718 
The Poisonous Power of Metals. [December, 
The sulphates he rejected as being too sparingly soluble for 
such experiments. In this manner he worked through mer- 
cury, copper, zinc, iron, cadmium, ammonium, potassium, 
nickel, cobalt, lithium, manganese, barium, magnesium, 
strontium, calcium, and sodium, and found their toxicity 
decrease in the order here given. Thus whilst the largest 
quantity of mercury in the state of perchloride, compatible 
with life, was 0*00029 § rm * P er litre, in sodium the limit 
rises to 24*17 grms. in the same volume of water. Several 
of the most poisonous metals, such as arsenic, antimony, 
and lead, have not been experimented with. 
The author calls attention to the fadt that, according to 
his results, there is no definite relation between the atomic 
weight of a metal and its toxicity. Copper, according to his 
table, is 600 times more poisonous than strontium, though 
its atomic weight is less. Lithium, whose atomic weight is 
only one-twentieth that of barium, is still three times more 
poisonous. Even in metals of the same chemical group the 
relation between atomic weight and toxicity does not appear. 
Thus cadmium (atom, weight 112) is by one-half less poi- 
sonous than zinc (atom, weight 65). Lithium (17) is seventy 
times more poisonous than sodium (23). 
The first point which strikes us in examining the results 
of M. Richet is their want of agreement with the conclu- 
sions reached by Dr. James Blake (“ Journal of Science,” 
1881, p. 318). This experimentalist found that, as regards 
bodies belonging to one and the same isomorphous group, 
the higher the atomic weight the more intense is the physi- 
ological adtion. M. Richet finds no such relation, as may 
be seen from the last two instances given above. He de- 
clares also that there is no relation between the chemical 
fundtion of a body and its toxic power, giving as an instance 
the experimental result that potassium is 250 times more 
poisonous than sodium. 
But it would be, we think, premature to pronounce any 
decided opinion concerning these and other discrepancies 
which might be particularised. The methods of experi- 
mentation in the two cases are totally different. M. Richet, 
as we have said above, immersed his subjects in poisonous 
solutions, and simply noted the fadts of death or survival, 
and the time in which the treatment proved fatal, if at all. 
Dr. Blake injedded known doses into a vein or artery, and 
recorded not merely the fatal quantity, but the nature of the 
symptoms produced, and noted the curves described by a 
kymograph connedted with the femoral artery. His re- 
searches, too, have extended to forty-one of the metals, and 
