THE ABSORPTION AND KMISSTON OF AIK. l-'O 



electrodes, taking for the cathode (which with all the other photographs had been 

 at the distant end of the tube) the electrode nearest the slit, that I got any better 

 results. It appears, therefore (and tbis is confirmed by all repetitions) that the 

 cathode-light of hydi'ogeu, under existing circumstances, contains more refrangible 

 and active rays than the anode-light. 



Why do my hydrogen photogi'aphs end at this place? For all essays to 

 photograph more refrangible rays with a 00° prism came to naught. I can find the 

 reason for it neither in the want of energy of the rays nor in the insufficient sensi- 

 tiveness of the ultra-violet plate. Nov can I attribute it to metallic reflection, since 

 this concerns decidedly shorter waves than these with which we have here to 

 do. I suspect i-ather that this limitation of the effect of my spectroscope is due to 

 the approach to grazing incidence of the rays upon the prism. Foi- I met with 

 a similar reduction of energy with my former 70° pi'ism ; yet afterwards when I 

 came to repeat the photographs with a 60° prism I found the spectrum considerably 

 strengthened. It is therefore to be expected that a still further diminution of the 

 refracting angle will lead to the photographing of still shortei- rays of hydi-ogen. 



OF THE WAVE-LENGTHS BEYOND 185 yW/'. 



Information as to the wave-lengths of the new region would be of particular 

 interest. The difficulties of measuring them are, however, so great that neither I 

 nor others have succeeded in overcomins; them. The wave-lens;th 162 /'/' to which 

 I arrived in 1892 ' was but the i-esiilt of a fii'st essay ; and the appai'atus I then had 

 was so far from perfect that this detei'mination is hardly sufficient. Last yeai', Mr. 

 F. F. Martens ^ took up the question, and resting upon the Ketteler-Helmholtz 

 dispersion-formula deduced from the angle of 21°, which is the difference in devia- 

 tion between the middle of ray most refrangible field and the last aluminum-line 

 at 185.4 f/jj, calculated the wave-length of that place in my last photograph to be 

 125 /</<, but this should pi'obably be 2 /'/< higher. It is inaccurate to say, as Mr. 

 Martens does, that this is the wave-length of the most refrangible ray photogi'aphed 

 by me ; since the deviation of 21° refei's to the middle of spectral field in question, 

 and the most refrangible line is deviated about 6° more ; which should have been 

 taken into account in the computation. My most refracted line has therefore a 

 wave-length decidedly less than 127 w- It must, however, be confessed that my 

 estimate of 1893 ^ of 100 /'/' was too hasty ; since it appears that even now I have 

 not got down quite to that point. 



OF THE VACUUM APPAKATUS FOU THE MEASUREMENT OF WAVE-LENGTHS. 



If, in Nos. 1 and 9, Plates I-III, of the spectrogi'aph, the prism be replaced by a 

 plane reflecting grating and the collimator and camera by two air-tight tubes arranged 



' Sitzutigsberichle d. kaiser/. Akademic d. Wissensch. in IVicn ; math.-natiinv. Classe, vol. cii, 

 Part I la, pp. 625-694, 1893. 



' Anna/en der Physik, 1901, Heft II, p. 619. 



' Sitziingsber. d. kaiserl. Akad. d. IVissensc/i. in IVieii ; maih.-natiii-iv. Clnsse, vol. cii, Part Ila, (i. 



677. 1893- 



