GENETIC CHANGE AND EVOLUTIONARY CHANGE 415 



the rate of increase in height might have been lower than the rate of in- 

 crease in length, in which case E would have had a slender head. The third 

 dimension, not represented in our flat diagram, may also change at a rate 

 difi'erent from the rates by which the other two dimensions change. Little 



B 



FIG. 18.4. A, negative allometry of the nasal 

 horn of a hypothetical rhinoceroslike mammal. 

 B, positive allometry in length of nasal horn cou- 

 pled v/ith isometric growth in v/idth (cf. E, Fig. 

 18.2). 



imagination is needed to picture some of the great variety of shapes and 

 proportions possible when growth in these three dimensions varies. 



In closing this phase of our discussion we may well mention the Carte- 

 sian coordinate method utilized by D'Arcy Thompson (1942) for represent- 



