December 11, 1914] 



SCIENCE 



835 



the problem in which he is interested. But 

 the results which he obtains are permanent; 

 these in turn he may use to arrive at tentative 

 conclusions concerning the other parts of his 

 problem. 



In like manner it may be necessary that a 

 theory in esperimental science should restrict 

 itself to a certain point of view in order to 

 remain scientific. The range of phenomena, 

 even in a restricted field, may be too great 

 to be taken account of at once. Therefore 

 some elements are left entirely out of mind 

 until considerable progress has been made with 

 the investigation. This was done in the case 

 of the kinetic theory of gases, the size of the 

 molecules being taken into consideration only 

 after extensive investigations had been made 

 in which this element was ignored. 



Such a plan of procedure will cause us no 

 uneasiness if we remember the guiding pur- 

 pose of physical science. It does not attempt 

 to afford us an explanation of the essence of 

 things; if it did so it would find itself amidst 

 inexplicable difficulty from the beginning. 

 Its purpose, on the other hand, is to give 

 direction to our researches into details and to 

 afford us the best means of acting on things 

 and of predicting phenomena. 



It may very well happen that a " false " 

 theory will serve this purpose better than a 

 " true " one. In other words, a theory which 

 is in agreement with only a narrow range of 

 facts may be better for us at a given time than 

 one which agrees with a much wider range. 

 The one more nearly perfect, in the absolute 

 sense, may be out of reach of our proper under- 

 standing or at least beyond any means of ex- 

 perimental verification at our command. 



As a first example of this let us consider 

 the case of a savage who has been accustomed 

 to take the animistic view of nature. It may 

 very well be true that his primitive theory 

 brings helpful ideas and enables hira to get 

 around in his world and interpret it in a satis- 

 factory way. His observations have little of 

 precision about them and consequently they 

 do not clash with his theory. To this creature 

 the Newtonian law of gravitation would be 

 meaningless and useless. For him it could 



serve none of the ends for which we employ 

 that or any other scientific theory. For him 

 to make such a hypothesis as this would be 

 distinctly unscientific. 



Another case in point is the old Greek theory 

 of which we spoke a few moments ago. Ac- 

 cording to it all matter had a unique origin, 

 and a primary task of the philosopher was to 

 discover what substance gives rise to all others 

 by the combination of its parts. None of the 

 answers which they were able to arrive at, as 

 we have seen, were of such character as to give 

 them greater power to act on things or to pre- 

 dict phenomena. In accordance with a true 

 scientific instinct the theory was therefore al- 

 lowed to drop out of mind. Nowadays it has 

 been revived in a different form because in 

 this form it now seems capable of being sub- 

 jected to experimental examination. 



Probably the best example of the difficulties 

 of a position where speculation has outrun ob- 

 servation is afforded by the atomic theory of 

 the ancients, a theory which is very close in 

 its general aspects to that which is usually 

 accepted at the present day. In recent 

 times this theory has given rise to the most im- 

 portant and far-reaching investigations. It has 

 in a remarkable degree all the characteristics 

 of a useful theory, which we enumerated 

 above, and in many ways has proved itself 

 vital in experimental investigations. Among 

 the ancients, however, it seems to have 

 led to nothing but speculations and disput- 

 ings. It was too far in advance of other parts 

 of scientific theory to be amenable to experi- 

 mental investigation. Though essentially in 

 agreement with facts, as we understand the 

 matter to-day, it yet led to no scientific con- 

 quests in ancient times. 



Such examples as these remind us that we 

 should not set ourselves the task of finding 

 the " true " explanation of things. From a 

 scientific point of view our plans should be 

 far less ambitious. This is a point, it seems 

 to me, which we should be careful not to lose 

 sight of. What we want to do is to frame 

 general laws which to us appear to be the 

 simplest we can find and which have the fol- 

 lowing three properties: they are in accord- 



