J. M. Blake—Notes on Diffraction Gratings. 37 
_ If we place twq whole copies of a grating, as would occur 
In respect to parallelism if one was bent on itself, making a 
folded edge parallel to the ruled lines, the resulting bands 
with the irregularities will be symmetrical about the central 
line of the ruling. The same position will also give symmetry 
about a line at right angles to the original lines, provided 
these lines are ares of circles. This will be seen to be the 
case in fig. 8, and on counting the whole bands, above and 
below the point where the ares are in effect parallel for a short 
distance, we find nine which represent a curvature of two and 
with two gratings slightly overlapping, gave the curvature two 
] es utherfurd’s ruling gave no evidence of 
curvature in the lines when the latter were brought parallel. A 
curvature of the bands occurs in both cases, which indicates a 
gradual increase in the distance of the lines in going from one 
side of the ruling to the other. The inference from fig. 1 
above and below the center. If we consider the greater curv- 
ing of the bands in fig. 3 as compensated for by the width of 
the ruling, exceeding by one-third that represented in fig. i, 
we have yet to take into account that the lines in the former 
ruling are three times as far apart, and hence it follows that. 
