698 RALPH S. LILLIE 
lem of inducing parthenogenesis in unusually resistant eggs re- 
solves itself into a choice of means by which the requisite rapid 
and reversible increase of permeability may be effected. I have 
considered the conditions in Asterias in more detail elsewhere.® 
In the present paper the description will be confined to experi- 
ments with Arbacia eggs. 
The following experiments were suggested by the foregoing 
theoretical considerations. If the essential action of the above 
salts, acting as parthenogenetic_.agents, is to cause a rapid initial 
increase of permeability, membrane-formation and the initiation 
of cleavage should be prevented by the addition of other salts 
that oppose the permeability-increasing action. In studying the 
toxic and antitoxic action of various salts on the pigmented larve 
of Arenicola I had observed that the effect produced by an anti- 
toxic salt like calcium chloride in diminishing the toxicity of pure 
solutions of sodium salts was always associated with a prevention 
of the marked and rapid increase of permeability,—indicated by 
loss of pigment—which the pure solution typically produces.® 
Now it seems clear that any marked and unchecked increase of 
permeability, sufficient to destroy the normal osmotic equilibrium 
and allow abnormal diffusion of substances into and out of cells, 
must derange and eventually destroy the chemical organization 
of the latter. It is, I believe, essentially for this reason that 
pure solutions of sodium salts are highly toxic; but if to such solu- 
tions small quantities of calcium or other favorable salts are added, 
the permeability remains normal or undergoes only gradual alter- 
ation so that the cell is enabled to retain its normal properties 
for greatly prolonged periods. The antitoxic salt thus acts by 
preventing or checking increase in permeability. I have already 
mentioned that pure isotonic solutions of neutral sodium or potas- 
sium salts acting on unfertilized Arbacia eggs produce a similar 
increase in permeability with loss of pigment, which is most rapid 
in iodide and sulphocyanate solutions; this effect, together with 
the associated toxic action, is greatly checked by adding a little 
* American Journal of Physiology, 1911, vol. 27, p. 289. 
® American Journal of Physiology, 1909, vol. 24, pp. 23-25. Other observations 
not yet published confirm these results. 
