1890.] 



On the Heat of the Moon and Stan 



489 



September 13th. No effect from Jupiter, but he was badly placed 

 near the S.W. horizon, and it was too windy for a satisfactory nega- 

 tive observation. Several observations of the moon showed the 

 greatest heat to be close to the limb ; the deflection for this part 

 ranged from 175 to 200 mm. 



September 17th. Night clear and quiet; heavy dew. I now deter- 

 mined more accurately the variation of heat from point to point across 

 the moon by arranging that it should transit centrally (or near a pole 

 if desired) over the disc and taking readings every ten seconds. The 

 first five curves in fig. 3 are the results of five consecutive transits 



Fig. 3. 



over the central part of the moon. The sixth curve was taken over 

 the south end of the moon and the seventh curve over the north end. 

 They were all taken when the moon was not far from the meridian. 

 These are the first observations which clearly indicated a maximum 

 of heat within the disc and not on the limb. This is evidently about 

 the position at which the Sun is vertical, being approximately at right 

 angles to the terminator. To save space and confusion, these curves 

 are all drawn superposed but separated in time to a small extent. 



September 19th. Full moon. The first curve, fig. 4, was taken 

 when the heat of the moon was largely absorbed by a piece of clean 

 window glass fixed across the mouth of the radio-micrometer. The 

 second curve was taken immediately afterwards (9 5 p.m.) without 

 glass. The heat transmitted by the glass, almost exactly 25 per 

 cent., is somewhat different from the proportion 173 percent., which 



