Examination of Photographic Lenses at Kew. 423 



focal length, thus introducing an error equal to SNi in the result ; 

 the focal length given is, in factj the distance from the principal 

 focus to the principal centre of similitude for the part of the plate 

 where the observation is made. But here again, since BN" 2 C, the 

 cone of incident rays, is spread over a disc on the plate of which 

 B"C" is the diameter (and not B'C'), the mean intensity of illumina- 

 tion of the plate between these points will vary inversely as (SF) 2 ; 

 and, if the plate covers an angle larger than BN 2 C, the C.I. numbering 

 of the stops will give a better indication of the relative exposure on 

 the assumption that SF is the principal focal length than if the true 

 value NjF is introduced into the calculations. Thus, what has been 

 given in the Certificate of Examination will again be nearer what is 

 practically required by the photographer than if the true principal 

 focal length has been recorded. If, however, the lens is intended to 

 be used for enlargements or reductions, and the final adjustment of 

 the distance of the object is to be made by reference to tables, then, 

 no doubt, the true principal focal length must be accurately given ; 

 but no photographer would ever use a lens showing sensible distor- 

 tion within 15 of the axis, for such purposes, for, if he did, the ratio 

 of the enlargement or reduction would vary sensibly in different 

 parts of his plate ; and, if there is no distortion within this distance 

 from the axis, S and NI will be coincident, and the Kew method will 

 give accurately and truly N\F as the principal focal length. Thus, in 

 th% only circumstances under which the principal focal length is 

 practically wanted with theoretical truth and great accuracy, it is 

 seen that^the results given in the Kew certificate do answer these 

 requirements. 



The third condition that has been laid down as being necessary 

 before the Kew method gives theoretically correct results is that the 

 nodal point should be the same for white light as for photographi- 

 cally actinic rays. This may be hypercritical, but if, in fig. 5, C" 

 and B" represent the images as seen on the photographic plate, C' 

 and B' those seen by the eye on the ground glass, ~Ni the mean posi- 

 tion of the nodal point of emergence for visible rays, and S the 

 mean position for actinic rays, then it is evident that FNx will be the 

 principal focal length found by the observation, whereas SF will be 

 the quantity required in calculations with regard to enlargements or 

 illumination. If the lens gives any distortion, NI would represent 

 the centre of similitude for visible rays and S that for photographi- 

 cally actinic rays ; the condition might, therefore, have been more 

 rigidly defined by stating that the point of similitude for visible rays 

 and that for actinic rays must occupy identical positions for parts of 

 the field between the points of observation. As far as can be judged, 

 this is a negligible source of error in all cases. 



A fairly large angular movement of the swinging beam, about 



2 F 2 



