SECTION 13. GENERALISATION OR EXTENSION. 105 



than creating, generalisations. Moreover, without an accepted 

 principle of classification to guide us, and a copious number 

 of generalisations to mark out for us the limits of our enquiry, 

 the process of generalisation would be probably to all intents 

 meaningless and vain, indeed impossible. (See Section V and 

 Conclusion 33.) 



41. We ought to distinguish at least three classes of gene- 

 ralisation. First, simple generalisation, where what js asserted 

 of a phenomenon, say a new fact about a certain colour, is 

 extended to a second colour or to its own highest class, the 

 sense of sight. Secondly, compound generalisation, where what 

 is asserted of a class, e.g., a novel fact concerning the sense 

 of sight, is extended to more or less closely related classes to 

 some or all the remaining senses, etc. And, lastly, universal 

 generalisation, where we extend to what is remotely related, 

 as reasoning from the nature of the senses to the memory, 

 and thence to the brain and, beyond, to life and matter in 

 general. 1 



42. It is also not unconditionally true that science is only 

 concerned with general facts. The chemist frequently reasons 

 from one substance to a second substance or to a small group 

 of substances. The physicist not seldom endeavours to connect 

 one force with one other, as magnetism with electricity, light 

 with heat, and light with electricity (as in Clerk-Maxwell's electro- 

 magnetic theory of light), or argues from the existence of a 

 magnetic field to the existence of a gravitational field. And, gene- 

 rally speaking, there are myriads of occasions when a scientific 

 extension does not pass beyond a second or a few facts. More- 

 over, astronomers will observe a single eclipse or a single star ; 

 physicists reduce one gas after another to the liquid and solid 

 state; 2 chemists add one element to another; seismologists will 

 inquire into the causes of a certain earthquake or volcanic 

 eruption ; geologists will describe the strata of a certain region ; 

 anthropologists will compose a monograph on a single tribe; 

 and economists will investigate the economic condition of a 

 particular district of a particular country at a particular time. 

 There is, in other words, a cumulative as well as a generalising 

 aspect to scientific enquiries, the former of which is well illus- 



1 Here is a broad generalisation, summing up the general nature of all 

 waves, whether connected with light, heat, sound, etc.: 



"1. The disturbance takes time to travel from one point to another. 



2. The disturbance is propagated through a medium. 



3. On meeting an obstacle the waves are reflected back, and the angles 

 of incidence and reflection are equal. 



4. The course of the waves is changed, i.e., they are refracted, when 

 they pass from one medium to another in which the rate of travel is different. 



5. The disturbance of a particle of the medium is alternating and not 

 continuous in one direction." (J. H. Poynting and J. J. Thomson, Sound, 1913, 

 pp. 3-4.) 



2 This has now been accomplished as regards all gases. 



