Temperature of the Moon. 43 



gave it (213*2). The siderostat-mirror is moved to throw on 

 the image of the moon. Let us call the resultant deflexion 

 C. The moon appears in this example very slightly warmer 

 than the sky, for the image moves on to 213*4. Next, by 

 another adjustment of the siderostat-mirror, the sky west of 

 moon is thrown on. Call this result D. The image on the 

 galvanometer-scale moves back to 211*9, indicating cold. 

 Finally the screen is interposed the second time. Call this 

 second interposition E. In an ideally constant apparatus, 

 the second interposition of the screen should give the same 

 reading as the first. Actually 210 divisions is obtained, 

 instead of 215 as before, owing to the so-called ••drift'''' of 

 the needle during observation. The mean of the two readings 

 for sky-east and sky- west is now subtracted from that for the 



"p _i_T) 



moon and gives, under the column C ■= — » the difference 



between the temperature of the moon and the sky in our 

 arbitrary degrees. It will be seen from a comparison of all 

 the numbers in this column that there are fairly accordant 

 indications of a maximum near the prismatic deviation of 

 39° 30', which corresponds with the wave-length of 2 M *1, and 

 approximately with the maximum of the solar-heat curve. 

 There is another maximum of far greater magnitude near 

 37° 30 / (wave-length about 14^), corresponding to the 

 maximum known to exist in the radiations of bodies at a 

 temperature of about 0° Cent., and due, it would seem almost 

 beyond doubt, to radiations from the sun-heated lunar soil. 

 It will be seen also that all numbers in this column have one 

 sign, i. e. the positive, indicating that throughout this series, 

 without exception, the moon has been found warmer than 

 the adjacent sky. This, indeed, is to be expected, since, 

 without the atmosphere, the temperature of this sky would be 

 nearly that of the absolute zero, and at any rate lower than 

 that of the moon. The difference is so considerable that even 

 the temperature fluctuations due to the interposition of the 

 intervening atmosphere in no case this evening hide the fact. 

 In order to compare the fluctuating radiation of the atmo- 

 sphere with a constant source, we take the mean of the sky 

 observations and compare it with the mean of the readings on 



,, T ,,. ,. . B + D A + E . rf 



toe screen. In this particular case, — - — — — -r — gives a dit- 



ference of but 0*1 division; but this is, as we have already 

 observed, when the bolometer is directed towards the orange 



