592 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



diagram shows the arrangement for reflection. The dotted figure shows 

 the position for a measurement of the direct beam. All measurements 

 of direct reflection were made for an angle of incidence of 12°. 5. 



The method of observ- 

 ... ' ins will be seen from the 



notebook record of a 

 single series of measure- 

 ments given in Table 

 IX. In the. table, D and 

 R indicate direct and 

 reflected beams, respec- 

 tively. The first and 

 second columns contain 

 the zero points and end 

 of swings of the gal- 

 vanometer Gi, and the 

 third column, the deflec-, 

 tion. The remaining col- 

 umns, in order, contain 

 the lamp galvanometer 

 deflection ; the deflection 

 of Gi reduced to constant 

 lamp ; the means of each 

 pair of D or R values ; 

 the means of alternate 

 readings ; and the final 

 column, the quotients 

 of the two preceding columns which are the reflection coefficients 

 sought. In all, three series of measurements were made on the silver, 

 and two series on the glass-silver faces of each vane. To get average 

 coefficients which would represent the range of condition of the mirrors 

 during the pressure measurements, the vanes were cleaned and new silver 

 coatings deposited between each two series on the same vane. The 

 reflection coefficients are collected in Table X. For each surface studied 

 the diffused reflection for a beam which had traversed air was determined 

 by setting the mirror holder for normal incidence. The diffuse energy 

 reflected at an angle of 25° falling on the full aperture of the mirror M 

 was measured, and the total diffuse energy for the hemisphere computed 

 on the basis of the cosine law. If I 9 A is the amount of diffuse radia- 

 tion falling normally upon the area QA, distant r from the vane and at 



M 



Figure 7. 



