Apparent Magnitude of the horizontal Moon. 243 



24 laches focal distance, and fixed them in a frame close to 

 each other in the same plane. This instrument I placed in 

 a perpendicular direction, with a white screen behind it, to 

 receive the images of a lighted candle that stood before it 

 at some distance. The screen being moved until the two 

 images of the candle were seen very distinctly upon it, thev 

 appeared exactly of the same magnitude; but when the 

 aperture of one of the glasses was contracted to an inch in 

 diameter, the image of the candle formed by that lens ap- 

 peared three or four times less than the image formed by the 

 other ; and as these two images were seen at the same time, 

 the eye could judge more accurately of their dimensions 

 than of the apparent magnitude of the moon near the ho- 

 rizon and on the meridian. But when I moved the screen 

 nntil the circular spot of light or spectrum formed by the 

 lens of four inches aperture measured one inch in diameter, 

 the spectrum formed by the other lens measured only one 

 quarter of an inch ; and as the screen was moved from the 

 glasses by slow degrees, the spectrums decreased in the same 

 ratio, until two perfect images of the candle were formed. 



Experiment II. 



In order to exhibit a more exact representation of the 

 phaenomenon, I had a frame made to contiin seven convex 

 lenses fixed in the form of a semicircle. They were all of the 

 same focal distance, but not of the same aperture. The two 

 largest apertures measured each 1-2 inches; these were to 

 represent the pupil of the eye when viewing the moon at 

 her rising or setting ; the two next to these had each an 

 aperture of Of6 inches; the other two were still less; and 

 the single lens at the top of the curve had the least aperture, 

 to represent the pupil when the eye is exposed to the moon 

 on the meridian. 



This instrument being fixed in a perpendicular position, 

 with a candle before it, as in the preceding experiment, and 

 ihe screen brought to a certain distance within the focus of 

 the glasses, the spectrums upon it appeared in the same 

 ratio to each other as the glasses that formed them. For a 

 representation of these spectrums sec Plate VI. 



0^2 Hence 



