TEMPEEATUEE OF THE SUEFACE OF THE MOON. 25 



unequal circunistauces by processes of inatbematical computation. Even for the relative total 

 brightiioss of the sun and moon very discrepant results have been reached, which may be best 

 exhibited in the form of a table of the principal detenninations. 



300,000:1 (Lambert.) 



400,000 : 1 (Lanibort, allowance made for various errors.) 



300,000 : 1 (Bonj,'uet, Essai d'optiqiic, &,c.) 



801, 000 : 1 ( Wollaston, Thil. Trans. (1&29), vol. 8.) 



480,000 : 1 (Bond, Memoirs American Academy, vol. 9.) 



618,000 : 1 (Zijllner, Photoraetrischo Uutersuchungeu, p. 10,5.) 



350, 000 : 1 (W. II. Tickering, Pr. Amer. Academy, 1880.) 



70, 000 : 1 (Sir William Thomson.) 



There has been no essential improvement in such photometric processes as are here in ques- 

 tion since the early measures by Bouguer and Lambert. Zolluer's ai-e perhaps made with more 

 care than others, but giving all these values equal weights, we have 405,800 : 1 as the mean ratio. 

 It is sufficiently evident that the limits of error are here wide, and wo shall adopt 40U,000 to 1 as 

 the mo.st probable value. 



Eecurriug now to the comparison of separate spectral rays in sunlight and moonlight, we 

 find that investigations have been independently made by two competent observers. 



In those of W. H. Pickering,* in which light from various sources was compared with that 

 from a standard Argaud gas-burner at four different j)arts of the spectrum, there is a very great 

 preponderance of violet in the solar rays as compared to the lunar. It is possible that the differ- 

 ence is too great, but we have already remarked upon the exti'eme difficulty of real accuracy in 

 such determinations, and our own earlier observations are of a like order of discrepancy. 



Dr. H. C Vogeljt compared, by means of a spectrophotometer, in which a petroleum lamp 

 served as a standard, moonlight and sunlight which had been reflected from various kinds of rock. 

 As a result he fouud that a selective absorption of the more refrangible rays of the spectrum took 

 place on reflection of the solar rays by the surface of the moon, although not sufficiently pronounced 

 to indicate any very decided color in the substance of which it is composed. The moonlight agreed 

 best with sunlight reflected from yellowish gray sandstone. 



The spectrophotometer used at Allegheny in the later ob.servations (in 1884)f to determine 

 the amount of the selective reflection under consideration was the result of the experience ob- 

 tained in 1883, and at the same time not dissimilar in principle to those employed in the researches 

 of Yogel and Pickering, the brilliancy of the two spectra being comjjared at dittereut i^oints by 

 means of an artificial source of light of supposed constant inten.sity. This artificial source was a 

 kerosene lamp in which the oil was kept at a constant level, with Argand burner, and screens so 

 placed before the glass chimnej- as to limit the eftective part of the flame to a cylindrical portion 10 

 millimeters high, taken where it was brightest. The lamp was trimmed and cleaned before each set 

 of observations, and although the constancy of its light seemed to be all that was desired, the 

 quality was so dittereut from that of either of the two heavenly bodies to be compared that the 

 accuracy of the observations was not so great as would have been obtained if a source like the 

 electric light, for example, had been used. 



In order to carry out the measurements, the intensity of the sunlight had to be diminished, 

 and that of the moonlight increased until they were both compai'able with the standard. In doing 

 this no attempt was made to determine the amount of diminution or increase, although a rough 

 approximation to this is possible, but as only relative brilliancies in difierent parts of the two 

 spectra were desired, attention was mainly paid to securing a convenient intensity in the light to 

 be compared. 



Plate 2 represents the arrangement of the apparatus. The light, reflected horizontally by the 

 12-iuch silvered mirror of the siderostat, enters the dark room by an aperture, A, in the north wall. 

 Here, if it is sunlight which is being compared, all is stopped except what passes through a small cir- 



' Proc. Amer. Academy, 1880, i). 236. 



t Monatsberichte d. Konigl. Akademie d. Wissenschaft z. Berlin, Oct. 21, 1880. 

 {These observations were conducted by Mr. J.E. Keelcr of this observatory. 

 S. Mis. 69 4 



