GREGORY: FISH SKULLS 445 



the ripening of the male gonads brings about marked proportional differences in certain 

 parts, we have an indication that the interaction of the various endocrine glands plays a 

 large part in the regulatory mechanism, as it is known to do in other classes of vertebrates. 

 Moreover, in fishes we have many examples of extreme elongation of the body, as in the eels, 

 and of extreme broadening, as in Lophius; these are at feast analogous with Stockard's 

 "linear" and "lateral" types of men and dogs; they are likewise probably due to contrasting 

 activities of whatever agencies may correspond in function to the thyroid and pituitary 

 extracts of mammals. 



The studies of D'Arcy Thompson ("On Growth and Form," 1917), Julian Huxley 

 (1931) and others have shown that in many animals the growth rate of a given part during a 

 given period is a logarithmic function of the increase in mass of the entire organism, and 

 that the growth of a localized excrescence again bears a logarithmic relation to the mass of 

 the structure that bears it. The fact that the dependence of growing parts upon their 

 sources of supplies may be expressed in tables like interest and compound interest, is 

 enlightening as regards the processes of individual development and growth. But the 

 chief problem of the student of the evolution of the fish skull is to discover how changes 

 from one schedule of rates of interest to another have been induced during the millions of 

 years which appear to have been necessary for the transformations from one feeding type 

 to a radically different one. 



In many, if not all, teleosts the changes in the proportions of the jaws and skull from 

 the embryonic to the larval stages and from the later to the adult skull are very striking. 

 Thus in the larval stages of many teleosts, as figured by Bigelow and Welsh (1925), the 

 mouth and jaws are small and upturned, the muzzle very short, eyes very large, braincase 

 large and swollen. Here as in later stages the bones of the skull simply share in the rates of 

 growth that affect whole regions. As growth proceeds the jaws lengthen, the snout 

 lengthens, the eyes are retarded, locomotor muscles overgrow the occiput, crests and ridges 

 appear upon the cranial vault, etc. 



These changes are doubtless brought about by the interaction of growth forces (lasting 

 from embryonic and cleavage stages) with nutritive and other environmental influences. 

 Thus Stockard (1931, pp. 108, 135) found that in young Fundulus cyclopean monsters could 

 be produced by the reaction of the hereditary mechanism with several different chemical 

 and physical agents, while Hubbs (1926) has noted many factors (including temperature, 

 salt concentration, etc.) that induce changes in the proportions of parts in the growing 

 fish, and which may under certain circumstances cause a departure from the adult norm. 



From the consideration of hundreds of species of fishes of all orders, recent and fossil, 

 it would seem that the factors that determine the presence of the standard parts of the fish 

 skull have been on the whole amazingly constant for perhaps several hundred millions of 

 years; but that the factors that determine departures from normal proportions in the adult 

 are, in geologic time, more or less highly variable. Perhaps this might be because the 

 embryological factors for the presence of mouth-parts, paired eyes, otic capsules, branchial 

 arches, etc. bring the Anlagen of these parts into existence during the early embryonic 

 stages, while their adult proportions are approached only in late larval stages. Conse- 

 quently the parts once established are seldom crowded out, but are subjected during late 

 embryonic, larval and immature changes to disturbances in growth rates that can hardly 

 fail to affect their adult dimensions. 



