42 Transactions of the Society. 



Employing three thermometer-balls, each 0*24 in. diam., and 

 their centres 0*3 in. apart, the separation of images of the sun 

 reflected from them could not be detected quite so far, and the 

 vanishing-distance when the reflection was from a cloud or a lamp 

 was much shorter. A fourth ball, 0*14 in. diam., being added, 

 with its centre 0*3 in. from that next it, its image was seen 

 separated only as far as the rest ; and with two balls only an indent 

 in the spot representing the two images gave an indication of 

 separation at nearly as great a distance as the two stripes before 

 mentioned. 



When the spots were best defined they had around each a well- 

 marked black circle, which will be presently reverted to ; their 

 finest appearance was through the a]3erture 0*02 in., and the next 

 through 0-017 in. 



In the above experiments, in which the picture is formed at 

 once upon the retina, we see that no details are distinguished 

 beyond a certain degree of minuteness, however large the opening 

 of the pupil, and this from the constitution of the eye. But when 

 with the telescope or the compound Microscope we view the image 

 formed by the pencils of rays inflected at its aperture, we may 

 magnify this image sufficiently to show us the full distinctness 

 belonging to the aperture, except as it may be impaired by the 

 glasses or speculum. 



For trials with the telescope it was found requisite that the 

 weather and light should be favourable, and the atmosphere free 

 from undulation ; and from these and other impediments the 

 observations were not numerous, but as far as they went they 

 were satisfactory. 



A good 5 ft. telescope by Dollond was chiefly used, to which 

 caps of different apertures were applied. And as at the distance 

 of 400 feet (the greatest that could conveniently be got in my 

 garden) the patterns when set up were too large to vanish with 

 the larger apertures, a convex mirror was also fixed so as to give- 

 diminished images of them ; the sizes of these images were care- 

 fully estimated, and they were found to be seen as distinctly in 

 proportion to their magnitude as their originals. 



A very good superior reflector by Watson, its speculum 4 in. 

 diam. and 11 1 in. focus, was at least equal in its defining-power 

 on the patterns to the 5 ft. Dollond with its whole aperture 

 of 3-8. 



Owing to deficient arrangement in making these observations, 

 which are a few selected from others of the same character but 

 less definite, most of them do not give the precise vanishing dis- 

 tances, yet their general result exhibits the same law that obtains 

 in vision by the naked eye through apertures of 0*02 in. and 

 smaller ones ; and the experiments to be detailed with the Micro- 

 scope will show it also to prevail with intermediate apertures. 



