KJG FROM lU.OMIDDX TO t^MOKY. 



captives t:i])]io(l a great tleal, and I gave them 

 a variety of things to phiy upon, as, for example, 

 a sweet-toned glass tumbler, thin sheets of zinc, 

 and resonant pieces of wood arranged to give 

 out various tones. They tested these things, but 

 seemed to prefer the sides of their cage, espe- 

 cially portions walled with clapboards, which 

 yielded a great volume of sound to their blows. 

 I spent many hours in noting down the number 

 and order of their taps, iti order to see whether 

 they constituted distinct signals. At first it 

 seemed to me that Number One liked to tap in 

 twos and fours, that Number Three was more 

 apt to make threes, or threes and fives, than 

 other combinations, and that Number Two 

 mingled fives and twos. The longer I listened, 

 the more combinations I found them making, 

 and at last I decided that with these birds it was 



mere chance whether they said or 



or They seemed to pay no attention to 



each other's performances, and to mean nothing 

 by their own tappings. If they tapped at all, it 

 was necessary to make some number of taps and 

 to space those taps in some particular way. If 

 in a large number of such series, ones, twos, 

 threes, fours and fives came equally freely and 

 frequently — as they seemed to — there ap- 

 peared to be no ground for imagining that the 

 different combinations indicated different feel- 



