preceded the adaptive structure, or else the structure preceded and gave origin to the use . . . Many facts render the first of these propositions much the more probable of the two" (1878, in 1887, p. 352). Acceleration must still make room for a new character by pressing earlier ones back, but the primary impetus for its origin is the animal's own activity. In early works, however, Cope seems to have held that the sequence of characters added in progressive evolution is fore-ordained and out of the animal's control: "Genera have been produced by a system of retardation or acceleration in the development of individuals; the former on pre-established, the latter on preconceived lines of direction" (1869, in 1887, p. 123). In direct contradiction to his 1878 statement, quoted above, Cope had argued in 1870 that the introduction of a feature precedes its use:

We look upon progress as the result of the expenditure of some force fore-arranged for that end. It may become, then, a question whether in characters of high grade the habit or use is not rather the result of the acquisition of the structure than the structure the result of the encouragement offered to its assumed beginnings by use, or by liberal nutrition derived from the increasingly superior advantages it offers. (1870, in 1887, pp. 145–146)

Acceleration, in this early reading, is the true motor of evolutionary progress. New features passively await their turn for expression; when acceleration has "made room" by crowding back the previous adult characters, these foreordained improvements make their automatic appearance.

The Extent of Parallelism

Cope granted his law of parallelism a much wider scope than Haeckel attributed to his own biogenetic law; for Cope took Haeckel's exception—cenogenesis—and tried to ignore it or render it as a variety of parallelism. Cope did consider the two most important aspects of cenogenesis: embryonic adaptation and heterochronism.23 In later works, he simply admits the existence of embryonic and juvenile adaptation (1896, pp. 202–203), but in his early articles, he attempts to explain it away by an argument that seems sophistic even in its own context.* How can we say that a human embryo represents an ancestral fish; after all, it displays so many definite adaptations to the fetal state that it closely resembles, in toto, no fish living or extinct. Cope

* It is rendered more intelligible by a consideration of the long debate that taxonomists have endured and propagated about "key" characters in the definition of groups. Are orders defined, for example, by differing states of a designated "ordinal" character (rather than by some assessment of overall morphology). Cope, in his statement on