lO 



SMITHSONIAN MISCELLANEOUS COLLECTIONS 



VOL. 80 



that these definite uniform periodicities continue in solar variation, 

 we shall be encouraged to predict the radiation of the sun for years 

 in advance. If successful in such predictions, all that may hang 

 upon solar variation will become equally predictable. 



Contemplating the variation of the sun, one is inclined to ask 

 Avhether all wave lengths take part proportionally in producing it, 

 or whether, as one would naturally expect, the variation grows 

 greater and greater towards shorter wave lengths. This question we 

 answered in the latter sense several years ago, by the curves of figure 6. 

 This indicates that, in fact, the red and infra-red vary almost not at 



1.30 



1.20 



Spectral Va 

 a,b, hargua 



C, MOIMTtZUMJA 



IATI0N5 OF THL 



. Solar cha 



SHORT-; 



Hala 



Sun 

 ncl5,2.3andIj 



NTERVfAL CHANGES, ;, 



0/0 



A 5 



1.00 



J**^ 



0.^50 0^60 \\70 0-pO 1-^00 



M£^: 



W^o-a^ 



WAVE 



0.,3^ lln&thI 0.^40 



3^: 



1.50 



B.QO/l 



ZOO' 150' IOC 50' 0' 



Fig. 6. — Solar variation localized in the violet and ultra-violet. 



all ; but that the solar variation keeps increasing, and very rapidly, as 

 we go to the shorter and shorter wave lengths. With a range of 

 only 2.3 per cent in total radiation, the ultra-violet, at wave length 

 0.35 micron, shows a variation in figure 6 of about 30 per cent. It 

 would be supposed, in view of this, that if our observations should 

 be continued to the limit of the solar spectrum, at 0.29 micron, we 

 should find there, perhaps, as much as 100 per cent change. In other 

 words, if the eye were sensitive to these extremely short wave 

 lengths, it would see the sun twice as bright on some days as on 

 others. 



This expectation is confirmed by the observations of Dr. Pettit at 

 Mount Wilson Observatory. By silvering a quartz lens, which 



