I'lIK .\I!SOI{I>IIOX AND KMISSIOX OV MR. 23 



In the spectra shown in Phxtes I-III (spectra 1 to 12) the refrangihility in- 

 creases from left to rijj^iit and witli the numbering of the strips. It begins, at the 

 least refrangible end, with a continuous spectrum of weak intensity, after which 

 come, first, some faint lines, but then numerous and remarkably strong lines. The 

 contiiuious spectrum, wliicli is a i)eculiarity of hydrogen, reaches down without in- 

 terru[>tion from here to the line 369.9 /(/< (Ames). The few lines that are found in 

 it probably belong, not to hydrogen, but to slight impurities of it, among which 

 mercurial vapor from the air-pump may well be the cliief. Some spectroscopists in- 

 cline to regard the continuous spectrum itself as a consequence of these impurities, 

 but I am unable to bring myself to this view because of many experiments with very 

 pure gases and with different tubes. It must not, however, be concealed that the 

 intensity of the continuous spectrum is connected with the purity of the filling of 

 the tube. The more scrupulously the gas within the Geissler tube is purified, the 

 weaker becomes the continuous spectrum. But I have never been able to l)ring 

 about its complete disappearance provided the exposure was long enough. 



The numbers in the table on p. 24 will sei've as an approximate measure of 

 the photogi'aphic activity of these rays on the one hand and of the sensitiveness 

 of the ultra-violet plate on the other. 



The first column refers to numbers in uiy spectroscopic journal. 



The " camera-angle " is the angle between the plane of the photogi-aphic plate 

 and the optical axis of the camera-lens. 



The lines of exposure refer to the taking of the original negatives of the four 

 parts of the plate. 



The width of the slit was altered by movement of one jaw oidy, and the effect 

 of this asymmetry is plainly visible in the photogi-aphs. 



The Values given for the current are merely the values without the coil, when 

 it was short-circuited. These, as well as the times of exposure, are merely given so 

 that differences in the different photographs should not be atti'ibnted to variation 

 either of the energy of the rays or of the width of the slit. 



The fii'st eight enlargements correspond to only four oi-iginals, as each of these 

 originals is twice as long as those of the last four. This is due to the fact that the 

 image for the first two thirds of the spectrum is much flatter tlian for the last third 

 and so permits photographs twice as long. 



The enlargements were performed with a projection-combination with an 

 aperture of 75 mm. by Carl Zeis of Jena. It is the most perfect objective for this 



purpose in the world. 



The dots and little si)ots are on the originals, being due to slight defects of the 

 coating of the plates which belonged to the method of making them that I still 

 employed in 1895, when these spectra wei-e photographed.' 



■ V. Schumann: On an improved process for making sensitive plates for the ultra-violet. 

 Annalen der F/iysik, vol. v., pp. 349-374, 190'- 



