246 Dr. Miiller on the Thermal Effects 



scale as fig. 1, p. 242. From the points H, G, F, D and B, 

 which correspond to the Fraimhofer's hnes occurring at these 

 points, perpendiculai's are drawn whose lengths are propor- 

 tional to the uudular lengths. This proportion is such that a 

 difference of 5 millims. in height corresponds to a difference of 

 00001 millim. in the undular length. The curve a b, drawn 

 through the extremities of the perpendiculars H, G, F, D and B, 

 represents, then, the law connecting the undular length with the 

 refractive index in crown-glass. 



The curve o 6 is continued beyond b in such a manner that 

 the course of the continued portion b c joins in as continuous a 

 manner as possible the portion a b which has been derived from 

 observations ; in other words, the law, according to which the 

 undular length and refractive index are connected in the visible 

 portion of the spectrum, is preserved in the graphic continuation 

 of the curve into the ultra red dark thermal rays. 



If, now, at S, that is, at the point which corresponds with the 

 extreme limits of the dark thermal rays of the solar spectrum, a 

 perpendicular be erected, this will cut the curve in a point C, 

 whose height above the axis of abscissae R S corresponds to an 

 undular length 0-0019 ; that is, the undular length of the extreme 

 dark thermal rays of the solar spectrum is 



0-0019 millimetre. 



This value cannot of course be considered as an exact, but only 

 as an approximate one, because the extension of any empirical 

 law beyond the limits of the observation s upon which it is based 

 can never lay claim to any great degree of exactitude, whether 

 such extension be performed algebraically, or, as in this case, 

 graphically. 



The mean of the two values, the one 0*00177, obtained by 

 calculation, the other 00019, graphically gives us 



m; = 000183 millim. 



for the undular length of the extreme dark rays of the solar 

 spectrum. 



The undular length of the extremest fluorescent rays derived 

 from sunlight, is, according to Esselbach*, O'OOOS millim. 

 The next lower octave to these rays, which ai'e the most refran- 

 gible of all, gives the undular length 00006 millim., which, as 

 is seen from fig. 2, nearly corresponds to the Fraunhofer's band 

 D in the orange. 



The second lower octave of the most refrangible rays, with the 

 undular length 0*0012 millim., falls in the midst of the dark 

 thermal rays of the solar spectrum. The third lower octave, with 



* See my ' Physics,' 5th ed. vol. i. p. 698 ; also Pogg. Ann. xcviii. p. 513. 



