388 THE POPULAR SCIENCE MONTHLY. 



possibility which it opens of asking and answering questions by 

 the deductive method. It might be supposed that a theory with- 

 out a mathematical basis would exhibit comparatively little 

 power of prediction, although its validity might never be ques- 

 tioned. Quantity of effect can not be measured, much less pre- 

 dicted, by deduction from theories susceptible only of historical 

 treatment. All that can be expected is the power to indicate the 

 presence or absence of things that are still unknown. But the 

 laws of heredity, variation, correlation, etc., although still unde- 

 fined and perhaps undefinable, furnish peculiar opportunities for 

 brilliant deduction. 



Paleontology would be a sorry science without the power of 

 restoration afforded by the principle of correlation. Its frag- 

 ments of bones and teeth and stumps and leaves would be almost 

 absolutely worthless. But from the standpoint of logic this is as 

 truly prediction by deduction from known laws as the minute 

 predictions for the nautical almanac. Perfect heredity would 

 place the principle on a basis of certainty. Any one character of 

 ruminants indicated to Cuvier the presence of all the rest. But 

 the generalized types of paleontology are transitional forms pos- 

 sessing combinations of ruminant characters, with others belong- 

 ing to the carnivora, such as he never dreamed of. But the recog- 

 nition of secular change in the correlations of organs, instead of 

 weakening, has strengthened the possibility of anticipating un- 

 known facts. 



In recent years the progress from deduction to verification has 

 been so rapid that frequently the latter has followed at the heels 

 of the former, so that the element of time has hardly entered be- 

 tween them to make them both more striking. Many of the de- 

 ductions from the theory of descent, afterward verified, are com- 

 monplaces to the scientist, but their logical force is not sufficiently 

 emphasized when the nature of the evidence is considered. 



The doctrine of descent required the belief that ruminants 

 once had upper incisors and canines. The belief was made almost 

 a certainty by the presence of partially developed fcetal teeth 

 where they are absent after birth. The confident expectation was 

 justified by the discovery of generalized ruminants with full sets 

 of teeth. If man were descended from lower forms, an explana- 

 tion was required for the absence of the os centrale as an inde- 

 pendent bone from the human wrist, for it is almost constantly 

 present in amphibians, reptiles, and mammals. Rosenberg looked 

 for it in the human embryo and found it. Wiedersheim * was 

 moved to declare that this was one of the greatest triumphs that 

 morphology, based on the theory of descent, had yet won. The 



* Wiedersheim, Lehrbuch der vergleichenden Anatomie der Wirbelthiere, p. 223. 



