THE UPWARD TRANSLOCATION OF FOODS IN WOODY PLANTS 289 
The relative root and top weights would, of course, vary with the soil and 
the climate, but there seem to be good indications that tree roots may not 
greatly exceed 50 percent of the top weight. Therefore, though the roots 
may have a carbohydrate content greater than the tops when measured as 
percentage of dry weight, the total quantity of carbohydrates in the 
roots is much less than that in the tops, and, since the roots must need 
quantities of food for their own use, it seems doubtful whether any is 
normally carried to the tops for shoot growth. The indications that root 
growth commences in the spring before shoot growth, as discussed later in 
this paper, may be considered as further proof that the food stored in the 
roots is used primarily by the roots. 
Data obtained from experiments designed to determine the path of 
upward translocation, a subject reported in a recent paper by the writer 
(Curtis, 1920), offer evidence that little or none of the food stored in the 
trunks or roots of trees is normally moved up to be used by the developing 
shoots and leaves. 
In one group of experiments, large numbers of twigs and branches 
were ringed early in the spring while the buds were still dormant or were 
just beginning growth. These rings were made at different distances from 
the tip in order to determine from how far back food was withdrawn for 
shoot growth. 
Since, as was shown in the previous paper, no appreciable upward 
movement of foods occurs through the xylem, the growth of a shoot above 
a ring would serve as an approximate measure of the amount of food avail- 
able. If the ring were back far enough from the growing tip to allow for 
growth practically as great as that on unringed twigs, it would seem that 
these twigs need not draw on the food stored at greater distances. 
A large number of stems of Acer saccharum were ringed on April 5 at 
different distances from the tips. In one series the rings were in the first- 
year wood, in another in the second- or third-year wood, and in another 
the rings were in that part of the stem ranging from five to fifteen years old. 
Some of the stems had made terminal growths in the previous year of from 
only I to 10 centimeters, while others had made growths of from 20 to 40 
centimeters. In each case a check stem was chosen as nearly matching the 
ringed one as possible. The check and ringed stems were usually the two 
terminals of a pair produced by dichotomous branching. Such a variety 
of branches was used that no attempt will be made to give more than a 
brief summary of the results. 
Of 15 twigs ringed in the one-year-old wood, the average terminal 
growth on May 6 was 0.84 cm. That of the corresponding stems not ringed 
was 2.22 cm. Of those ringed in the two- and three-year-old wood the 
average terminal growth was 2.03 cm., while that of the corresponding 
checks was 2.25 cm. The leaves of the ringed stems in these cases did not 
show the bronze tinges that were common in the normal young leaves, but 
