GREGORY: FISH SKULLS 453 



cyprinoid genus Cycleptus (Fig. 300^/) the head is remarkably small in proportion to the 

 body, while in the percoid Hoplopagrus guntheri (Fig. 3005), on the contrary, the head is 

 remarkably large. And there is the greatest diversity in the association of jaw length with 

 snout length (Fig. 301). In short, every region of the head, while sharing more or less in 

 the general tendencies of the body as a whole, seems to have its own intrinsic tendencies to 

 attain specific characteristic proportions. 



If different general tendencies toward length, height or thickness of skull and body be 

 correlated in part with different amounts of enzymes received from the different ductless 

 glands at different periods of growth, then each region or part of the skull must have its 

 average specific rate of reacting to the enzymes that are distributed by the circulatory 

 system. At any rate, each part seems to respond differently to a given enzyme, so that, for 

 instance, the ripening of the gonads in the male hook-jawed salmon produces something that 

 accelerates the growth of the tip of the lower jaw and surrounding parts without producing 

 any equally conspicuous changes in other parts of the skull. 



The Interaction of Growing Parts 



Besides these general and regional responses perhaps to growth-stimulating enzymes, 

 we must take into account the lively reactions of the growing parts to each other. Osseous 

 tissue seems always to retreat from or grow around nerves, sense organs, brain, blood-vessels 

 and glands and to strengthen itself against tensions transmitted by tendons, ligaments, 

 muscles or connective tissue, and against compression or shearing stresses received from 

 these or from other bones. For example, all the bones that come near the eyes have con- 

 cave surfaces toward the eyes, while bones that carry the organs of the lateral line system 

 develop pits and tunnels for them. On the other hand, bones that give attachments to 

 muscles or tendons build themselves out into knobs or crests or develop ribs, trabeculse or 

 other forms of stiffening. , 



The principal muscle crests appear late in development and are dependent in part upon 

 growth stimuli transmitted from surrounding parts. Nonetheless they are clearly deter- 

 mined in part also by the reaction of hereditary factors with the bodily environment. 

 Moreover, the presence and size of certain crests may be just as specific as color spots or 

 number characters. Every crest on the skull roof of a certain Cretaceous teleost figured by 

 A. S. Woodward, foi" example, compares very closely with that of its modern descendant. 



Evolution through Differential Growth Rates 



Yet if we compare the skull of any highly specialized modern percoid with that of its 

 primitive early Tertiary or Cretaceous ancestor, we shall see at once that there has been a 

 progressive increase in certain parts and a relative decrease in others, so that, for example, in 

 the case of the sheepshead {Archosargus) the jaws have become smaller but stouter, the 

 lateral teeth have become molariform, and there have been many correlated adjustments for 

 the support of the crushing dentition. Hence arises a paradox: for while adjustments have 

 evidently taken place both in response to changing internal conditions and in response to 

 changing reactions toward the environment, the experimental evidence weighs most strongly 

 against any crude form of the Lamarckian hypothesis of the inheritance of "acquired 

 characteristics." The immense extent of geologic time that has been required to overcome 



