﻿312 Prof. R. W. Wood on Diffraction 



informed me that lie bad found a diamond point which ruled 

 gratings showing anomalies similar to those which I described 

 in the earlier paper, and he has ruled me a number of grat- 

 ings upon a thin silver surface deposited upon a speculum 

 plate. The lines are ruled at the rate of 10,000 to the inch 

 on the so-called 20,000 machine. The first order spectrum 

 on one side is very bright, while that on the other is quite 

 faint, and it is in this faint spectrum that the abnormalities 

 are found. This grating appears to be of such good quality 

 that it seems probable that excellent optical gratings of very 

 large size can be made in this way. We can choose our 

 metal according to the region in which Ave wish to work. 

 Gold, for example, would be far better than speculum metal 

 for the region between 5500 aud 8000, for gold reflects 90 

 per cent, in the red, while speculum metal reflects only 

 65 per cent., a matter of some importance for example in 

 the study of the resonance spectra of iodine, which are not 

 very bright and fall exactly within this region. 



Gratings made in this way would have to be handled very 

 carefully however, for I have found that the lightest possible 

 rubbing with soft chamois skin alters the distribution of the 

 light in the spectra in a most remarkable manner. It is 

 possible iu this way to raise the intensity of the faint first 

 order spectrum four- or fivefold and cause the dark bands 

 to disappear entirely. It is quite interesting that the 

 groove form can be thus "figured" after the grating is 

 ruled, though I doubt whether the efficiency of a grating 

 could ever be improved in this way, since the tendency 

 seems to be always to give a more uniform distribution 

 of light among the specira, which is precisely what we do 

 not want. This " figuring " of the groove has, however 

 proved of value in the study of the bright and dark bands 

 as they can be developed in intensity in this way, and caused 

 to change their position in the spectrum. 



While the present study has not completely solved the 

 question of the cause of the remarkable anomalies, many 

 additional experimental data have been obtained, and it is 

 possible that the results may point to the modifications 

 necessary to make Lord Rayleigh's theory cover the 

 ground. 



I shall now describe in detail the observations made with 

 one of the gratings. The groove form is probably quite 

 similar to that used in the case of the echelette gratings, one 

 side very steep, the other making only a small angle with 

 the original surface. This is similar to that of the grating- 

 described in 1902, from which I made a cast at the time 



