glass of a camera obscura, receives the images of visible objects. This membrane is an expansion of the optic nerve O, or A in Fig. 5, A, which passes to the brain, and, by a process of which we are ignorant, gives us vision of the objects whose images are formed on its expanded surface. The globular form of the eye is maintained by two fluids which fill it,—the _aqueous humour_, which lies between the crystalline lens and the cornea, and the _vitreous humour_, ZZ, which fills the back of the eye. [Illustration: FIG. 5, A.] [Illustration: FIG. 5, B.] But though we are ignorant of the manner in which the mind takes cognizance through the brain of the images on the retina, and may probably never know it, we can determine experimentally the laws by which we obtain, through their images on the retina, a knowledge of the direction, the position, and the form of external objects. If the eye MN consisted only of a hollow ball with a small aperture H, an inverted image, _ab_, of any external object AB would be formed on the retina _r_O_r_, exactly as in Fig. 4. A ray of light from A passing through H would strike the retina at _a_, and one from B would strike the retina at _b_. If the hole H is very small the inverted image _ab_ would be very distinct, but very obscure. If the hole were the size of the pupil the image would be sufficiently luminous, but very indistinct. To remedy this the crystalline lens is placed behind the pupil, and gives distinctness to the image _ab_ formed in its focus. The image, however, still remains inverted, a ray from the _upper_ part A of the object necessarily falling on the _lower_ part _a_ of the retina, and a ray from the _lower_ part B of the object upon the upper part _b_ of the retina. Now, it has been proved by accurate experiments that in whatever direction a ray AH_a_ falls upon the retina, it gives us the vision of the point A from which it proceeds, or causes us to see that point, in a direction perpendicular to the retina at _a_, the point on which it falls. It has also been proved that the human eye is nearly spherical, and that a line drawn perpendicular to the retina from any point _a_ of the image _ab_ will very nearly pass through the corresponding point A of the object AB,[31] so that the point A is, in virtue of this law, which is called the _Law of visible direction_, seen in nearly its true direction. [31] _Edinburgh Transactions_, vol. xv. p. 349, 1843; or _Philosophical Magazine_, vol. xxv. pp. 356, 439, May and June 1844.