NO. 8 



WATER-VAPOR TRANSPARENCY — FOWLE 



i; 



ROCK-SALT PRISMS 



For the work down to about lo /x two rock-salt 60° prisms were 

 available, one with faces 18 cm. tall by 15.5 cm., the other 19 by 

 12.7 cm. The former was cut from an especially clear crystal and 

 was generally used. For the work extending to greater wave-lengths 

 two rock-salt prisms of about 15° were used with faces 13.4 tall and 

 8.3 wide, 0.4 cm. thick at the refracting edge, 2.7 cm. thick at the 

 base. Further the front of the bolometer case was closed b)^ a 

 rock-salt plate i cm. thick. Table I gives in line (i) the coefficient 

 of transmission for rock salt (a, in the formula e''^^ where d is the 

 thickness in cm.), (2) the mean transmission for the smaller prism, 

 using the mean thickness, 1.5 cm., (3) the transmission for the rock- 

 salt plate I cm. thick, (4) the transmission for the two combined, 

 (5) approximate values for the amount of light reflected from the 

 4 rock-salt surfaces and (6) the complete transmission factor. 



Table !.■ — Transmissibility of Energy Through Rock-Salt 

 Prism and Plate 



I9M 20^ 



2.34 5.1 



. 030 . 000 



. 096 . 000 



. 003 . 006 



.107 .090 



. 003 . 009 



To protect the prisms from hygroscopic fogging of the refracting 

 surfaces, they were covered with a thin film of asphalt varnish. A 

 very dilute solution of asphalt in benzol was made, the freshly 

 polished surfaces were dipped in this and the excess allowed to 

 rapidly flow off. An even, very thin, nearly colorless film could thus 

 be obtained. Its absorption for the long-wave radiation used was 

 found to be inappreciable. Despite these protecting films the prisms 

 fogged and were often repolished before each day's observation, and 

 sometimes twice a day. A tube of phosphorus pentoxide (P2O5) 

 kept close to the little-exposed and well-varnished rock-salt plate 

 which closed the bolometer case entirely prevented fogging. 



The following table gives the data relative to the dispersion of the 

 various prisms used. The wave-lengths. A, are given in microns or 

 millionths of a meter, fi. The deviations, A6, are in minutes of arc 



^ Rubens and Trowbridge, Wiedemanns Annalen, 60, p. 7S7, 1897. 



