VOL. LXXXVII.] PHILOSOPHICAL TRANSACTIONS. 20Q 



the thinnest plate makes the broadest ring, the diameter of the rings being in the 

 inverse subduplicate ratio of the plate's thickness; just so is it with fringes by 

 flexion; nearer the body the fringes are broadest, and their diameters increase in 

 the same ratio with the diameters of the rings by plates whose thickness is uniform; 

 each distance from the bending body therefore corresponds with a ring or fringe of 

 a particular breadth, and the alternate distances correspond with the dark intervals: 

 the question then is, what becomes of the light which falls on or passes at these al- 

 ternate distances? In the case of thin plates, this light is transmitted in other rings; 

 we should therefore be led to think that in the case of the light passing by bodies, 

 it should be at one distance inflected, and at another deflected; and in fact the 

 phenomena agree with this, for fringes are formed by inflexion within the shadows 

 of bodies; they are separated by dark intervals; the fringes and the intervals with- 

 out the shadow decrease in breadth according to the same law; so that the fringes 

 and intervals within the shadow correspond with the intervals and fringes without, 

 respectively. Nor will this explanation at all affect the theory formerly laid down ; 

 it will only, if found consistent with further induction, change the definite spheres 

 of inflexion and deflexion into alternate spheres. At any rate, the facts here being 

 the same with those described by Newton, but in different circumstances, teach us 

 to reconcile the difference, which we have attempted to do, as far as is consistent 

 with strictness; and what we have seen not only entitles us to conclude that the 

 cause is the same, but also inclines us to look for further light concerning that 

 cause's general operation: and I trust some experiments which I have planned, with 

 an instrument contrived for the purpose of investigating the ratio of the bending 

 power to the distances at which it acts, will finally settle this point. 



II. Another conclusion follows from the experiments now related, viz. that we 

 see the great importance of having specula for reflectors delicately polished ; not 

 only because the more dark imperfections there are on the surface, the more light is 

 lost, and the more colours are produced by flexion (these colours would be mostly 

 mixed and form white in the focus), but also because the smallest scratches or 

 hairs, being polished, produce colours by reflexion, and these, diverging irregularly 

 from the point of incidence, are never collected into a focus, but tend to confuse 

 the image. Indeed it is wonderful that reflectors do not suffer more from this 

 cause, considering the almost impossibility of avoiding the hairs we speak of: 

 however, that they do actually suffer is proved by experience. I have tried several 

 specula from reflecting telescopes, and found that though they performed very well, 

 from having a good figure, yet from the focus, when they were held in the sun's 

 light, several streaks diverged, and were never corrected ; others had the hairs so 

 small, that it was very difficult to perceive the colours produced by them, unless 

 they fell on the eye. Glass concaves were freer from these hairs, but they were 

 much more hurt by dark spots, &c. In general, the hairs are so small in well 

 wrought metals, that they do little hurt ; but when enlarged by any length of ex- 

 posure to the light and heat in solar observations, they produce irregularities round 



vol. xvm. E E 



