But that which is seen without a limiting outline may be thus affected without our perceiving the change. On the other hand, if what we look at has a visible termination, we have an evident indication that a displacement occurs. If, therefore, we wish to ascertain the relation or degree of such a displacement, we must chiefly confine ourselves to the alteration of surfaces with visible boundaries; in other words, to the displacement of circumscribed objects.
192.
The general effect may take place through parallel mediums, for every parallel medium displaces the object by bringing it perpendicularly towards the eye. The apparent change of position is, however, more observable through mediums that are not parallel.
193.
These latter may be perfectly spherical, or may be employed in the form of convex or concave lenses. We shall make use of all these as occasion may require in our experiments. But as they not only displace the object from its position, but alter it in various ways, we shall, in most cases, prefer employing mediums with surfaces, not, indeed, parallel with reference to each other, but still altogether plane, namely, prisms. These have a triangle for their base, and may, it is true, be considered as portions of a lens, but they are particularly available for our experiments, inasmuch as they very perceptibly displace the object from its position, without producing a remarkable distortion.
194.
And now, in order to conduct our observations with as much exactness as possible, and to avoid all confusion and ambiguity, we confine ourselves at first to