April 17, 1914] 



SCIENCE 



557 



at once that the issues between such doc- 

 trines as vitalism and a mechanistic account 

 of nature appear, from the point of view of 

 Maxwell's classification, in a somewhat un- 

 familiar perspective. For one need no 

 longer merely contrast two views, the me- 

 chanical and the vitalistic. One now has a 

 third and a mediating point of view to com- 

 pare with both of them. The result is in- 

 structive. 



Vitalism, whatever else it involves, al- 

 ways makes prominent some aspect of na- 

 ture, and in particular some aspect of or- 

 ganic nature, such that this aspect is sup- 

 posed to be, in some individual case, strictly 

 historical. If an organism is due to a pur- 

 posive process, if the reactions of an or- 

 ganism are, in any instance, events of the 

 nature of conscious or of subconscious 

 deeds — then something unique, historical 

 and novel occurs whenever one of these 

 vital processes is exemplified by an individ- 

 ual event. 



On the other hand, if the mechanistic 

 view of nature can exhaustively express 

 the real facts, then the only natural events 

 are of the type which the eclipses exem- 

 plify. The single events are, so to speak, 

 points on a curve, selections from an ideal 

 continuum whose constitution is definable 

 in terms of an invariant differential equa- 

 tion. 



But the third or statistical mode of view- 

 ing nature takes account of another aspect 

 of the processes of nature. The world of 

 the statistical view still contains events sup- 

 posed to be unique and individual; but 

 from the statistical point of view the main 

 interest lies no longer in each event as it 

 occurs, nor yet in its unique character. 

 The statistical interest is directly con- 

 cerned with a set or aggregate of events, 

 with a discrete multitude of occurrences. 

 These occurrences may prove to be ex- 

 amples of law. The statistical view is 



deeply interested in finding that they are 

 examples of law. But the law for which 

 the statistical method seeks is no longer a 

 law that is ideally statable in terms of an 

 invariant differential equation or in terms 

 of any other timeless invariant. When 

 found, the statistical law is an account of a 

 collection of facts in terms of averages in- 

 volving many events. 



This account takes some such form as 

 saying: "The average magnitude or ve- 

 locity or size or range of the events 

 of the class C is approximately F." Or, 

 again, the statistical view succeeds when 

 we can say: "A proportion which is 

 approximately p of the events of the class a 

 have the character 6." Or finally, one ex- 

 presses the statistical view when one is able 

 to assert: "There is a probability q that c 

 differs from d by not more than such and 

 such an amount, — ^say X." All such gen- 

 eralizations, where the objects in question 

 are living organisms, relate to events, but 

 neither to merely historical single events 

 nor to events subject to fixed laws. The 

 statistical laws are probable and approxi- 

 mate laws about numbers of events. 



Laws and probabilities, stated in some 

 such form as the one just suggested, con- 

 stitute the characteristic formulas of the 

 statistical view of nature. 



It is easy to illustrate how the statistical 

 view contrasts with both the mechanical 

 and the historical point of view by consid- 

 ering how each point of view applies to an 

 event such as is expressed by the assertion : 

 "A killed B." 



For a strictly historical point of view 

 this event, this homicide, is an unique oc- 

 currence — possibly a free-will act. It falls 

 under moral and criminal laws, but these 

 relate only to its value and its legal conse- 

 quences. The interest of the case for a 

 judge or a jury lies in its novelty — and in 

 its uniqueness. For a strictly mechanical 



