212 The Carnivora. 
full charge. But firing a full charge, without any warning, 
near a nervous dog is likely to set up a terror which he will 
not entirely overcome. His confidence in man is such that 
he has no fear of anything which he thoroughly believes to 
be originated by him. 
There is much ground, of great interest, in animal psycho- 
logy still unexplored, although so much has been done to 
elucidate mental phenomena by such works as Dr. G. J. 
Romanes’ “Animal Intelligence” and “Mental Evolution in 
Animals.” Whether or not, as Comte thought, animals have 
some crude ideas of “fetishism” or “animism ”—a view 
strongly combated by Mr. Herbert Spencer in his “ Prin- 
ciples of Sociology ”—they behave in exactly the same manner 
as many human beings, when inanimate objects move acci- 
dentally, or are made to move experimentally, in some way 
inconsistent with their uniform experience of such objects. 
Dr. Romanes has made some experiments on this, which are 
of so much interest that I trust he will pardon my quoting 
them; especially as, having appeared in Nature (27th Dec., 
1877), the original account may be difficult of access to 
many readers: “The terrier used to play with dry bones 
by tossing them in the air, throwing them to a distance, 
and generally giving them the appearance of animation, in 
order to give himself the ideal pleasure of worrying them. 
On one occasion, therefore, I tied a long and fine thread to 
a dry bone, and gave him the latter to play with. After he 
had tossed it about for a short time, I took an opportunity, 
when it had fallen at a distance from him, and while he 
was following it up, of gently drawing it away from him by 
means of the long and invisible thread. Instantly his whole 
demeanour changed. The bone, which he had previously pre- 
tended to be alive, now began to look as if it really were 
alive, and his astonishment knew no bounds. He first ap- 
proached it with nervous caution, but as the slow receding 
motion continued, and he became quite certain that the move- 
ment could not be accounted for by any residuum of the 
