38 



MEMOIRS OF THE NxVTION^AL ACADEMY OF SCIENCES. 



tioiis nccessiiry for obtaiuiiig the coefficient of transmission by the atmosphere. This coefficient, 

 as has been explained elsewhere*, is found by means of the formula 



and from this may be found the original energy of the obser^^ed radiation before it entered the 



atmosphere, 



log iiJ^log rfi— ilf] /S, log f 



although these formuliB are strictly ai)plicable only to homogeneous rays, and hence give only ap- 

 proximate results. Each "detlection" is the mean of a number of observations made nearly at 

 the same time. 



Lunar heat ohncfMitions of Devemhcr 2. 1884. 



[Height of barometer /j:=7.34 dcciuieters. ] 



The comparison of observ^ations made at great and small zenith distances is also exhibited in 

 the form of a table. 



■Load, Log 



dz 



-Lo: 



/LogdA 

 \Log dij 



2.2625 :— .2467 

 2.2945 -.2147 

 2.4533 :-. 05.59 

 2.2625 -.2358 



9. 3921 

 9. 3318 

 8. 7474 

 9. 3726 



Log (^2 A 



1. 2790 

 1.21162 



0. 7551 



1. 2465 



Log 

 (Log t) 



Log t 



Logt. 



Log E. 



E. 



8. 1131— .0130 .971 

 8. 1256— .0134 ,.970 

 7. 9923i— . 0098 1.978 



8. 12611— .0134 .970 



I 



.1040 2.6132 410 

 . 1072 , 2. 6164 413 

 .0784 1 2.5876 i 387 

 . 1256 i 2. 6239 i 421 



The average value of t is 0.972, which is the fraction of the lunar radiation transmitted by a 

 column of air capable of supporting 1 decimeter of mercury. The fraction of a vertical beam 

 transmitted by the entire depth of the atmosphere would be ^'•'^=.806. 



The correction due to the change of the phase of the moon during the course of the night's 

 observations is taken from a curve based on the formula of Ziillner. 



CLA8.S 4. — CCMPARATIVE TIIANSMISSION OK GLA.SS FOR LUNAU AND SOLAU HEAT. 



The pieces of glass used were the same as those employed in the preliminary experiments. They 

 were A, B, and the "large window pane." A series of observations made by moving the sidero- 

 stat mirror so as to expose alternately to the adjacent sky and to the moon gave, as has already 

 been said, systematically different results from those obtained by the interposition of a screen and 

 other modes of observation. For reasons already given, the values found by alternate exposure 

 to the moon and sky are preferred. We give as an example the observations of December 3, 1884, 

 on the sun, and of November 20, 1884, on the moon, the general disposition of the apparatus em 

 ployed being that indicated in plate 1, and the glass, thoroughly dried and cleaned, being in ev- 



• American Journal of Science, Vol. 125, page 176. 



