4i8 ON INCREASE IN SIZE AND WEIGHT [pt. iii 



it follows that the velocity of change in an autocatakinetic system 

 progressively rises from zero hour to a maximum value when x = ^A 

 and afterwards constantly falls. When the data are plotted in this 

 way the existence of growth-cycles, whether real or not, comes out 

 much more clearly than when the absolute curves alone are used. 

 It is often said, however, that the increment curve emphasises small 

 fortuitous variations more than the absolute curve, and certainly 

 there are cases, notably the chick embryo itself, where the increment 

 curve shows up cycles which are but poorly shown on the absolute 

 weight curve. Brody & Ragsdale in their first memoir on this 

 subject dealt only with the growth of the cow, and concluded that 

 one complete growth-cycle was accomplished in the foetal condition. 

 A later paper considered the growth of the fowl, for which, on the 

 data of Card & Kirkpatrick for growth after hatching, two cycles 

 appeared, with maxima at 9 and 18 weeks respectively. For the 

 growth of the embryo, the data of Hasselbalch and of Lamson & 

 Edmond, re-arranged by Brody, gave two maxima also, but at slightly 

 different times, thus: 



Lamson & Edmond 11-5 and 16-5 days 

 Hasselbalch 10-5 and 15-0 days. 



Brody's figures for these curves are shown in Fig. 48. LeBreton & 

 Schaeffer,who subsequently published a series of chick embryo weights, 

 found maxima at 9 and 15 days, but in Murray's series there is no peak 

 at any time, except a doubtful one on the i6th day, and a plateau be- 

 tween the 1 2th and 15th days. Schmalhausen's data again, when the 

 increments are calculated, show in the case of both series peaks at 

 10 and 12*5 days, with an additional one, in the case of his 1927 series, 

 at 17 days. 



In the face of this consensus of evidence, it is not altogether easy 

 to conclude with Murray that "Brody's rhythmic growth curves were 

 due to chance variations". It is true that Lamson & Edmond; 

 Hasselbalch; and LeBreton & Schaeffer used too few embryos 

 in their work, but even in Murray's own work, where about 650 

 embryos were used, there is an unexplained drop between the 17th 

 and 1 8th days, as well as a plateau between the 12th and i6th days, 

 both quite outside the probable error. Murray did not, it is true, 

 get the i8th-day drop in all his experiments. In Schmalhausen's 

 two series about 400 embryos were used. 



