296 CONCERNING THE EFFECT OF tJUAVlTV 



(3) The substance miist lend itx'lf iTiidilv id llic detec- 

 tion :)f change in concentration. 



]r <oenis likclv that sonic oi'iinnic volnlcs and solvents can 

 be found ^vhi(•h will have all the-e factors at a niaxiniuni. Data 

 on the first two factors can he found for some substances in 

 Landolt Bornstein's tables, and of these cane sugar seems to be 

 one of the most satisfactory. It has a large diifusion constant, 

 and can he obtaiuf^l very ])ure. and its concentration can 1)C 

 accurately estimated by the ])olarisco}>e. 



The following calculations are made fur a tube '5 meters in 

 length, and I) for caiu^ sugar is taken as .300 as about the mean 

 of the results of (rrahni and Arrhenius. 



D = .30O where the cm. is the unit of length, the gram the 

 iniit of mass, and rhe day the unit, of time. 



('alcuhition of the f;dl for one year: that is, f<u- n at x ^ 0, 

 and t = 365. 



(- 10 X. 3x365 -9x10x0.3x365 

 90000 xT 90000x1 ~ \ 



„ ' +^- ; 



X -0.012 -0.108 X X 



= -A^ +'y' ) 



= -V! ( •»-988. -f 0.098 \ 



= o.^Ja approximately. 



So that in a tid)e of this length the fall in concentration of 

 the sugar solution at the end of one year would only he ab:)ut 

 one-tenth of the total fall after an infinite time. For a con- 

 centration oi" 1 to 4 tlii< would be about ten per cent, of 



—7 . — G 



10x10 X 1 .■)»> r^ 1.") X 10 gr. which is a change of con- 

 centration thai in a -'ilution of that strength would defy iletec- 

 t ion. 



