and Energy Distribution of Diffraction Gratings. 541 



§5. 



There are several methods of determining the shape of* the 

 groove in the gratings employed. If the main plane of the 

 groove be a perfectly plane mirror the grating, mounted in 

 a system with fixed collimators, may be turned through the 

 angle between the principal plane of the groove and the un- 

 disturbed surface ; this will bring an image reflected from 

 the plane of the groove into the observing telescope, and the 

 angle is easily read. This image may be designated as the 

 " oblique " image. However, if the surface of the groove is 

 a poor reflector for light the definition is liable to be so bad 

 that the above described method will not give consistent 

 results. In place of the observing telescope a direct-vision 

 spectroscope may be used, and the position of the oblique 

 image determined by setting so that the brightest and most 

 complete spectrum is seen; this of course postulates a similar 

 behaviour of the grating in the visible region of the spectrum 

 to that observed in the infra-red. A third method consists 

 in making a replica from the grating, using collodion stained 

 with a very dark colour ; these replicas are embedded in a 

 block of wax and sectioned through, in planes at right angles 

 to the direction of the grooves, by a microtome. These sec- 

 tions may be mounted on slides, and the shape of the groove 

 may be determined visually with a microscope, or from a 

 microphotograph of the section. 



Optically, all the gratings used gave very unsatisfactory 

 definition of the oblique image; the replica process was 

 resorted to in order to find out approximately the shape of 

 the grooves*. This brought out the fact that all the gratings 

 with the exception of the Wood No. 7 had grooves the sides 

 of whose profile were irregularly curved, but the AVood 

 grating No. 7 showed as a groove profile a triangular shape: 

 the sides of this triangle being fairly straight lines. Since 

 this grating gives the simplest form of groove, it will be 

 used in all calculations in which the geometrical properties 

 of the groove are involved. Unfortunately, due to the 

 shrinkage of the collodion film, and the fact that the section 

 was not taken exactly perpendicular to the direction of the 

 groove, the value of the angle taken from the micro- 

 photograph may not be relied upon ; however, the ratio 

 of the projections of the two slanting sides of the groove 

 upon the undisturbed surface is a constant quantity, 

 independent of the angle at which the section is taken : 



* The authors are indebted to Prof. C. F. W. McClure and Mr. Otto 

 Kampmeier, of the department of Biology of the Unrver&itv, for tailing 

 and mounting the sections and making microphotographs. 



