authority. Sucli sciences as Botany depended chiefly on 

 observation^ others like Metallurgy on experiment. The latter 

 led to more rapid progress, our knowledge of the constitution of 

 the stars and nebulae being due to the quasi-experimental 

 work of the spectroscope. Nature herself occasionally so to 

 speak performed the experiment, as in transits and eclipses. 

 Success in observation and experiment depended (1) on the 

 absence of bias, e.g., on kinetical principles the relative 

 superiority of a flea's leap was no proof that it was com- 

 paratively stronger than a man. Again, it did not follow that 

 the darkness of a sun-spot implied a colder interior to the sun. 

 (2) The operation should be performed with a distinct object in 

 view ; aimless experiment was of no value. 



Having dwelt on the importance of scientific classification, 

 nomenclature, and terminology to secure a clear co-ordination of 

 knowledge, and -words devoid of ambiguity to name the things 

 classified and describe their attributes, Mr. Petitfourt explained 

 the Inductive Methods of John Stuart Mill, which consisted of 

 five canons : — 



SUMMARY OF THE INDUCTIVE CANONS. 



A. Quantitative Methods. 



I. Method of Agreement (chiefly used in observation ; 

 imperfect, owing mainly to the plurality of causes). 



The sole invariable antecedent of a phenomenon is probably 

 its cause. 



Example. — The iridescent colours of (a) mother-of-pearl ; 

 {h) impressions of mother-of-nearl on wax ; (c) diffraction gratings, 

 are due to the only thing common to the three cases, viz., the 

 nature of the surface. Instances of its application — the Theory 

 of Dew, geological stratification, diagnosis of disease, bacteriolog}', 

 etc. 



II. Method op Difference (chiefly experimental). 



If an instance of a phenomenon and an instance of i<s 

 absence differ only in one circumstance (that one occurring in the 

 former case), this circumj-taiice is causally connected with the 

 phenomenon. 



Example. — When the gas of a soda-water bottle expels the 

 cork, it performs work and thus consumes heat ; this may be 

 observed by the deflection of the thermo-electric pile. (Tyndall.) 



Instances. — All chemical tests and " crucial instances " for 

 deciding between rival hypothesis. 



Caution. — One factor only must be varied at one time. 



