So far we have merely discussed what may be described as the sensory elements in the picture, and the responses in emotion and attitude due to these elements. But in most painting there are further elements essentially involved. It has been asserted that all further elements are irrelevant, at least to appreciation; and as a reaction to common views that seem to overlook the sensory elements altogether the doctrine is comprehensible and perhaps not without value. For too many people do look at pictures primarily with intent to discover what they are ‘of’, what they represent, without allowing the most important thing in the picture, its sensory stimulation through colour and form, to take effect. But the reaction goes too far when it denies the relevance of the representative elements in all cases. It may be freely granted that there are great pictures in which nothing is represented, and great pictures in which what is represented is trivial and may be disregarded. It is equally certain that there are great pictures in which the contribution to the whole response made through representation is not less than that made more directly through form and colour. To those who can accept the general psychological standpoint already outlined, or indeed any modern account of the working of the mind, the assertion that there is no reason why representative and formal factors in an experience should conflict, but much reason why they should co-operate, will need no discussion. The psychology of ‘unique æsthetic emotions’ and ‘pure art values’ upon which the contrary view relies is merely a caprice of the fancy. The place of representation in the work o different masters varies enormously and it is not true that the value of their works varies correspondingly. From Raphael and Picasso at one extreme to Rembrandt, Goya and Hogarth at the other, Rubens, Delacroix and Giotto occupying an intermediate position, all degrees of participation between non-representative form and represented subject in the building up of the whole response can be found. We may perhaps hazard, for reasons indicated already, as a principle admitting of exception, that what can be done by sensory means should not be done indirectly through representation. But to say more than this is to give yet another instance of the commonest of critical mistakes: the exaltation of a method into an end.