PuKNELL. — On True Instincts of Animals. 27 



tives of each species to Eesolution Island if the work is taken 

 in hand at once, and the island placed under the care of a 

 skilful curator. If it he postponed for any length of time, 

 who can say what may occur ? It would require a very short 

 time indeed to destroy every land-bird on Antipodes Island, or 

 on the Snares ; and, now that attention has been drawn to 

 their interest, their value, and to their linnted power of flight, 

 the danger has become urgent. 



If this address should be instrumental in drawing atten- 

 tion to the danger and accelerating the adoption of protec- 

 tive measures it will not have been given in vain ; but I ven- 

 ture to hope that it may be productive of still greater benefit 

 in leading some of those present to investigate the phenomena 

 of change and replacement which are now in progress, and in 

 the results of which we are so deeply interested, before the 

 opportunity has passed away for ever. 



Art. II. — True Instincts of Animals. 

 By Charles W. Purnell. 



[Read before the Pliilosophical Institute of Canterbury, 1st Jfilay, i6'9J.] 



The definition of the term "instinct" has been greatly nar- 

 rowed of late years by scientific thinkers. Formerly, every 

 action of an animal betokening intelligence was attributed to 

 instinct, but latterly the term has been restricted to actions 

 like that of cell-making in the bee, the construction of dams 

 and canals by the beaver, and so forth — actions which are per- 

 formed in an apparently mechanical manner by one generation 

 after another, and seem to be prompted by some other faculty 

 than intelligence. It is now admitted that many acts done 

 by the higher animals must owe their origin to a faculty akin 

 to, if not identical with, human reason ; but the apparently 

 unchanging and invariable nature of such actions as those just 

 mentioned — as the construction of webs by spiders and nests 

 by birds, and the migration of birds — seem to mark off these 

 actions from the variable acts which are done upon the spur 

 of the moment at the bidding of the animal's intelligence. 



I think we can restrict the definition still further. Writers 

 upon this subject have not taken sufficiently into account how 

 much the young animal may be taught by the old, and how 

 much it can learn through imitation and from its own observa- 

 tion. The migratory habits of certain birds, for example, are 

 always set down to " instinct "; but birds usually migrate in 



