270 
Pearsall.—Studies in Growth . 
may be pointed out in passing that the subsequent development of new 
shoots quickly reduces the rate of root growth to the original lower level. 
The establishment of this negative correlation between shoot and root 
growth may apparently be extended much farther as a result of Townsend’s 
work, although this author attributes his results to some stimulus caused 
by injury. He found that in seedlings the removal of the root tip accele¬ 
rated the rate of stem growth, and that the removal of a leaf in a young 
Phaseolus resulted in an increased growth rate of the remaining leaf. The 
hypothesis that food supply is the factor limiting growth in these cases 
appears to be too suggestive to be lightly dismissed. 
Correlation in the Cotton-plant. 
* 
Further, this hypothesis appears to have equal value in explaining the 
form of the flowering and fruiting curves in the cotton-plants. These 
curves have been exhaustively investigated by Balls and Holton ( 1 , 2 ) and 
by Harland ( 3 ), and their papers form a valuable contribution to the study 
of growth. They conclude that edaphic factors limit the growth rate of the 
cotton-plants studied (in Egypt and St. Vincent), and according to Balls ( 2 ) 
the limiting factor is probably the supply of nitrates. With a limited rate 
of food supply the rapid growth of one set of organs should therefore mean 
a correspondingly decreased growth rate in other parts of the plant. 
Thus in Fig. 5, where S = stem, B = branch, F = flower, and Fr = 
fruit, the sum of the growth of S, B, F, and Fr, will be a constant when the 
rate of food supply is constant. If, then, the growth of S ceases, the 
growth rate of B, F, and Fr will increase proportionally ; or conversely, if 
the growth rate of any organ, say Fr, increases, the rate of growth of the 
others (S, B, and F) should decrease. 
If it is assumed that the supply of substance from the roots is a limiting 
factor in the total growth of the stem, &c., then in the case figured the 
growth of the fruit (Fr) would proceed at its maximum speed, since it draws 
first on the supply stream, while some or all of the organs B, F, and S 
would grow at a reduced rate, the rate being least in those farthest from the 
source of supply X. Thus the rate of growth in the stem apex (S) would, 
other things being equal, be the lowest in the case cited. 
In Egyptian and Sea Island varieties of cotton, the development of 
stem apex, branches, flowers, and fruits occurs essentially as in the figure 
given. The exhaustive data of Balls and Holton ( 1 , 2 ) and of Harland ( 3 ) 
are therefore available for a further analysis of these assumptions. 
Their results are in all cases expressed as the average plant of a large 
crop, the number of plants being sufficient to eliminate most sources of 
error. They give the average number of flowers opening per plant per day, 
and the average number of bolls (fruits) maturing per plant per day. For 
