In rats all topical anesthetics tried, whether surface active or not, induced mydriasis, whereas apostesine, a local anesthetic not applicable topically, was ineffective. Section of the ophthalmic branch of the trigeminal nerve in 3 rats and ablation of the gasserian ganglion in 2, produced complete anesthesia and moderate mydriasis in the eye on the operated side. Injection of 2% butyn solution into the gasserian ganglion in 5 rats caused 1 hr.''s moderate mydriasis and complete anesthesia of the eye on the same side. Section of the cervical sympathetic nerve in 10 rats did not interfere with the pupillary dilatation produced by instillation of hexylresorcinol, indicating that anesthesia of the cornea does not reflexly inhibit the sympathetics. Apparently pupilloconstrictor tone in the rat is maintained, partially, at least, as a reflex involving afferent components of the optic nerve. The iris control in the rat differs from that of the dog, rabbit, and man in having a mechanism the tone of which is controlled, in part, by sensory impulses arising in the cornea, conjunctiva, and periorbital tissues. The effect of surface active substances in causing mydriasis after local application is due to their anesthetic properties.