-34- 



Grov/th-Teinpera ture Relations « 



The considerations presented in the preceding 

 paper led to the suggestion that all the mean grov/th rates 

 for each separate series might be averaged to give an 

 average growth rate for the given test and series, that 

 these averages for tv;o corresponding tests vfith the same 

 temperature might be themselves averaged to give a single 

 value representing the growth rate for the given temperat^^re 

 and solution t^^pe, and that all the six type averages might 

 be averaged for each temperature to give a single growth 

 index for each temperature considered. The logical basis 

 for this mode of treatment may be stated as follows: Since 

 the data at hand do not establish any relation between 

 solution composition and grovfth rate for any temperature, 

 all solutions may be treated as though they were physiolog- 

 ically alike, within the limits of precision set by the in- 

 nate variability of the seed used, etc. 



An inspection of the 42 tables obtained for the 6 

 solution types tested v;ith 7 different ns.intained temperatures/ 

 for the entire culture period/ as well as for the tvfo partial 

 periods (of nhich table I is an example), brought out very 

 clearly the fact that the temperature influence on growth rate 

 was pronounced and consistent, in spite of the great individual 

 variations of the seedlings and quite v;ithout regard to the 

 rasBceup of the solutions used. In the following pages the 



