respecting the Inflection of Light, 413 



periments recorded by Newton in the place referred to, it is 

 certainly most unfair and injurious to his memory to quote 

 them without bearing in mind his strong and peculiar remark 

 respecting them (at the end of observation 11.): " When I 

 made the foregoing observations, I designed to repeat most of 

 them with more care and exactness, and to make some new 

 ones for determining the manner how the rays of light are 

 bent in their passage by bodies for making the fringes of co- 

 lours with the dark lines between them. But I was then in- 

 terrupted, and cannot now think of taking these things into 

 further consideration ; and since I have not finished this part 

 of my design, I shall conclude with proposing only some 

 queries, in order to a further search to be made by others." 



And again, in the " Advertisement" prefixed to the book : 

 " The subject of the third book I have also left imperfect, 

 not having tried all the experiments which I intended when 

 I was about these matters, nor repeated some of those which 

 I did try until I had satisfied myself about all their circum- 

 stances. To communicate what I have tried, and leave the 

 rest to others for further inquiry, is all my design in publish- 

 ing these papers." 



These remarks of the illustrious author will speak for them- 

 selves; and it is evident he would be the last to urge his con- 

 fessedly imperfect trials in opposition to decisive results. But 

 even here I have, I think, sufficiently shown in my former 

 paper, how very small his inaccuracies were; and the only case 

 in which any real contradiction appears, is one, as I observed 

 above, in which it is almost certain that we do not accurately 

 know all the conditions. 



Next, with regard to curvilinear edges, I would observe 

 that no comparison can be fairly drawn between any experi- 

 ments with straight edges and those with curved. Mr. Bar- 

 ton, in adopting the latter, has chosen a method which in cal- 

 culation would involve extreme complexity, and it is a case to 

 which the formula in question does not apply. In the case of 

 the rectilinear parallel edges, an important simplification is 

 afforded, as we have only to calculate the effects in one plane, 

 viz. in that perpendicular to the length of the slit and to the 

 plane of the edges. Whereas with curved edges we must com- 

 bine with this the effect in the plane of the length of the aper- 

 ture. To make a fair application, then, of the theory, we ought 

 to follow out the calculation, and modify the formula?, so as to 

 include this case. . 



This, it will be readily acknowledged, by any one acquainted 

 with the nature of the formula?, will be a difficult investiga- 

 tion ; nevertheless it is essential to go through it before we 



