332 DEVELOPMENT AND PURPOSE CHAP. 



truths implied in the process of demonstration. What on 

 the surface we call a sound generalisation from experience 

 is not, of course, a truth which merely sums up in a single 

 formula a set of observed facts. It states a connection as 

 holding generally, and if it is thoroughly scientific, uni- 

 versally ; that is to say, as holding not only for certain 

 things that we have observed, but for others that we may 

 observe or that may never fall within the sphere of our 

 own experience at all. Hence the scientific use of experi- 

 ence is a process which goes beyond observed and recorded 

 facts, and makes a certain use of these facts for purposes of 

 inference. Suppose that we can analyse this usage, and 

 write down certain propositions which, if true, justify it, 

 and, if false, destroy it. These propositions must then be 

 regarded as the tacit assumption of the scientific use of 

 experience. If they are true, the results of science are 

 trustworthy, and if not, not. That being so, it is clear that 

 whatever validity attaches to the results of science, what- 

 ever confidence we can legitimately place in its generalisa- 

 tions, must attach equally to these assumptions. Now it is 

 open to thinkers to question whether science itself is valid, 

 and if the answer be in the negative, this argument for the 

 validity of its assumptions falls with the structure of science 

 itself. I do not propose here to add anything to that 

 which has been already said on the ultimate ground of 

 rational thought, but assuming for present purposes the 

 general validity of the scientific method of enquiry, I 

 propose to consider some of the assumptions of science, 

 and to see what light they throw on the problem before us. 



2. As has been shown above, the impulse of rational 

 thought stated generally is to weave its experiences into a 

 systematic whole. The isolated unconnected experience is 

 as it stands non-rational. It requires to be connected, to 

 be shown as dependent upon some other element in experi- 

 ence, and as itself in turn a factor in the determination of 

 other elements. That is to say, the work of thought, when 

 summed up, is to build up a Whole of experience, in which 

 each part is a condition and is also conditioned. Every 

 element of the whole would have a reason rendered for it 



