ARISTOTLE S METHOD. 23 



i. c. 5, 645a. In that chapter he says that these ought to 

 be carefully studied, not omitting even the lowest forms of 

 life, which, even if not attractive in themselves, show 

 Nature s handiwork, and afford pleasure to those who inquire 

 into the causes of phenomena and are interested in philo 

 sophy. We ought not, he says, to turn away from an 

 investigation of the lower animals, for every part of Nature 

 reveals something to admire, and, just as Heraclitus, warm 

 ing himself by his kitchen fire, was reported to have told 

 the strangers who called to see him not to be afraid to enter, 

 for gods were present even in his humble dwelling, so 

 Aristotle invites us to study every kind of animal, without 

 being ashamed, for all of them show something natural and 

 beautiful. 



Then, with respect to the manner of reasoning on the 

 facts obtained, Aristotle seems to proceed on principles 

 equally sound. He asserts that we commonly conduct an 

 inquiry, not with reference to the question discussed, but 

 with reference to the opponent who argues the question 

 with us, and that, if there is no opponent, we conduct the 

 inquiry until we can satisfy our own objections. Therefore, 

 he proceeds to say, he who intends to investigate completely 

 any subject must take care to satisfy himself on all diffi 

 culties arising out of the subject, and this can be done only 

 after he has examined all differences of opinion on the subject 

 of inquiry.* 



The above is a brief account of Aristotle s method, as it 

 is set forth in his writings. It might be expected that, after 

 laying down such excellent rules, the results obtained by 

 him would have been uniformly trustworthy, but this was 

 not so. His own practical application of the method was 

 defective. He recognized the importance of a preliminary 

 ascertainment of facts, but he did not appreciate that there 

 were many natural phenomena about which very numerous 

 observations must be made before any generalized statement 

 of them, or any theory explaining them, could be formulated. 

 It must have been necessary for him, just as it has been for 

 investigators since his time, to decide how many observations 

 ought to be made before the generalizing or theorizing pro 

 cess could be safely carried out. There are many indications 

 in his writings on the Natural Sciences that he erred in 

 being satisfied with an insufficient number of observations. 

 Further, he was unaware how necessary it was to make 



* De Coelo, ii. c. 13, 294&. 



