14 FREER. 



from a Zeiss ultra-violet photomicrographic apparatus. The three are mounted 

 in a line upon a prism base, and the image of the spectrum is projected upon 

 a camera placed at one side, and in the plane of the slit. The necessary motions 

 of camera, lens and grating are provided, but, unfortunately, we can not work 

 with curved plates, so that only a small portion of the spectrum is in focus at 

 one time. The whole apparatus is inclosed in a light-tight bos and secondary 

 spectra are excluded by an appropriate shield within the box. (See fig. 1.) 

 The use of the heliostat can be avoided by pointing the apparatus directly at 

 the sun. 



The spectra obtained at noon are shown by Plate I. They do not 

 probably extend beyond A=291 /i fi, and therefore not much farther than 

 has been observed by others. Measurements undertaken by Miethe and 

 Lehmann 20 in Assuan, Berlin, Zermatt, Gomergrat, and Monte Rosa 

 give practically identical numbers, namely 291.55 fifi to 291.31 fifi during 

 the latter part of August and the first part of September; these authors 

 finding, in contradistinction to the observations of Cornu, 21 that altitude 

 above the sea level makes no great difference. As one of these places 

 is at 24° 30' north latitude, where the others are in northern climates, 

 it is evident that as the extent of the ultra-violet field does not change 

 materially, the intensity factor in the solar spectrum must vary to a great 

 extent in different places. However, it is possible that a considerable 

 range of ultra-violet is absent at present (March 1) from our sunlight. 

 Probably this area will increase as the angle of the sun diminishes and 

 as the season advances; and it may reach a maximum in April, although 

 these recent results would seem to indicate that even here we will not 

 get below 288 fi.fi. 



Enough has already, been demonstrated to prove that the emphasis 

 placed upon the ultra-violet spectrum in the preceding part of this paper 

 is justified. 



GENERAL CONSIDERATIONS. 



If we consider the equator as being in one plane and the poles in 

 another, normal to the incident rays of the sun, it is obvious that the 

 relative distance between these two points is practically infinitely small 

 as compared with the distance of the sun from the earth. If we regard 

 this phase alone, obviously the insolation at the equator and the poles 

 would approximately be equal. However, these two positions are on 

 different points of a sphere, which is surrounded by the atmosphere, so 

 that not only is the angle of incidence of light rays different in different 

 regions, but also, as they pass from the ether into the atmosphere at 

 different angles, they suffer refraction in differing degree. The shorter 

 wave lengths, in this respect, are influenced to a greater extent than the 

 longer, and, under proper conditions, are even finally totally absorbed or 

 reflected, so that none would reach the earth. This condition would 



20 Sitzungsler. d. k. Pr. Akad. d. Wissensch. (1909), 8, 268. 

 - l Compt. rend. Acad. Sci. (1879), 88, 1107; ibid 89, SOS. 



