1 8 E VOL UTION OF BIRD-SONG 



lungs by an action akin to swallowing, and retain- 

 ing it there by means of the closing of the glottis. 

 After spending some months in the water, they 

 emerge once more, to stay on land for the re- 

 mainder of the year. But during all this time they 

 never utter any sound which can be considered as 

 vocal, though, in the act of breathing, their jaws 

 sometimes close with a sort of snap. Persons who 

 have kept newts in aquaria are generally aware 

 that, when a newt is held by the tail it wriggles 

 violently, and often, when so struggling, utters a 

 brief but distinct croaking sound. The noise is, no 

 doubt, caused by the body of the creature striking 

 against the fingers, by which action the lungs are 

 compressed, and consequently the contained air is 

 forced through the tightened glottis, and thus pro- 

 duces a croak. Perhaps the gigantic ancestors of 

 our newts did not struggle quite so actively, but 

 any sounds, proportionately louder, thus produced 

 in their contests, may have been of service by 

 occasioning the release of amphibians from the grip 

 of predacious enemies to whom such outbursts were 

 unfamiliar. When a newt is seized towards the 

 hind quarters by a snake the croak is heard ; but 

 I cannot say that I have ever known it to be of 

 any avail against such an attack, and I have 



