IQISI 



/. Arthur Harris and Ross Aiken Gortner 



63 



For wall, 0.000903 + 0.000182, d/Ei = 4.96 ; 



For prolification, 0.000988 + 0.000194, d/Ea = 5.09 ; 

 There can be no reasonable question of the trustworthiness of the 

 diff erence. Compare the same constants for exp. C and D : 



For wall, 0.000316 + 0.000135, c?/.Ed= 2.34; 



For prolification, 0.000314 + 0.000138, (//£d = 2.27; 

 Note that the absolute differences are much lower and that the prob- 

 able errors are relatively higher. 



The ranges of Variation observed in the concentration of solutes 



are: 



Wall Prolification 



0.0377 — 0.0429 = 0.0053 0.0342 — 0.0379 = 0.0037 



0.0400 — 0.0482 = 0.0083 0.0357 — 0.0462 = 0.0105 



0.0376 — 0.0463 =: 0.0087 0.0329 — 0.0437 = 0.0107 



0.0374 — 0.0412 = 0.0038 0.0347 — 0.0391 = 0.0044 



0.0358 — 0.0490 = 0.0132 0.0333 — 0.0482 = 0.0149 



Experiment 



A 

 B 

 C 

 D 

 E 



Or, expressed in terms of the Square root of mean square deviation 

 from the mean (Standard deviation), the fluctuation is 



Series E shows a greater variability in concentration, just as was 

 found to be true for specific gravity. 



The explanation of these results probably lies in the wider ränge 

 of external environmental conditions and internal physiological 

 States prevailing during the collection of certain of the series as 

 explained in a preceding section of this paper. 



When one turns to a comparison of wall and prolification he 

 finds that the entries in the individual tables show that there are 

 55 cases in which the specific gravity of the sap of the prolification 

 is lower than that of the wall, to 2 cases in which it is higher. In 

 one instance they are identical. The means for the three series in 

 which the number of observations is large enough to justify the 

 calculation of a probable error are : 



