Temperature of the Moon. 41 



The above conditions are still of ideal simplicity. The 

 great, the almost insuperable difficulty of actual observation, 

 lies less in the minuteness of the actual radiation, or even in 

 its two-fold character, than in the fact that it is masked to us 

 by the changes of an always intervening atmosphere. The 

 case of observations on the sun is totally different from the 

 present one, and would be so even if the sun were withdrawn 

 till it emitted no more heat than the moon; for in this latter 

 imaginary case, the greater portion of the solar radiation 

 would still lie in a spectral region totally distinct from that 

 in which the radiations proper to the obscuring atmosphere 

 are found, and it is the peculiar, unavoidable difficulty, at 

 every stage of this long investigation, that since the moon 

 and the air are both alike cold bodies, their invisible spectra 

 are, in general, superposed in the same field. Let us add to 

 this, that the (invisible) spectrum of the air is usually not 

 fixed but fluctuating, and we shall see the desirability of 

 having some separate standard with which to compare it from 

 night to night. This we obtain most conveniently by filling 

 a vessel of proper shape and size, either with water or with a 

 freezing mixture at a temperature constant for the series, 

 and making this vessel itself the screen which is interposed 

 between the bolometer and the lunar rays. The bolometer 

 must remain unmoved, and the direction of the heat-receiving 

 apparatus to the east or west of the moon must always be 

 understood to be obtained by a slight motion of the siderostat 

 mirror. 



That the minute change in the angle of presentation of the 

 face of this mirror does not affect its own radiations appre- 

 ciably might well be anticipated, but is a fact which has not 

 been left unproven by direct experiment. The bolometer, 

 then, enclosed in a non-conducting case which cuts off 

 radiations from every object but the mirror or prism imme- 

 diately in front of it, practically feels only the radiations 

 from the moon, or from the sky immediately on each side of 

 it, except when the screen is interposed. Mirrors and prisms 

 do indeed radiate heat to it from their own substance, but 

 these radiations may be considered as absolutely constant, 

 and as therefore absolutely negligible during the brief cycle 

 of a single observation. 



We confine ourselves here to the above general explanation, 

 referring the reader who may be interested in the details of 

 the observations to the original memoir, remarking, however, 

 that the actual spectral position of a ray is given by a circle, 

 reading to 10" of arc, and that previous measures of our 



