ANIMAL INTELLIGENCE. 11 
unconscious automata, they continue so for more than a very 
brief period—that they are, as it were, delicate and ingenious 
pieces of clockwork, performing with regularity those func- 
tions for which they are designed and adapted, so long as 
they are regularly wound up, i.e., fed. Experience, whereof 
the effects are manifest in every animal sufficiently highly 
organised for man to interpret its behaviour, and which may 
exist in the grades of life so low as to baffle human scrutiny— 
experience, I say, and instruction, whereof very few, if any, 
vertebrate animals are insusceptibie,! are undoubtedly agents 
upon animal behaviour predicating a mental process such as 
could be implanted in no mere machine. To take a very 
homely illustration: no amount of repeated battering will 
prevent a humming top bumping itself against furniture and 
other obstacles when it is set spinning; but one recognises 
the effect of experience upon the conduct of animals so low 
in the scale of life that it is difficult to believe that any sentient 
creature can be totally devoid of conscious volition. 
In 1873 Dr Mobius reported to the Society of Natural 
Science for Schleswig-Holstein some observations by Herr 
Amtsberg of Straisund on the behaviour of a large pike. 
Being confined in an aquarium, this fish wrought such havoc 
among other fish in the same tank that Herr Amtsberg 
caused it to be separated from them by a sheet of plate-glass. 
Thereafter, every time the pike made a dash at one of its 
neighbours, it received a severe blow on the nose. The pre- 
datory instinct was so strong that it took three months to 
convince the pike that every attempt upon the life of these 
small fish resulted in pain to itself. Thereafter it let them 
alone, even when, after six months, the glass partition was 
removed. Experience had taught it that these particular 
fish could not be attacked with impunity, whereupon its in- 
telligence came into play to control its predatory instinct, 
1 It is a popular belief that guinea pigs are not susceptible to 
instruction, and evince no recognition of one human being as more 
familiar than another. Probably this is no more than sheer asser- 
tion, founded on the phlegmatic behaviour of the animal in cap- 
tivity, and not put to the test of experiment. 
