RELATIVE PEKCEPTION OF SIZE, ETC. 45 



our heaviest known metal, struck into medallions by 

 means of a powerful lever and die, we would conclude 

 that if a pair of human fingers were as strong and as hard 

 as such a lever and its die, we would cease to call these 

 metals hard to such hands. But we cannot make the 

 strength or hardness of our hands an absolute or a general 

 standard, for no two pairs of hands are exactly alike in 

 strength. Therefore we must for general purposes resort 

 to some conventional and generally agreed on standard. 

 If we take metal, the standard will be very little more 

 reliable than the human hand, for no single metal we may 

 adopt can be found of uniform purity and hardness every- 

 where, even with the same tempering and under the same 

 temperature ; so that a fixed standard of hardness is per- 

 haps one of the most difficult of all conventional standards 

 to establish. And after all it would, when fixed, only enable 

 us to say that bodies to which it was applied were so 

 much harder or softer than it, but not that they had an 

 abstract and positive hardness, independent of comparison ; 

 for to some bodies they would still be relatively soft, 

 and to others relatively hard. 



In like manner size is perhaps not seen exactly alike 

 by any two pairs of human eyes, for eyes differ just as 

 much as faces, and that in many particulars too minute 

 to be mentioned here. The retina of a large eye 

 must portray a larger image of objects than the retina of 

 a small eye. For the accuracy of the eye does not depend 

 on its seeing objects as they really are in point of dimen- 

 sions, but only on its seeing them as they appear to be in 

 comparison with each other ; so that a small bird whose 

 eye is no larger than a cabbage seed sees the relative sizes 

 of objects to each other, and to any part of its own 

 body under its range of vision, as accurately as an ox 

 whose eye is more than twice the diameter of a man's ; and 

 yet the details in the image of a landscape of thirty miles in 



