6 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 'J2 



stant becomes 2.66 for the rays of short wave-length for which glass 



is highly transmissible. 



Test experiments of three kinds were made with Melikeron No. i : 

 (i) Using an incandescent lamp source, comparisons were made 



by interchanging Melikeron No. i and Pyranometer A. P. O. No. 5. 



Table I 



Date, 1920 



Feb. II. 

 Feb. 12. 

 Feb. 13. 



Feb. 13. 



Feb. 14. 



Feb. 14. 



Conditions 



Carbon lamp, 60 cm directly above. . 

 Same, except angle 25° from zenith. 

 Mazda "Daylight" lamp, 30 cm. 



above, 13" from zenith. 

 Same source, 8 cm. above and 25 cm. 



east, 72° from zenith. 

 Same source, 30 cm. above, 7J° from 



zenith. 

 Same source, 30 cm. above, 4*° from 



zenith. 



Calories 

 (by melik- 

 eron No. i) 



•0538 

 .0522 

 •0558 



•0245 



.0502 



•0435 



Calories 

 (by pyran- 

 ometer 

 No. 5) 



.0648 



•0537 

 .0562 



•0237 



•0517 



.0512 



Ratio 

 melikeron 



pyranom- 

 eter 



.830 



.972 



•995 



I -033 



.972 



.850 



Both instruments were leveled, and the source placed at varying angles 

 and distances. To insure constancy of the source, storage batteries 

 were used. The results are summarized in table I. 



As would be expected, since in Melikeron No. i the rear mirror is 

 perpendicular to the honeycomb tubes, the instrument does not mea- 

 sure the full amount of radiation falling normally or nearly so, for 

 this passes through the tubes to the rear mirror and is reflected by 

 it directly back without being absorbed. For incidence greater than 



Table II 

 (February 18, 1920) 



Mean constant No. 22 with R. S. =10.78. 

 Mean constant No. 22 no R. S. = 12.18. 



