674 PROCEEDINGS OF THE AMERICAN ACADEMY. 



lamp, although not negligible under extreme conditions, have been 

 very much exaggerated. These observations in fact quite confirm the 

 recent statements of Birch- Hirschfeld ^ on this subject. The ordinary- 

 care which one exercises in keeping intensely brilliant lights out of 

 the eye at short range seems also sufficient to avert any trouble from 

 the specific effects of extreme ultra-violet radiations. 



The accompanying plates, all taken with a rather wide slit, about 

 .2 mm., summarize the results obtained. 



Figure 1, Plate 1, shows the ultra-violet spectrum of the quartz lamp 

 from about w. 1. 410 /x/x to about w. 1. 310 /x/x, as shown by the approxi- 

 mate wave-length scale above. It will be noted that, save for the 

 brilliant triplet near w. 1. 365 /x/t, the region of the ultra-violet, down to 

 the double at about w. 1. 313, has very few strong lines, and the region 

 between the two groups mentioned is especially barren. There is in 

 fact much less total strength apparent in the part of the quartz-lamp 

 spectrum shown on these plates, than appears in the spectra of the sun 

 and of ordinary electric arcs, the strong mercury lines being few. It 

 is only in the extreme ultra-violet beyond the range shown in Plate 1, 

 that the mercury spectrum is particularly brilliant. 



Figure 2, Plate 1, shows the same spectrum when cut down by a 

 polished plate of A.O. crown glass 6.2 mm. thick. The weakening of 

 the further ultra-violet is here very marked, especially in case of the 

 doublet at w. 1. 313 fifi. The absorption near the other end of the strip 

 is comparatively small and, in fact, the glass may be said to be moder- 

 ately transparent to near w. 1. 313 /xft. No lines of the mercury spec- 

 trum of less w. 1. than this appear on this plate or on any other of the 

 plates reproduced herewith. In other words, all the glasses tested and 

 shown in these spectrograms cut off completely all radiation of w. 1. 

 less than that stated. The A.O. crown is one of the glasses commonly 

 used for spectacle lenses in American practice. 



Figure 3, Plate 1, is the quartz arc spectrum as reduced by a medium 

 tint of ordinary commercial " smoke " glass 5.6 mm. thick. In this 

 case everything of less w. 1. than about 360 mm is completely obliterated. 

 The strong group near w^. 1. 365 fifi gets through with a fair degree of 

 brilliancy, and a similar moderate absorption extends into the violet 

 and elsewhere into the visible spectrum. 



Figure 4, Plate 1, is the spectrum as reduced by a plate of B. & L. 

 crown 4.2 mm. thick, a glass also much used for spectacles and com- 

 parable with Figure 2. It has, however, a slightly higher refractive 

 index and shows more absorption in the extreme lines and not quite as 



* XI Congrcsso Intcrnazionalc di Oftalmologia, p. 623. 



