IMPERFECTIONS OF SENSATIONS 51 



other point of view it appears to be more or less 

 oval. We judge it to be circular, not from its 

 actual appearance but from our memory of 

 previous similar impressions of it, coupled with 

 experiences of touch : we infer that it is circular, 

 although to our sensation it may appear oval. So 

 also with colours, the actual appearance of which 

 varies very greatly according to the amount of 

 light that falls upon them and the direction from 

 which it falls. Our real guide is not sensation, but 

 perception, which is sensation adjusted in the 

 light of previous experiences. Not only, then, is 

 the brain, as we have seen, the origin of sensation, 

 but it needs processes of the brain to render 

 sensory impressions useful to us. Practice enables 

 us to adjust with instantaneous rapidity. But 

 babies that reach for a bright object, regardless of 

 its distance, prove that the faculty of adjustment 

 rests upon experience. Having made these ad- 

 justments habitually, we can hardly disembarrass 

 ourselves of their guidance. It is quite difficult 

 to draw things, not as we infer them to be, but 

 as they actually appear : children's pictures, 

 primitive art and oriental art all aim at truth by 

 representing not the seen but the inferred. It 

 requires, indeed, an artist to set down the simple 

 impressions of the eyes. These are in perspective, 

 and those who look at the pictures are convinced 

 of solidity by their own powers of inference. 



Our sensory impressions need, then, to be 

 adjusted before they will guide us trustworthily, 

 and we correct each one of them, as it comes, by 

 investing it with attributes, such as roundness or 

 solidity, which it does not appear to possess, but 

 which we infer from the memory of previous 

 impressions and tactile experiences that it does 

 possess. In other words, the key to our impres- 

 sions is a stock of memories and ability to infer 



