and Energy Distribution of Diffraction Gratings. 539 



the variation of the percentage of the energy in the first 

 order with increasing wave-length. It will be noted that 

 the curve goes through a, minimum at wave-length 2 fi. 

 This is due either to the absorption band at 2*1 //, affecting 

 the first order rather more than either of the others, or, as is 

 more likely, the order in this position is a weak order on 

 account of a destructive diffraction due to the shape of the 

 groove ; this effect will be fully considered in § 6. As will 

 be seen later, the shape of the groove is not accurately 

 known for this grating, but it is evident that the distribution 

 of the energy from it is what might be expected from a 

 grating that would naturally reflect energy from one side 

 of the groove in the direction 20°-25° out from the central 



§4. 



The wave-lengths accompanying the various curves are 

 only approximate, as there was a lost motion in the vacuum 

 spectrometer which was not detected until the experimental 

 work had been completed. This point is not vital, however, 

 as the intention of the authors was not to make accurate 

 wave-length determinations, but to take serial readings and 

 determine the quantity of energy thrown off with the grating 

 in the different positions. 



By referring to the curve for the energy from the INernst 

 lamp after it has undergone reflexion from a gold mirror 

 (fig. 3, p. 540), it will be seen that the maximum of the 

 energy distribution for a given wave-length will be shifted 

 in angular reading if the wave-length correspond to a point 

 on either of the steep portions of the curve. Paschen * has 

 worked out this correction, and it was applied to the curves of 

 the Wood grating No. 7. The true location of the maximum 

 is shown by the dotted curve slightly displaced from the 

 observed maximum, this displacement being to greater or 

 less wave-length, according as 



\>2'6yu, or \<2'6 fi. 



This correction is of importance only in one respect : it 

 enables the exact wave-length to be calculated from the true 

 position of the maximum of a given order and the constant 

 of the grating ; this accurate value of the wave-length is 

 essential when the questions of retardation over a single 



* Wied. Ann. vol. li, p. 8 (1894). 



