6 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. Jl 



the prism is slowly rotated, causing the spectrum to advance from 

 infra-red toward ultra-violet across the blackened strips. The 

 thermo-couple of the pyranometer causes the galvanometer needle 

 to deflect, and this reflects a beam from an auxiliary source on 

 to the photographic plate so that the spot deflection is at right 

 angles to the direction of motion of the plate, thus automatically 

 tracing an energy curve on the plate. 



The observations were made from about wave length 2.2/a in the 

 infra-red, to 0.33^1 in the ultra-violet. Three curves were run on 

 the plate, the first two being with the pyranometer, and the third 

 with the regular bolometer. During the first curve the spectroscope 

 slit was opened 1.37 mm., during the second 2.85 mm. and during the 

 regular holograph 0.35 mm. The time of run was 7 m., 20 s. Sectors 

 were changed at four places in the run in order to give a good 

 deflection and not allow the curve to leave the plate. The last 

 section in the ultra-violet was run with full intensity of solar beam. 



Of course it could not be expected that much definition could 

 be obtained with a slit of such width, but an examination of the 

 plate will show that the various wave lengths are recorded in about 

 their true magnitude, through practically the entire range of the. 

 spectrum where there is an appreciable amount of energy. It also 

 shows that the glass hemisphere employed transmits these waves 

 in about their true relation. By making due allowance for sector, 

 slit widths, and width and height of paper diaphragm as compare^ 

 with the bolometer and corresponding deflections for equal wave 

 lengths on the plates, it is seen that the pyranometer is approximately 

 one-fourth as sensitive as the bolometer. 



SKY RADIATION MEASUREMENTS 



Practically all pyranometer observations made at Hump Mountain 

 were taken with the instrument resting on the roof platform — 

 as heretofore mentioned. As stated before, there was approximately 

 5.1 per cent of the total area of the hemsiphere cut off by the hill to 

 the west, but as this all lies next to the horizon from whence the 

 rays strike the pyranometer strips at a very unfavorable angle, and 

 since as will be shown later, the hill itself offers radiation to help 

 offset that cut off from the sky, it is perhaps well within the general 

 accuracy of the results to ignore the hill entirely. 



