274 Raymond H. Pond. 
with the plain reagent, indicating that the concentration of the 
enzyme product is too slight to displace the point of neutralization. 
A comparison of the figures for the boiled solution in this test with 
those for the control in preceding zinc tests shows that the latter 
tend to run a little higher. This must be due to slight acidity of the 
.ethylbutyrate rather than to any enzyme action of the boiled extract. 
By way of recapitulation it may be said that the tests show a very 
marked grouping of the metals according to toxicity, and that the 
limits are sharp and do not overlap even when the extreme allowance 
for error is made. Thus it does not seem quite certain that lead and 
zinc are of equal toxicity; however, copper can hardly be classed by 
itself and certainly not as approaching equality with silver. On the 
€quinormal basis mercury is certainly eight times more toxic than 
silver and probably sixteen times. Silver is sixteen times as toxic as 
copper, lead, or zinc. Cadmium and cobalt are only one fourth as 
toxic as copper, lead, or zinc, and only twice as toxic as barium, 
Strontium, magnesium, or ammonium. This arrangement of the 
metals according to toxicity disagrees with that of Mathews, 
McGuigan, and Caldwell in several particulars, most notably in the 
relative toxicity of mercury and silver. All three of those men found 
silver to be more toxic than mercury. 
If the degree of toxicity of a series of metals varies inversely with 
the solution tension of those metals, the relative toxicity of any pair 
would be theoretically and approximately, at least, the reciprocal of 
the relative solutién tension. Thus, copper, having about one fifth 
the solution tension of lead, should be five times more toxic. In the 
following table, in the column headed “ Theory,” are given a few of the 
ratios derived from the table of solution tension as quoted by Mathews. 
In the column headed “ Found” are given the ratios resulting from 
my own experiments. The figures are based upon equinormal con- 
centration, since Mathews’s table of the values for the respective 
Solution tensions of the metals is for those metals in equinormal 
solutions. 
weeny UN GHEE CS ee ee pe ee eS Bc 
Copper ann Wan 0 SS ee a ORE ee 
Seed abd COUN a a act 
LemOn and GAO) oc es ee ee ee Bet 174 
ane inl cadmniths 6 6 3 2 a 64:1 
