THE MUTATED GENE 211 



a clue to the possibility of linking gene action with the produc- 

 tion of growth patterns: any gene that controls quantitatively the 

 rate of any one process, which is represented by the constants of 

 the formula, automatically produces a definite result in terms of 

 form. Form, then, would also be of the type of automatic 

 pattern formation: something connected with growth produced 

 at a definite time and place and in a definite quantity sets the pace 

 for the working of differential growth by fixing the numerical 

 value of the constants in the formula; it acts in this sense also as 

 an evocator of pattern. 



This important conclusion had already been derived by Gold- 

 schmidt (19206) from D'Arcy Thompson's (1917) classic work 

 on growth and form (including also the phylogenetic significance, 

 which recently has been emphasized again by Huxley, 1932, 

 and tested in an admirable way by Hersh, 1934a). Goldschmidt 

 wrote : 



Let us suppose that we succeeded in reducing all processes of morpho- 

 logical differentiation to simple physical and mathematical conditions, 

 to surface tension and to all the geometrical consequences of differential 

 growth, etc., the problem of heredity remains the same: to find why, at 

 a given moment in given rhythm and localization, the specific chemical 

 and physical situation is produced, the consequence of which may be 

 stated as the mathematical law of development. . . . [Here follows a 

 statement of the theory of rates of reactions in time.] . . . This quanti- 

 tative conception makes it possible to refer extraordinary differences in 

 the end result, i.e., evolution within considerable limits, to very small 

 causative processes. One may find in Thompson's brilliant book how 

 the production of all forms of shells may be reduced to small changes in 

 the values of the terms of the underlying equations for the actually found 

 curves ; and therefore how then a series of morphological differences may 

 be shown to be produced by changes in differential growth within simple 

 mathematical laws. Remembering our proof (in the work on inter- 

 sexuality) that very specific differential growth may be started by the 

 production of specific hormones [here in the sense of determining stuffs, 

 evocators] at a definite time, one might realize the number of evolu- 

 tionary processes which might be caused by small changes in the under- 

 lying genes followed automatically by numerous shifts in the interplay 

 of properly timed coordinations. 



The genetic facts that permit the derivation of such views are 

 only beginning to be studied quantitatively. In earlier chapters, 

 we have mentioned a number of facts that have to do with 



