190 BOTANICAL GAZETTE [SEPTEMBER 
The plants are shown in jig. 13, and the curves in fig. 14. Those 
for weight, transpiration, and leaf area approximate each other very 
well. With increasing fertility of the soil the transpiration increases 
somewhat more rapidly than the area, while the same function 
increases somewhat less rapidly than the weight of tops. Thus the 
curve of transpiration per unit area rises slightly, while that for 
transpiration per gram falls to about the same degree. The last Imo 
curves, however, both approximate horizontal straight lines. 
245 Series XI—this ® 
Jom study of soil a 
ir T from the soils used 2 Be 
X. The plants ie 
in bottles, as has bet pottle 
already stated. One 0S 
es, containing four ee 
Sts ae used for each culture. from 
San experiment nae 
2 3 a 6 November 11 t0 
Fic. 14 I, 1904. The extent 
, ken at the sam 
changed every four days, and the transpiration was ta tthe i 
intervals, beginning November 15. The plants were abou 
old when placed in the bottles, the seeds having been 8 
sand. The data are given in the table. in fig: 16 
A photograph of the series is given in fig. 15, 4 curves In tion dis: 
It will be observed that the criteria other than aa 
