146 DEVELOPMENT AND PURPOSE CHAP. 



between mechanical abstraction on the one side and mysti- 

 cism on the other, the one, to adapt a famous antithesis, 

 relatively void, and the other blind. 



In point of fact there always is in experience more than 

 thought can render in articulate terms. This holds of a 

 very simple experience. Even one of the colour sensations 

 to which we referred above has a quality which it is 

 difficult, if not impossible, to render quite perfectly in 

 abstract terms. The green of the oak leaf is a green of 

 a particular shade and quality. We express this quality 

 as far as we can by calling it a darkish green, shading to 

 a slight suggestion of blue when the leaf is fully out. But 

 it is difficult to give it its precise quality without calling 

 it the green of the oak leaf, which is after all a definition 

 in a circle. * Green,' c dark green,' ' bluish green,' are, in 

 fact, general terms which, with a varying measure of 

 accuracy express the character of the colours that we see. 

 By attention and comparison, by trained perception and 

 analysis, we can keep on increasing this accuracy so that 

 it approximates to the limit of the concrete sense datum, 

 Hence the painter's colour vocabulary is richer, and more 

 diversified with shades of perception, than that of ordinary 

 language. As the process of analysis advances so the 

 rendering of experience becomes more perfect, and the 

 element of error inherent in the translation of experience 

 into thought becomes less and less material. Naturally, 

 the more complex and subtle the object which we are 

 approaching, the more backward we are in this process. 

 When we are dealing with something like the sense of 

 duty in which a thousand subtle threads of feeling are 

 involved, but which is always pre-eminently a unity and 

 destroyed by any breaking up of its elements, the task of 

 analysis is of far greater difficulty. When, again, we are 

 dealing with the nature of life, we are attacking that which 

 for the most part is only known to us directly by certain 

 superficial effects. Direct observation of the inner pro- 

 cesses fails, and any conception that we form can only be 

 the result of a prolonged effort of synthesis applied to very 

 diverse and always insufficient data. Lastly, when we 

 consider religious conceptions, we are dealing with the 



