244 Transactions of the 



is almost balanced and rendered ineffectual by the contrary attrac- 

 tion of the liquor ; but if they go out of it into a vacuum, which has 

 no attraction to balance that of the glass, the attraction of the glass 

 either bends and refracts them, or brings them back and reflects 

 them." 



In this sentence, the word " vacuum " is in italics, as if the 

 author doubted the influence of air, and both here, and subse- 

 quently, speaks more definitely of bodies in close proximity to the 

 reflecting surface " attracting " the ray of light. 



I must ask to be excused for this episode, on the plea of a desire 

 to consider and discuss any obscure point in the principle under 

 present attention. 



Addendum. 



Wishing to ascertain the distance at which objects on a total 

 reflecting surface were rendered visible by light transmitted in con- 

 sequence of their presence or contact thereon, I yesterday tried the 

 simple expedient of Newton, and pressed a lens of long radius, on 

 the back of a right-angled prism, in order to measure the distance 

 by the colour rings. These were distinctly shown at all degrees 

 below the angle of total reflexion ; but as soon as the angle exceeded 

 40^, and the bow that marks the boundary had passed, the colour 

 rings vanished, and the so-termed " black spot " alone remained, 

 showing with peculiar distinctness. This is strictly a transparent 

 circle through which all light is admitted and none reflected. 



Under all incidences within the angle of total reflexion no colour 

 appeared round the margin of the spot, except a very faint band 

 seen by polarized light ; but this was entirely due to the strain and 

 compression of the glass, and had nothing to do with light trans- 

 mission. The black spot was at its maximum size just after it had 

 passed the boundary into the field of total reflexion ; then, as the 

 obliquity of the light was increased, the diameter of the spot 

 diminished and became of less intensity, till at an incidence of about 

 5' it was almost invisible. 



From this it may be inferred that no colour efiects or decom- 

 position of light can be produced in abstracting it from a total 

 reflecting surface, for the reason that the object withdrawing that 

 light must be brought into such close contact as to be within the 

 distance of any undulation that can produce colour. This would be 

 less than the ten millionth part of an inch ; any distance beyond 

 this will not affect the total reflecting surface. Now this distance 

 is so small, that the question may be raised whether even the 

 minutest atoms in the form of diatoms can lay sufficiently close, on 

 account of their irregularities, to abstract the light, and whether 

 their visibility, and also that of insect scales, must not depend 



