EDUARD UHLENHUTH 



475 



this drop occurred in the 24th week, though less distinctly than in 

 most of the other series. From the average sizes reached in each series 

 when the drop occurred and from the time when the drop took place, 

 the rate of growth was calculated in day-millimeters, as indicated in 

 Column 3 of Table I. 



In comparing the length of the larval period for each series with the 

 corresponding rate of growth it becomes evident that the smaller 

 the rate of growth observed, the later metamorphosis took place. 

 Furthermore, the rates of growth appear to be proportional to the 



TABLE I. 



Rate of Growth (R) in Worm-Fed Ambystoma opacum during Larval Period of 

 Active Growth {Calculated from Averages). 



velocity of metamorphosis; for the product of the rate of growth into 

 the duration of the larval period (R X A) is constant, as may be seen 

 from Table I, the average value of R X A being 60. 



This means that during the process of growth a substance is 

 evolved which when present in a definite amount induces metamor- 

 phosis, provided that the larvae have been fed on normal food which 

 apparently contains enough iodine to furnish the other substance 

 (iodine) required for metamorphosis in a sufficient quantity; the 

 greater the rate of growth the quicker that quantity of the first sub- 

 stance is formed which is required to induce the secretory action of 



