SECTION 13. GENERALISATfpN OR EXTENSION. 109 



In certain departments of knowledge it is hazardous to reason 

 from a few particulars to the general, and in those departments 

 enumerations tend to partake of completeness. For instance, 

 the population of a town or country, at a given date, as well 

 as many other social facts, are determined by complete enume- 

 ration. In the same way rain gauges, thermometrical and wind 

 records, graphs, questionnaires, and similar means are employed 

 for the purpose of obtaining a knowledge of general facts which 

 may afterwards form a basis for deductions. So, too, the number 

 of species of discovered animals and plants is counted, and a 

 census is taken of the host of stars. A related method is ex- 

 pressed by the law of averages where series of facts diverging 

 within certain limits are reduced to unity by extracting the 

 average of the series. Complete inductions play also an im- 

 portant part in textual criticism. The various Bibles and Classics 

 of the world for instance, the sacred books of the East have 

 been thus subjected to a treatment where almost every detail 

 is exhaustively enumerated, and nothing is taken for granted. 

 In these several cases the end general statements is identical 

 with ihat in incomplete inductions (which, after all, often tend to 

 approach completeness), and both kinds of inductions represent 

 methods utilised in science. Individuals are to be met with in 

 every walk of life who, if a moderate additional effort will secure 

 it, prefer the certain to the uncertain, and frequently perform 

 complete inductions, or the most complete ones practicable, 

 where they might have been satisfied with relatively incomplete 

 ones. Their procedure is to be commended methodologically. 1 



45. A generalisation may, therefore, only be legitimately 

 attempted where appreciable time and thought would be saved 

 by its being posited. If, for example, some one desired to under- 

 stand the fundamental nature of the sensations, he would do 

 well to draw up, before venturing on a comparison between the 

 sensations with a view to learning how far they are distinct, 

 very complete lists of the leading characteristics of the various 

 senses, afterwards treating them synthetically in accordance with 

 Conclusions 14 to 35. In a preliminary survey utilising 172 

 more especially, and seeking pertinaciously by observation, 

 experiment, and comparison, for new points until no further 

 points reveal themselves- he might enumerate the peculiarities 



1 On complete induction, see Bradley's Logic, pp. 329-330. 



- Here is the vital test of a methodology. Granted fair acquaintance with 

 the subject of Sound as a whole, most of the material should be directly 

 derived through the application of methodological rules. E.g., Conclusions 19 

 and 20 would inform us where to find our material; the second Table 

 of the Primary Categories would direct us to other important aspects; and 

 Conclusions 27 and 28 would further ensure exhausting the subject. Also, 

 since the aim is to compare, comparison will be a fertile source of sugges- 

 tions. The task should be thus completed in one-tenth of the time and be 

 of ten times greater scientific value. This is presumably what Bacon meant 

 by "helps" for the understanding. 



