112 
tion. Five years later Sumner published his 
observations that the action of copper salt can 
be deferred through the addition of cane- 
sugar, which is of course similar to the ob- 
servation by Gies and myself. In the ease of 
the antagonization of ZnSO, by another 
electrolyte we are, however, dealing with the 
action of both electrolytes on the same col- 
loid. 
2. Sumner states also that distilled and fresh 
water are toxic for Mundulus and that there 
exists an antagonism between distilled water 
and salts for these fish. The fact that a 
number of Fundulus can live a long time (if 
not indefinitely) in distilled water and that 
these fish, if they become landlocked, can live 
indefinitely in fresh water indicates that the 
distilled or fresh water are not in themselves 
toxie for these animals but that the toxic ef- 
fect occasionally (but not always) observed is 
due to an inconstant or quantitatively vary- 
ing constituent of the water. This constituent 
may be a parasite, or it may be a substance 
given off by the fish itself, e. g., CO.. 
Wasteneys and I have recently found that 
CO, may produce the same changes on the 
skin and the gills of the fish as those produced 
by mineral acids; and that, as in the latter 
ease, the etching effects of the CO, may be 
counteracted through the addition of a neu- 
tral salt. The beneficial effect of the addition 
of some salt to the fresh or the distilled water, 
therefore, indicates that the salt either kills 
certain parasites contained or developed in the 
distilled water, or antagonizes the toxic ef- 
fects of some electrolyte, e. g., carbonic acid, 
if its concentration exceeds a certain limit, as 
it possibly did in some or all of Sumner’s ex- 
periments. 
Jacques LOEB 
ROCKEFELLER INSTITUTE 
THE PERMEABILITY OF PROTOPLASM TO IONS AND 
THE THEORY OF ANTAGONISM 
EVIDENCE was recently presented which 
showed’ that a great variety of salts readily 
enter living cells and that antagonism between 
salts may be due to the fact that they mutu- 
1 Scrmnce, N. S., 34: 187, 1911. 
SCIENCE 
[N.S. Vou. XXXV. No. 890 
ally hinder or prevent each other from pene- 
trating the protoplasm. 
In these experiments plasmolysis was the 
criterion of penetration. Plasmolysis shows 
which salts enter and how rapidly, but does 
not indicate whether it is the ions or only 
undissociated molecules which penetrate the 
cell. To decide this question experiments 
were performed to test the electrical con- 
ductivity of living tissues in various solutions. 
The results agree in showing most conclu- 
sively that ions readily penetrate living proto- 
plasm and that many ions which penetrate 
quite rapidly in pure solutions may be hin- 
dered or prevented from going in by the addi- 
tion of small amounts of CaCl, or other salts. 
To obtain reliable results in conductivity 
experiments material should be used which is 
not injured by weak currents or by other 
experimental conditions. It is desirable that 
the amount of space between the cells be 
constant so that the current which passes 
between the cells may be a constant frac- 
tion (as small as possible) of that which 
actually traverses the living protoplasm. The 
current should pass through a large num- 
ber of thin sheets of living tissue, sepa- 
rated by thin films of solution. The pene- 
tration of various ions may then be studied by 
merely changing the solution. If the material 
is in thin sheets the ions are forced by the 
alternating current to pass in and out of a 
great extent of protoplasmic surface; this is 
of great importance, since the larger the sur- 
face the more reliable the measurement. The 
sheets of living tissue should be sutticiently 
rigid to permit manipulation and to endure 
without injury pressure sufficient to pack them 
firmly together so that the films of solution 
which separate them may be as thin as possible. 
All these conditions are admirably fulfilled 
by the common kelps of the Atlantic coast 
(species of Laminaria). This material was 
accordingly used throughout the investiga- 
tions. 
Disks about 13 mm. in diameter were cut 
from the fronds by means of a cork-borer 
(the average thickness of the frond was about 
0.5 mm.). From 100 to 200 of these disks 
