324 AIM AND PEOGRESS OF PHYSICAL SCIENCE. 



me but recently, and that too in the course of my summer 

 holiday in the mountains. The gaps which I may leave will 

 at all events be abundantly tilled up by the proceedings of the 

 Sections. 



Let us then proceed to our task. In discussing the progress 

 of physical science as a whole, the first question which presents 

 itself is, By what standard are we to estimate this progress ? 



To the uninitiated, this science of ours is an accumulation 

 of a vast number of facts, some of which are conspicuous for 

 their practical utility, while others are merely curiosities, or 

 objects of wonder. And, if it were possible to classify this 

 unconnected mass of facts, as was done in the Linnean system, 

 or in encyclopaedias, so that each may be readily found when 

 required, such knowledge as this would not deserve the name 

 of science, nor satisfy either the scientific wants of the hviman 

 mind, or the desire for progressive mastery over the powers of 

 nature. For the former requires an intellectual grasp of the 

 connection of ideas, the latter demands our anticipation of a 

 result in cases yet untried, and under conditions that we 

 propose to introduce in the course of our experiment. Both 

 are obviously arrived at by a knowledge of the law of the 

 phenomena. 



Isolated facts and experiments have in themselves no value, 

 however great their number may be. They only become 

 valuable in a theoretical or practical point of view when they 

 make us acquainted with the law of a series of uniformly 

 recurring phenomena, or, it may be, only give a negative result 

 showing an incompleteness in our knowledge of such a law, till 

 then held to be perfect. From the exact and universal con- 

 formity to law of natural phenomena, a single observation of a 

 condition that we may presume to be rigorously conformable to 

 law, suffices, it is true, at times to establish a rule with the 

 highest degree of probability ; just as, for example, we assume 

 our knowledge of the skeleton of a prehistoric animal to be 

 complete if we find only one complete skeleton of a single 

 individual. But we must not lose sight of the fact that the isolated 

 observation is not of value in that it is isolated, but because it 



