3 2 4 



GEORGE F. BECKER AND ARTHUR L. DAY 



lightened (74 gm. to o. 7 gm.) and differs in no detail from Bruhns 

 and Mecklenburg's series (Table II) save that our rate of evapora- 

 tion from the solution was probably faster than theirs. The 

 measurements leave nothing to be desired in the simplicity and 

 directness of the proof offered that a loaded crystal may also lift 

 its load in a solution containing an unloaded one (as unsuccessfully 

 attempted by Bruhns and Mecklenburg), though in the light of the 

 foregoing analysis it must be clear that the conditions there are 

 least favorable for such growth. Nevertheless, growth will occur 

 there also if the rate of evaporation is sufficiently high. It may not 

 be inferred, however, that the supporting rim of the unloaded 

 crystal carries no weight ("Er muss ja vornehmlich den Kristall 



TABLE V 



Conditions as Before 



tragen," Bruhns and Mecklenburg, p. 105) ; r it merely carries less 

 weight than the corresponding portion of the loaded crystal. The 

 conditions in the liquid layers adjacent to the two crystals there- 

 fore differ in degree only, and one crystal may grow, or both may 

 grow, according to the degree of supersaturation developed by the 

 conditions of evaporation. Since the process of distribution of the 

 crystalline molecules depends upon diffusion, a small difference of 

 load means but a small difference in concentration at unit distance, 

 and consequently a sluggish molecular flow. Diffusion would be 

 similarly delayed by increasing the horizontal distance between a 

 heavily weighted crystal and an unweighted one, but we have not 

 put this evident inference to the test of experiment. 



Had Bruhns and Mecklenburg chanced to place their unloaded 

 crystal beneath the overhanging load upon its neighbor, it would 



1 See footnote, p. 320. 



