546 



ON INCREASE IN COMPLEXITY 



[PT. Ill 



"No satisfactory method", Murray said, "of measuring changes 

 in form quantitatively is known, so that it was necessary to resort to 

 the expedient of selecting forms spaced by a visual impression so as 

 to represent approximately equal degrees of gross change. In other 

 words, from a series of drawings made at frequent intervals, certain 

 ones were chosen which seemed by inspection to be equally spaced 

 from one another in respect to their relative complexity of form. The 

 test is thus necessarily arbitrary and open to criticism because of its 



<6 

 O 

 O 



o 



Ds.ys z 



6 8 10 



Incubd^tion 5.ge 



Fig- 91- 



a 



subjective nature. By taking the average result of many eggs it was 

 then determined what were the exact incubation ages of the embryos 

 with heart forms such as those selected." The illustration shows these 

 clearly (Fig. 90). The reciprocals of the time intervals between suc- 

 cessive drawings were used as rough criteria of the rate of form 

 development. The graph showing this done is reproduced in Fig. 91, 

 from which it is clear that the rate of evolution of external form falls 

 precipitously at first, and then ever more slowly, essentially resembling 

 in this way the instantaneous percentage growth-rate, according to 

 Schmalhausen, or the averages of the steps in the fall of the in- 

 stantaneous percentage growth-rate, according to Brody. Growth 



