The idea suggested in this paper, occurred to the author three years ago, from reflecting that the destruction of spherical aberration in the eye might be effected by a modification of the curvature of the lens, as well as by the variable density which it is known to possess, and which has usually been accounted for as intended for that purpose.

The author considering the probability to be almost infinite against the sphericity of the surfaces (a necessary evil in our instruments, but inexplicable in a natural organ), a conviction which he afterwards found to be reduced to certainty by experiments which have actually been made on the figure of the lens—he conceived that the variable density of this part of the eye must have some other cause.