692 Profs. H. Rubens and 0. von Baeyer on Extremely 



the evidently much longer-waved radiation of the mercury 

 vapour, The extremely high transmission of quartz is of 

 particular interest for these kinds of radiation. On calcu- 

 lating the coefficient of absorption 



from the transmission for the 41* 7 mm. thick quartz plate, 

 cut perpendicularly to the axis (d being the thickness of the 

 plate in mm,, D r ' the transmission, corrected on account of 

 the loss by reflexion), we obtain for the here investigated 

 radiations the following values of q r : — 



g 1= =(M)44; ? 2 = 0-026; ^ = 0*0089 ; g 4 = O0057. 



It is evident that the rays of the mercury-lamp filtered by 

 black cardboard must penetrate about 8 times as thick a 

 quartz-layer as the rays emerging from the Welsbach mantle, 

 before being attenuated to the same fraction of their primary 

 intensity. Similar circumstances prevail with amorphous 

 quartz, but here the absorption-power for the four investigated 

 radiations is about 20 times as great as at the natural 

 modification. 



Glass and mica seem, like fluorite, rock-salt, and sylvine, 

 to belong to the substances whose main region of absorption 

 lies among the wave-lengths below 100 /x. The high trans- 

 mission of paraffin, hard rubber, and amber, well-known as 

 good isolators, is not surprising ; neither is the small absorp- 

 tion of the elements diamond and selenium. 



Water shows a far smaller absorption power for the radia- 

 tion emitted by the mercury-lamp (particularly after its 

 filtration through quartz or black cardboard) than for the 

 rays emanating from the Welsbach mantle. The reflexion 

 from the water surfaces can also not be considerable, as the 

 values of the coefficient of absorption calculated from both 

 layers of different thickness agree satisfactorily without con- 

 sideration of the reflecting power. This would not be so in 

 case of a considerable loss by reflexion. We may, therefore, 

 assume that in these spectral regions water still possesses a 

 refractive index of small magnitude, lying far closer to the 

 value observed in the visible spectrum than to the square- 

 root of the dielectric constant for slow vibrations. 



As our measurements of absorption cannot give quan- 

 titative determinations as to the average wave-length of 

 the investigated radiations, we have attempted to measure 

 the wave-length by aid of the previously employed inter- 

 ferometer*. The interference-curves obtained with the 



* PI. Rulbens & II. Hollnagel, Phil. Mag. [6] xix. p. 761 (1910). 



