Light in a Diffraction Grating Spectrum. 401 



It is useless to attempt to fully explain the very complicated 

 sequence o£ events winch I have outlined, until some working- 

 hypothesis is established which will explain some one of them, 

 and it appears to me that the first thing to do is to make some 

 assumption which will explain the very remarkable fact that 

 a change of wave-length of one part in a thousand is sufficient 

 to change the illumination from a maximum to a minimum. 



We know that this can take place if we are dealing with 

 interference with a large difference of path. Hamy's 

 " extincteur "* is a piece of apparatus which illustrates this 

 better than anything with which I am familiar. It occurred 

 to me that possibly the anomalies were to be referred in 

 some way to the interference between disturbances coming 

 from widely separated lines, though I had no very definite 

 idea as to just how it could produce any of the anomalies, or 

 how it was to be connected with the polarization effect. It 

 seemed worth while, however, to investigate the matter, and 

 I accordingly covered the grating with a thin sheet of black 

 paper, leaving exposed only a strip about 0*3 mm. wide along 

 one edge. By bringing the eye close up to this small strip 

 the spectrum could be distinctly seen, but the sharpness of 

 the dark bands seemed to be undiminished. As there were 

 only about 200 lines acting there could not have been any 

 very considerable difference of path between even the 

 extreme rays. In consequence of this I am compelled to 

 refer the matter to the form of the groove. The important 

 fact which must be taken into account in any endeavour to 

 explain the action of the grating is, that the anomalies only 

 occur when the electric vector lies across the ridge. We can 

 speculate about the action of the narrow ridges on the light 

 waves, assuming, perhaps, something of the nature of reso- 

 nance taking place across the ridge, or we can seek for 

 the explanation in the behaviour of the transverse vibrations 

 in between the ridges, but in any case we are confronted 

 with the difficulty of explaining the tremendous change in 

 the intensity of the illumination with the exceedingly small 

 change of wave-length. 



The study of this grating has been limited to the two 

 or three days immediately preceding the closing of the 

 laboratory for the summer, consequently I have been unable 

 to give a very exhaustive account of its behaviour under 

 other conditions, or secure any very satisfactory photographs 

 of the peculiar spectra. The few photographs which I have 

 taken and which are reproduced, were made on some old ortho- 

 chromatic plates, without any especial appliances, the plates 

 * M. Haniy, Compt. Rend. cxxv. p. 1092 ,(1897). 



