POWER OF CALCULATION. 457 



own feast began. This incident contrasts with the other, 

 first mentioned, of the sparrows and other birds that were 

 candidates for the morning charity of a lady in Edinburgh. 

 In the latter case the pensioners, who were as regularly ex- 

 pected and provided for as various human pensioners by the 

 same benevolent heart, showed their confidence in her, their 

 absence of fear, by feeding in her presence. 



Ducks, dogs, donkeys, and many other animals are simi- 

 larly regular or punctual in their attendance at man's dinner 

 hour. 



Macaulay mentions a Newfoundland dog that visited a 

 baker every morning, save Sunday, as the clock struck eight. 

 One of the commonest tricks which the dog is taught is to 

 call (or awake) its master at a fixed hour in the morning. 



The mules employed in loading and unloading vessels on 

 the quays of New Orleans know the duration of their work 

 the length of time, the number of hours, during which they 

 are called upon to labour neighing to their masters as a 

 signal when it is time to be unyoked (Houzeau). 



A London barrister one of the staff of a well-known 

 provincial newspaper told me of a certain cat of his that 

 used to meet him regularly at a certain hour on a certain 

 road, on his way home from office. In such a case as this it 

 is difficult to resist the conclusion that it must have had some 

 means of reckoning time. Wild dogs in Peru meet at certain 

 hours, at certain places, for certain purposes (Pierquin) . The 

 tame orang has, or gets into, regular hours for going to bed 

 at night and rising therefrom in the morning (Houzeau). 

 Blind beggars' dogs visit church-doors at the proper hours 

 (Low). 



It has been alleged by Houzeau- with what degree of 

 truth does not quite appear that certain hours are observed 

 by the cock in its crow, whether it be light or dark in other 

 words, that unnatural and artificial continuity of light or 

 darkness does not affect its periodicity of work and rest. 

 Statements, however, of an opposite kind have been made, 

 and are much more probably true. It has been abundantly 

 proved, indeed, that artificial light and darkness produce 

 certain curious results that are to be considered errors of 



