December 22, 1911] 



SCIENCE 



885 



jects as the apple in very wide limits without 

 injury or complicating alterations. 



We found the following method most satis- 

 factory. An apple with perfect skin was put 

 into a glass dish containing a small amount 

 of a liquid a. On the opposite side of the 

 apple a piece of the skin and the underlying 

 tissue was removed and into the hole was put 

 a small quantity of a liquid h. The latter re- 

 mained constant throughout the experiment, 

 while a was changed according to the nature 

 of the experiment. Both liquids a and 6 were 

 connected with calomel electrodes and the 

 E.M.F. was determined by Poggendorf's com- 

 pensation method (capillary electrometer). 

 The temperature was kept practically constant 

 (about 19° C). 



We, therefore, were studying the E.M.F. of 

 the following system: liquid a; apple; liquid 

 h, the membrane of the apple being the solid 

 phase between liquids a and h. According to 

 the theory a fivefold dilution of a should al- 

 ways give a constant decrease of E.M.F., 

 namely, .040 volt; and a tenfold dilution 

 should decrease the E.M.F. always by a value 

 of .058 volt. These values may be expected 

 to be smaller if the ideal conditions of semi- 

 permeability are not fulfilled. 



In the following experiment liquid h (in 

 the apple) was m/10 KCl. Liquid a, on ex- 

 ternal surface, varied in concentration. The 

 calomel electrodes contained m/10 KCl. The 



TABLE I 



Concentration of 



Liquid a. E.M.F. 



m/10 KCl + .040 volt 



m/50 KCl -I- .068 volt 



m/250 KCl -I- .103 volt 



m/1,250 KCl -1- .137 volt 



m/6,250 KCl +' .169 volt 



sign -|- means that liquid a was positive to 

 liquid 6. In this experiment each successive 

 liquid was five times as diluted as the previous 

 one and the theory demands that the differ- 

 ence of E.M.F. between two successive solu- 

 tions should be identical. If we compare the 

 interval from m/50 to ot/6,250 this is true. 

 A dilution from wi/50 to ot/250 increases the 



E.M.F. by .035 volt; from m/250 to m/1,250 

 by .034 volt ; from m./l,250 to m/6,250 by 0.32 

 volt. By diluting from m/10 to m/50 we find 

 .028 volt, which is a little too small. We have 

 found it to be generally true that as soon as 

 we work with more concentrated solution than 

 m/50 the differences are a little smaller than 

 the theory demands. Whether this is due to 

 the decrease in ionization or to an injurious 

 effect of the solutions of higher concentra- 

 tion upon the skin of the apple we are not yet 

 prepared to state. 



The values observed are all a little smaller 

 than we should expect. According to Nernst's 

 formula the difference of potential for a dilu- 

 tion of five should be .040 volt, while we found 

 a difference of .033 volt. The fact that we 

 did not get the maximum potential difference 

 is, perhaps, due to the fact that the skin of the 

 apple is not completely impermeable for anions. 

 Experiments with other salts gave similar 

 results as far as the effect of concentration 

 was concerned. 



3. The sign of the E.M.F. of the system 

 electrolyte; apple; electrolyte was always in 

 that sense as if the membrane of the apple 

 were more permeable for kations than for 

 anions. In order to test this possibility the 

 electromotive effects of NaCl were compared 

 with those of Na^SO,. If our assumption 

 were correct the E.M.F. of a NaCl solution 

 should always be equal to the E.M.F. of a 

 Na„SO^ of half the concentration of the 

 former. The following example shows that 

 this is actually the ease. The internal liquid 

 h remained constant throughout the experi- 

 ment m/10 KCl. The external liquid a 

 varied according to Table II. 



TABLE II 

 Liquid a. E.M.F. 



m/10 NaCl -f .038 volt 



m/100 NaCl -|- .090 volt 



m/1,000 NaCI -|- .139 volt 



m/2,000 Na^SOi -|- .141 volt 



m/200 Na,SOi -1- .092 volt 



m/20 NajSOi -|- .044 volt 



This experiment shows that whether the 

 anion is CI or SO, as long as the eoncentra- 



