54 THE METHOD OF SCIENCE 



present, however, I desire merely, but as strongly as possible, to 

 fix the reader's attention on the theory of scientific inquiry and 

 obtain his assent to it. He must bear in mind that the method 

 of work on which I insist is no new thing. It is the method 

 which has been proclaimed as the only right one by every 

 authority who has both thought about and expressed himself on 

 the subject ; it is the one which has been followed by every great 

 scientific worker ; it is the one by which the truth of every law of 

 nature known to us has been established. At the risk of tiresome 

 reiteration, this method of work may bejndicated as follows: 

 When, as in the study of heredity and evolution, we wish to 

 establish the relations between facts that occur in sequences (i.e. 

 when we wish to discover chains of cause and effect), we must 

 formulate hypotheses. But hypotheses remain mere guesses until 

 they have been tested and found to be true. They can be tested 

 only by making deductive inferences of consequences from them 

 and then appealing to reality to ascertain if these consequences 

 actually follow. If a hypothesis enables us to predict correctly in 

 this way, it is in all probability true ; and it is the more certainly 

 true (and useful) in proportion to the number of different correct 

 predictions it enables us to make in proportion as it accords 

 with the whole universe of relevant truth. Moreover, such deduc- 

 tive appeals to reality not only enable us to test hypotheses and 

 so convert them into laws, they not only enable us to bring within 

 the range of each real law masses of data which had not been 

 used for the corresponding hypothesis, but they enable us to 

 perceive the relations between the several laws, and so, by linking 

 the laws together, to advance towards that harmony of knowledge 

 which is one of the principal aims of science. 



84. To return to the problem of recapitulation the facts are 

 that the offspring of each successive generation resemble their 

 parents closely but not exactly, and that, apart from the differen- 

 tial play of stimuli, this imcomplete resemblance is due to a 

 recapitulation with variations of the parental development. The 

 induction is that the development of the individual is a recapitula- 

 tion, with additions and omissions, of the evolution of the race. 1 



1 The thinking here is somewhat complex, and complex thinking is often 

 denounced as ' deductive/ especially by medical writers. But when we bear in 

 mind that individual progression depends on a complete recapitulation (as regards 

 the parts under consideration) of the parental development plus an additional 

 step, and retrogression on an incomplete recapitulation, and that racial pro- 

 gressions and retrogressions are nothing other than the sums of progressions and 

 retrogressions made by individuals during the evolution of the race, the thinking 



