Temperature of the Moon. 35 



value at — 1 42° Cent.) has no sensible existence, but may be 

 here treated as that of the absolute zero. 



There is next given a description of the apparatus, which 

 consists essentially of a siderostat, carrying an 18-inch mirror, 

 and capable of sending a lunar beam of corresponding capacity 

 horizontally into an adjoining dark room and keeping it 

 fixed there during the night's observations. In the path of 

 the beam can be interposed a large double screen of blackened 

 copper, ordinarily filled with water. The beam then falls on 

 a condensing mirror, whose ordinary aperture of 8 inches 

 does not, as may be seen, utilize the whole of the beam trans- 

 mitted by the siderostat, but has been selected in reference 

 to the capacity of the rock-salt train of lenses and prisms 

 wdiich forms the spectrum. This train is believed to consist 

 of pieces of salt of hitherto unapproached perfection in work- 

 manship, as at the time our investigation commenced no salt 

 prisms were procurable giving a single Fraunhofer line in 

 the solar spectrum ; while with the actual rock-salt train, D 

 is divided in the spectrum of the moon. The general con- 

 struction of the spectrobolometer, and of the special bolometer 

 employed with it, will be found given in previous papers*. 



There are three principal methods of investigation : — 



First, the measurement of the total heat of the moon 

 with a concave mirror, admitting the interposition of a sheet 

 of glass to rudely indicate the quality of lunar rays as com- 

 pared with those of the sun. This method, which was that 

 employed by Lord Rosse, has been very thoroughly practised 

 here with results which have been partly given in the previous 

 memoir. 



Second, a method, practised here for the first time, and 

 yielding quite peculiar results, has been to form, usually with 

 this same mirror, an image of the moon, but to now let this 

 fall upon the slit of a special spectroscope provided with the 

 rock-salt train referred to ; and after expanding this exces- 

 sively minute heat in this way, it has been found possible, 

 with late improvements in the apparatus, to measure by the 

 bolometer the different degrees of heat in the different parts 

 of this lunar heat-spectrum, both visible and invisible. The 

 doing of this, with its results, forms the principal subject of 

 the present memoir. 



Third, since such a mirror as that just mentioned, owing to 

 its short focus, forms an extremely small lunar image, in cer- 

 tain observations (carried on, however, only during a limited 



* See this Journal, xv. March 1883, and xxi. May 1886. For a descrip- 

 tion of the improved form of bolometer and galvanometer see vol. xxii. 

 August 1886. 



D2 



