Driesch’s analysis of the constructive instinct shows that these facts of adaptation or regulation fit in with the idea of sensory control rather than with that of a chain-reflex. In the supposition that the successive stages of instinctive construction are due to a chain-reflex, consisting of a series of elementary motor reactions _a_, _b_, _c_, etc., in which _a_ produces the external work A and, on terminating, releases _b_, which, in turn, produces external work B and releases _c_, etc., clearly _b_ could never appear before _a_, and the sight of A ready-made would not inhibit _a_, nor would the removal of A defer the advent of _b_. In other words, regulation would be impossible. But, if we suppose that not the elemental act _a_, but rather the sensory perception of A, the first state of the external construction, is the stimulus to _b_ and, consequently, to the production of the second state of construction B, then we understand why _b_ is released independently of _a_, when, for example, an insect discovers a ready-made substitute for A, the initial step in its construction, and we also understand why, in cases of accidental damage resulting in the total or partial removal of A, the reaction _b_ is deferred and the reaction _a_ prolonged, until the repair or reconstruction of A is complete; for, in this supposition, the addition of A will inhibit _a_ and release _b_, whereas the subtraction of A will inhibit the appearance of _b_ and consequently defer B, until the state of construction A, the sight of which is the stimulus to _b_, is complete. The fact of regulation, therefore, entails _sensory_ control of the serial responses involved in the constructive instinct. Hence, as H. P. Weld of Cornell expresses it: “We may safely assume that even in the lowest forms of animal life some sort of sensory experience releases the (instinctive) disposition and to an extent determines the subsequent course of action.” (Encycl. Am., v. 15, p. 168.)