32 Chapters on Animals. 



possible ; but the table was octagonal, and the dog found 

 the letters when her back was turned to her master as 

 easily as when she could look him in the face ; besides, 

 when M. du Rouil was seated, and I was the showman, he 

 did not look towards the dogs at all, but at the fire. What- 

 ever communication did take place must have been entirely 

 by intonations of the voice, but we could hear these as well 

 as the dogs could, and with all our listening we could detect 

 nothing like a regularly recurring and easily recognisable 

 signal. When he asked Blanche to turn_/(:7/ into the plural, 

 he did it exactly with the words and in the manner that 

 you would use to a child at school. He often encouraged 

 the dogs with such words as Allans, allons ! Cherches, 

 chenhez bien ! Vite, znte, vite ! but he went on with these 

 encouragements exactly in the same words and in the same 

 tone after the word was completed to put the dog's knowl- 

 edge to the test, and she went on seeking, and then whined 

 and rang a bell to say that there were no more letters 

 needed. I had been told that Blanche could, of course, 

 spell any word that her master could spell, because she 

 only took the letters he fixed upon, yet he said she could 

 not spell fire for me. This, however, may have been a 

 ruse on his part, and I do not insist upon it. 



If the dogs had appeared to know rather less we should 

 have believed that the knowledge was really theirs, but 

 then they seemed to know too much. Lyda showed us 

 some tricks with numbers, that are familiar to arithmeticians, 

 but clearly beyond the canine comprehension. This satis- 

 fied me that some communication existed, and yet I was 

 utterly unable to detect it. It is clear, therefore, that the 

 dogs understood and acted upon a system of signalling 

 which the intelligence of the human spectators was not 



Allans . . . Vite. Go on, go on ! fetch, fetch it ! quick, quick, quick! 



