82 SMITHSONIAN MISCELLANEOUS. COLLECTIONS [vOL. 45 



determined by an inspection of the rate of decrease of successive cor- 

 rected areas approaching the Hmits of the curves, and the correc- 

 tions were checked both b}- computing according to Wien's formula 

 the probable form of the solar energy curve corresponding to the 

 assumed solar temperature of 6000°, and by examination of the 

 normal energy curves outside the atmosphere as computed from 

 holographs and given in plate xxii. 



I have thought it worth while to give in addition to the general 

 mean, the means also of observations before and after March 26, 

 when, for some unexplained reason, a fall of about 10 percent was 

 noted in the computed solar constant. The observations of February 

 19,^ March 25, March 26, and April 29, 1903, appear to be entitled 

 to the greatest weight among those given, on account of the regularity 

 of the actinometric curves of those days and the closeness with which 

 the plotted points for determining the atmospheric transmission 

 coefificients lie upon straight lines, as shown for two of the days in 

 question on plate xx. Since May i it has been almost impossible to 

 get sufficient observations for computing a solar constant owing to 

 cloudiness, but interest attaches to further determinations and these 

 are to be made when practicable. 



FORM OF THE NORMAL SOLAR ENERGY SPECTRUM OUTSIDE THE 

 earth's ATMOSPHERE AND THE PROBABLE TEM- 

 PERATURE OF THE SUN 



The reader has no doubt noted that, by applying corrections for 

 atmospheric and instrumental absorption, the holographic spectrum 

 energy curves may be reduced in form as well as in area to represent 

 the distribution of energy in the spectrum" of the solar beam outside 

 the atmosphere. This has been done in several instances, and in 

 doing so the curves have been transformed from the prismatic to 

 the normal wave-length scale by taking account of the prismatic 

 dispersion, and several of these curves are platted in plate xxii. No 

 account is taken in the curves, shown in plate xxii, of selective 

 absorption bands whether solar or terrestrial, smoothed curves only 

 being given. 



It will be noted that there is a fair agreement in general form 

 between these independently derived curves, and that they unite in 



^ February 19, 1903, was the most extraordinary day as regards absence of 

 water vapor in the atmosphere which has ever been noted here. The great 

 water-vapor bands oi'il in the infra-red spectrum were nearly filled up, and 

 the long wave-length side of the band S2 presented an almost unrecognizable 

 appearance. 



