1881.] The Poisonous Power of Metals. 717 
replaced by limestone, which preserves, in a state of greater 
or less perfection, the form of the decaying root or branch, 
and becomes a complete pseudomorph. This has gone on 
to such an extent that the sand is here and there strewn 
with these pseudomorphs, while it is not difficult to find 
calcified roots in situ, even the most delicate filaments being 
perfectly replaced by brittle limestone. 
XVII. Conclusion. 
I have set down these miniature observations in the belief 
that, trivial as they seem, they have some educational value. 
I am firmly persuaded that they and their like educe and 
develop faculties of observation, and thus throw the mind 
into a fit attitude for observing on a larger scale should 
opportunity occur. 
III. THE POISONOUS POWER OF METALS. 
CH. RICHET has been experimenting on the 
comparative poisonous power of different metals, 
by observing what quantity of each may be dis- 
solved in a given volume of sea-water without proving 
fatal to fishes immersed therein in less than forty-eight 
hours. The maximum dose of each metal in the state 
of chloride, compatible with life for the time specified, he 
names the “limit of toxicity.” The species of fish experi- 
mented upon were Servanus Cabrilla , Crenolabvus Mediterra- 
neus, Iulis vulgaris, and I. Geoffredi. A previous set of ex- 
periments had shown that these fishes were capable of 
surviving for an indefinite time — i. e., longer than eight 
days— in 2 litres of sea-water, placed in a wide crystallising 
pan, open to the air. M. Richet seems to have proceeded 
in the following manner : — Dissolving in the water a large 
excess of the metallic salt in question, and, noting how long 
the fish survived, he gradually reduced the dose till he 
reached a quantity which did not prove fatal. In all cases 
chlorides were used, since a special trial had shown that 
they were less poisonous than the corresponding nitrates. 
