372 NEW ESSAYS 



nothings cannot make something. We never sleep so 

 profoundly as not to have some feeble and confused 

 feeling, and we should never be wakened by the greatest 

 noise in the world if we had not some perception of its 

 beginning which is small, just as we should never break 

 a cord by the greatest effort 64 in the world, if it were 

 not strained and stretched a little by less efforts, though 

 the small extension they produce is not apparent 6 \ 

 " These petites perceptions have thus through their conse- 

 quences 66 an influence greater than people think. It is 

 they that form this something I know not what, these 

 tastes, these images of sense-qualities, clear in combination 

 but confused in the parts 67 , these impressions which 



6 * G. reads ' effect.' 



65 Cf. Montaigne, Essais, bk. ii. ch. 14 : 'If we suppose a piece 

 of twine equally strong throughout, it is utterly impossible that it 

 should ever break. For in what part of it is the breaking to begin, 

 the flaw to appear? And for it to break in every part at once is 

 against all nature.' See Introduction, Part iii. p. 144 note. 



66 E. omits 'through their consequences.' 



67 Sense-qualities, according to Leibniz, are each clear as a whole, 

 that is to say, each can %e perfectly distinguished from others. 

 But they are not also distinct ; that is to say, we cannot perfectly 

 analyze their elements. Such an analysis is possible ; but we cannot 

 perform it, for it would involve an infinite process. Each sense- 

 quality ' contains infinity,' for it "has connexions with everything 

 else in the universe. Cf. Lettre a la Heine Sophie Charlotte (1702) : 

 'We know by what kind of refraction blue and yellow are made, 

 and that these two colours when mixed make green. But we 

 cannot yet understand, for all that, how our perception of green 

 results from our perceptions of the two colours which compose it, 

 nor how our perceptions of these colours arise from their causes. We 

 have not even nominal definitions of such qualities so as to explain 

 the terms for them. ... If I were to say to some one : You know that 

 green means a colour consisting of blue and yellow mixed, he would 

 not make use of this definition as a means of recognizing green 

 when he came upon it. But this is the function of nominal 

 definitions. For the blue and the yellow which are in the green 

 are not distinguishable or recognizable^ and it is only by chance, so 

 to speak, that we have found this by observing that this mixture 

 always makes green. Thus the only way to enable a man to 

 recognize green in future is to show it to him at present ; but this 



