PROPAGATION OF FISH. 239 
not swallow him. Being some hundreds together, they 
strnck at him as they would at a stick if thrown upon 
the w^ater, but immediately dropped and let him escape 
to land. Although I tried the experiment over and over 
again, they had found out what he was, and now would 
not touch him. 
To complete my enlightenment, I only wanted an 
experience which was destined soon to be furnished. I 
had au aquarium in which were kept, among a quantity 
of small minnows and other fish, a frog that had ex- 
panded from a tadpole, together with several tadpoles 
that were expanding rapidly. The frog had attained a 
respectable size, and was a great favorite, from the readi- 
ness and suddenness w4th which he seized and devoured 
flies offered to him, and was endeared to our hearts by 
several hair-breadth escapes, such as jumping out of the 
aquarium and being lost for days rouud the room, and 
even falling out of the window, through the area grating 
into the cellar. He was wondrously solemn, but had a 
way of darting on a fly that was invariably fatal. We 
began to observe, however, that our small fish disap- 
peared strangely, " leaving not a wreck behind," and 
always at a time when the frog seemed to have suddenly 
grown in circumference. This continued till my suspi- 
cions connected the two together, and one day we abso- 
lutely saw him seize and swallow a minnow half as long 
as his own body. His j^lan was to remain perfectly 
motionless till his desti^Pl V^^Y swam near, when he 
would make one spring, and devour it at a gnlp. In 
spite of his cannibal propensities, w^e kept him, although 
we could often see the shape of a fish under his extended 
