SEVENTH ANNUAL MEETING 



71 



the habitats. It has therefore seemed desirable to limit our 

 comparisons to this period of stress, and to establish the 

 rather arbitrary limit of the 10 weeks from the last of June 

 to the first of September. A summary of these weeks appears 

 in Table I. 



TABLE I. 



Wilting- coefficients and mean percentages of growth-water 

 in the various associations during the 10 midsummer weeks 

 of 1911-12: 



No one can realize more fully than does the writer the limita- 

 tions of the data or the desirability of having them supple- 

 mented by more numerous determinations, especially from 

 deeper strata. These limitations will make the conclusions 

 more or less provisional and subject to modification and cor- 

 rection in the future. The most interesting and profitable com- 

 parisons are doubtless those to be made from a consideration 

 of the growth-water data shown in Table I., and especially as 

 expressed in the sixth and seventh columns, which contain the 

 mean percentages of growth-water and the comparative 

 amounts, the beech-maple forest being taken as a standard 

 and its mean growth-water represented by 100. In the five 

 associations that form the succession, the mean growth-water 

 for midsummer of the two seasons will be found, if we except 

 the cottonwood dune, to form a progressive series, the most 

 mesophytic association having the largest amount. This was 

 exactly what was supposed to be the situation, but hitherto 

 no quantitive data of such moisture relationships have been 

 available, it has been impossible to tell how much an asso- 

 ciation differs in its water conditions from the preceding or 

 succeeding association. As has already been stated, the cotton- 

 wood dune, with a larger and more constant water supply than 

 the two succeeding associations, must owe its surplus to the 

 conserving power of its dust mulch and to the small outgo due 

 to the paucity of its vegetation. 



