446 TRANSACTIONS OF THE AMERICAN PHILOSOPHICAL SOCIETY 



The Paradoxes of Adaptation 



So far as I am aware, no subspecific difference in proportion from a parent type has ever 

 yet been shown to have an adaptive or selective value. But this negative finding by no 

 means offsets the positive evidence for the reality of adaptive changes. We must suppose 

 either that the number of cases already investigated thoroughly is not yet sufficiently large, 

 or that the time during which any given subspecies has been studied is not a sufficiently large 

 fraction of a geologic epoch to reveal, for example, sensible differences In food preferences 

 among related subspecies, or correlated differences in average size of jaw and size of food. 

 Such correlations must be facts because the end results of some process of differential selec- 

 tion are already facts of record in the well established "adaptive radiation" of such groups 

 as the percoids, the scorpaenoids, the cyprinoids, characins, siluroids and many others in 

 which the jaws are mechanically adapted for very diverse purposes. 



Thus the first paradox of adaptation in fish skulls is that they have as yet to be demon- 

 strated in their initial stages, as between related subspecies, but are perfectly patent as 

 between related genera. 



Another seeming paradox lies in the apparent conflict between "Wolf's Law" and the 

 non-inheritance of individually "acquired" characters. The facts cited in this paper 

 sufficiently prove that in the skull of fish, just as in the mammalian skeleton, bony trabeculae, 

 ridges, buttresses, etc., arise in response to specific stresses, such as those generated by the 

 thrusts of one moving part upon another, in other words, that bones are usually strengthened 

 in proportion to the loads they bear (which is essentially "Wolf's Law"). But, if we may 

 judge from analogous cases, any special modifications which might be induced in the bony 

 trabeculae of an adult fish under laboratory conditions could not reappear in the subsequent 

 generations unless the same stimulus were applied to each. On the other hand, we should 

 expect from numerous analogies that a young sparid would develop perfect incisors and 

 massive molar teeth, even though he might be deprived of the opportunity to use them. 



Thus the second paradox of adaptation, long well known to palaeontologists, is that 

 •adaptive changes take place as if they had arisen through the inherited effects of habit, use 

 or disuse, without having really so arisen. 



The student of the fish' skull as a natural mechanism must be impressed again and 

 again with the "purposive" or adaptive value of its individual parts. It is a fact of record 

 that the complicated parts of a natural mechanism, like the upper and lower jaws, their 

 teeth, their muscles and supporting elements, do evolve and have evolved with reference to, 

 or in causal relations to, each other, so that natural mechanisms often bear a surprising but, 

 as I do believe, spurious resemblance in function to the products of human design. Simi- 

 larly, two individuals may assume a "purposive" relation to each other, as in courtship, 

 mating, care of young, nest-building, and so forth. But it has been shown elsewhere 

 (Gregory, 1924) that all vital processes are essentially anticipatory in character: the trap 

 is made and set before it ever goes off; the teeth are made up before they come into use; 

 the needs of the next generation are anticipated long before the needs are felt by the yet 

 unborn. Since all vital processes are essentially anticipatory at least in part it is not 

 surprising that the incipient "rectigradations" of Osborn, which are said to antedate their 

 own usefulness, should also share this anticipatory character. 



Thus the third paradox of adaptations is that they have the appearance of being pur- 

 posive designs but are more probably the result of consistent action by an unconscious 

 selective mechanism. 



