is represented on the photographic 

 plate by a line more than three inches long. The scale of the instru- 

 ment is thus very suitable for the details of nebulge, and for nearly 

 everything except the centers of the densest clusters. For long ex- 

 posures on difficult objects, such as globular clusters, the telescope must 

 follow the stars in their diurnal motion with great precision. This 

 can only be accomplished with such an instrument by watching a star 

 visually and keeping it constantly bisected by the lines of a reticle. 

 Formerly, a secondary telescope was used for this purpose, but, due to 

 the flexure between the two tubes, and perhaps for other causes, really 

 fine photographs were not obtained with this telescope until a lens for 

 following was inserted into the field of the main instrument, so that 

 the other telescope was dispensed with. In all cases the mean move- 

 ment of the telescope is provided for by carefully devised and well- 

 constructed clockwork; a^d in the case of small and rigid instruments 

 this alone serves fairly well, unless the exposure is more than an hour. 



More than five hundred variable stars have been discovered by the- 

 writer in the globular clusters, by means of charts made with this 

 instrument. These constitute nearly one half of all the variable stars 

 known, but they all occur in only one thirty-thousandth part of the 

 sky. At the centers of some of these clusters, the stars are packed 

 together so densely that there are one hundred stars to the square 

 minute. If the stars were equally dense over the whole sky, their 

 number would exceed ten billions, and the sky would be so luminous 

 that there would be no real night. In one of these clusters. Messier 3, 

 one hundred and thirty-two variables were found. These are all 

 situated within a circle whose area is one fourth of a square degree, or 

 only one one-hundred-and-sixty-thousandth part of the sky. In this 

 cluster one star in seven is variable. The photographs used for this 

 investigation must be made with the greatest care, and must then be 

 enlarged, or else examined by a microscope, since the images of the 

 stars on the original plates resemble ihickly scattered grains of dust. 



The duration of exposure employed varies enormously, according 

 to the instrument and the object to be attained. They have been made 

 from one second up to twenty-four hours. With the great Bruce lens, 

 an exposure of one second is sufficient for the brightest stars, while an 

 exposure of four hours, or more than fourteen thousand times as long. 



