434 Major L. Darwin. On the Method of 



difference of intensity of illumination of the centre of the line and 

 the field is the minimum difference of shade discernible by the eye, 

 and this will be independent of the actual intensity of the field, and 

 will not vary much with different observers. But it has been shown 

 that the thickness of the line does vary with the defining power of 

 the lens, and it may therefore be concluded that the test adopted 

 at Kew is not open to serious objections on theoretical grounds. 



In the foregoing discussion, it has, however, been assumed that the 

 curve representing the image of the edge of a surface is such as that 

 which Helmholtz has shown .to be produced as an ocular effect by the 

 circles of diffusion being due to want of accommodation of the eye 

 itself :* it will be observed that no part of the curve is tangential 

 to the vertical. If, however, the curve is similar to that given by the 



FIG. 12. 



FIG. 13. 



same author as being due to dispersion in the eye, and illustrated in 

 fig. 12, it will be seen that the result of gradually diminishing the 

 thickness of a line will not be exactly as above described ; for, how- 

 ever thin the dark line on the bright ground becomes, the intensity 

 of illumination at its centre can never be more than twice ch and if 

 the ratio of twice ch to cd is less than a given ratio, the image of the 

 black line will remain visible until it is so thin that the eye cannot 

 perceive it. Therefore it might come about that two lenses giving 

 images of the edges of surfaces as different as flhf and nlhn', as 

 shown in fig. 13, might give equally good results under the Kew test 

 for definition, because in both cases the limit of visibility would be 

 due to the minimum size of the line visible by the eye, and would 

 have nothing to do with the definition of the lens. Helmholtz 

 remarks on the very little evil effect of a diffusion represented by the 

 curve shown in fig. 12, since the true edge is always visible. Hence 

 we may assume that the Kew method still gives in such cases a good 

 practical test for definition, though it does not test the amount of 

 lisperaed light over the image of fine lines, or, as a photographer 



' ' Optique Physiologique,' Helmholtz, Paris, 1867, p. 185. 



