344 
The Naturalist tn La Plata, 
hand that held the tail had been visible and in the 
cage, the bites would undoubtedly have been inflicted . 
on it ; but no enemy was visible ; yet the fury and 
impulse to attack an enemy was present in the 
animals. In such circumstances, the excitement 
must be discharged— the instinct obeyed, and in 
the absence of any other object of attack the illusion 
is produced and it discharges itself on the struggling 
companion. It is sometimes seen in dogs, when 
three or four or five are near together, that if one 
suddenly utters a howl or cry of pain, when no man 
is near it and no cause apparent, the others run to 
it, and seeing nothing, turn round and attack each 
other. Here the exciting cause — the cry for help — - 
is not strong enough to produce the illusion which 
is sometimes fatal to the suffering member ; but 
each dog mistakingly thinks that the others, or one 
of the others, inflicted the injury, and his impulse is 
to take the part of the injured animal. If the cry 
for help — caused perhaps by a sudden cramp or the 
prick of a thorn — is not very sharp or intense, the 
other dogs will not attack, but merely look and 
growl at each other in a suspicious wmy. 
To go back to Azara’s anecdote. Why, it may 
be asked — and this question has been put to me in 
conversation — if killing a distressed companion is 
of no advantage to the race, and if something must 
be attacked — why did not these rats in this instance 
attack the cage they were shut in, and bite at the 
woodwork and wires ? Or, in the case related by 
]\Ir. Andrew Lang in. Lonrpnan s Magazine some 
time ago, in wAich the members of a herd of cattle 
in Scotland turned with sudden amazing fury on 
