Nov., 192 1 ] 
RABER — ■ PERMEABILITY 
467 
found in general a greater increase in resistance with trivalent and tetra- 
valent cations (3) than with bivalent cations (4), but he did not compare 
the effects of the same ion with different charges. 
The choice of ions for this purpose is very limited. Antimony, arsenic, 
and tin can not be used because of their weak basicity which results in 
extremely acid solutions. The toxic action of copper is a sufficient reason 
for not using it. The insolubility of mercurous, chromic, and cobaltic salts 
almost prohibits their use. This leaves iron as a possibility among the 
common cations of variable valency. 
For this study FeCls and FeCU were used. The solution of FeCls was 
0.20 M and had a pH value of about 2.5. When the tissue was placed in 
such a solution having the same conductivity as normal sea water, there 
occurred a very sudden and temporary rise in resistance followed by a 
rapid fall. The following table and figure 2, A show the results: 
Time in 
Percentage of 
Probable Error 
Minutes 
Original Resistance 
(Three Experiments) 
2 
125 
1% 
5 
107 
I 
10 
71 
5 
20 
55 
9 
40 
47 
8 
60 
44 
7 
80 
42 
7 
The tissue becomes extremely hard and decidedly yellow in color, which 
seems to be due to the acid reaction of the solution. Inasmuch as the 
acidity of the ferric chloride lies outside the ineffective limit (2), a certain 
amount of the rise in resistance shown above is doubtless due to the hydrogen 
ions present. 
The FeCU solution of the same conductivity as that of FeCls used is 
about 0.28 M and has a pH of about 4, but in order to compare the effect 
of the valency the solution was acidified by the addition of HCl. The 
acidity of these solutions (since colored ions are present) is measured and 
compared in the same way as that of the solutions discussed in the previous 
section. The final solution used consisted of 45 g. FeCU per liter with 
enough HCl added to make a solution of the same conductivity and the 
same pH as the solution of FeCU. The results of this set of three experi- 
ments are given below and in figure 2, B. 
Time in 
Percentage of 
Probable 
Minutes 
Original Resistance 
Error 
2 
118 
4% 
5 
91 
5 
10 
68 
5 
20 
46 
4 
40 
31 
6 
60 
28 
3 
80 
27 
3 
