IV. Correlations in Development. 265 
although the stem continued to grow at a decreasing rate. Since the stem 
still continued to grow rapidly it is clear that food supply from the 
cotyledons was not the factor preventing the further growth of the roots, 
and we must then suppose either that the stem entirely absorbed the 
cotyledonary food supply, preventing any material from reaching the roots, 
or else that some other factor preventing root growth was in operation. No 
evidence has been found for the latter assumption. 
Possible objections to this series of observations were three in number. 
Fig. 2. Growth of stem and root (Series III) in peas at 2 §°C. Secondary roots appeared 
at R. Stem curve reduced to half the vertical scale. (Rate curves above.) 
(1) The variation of the plants and probable error of the determinations 
were so high that arguments based on the growth rate were barely justifiable. 
(2) The determinations were few in number after the appearance of secon¬ 
dary roots. (3) At the high temperature employed, the period of rapid 
growth was passed through so quickly that it was difficult to see whether 
or no any correlation existed between stem and root growth. Another 
series (IV) of determinations was therefore undertaken, the first objection 
above being met by the reduction of the variations as described in a previous 
paragraph (p. 263). The observations were carried out at lower temperatures 
