THE PERCEPTION OF SPACE. 209 



his fingers, must either be absolutely identical or absolutely 

 unlike. They cannot be similar in diversity, " for they are 

 simple notions, and it is of the essence of such to enter the 

 mind or leave it all at once, so that one who has a simple no- 

 tion at all, possesses it in all its completeness, . . . There- 

 fore, since it is impossijole that the blind should have of 

 the forms in question ideas completely identical with our see- 

 ing ones, it follows that their ideas must be rcidically dif- 

 ferent from arui ivholly irreducible to our own.'' * Hereu]3on 

 M. Dunan has no difficulty in finding a blind man who still 

 preserves a crude sensation of difi'used light, and who says 

 when questioned that this light has no extent. Having ' no 

 extent ' appears, however, on farther questioning, to signify 

 merely not enveloping any particular tactile objects, nor 

 being located within their outline ; so that (allowing for 

 latitude of expression) the result tallies perfectly with our 

 OAvn view. A relatively stagnant retinal sensation of difiused 

 light, not varying when diiferent objects are handled, would 

 naturally remain an object quite apart. If the word 'ex- 

 tent ' were habitually used to denote tactile extent, this sen- 

 sation, having no tactile associates whatever, would natu- 

 rally have ' extent ' denied of it. And yet all the while it 

 would be analogous to the tactile sensations in having the 

 quality of bigness. Of course it would have no other tac- 

 tile qualities, just as the tactile objects have no other opti- 

 cal qualities than bigness. All sorts of analogies obtain 

 between the spheres of sensibility. Why are ' sweet ' and 

 ' soft ' used so synonymously in most languages ? and why 

 are both these adjectives applied to objects of so many 

 sensible kinds. Kough sounds, heavy smells, hard lights, 

 cold colors, are other examples. Nor does it follow from 

 such analogies as these that the sensations compared need 

 be composite and have some of their parts identical. "We 

 saw in Chapter XIII that likeness and difi'erence are an ele- 

 mentary relation, not to be resolved in every case into a 

 mixture of absolute identity and absolute heterogeneity of 

 content (cf. Yol. I, pp. 492-3). 



I conclude, then, that although in its more superficial 



*P. 135. 



