44 Prof. S. P. Langley on the 



in the luminous part of the spectrum, and our cold screen, it 

 need hardly be said, is not emitting any rays of orange light. 

 We conclude, then, that the apparent deflexion of O'l div. 

 here has no real significance ; but, as we go down the spec- 

 trum towards the region of the radiation from cold bodies, we 

 find, beyond deviation 39° (k= 4^*3), values which indicate a 

 real excess of screen over sky radiations. 



We have selected the first number in our second series of 

 Feb. 9, 1887, as the example, because this follows the normal 

 course without marked instrumental disturbance. The nature 

 of these fortuitous disturbances and the methods adopted for 

 their correction are explained in the original memoir. In 

 many cases, as will be seen, they may cause the sky on one 

 side or the other of the moon to appear momentarily warmer, 

 although the mean of the two is colder. 



Comparisons of the sky and screen by directing the bolo- 

 meter to the sky at the zenith and at the east and west horizons 

 follow, and three double series at the most important points in 

 the spectrum, made with the moon at a high, a low, and an 

 intermediate altitude, give indications of the atmospheric ab- 

 sorption. These, however, are not given here, but only the 

 double series (No. 2), of which a literal transcript from the 

 book of original entry now follows : — 



Station, Allegheny. 



Date, February 9, 1887. 



Wetbulbat7h30m= + lP-2C. ) T , 



Dry bulb at7h30m=+16°-0C. f ±niemai * 



Temperature apparatus at 9h 45m= +19°'8C. 



External temperature, near the freezing-point. 



State of sky at 8h 30m clear. Very good sky. 



Aperture of slit =3 millim. =27'-9. 



Prism used L„ : A=60° 00' 28". 



Rock-salt lenses set at 37 centim. 



Spectrum thrown west. 



Galvanometer No. 3. 



Time single vibration = 12 sec. 



Deflexion per degree Centigrade = 17'8. 



Bolometer No. 1 ; aperture=3 millim. = 27' # 9. 



Setting on D 2 =41° 08' 30". 



Battery current = 0"036 ampere. 



Header at circle, J. P. 



Reader at galvanometer, F. W. V. 



