573 
Electrolytes upon Cells of Saxifraga uinbrosa. 
Sb 
protoplasm to take place earlier than the calculated time. This tendency 
had already been found in experiments in dilute solutions at room 
temperature, but it took place with more concentrated solutions at 25 0 C. 
than was the case at the lower temperature. The equations shown are 
therefore only suggested as applying for a time range of ten minutes to 
four hours at this temperature. Table V shows the constants collected for 
the series of potassium salts and for that of sodium salts. It will be seen 
that ' k' varies from 0-67 to 2*3, while K lies between 0*23 and 2-41. It 
is possible that the percentage error could have been reduced by still more 
care in selecting material so that there was greater uniformity in thickness 
of the cell walls, as this must have been a factor influencing the diffusion. 
It is interesting to examine the constants in the similar equation already 
found by other workers. In the work of Harvey ( 3 ) upon Chlamydomoncis 
immersed in resorcin of various concentrations, the criterion used was 
cessation of movement. In this case 'k* was 1 -2 1, while K had a value 
between 3-75 and 3*53. 
In the paper of Watson ( 2 ) the figures obtained by Chick for dis- 
infection of bacteria were examined. Here the arbitrary measure of change 
was loss of power to multiply. The liquids used were phenol, silver 
nitrate, and mercuric ^chloride. The co-efficient ‘/£’had values from o*86 
to 5-5 for different substances, while the constant K fell between 
3 and 6-8. 
In the paper of Harvey it is stated that the equation in question is one 
obeyed in a chemical reaction where one molecule of one compound reacts 
with ‘ h ’ molecules of another compound, which is present in great excess, 
when the concentration ‘ C * is made to vary and T is the time for the 
reaction to reach completion in each case. In the results recorded here 
it is clear that equimolecular solutions of two salts with monovalent kations 
and similar anions produce different effects, while marked differences occur 
if the kation be kept constant and the anion varied, so that simple chemical 
action does not seem to explain the changes produced. Of recent years 
great interest has centred round the absorption of ions by colloidal matter 
and consequent precipitations owing to electrical effects. The proteins of 
protoplasm are known to fall into the collodial category. 
Precipitations of colloids by electrolytes show many complications. 
There is a tendency for increased adsorption as the valency increases, but 
valency alone is not enough to consider. Wilder Bancroft ( 5 ) suggests 
there is ‘ specific adsorption 5 so that the concentration of a given electrolyte 
necessary to neutralize the charge upon a given colloid depends upon the 
nature of both kation and anion. According to Hardy (6) when we deal 
with living matter, or with organic compounds with large molecules, the 
influence of the electrolyte is determined not only by the charge carried 
by the ion but by the volume of the ion also. Further, precipitation 
