5°6 
Seifriz .— Observations on the Reaction of 
indicating lower surface tension, since the membrane readily loses its hold 
on the cell wall. 
A still more beautiful demonstration of a reduction in surface tension is 
to be had from the absence in alcohol-treated cells of those delicate proto¬ 
plasmic strands which radiate from a plasmolysed protoplast to the cell 
wall in an untreated Elodea cell ( 5 , p. 18). These long fine protoplasmic 
threads could only be formed by a protoplasmic surface of high tension value 
Czapek maintained that a substance must have a tension value less 
than 0*685—which he thought to be the surface-tension value of protoplasm 
(water being 1)—in order to affect the diosmotic properties of protoplasm. 
Czapek based this theory on the idea that, since a lower tension value 
means greater surface activity, then an entering substance must possess 
greater surface activity, consequently a lower surface tension, than the con¬ 
stituents of the plasma membrane in order to pass through the membrane. 
It will be interesting to see how the surface-tension values of the saponins 
and the alcohols used compare with the critical value given by Czapek. 
The surface-tension value of 1 per cent, ethyl alcohol is 0*933, °f ^ P er 
cent, alcohol it is 0*720, of 9 per cent. 0-700, and of 10 per cent. 0-682. 1 The 
surface-tension value of 1 per cent, saponin is 0*885, °f 1 P er cent, senegin 
0*819, an d of 1 per cent, smilacin o*68i. 2 It will be seen that while none of 
the surface-tension values of the three glucosides, with one exception, that 
of smilacin, are near the critical toxic value of 0*685 set by Czapek, yet the 
tension values of the glucosides and of all the more prominently toxic con¬ 
centrations of alcohol are considerably less than 1. 
It is of interest to note that 9 per cent, ethyl alcohol, which we found 
to be a critical concentration in the toxicity of this alcohol for protoplasm— 
since above this percentage cells are killed in relatively few minutes, while 
below it many hours are necessary to kill—has a tension value of very 
nearly the critical value set by Czapek. 
We have seen, however, that a much lower percentage of ethyl alcohol , 
than 9 is sufficient to effect osmosis of the cell contents. As low as a 3 per 
cent., even a 1 per cent., solution will, in several days, materially affect the 
diosmotic properties of the cell. We can only explain the effect of the low 
percentage of alcohol, on the basis of Czapek’s theory, by assuming that 
there is a concentration of the alcohol at the surface of the protoplast. In 
time the alcohol thus concentrated on the plasma membrane will be of 
sufficient strength to have a surface-tension value low enough to cause 
a reduction in the tension of the protoplasmic surface, and consequently 
increase the permeability of the membrane. While such an assumption is 
1 These values are taken from Duclaux (6, pp. 22-3). 
2 These values were obtained by comparing the number of drops of water with the number of 
drops of reagent, of the same volume and temperature, from a Traube stalagmometer. The value 
of water is taken as unity. 
