THE EFFECT UPON PERMEABILITY OF POLYVALENT CAT- 
IONS IN COMBINATION WITH POLYVALENT ANIONS 
Oran L. Raber 
(Received for publication January 20, 192 1) 
In most of the studies carried on with a view of ascertaining how bi- 
valent ions (such as calcium and magnesium), and trivalent ones (such as 
aluminum) affect permeability, the anions employed have been monovalent 
ones (chloride, nitrate, etc.). In studies in which Laminaria has been used, 
it has been found by Osterhout (i, 2) that wherever bivalent and trivalent 
cations were employed, the first effect of the salt has been to cause an 
increase in the resistance of the tissue (followed by a decrease), while if 
monovalent cations are used (with exception of acids and sodium taurocho- 
late) the first effect is a decrease in resistance. 
Since the anions, as shown in previous papers (3, 4) seem to play an 
important part in determining the permeability of the tissue, it seems of 
interest to investigate the effect of a salt composed of a polyvalent cation 
and a polyvalent anion. 
The number of such salts which are sufficiently soluble for the present 
experiments is exceedingly limited, because of the pronounced relation 
between valency and solubility. Magnesium citrate, magnesium sulphate, 
and aluminum citrate were finally selected for study since they possess 
more than most others the requisite characteristics for this work, in respect 
to acidity, solubility, osmotic pressure, etc. 
The solution of magnesium sulphate used was about 1.09 M and had the 
same conductivity as sea water. Its pH value is about 8 and it is hence 
only very slightly alkaline. Figure i (curve A) shows the average result 
of the six experiments performed. The probable error of the mean (as 
based on Peter's formula) is under 5 percent of the mean. The rise in 
resistance at the start is seen to be very small and temporary, and at the 
end of five minutes the resistance has dropped to 84 percent of the original 
value in sea water. 
With magnesium citrate the difficulty of solubility was encountered. 
Since it was impossible to get a solution of the same conductivity as sea 
water, the practice of diluting with chloride was resorted to. When a 
solution composed of one sixth magnesium citrate 0.16 M and five sixths 
magnesium chloride 0.28 M is used, the results are as shown in figure i, B 
(three experiments; probable error of the mean under 5 percent of the mean). 
If this experiment is repeated, using magnesium chloride of the same 
electrical conductivity as the previous mixture, viz., about 0.24 M, the 
382 
