on the Affections and Properties of Light. 377 
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 farther 
light concerning that cause’s general operation : and I trust 
some experiments which 1 have planned, with an instrument 
contrived for the purpose of investigating the ratio of the bend- 
ing 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 spe- 
cula 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 co- 
lours 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 re- 
flectors do not suffer more from this cause, considering the 
almost impossibility of avoiding the hairs we speak of : how- 
ever, 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 fi- 
gure, 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 en- 
