in the Ultra-red Region of the Spectrum of the Sun. 245 



If in this way we should find a substance possessing the 

 desired property, and were to insert it in the path of the rays 

 in our arrangement, it would separate for us the thermal 

 spectra of the first and second orders just as the ebonite plate 

 divided the luminous from the dark spectrum. 



A second circumstance that might set a limit to the further 

 investigation of the extreme ultra-red rays after the method 

 we have chosen would be that of the lampblack on the 

 radiometer-vane not possessing the requisite absorptive power 

 for rays of so great a wave-length. If this should turn out 

 to be the case, another substance will have to be selected, 

 instead of lampblack, for the radiometric substance, possessing 

 greater sensitiveness for the rays in question. Thus it is, 

 perhaps, not impossible to advance beyond these practical 

 limits of the investigation. 



The case stands otherwise, however, with a limitation re- 

 sulting from the theory, which follows directly from the 

 formula (p. 240). Since, namely, 



\=d [sin S + sin (S + SJ, 

 and sin 8 + sin (S + o\) can only become equal to 2 at the 

 most (a limit, moreover, never to be reached in practice), X 

 can never become greater than 2d. If still greater wave- 

 lengths occurred in the sun's radiation, these rays would not 

 appear in the interference spectrum, and to verify their exist- 

 ence by our method would be impossible. Now, as the 

 greater d is made the smaller does the breadth of the 

 spectrum become, this limitation of the method is a very 

 serious one, because the accuracy of the determination essen- 

 tially depends on the breadth of the spectrum. With the 

 grating-mirror employed by us, this maximum value of \ 

 would amount to 



\ m = 0-002970, 



and therefore would be a quantity of absolutely the same 

 order as the value of \ found from observation. 



If, then, rays of so great a wave-length really do occur in 

 the sun, this method, like every other that rests on the ex- 

 amination of a diffraction-spectrum, would not be capable of 

 demonstrating them. Provisionally, however, it "will be 

 possible, by perfecting the means in the way above indicated, 

 to demonstrate by our method the existence in the spectrum 

 of the sun of rays of still greater wave-length than those found 

 by us. 



Phil. Mag. S. 5. Yol. 15. No. 94. April 1883. 



