Especially interesting is the list of substances which were recognized and correctly named by both of the normal women in these experiments, but which the anosmic was unable to identify. Patrick enumerates twenty-seven such common substances. Among them, by way of example, were vanilla extract, pineapple syrup, banana, grape, quince, strawberry, tea, chocolate, sour milk, kerosene, claret, rhubarb, onion, eggs, and boiled turnips. The results suggest that these substances, although they seem to have very characteristic tastes, are actually differentiated and recognized on the basis of their olfactory rather than their gustatory or tactual qualities. One justification of this olfactory sacrifice is suggested by the fact that biologically one of the most important functions performed by smell is that of aiding in the discrimination of food. Through smell the animals perceive at a distance a substance which may offer itself as possible food. Biologically, the immediate guide to the acceptance or refusal of food is the sense of taste. In so far as smell is in part a subordinate servant in this matter, and hence becomes easily associated with such reactions as “eating” or “not eating,” no injury is produced through the occasional confusion of the two modes. We have thus reduced the rich manifold of taste to the qualities of sweet, salt, sour, and bitter,—four meager qualities as compared with the numerous unanalyzable qualities of various of the other sensory modes. We have now to show by what logic, through what technic, and on the basis of what evidence, we are compelled to grant to taste four qualities rather than two or twelve, and why the final grant consists of these particular four rather than others. _Psychological Analysis of the Taste Qualities_ As Chatin long ago observed, “The three senses,—taste, touch, and smell,—are so intimately combined that they seem to refuse to yield themselves to minute analysis.” These associations seem to be even stronger than those between the various taste qualities, of which Ladd and Woodworth have remarked, “On the whole it appears as if the four tastes were rather isolated from each other, each representing almost an independent sense. There is much blending, to be sure, but the amount is apparently no greater between one taste and another than between tastes and odors.” We may now fairly ask how these four qualities may be made to reveal their elementary and independent character, once we have eliminated the complicating factors introduced by the intrusion of qualities from other senses.