BOOK II.] lANCING Ilf STANCES. 527 



Btancea ; for flame increases over the exlialation of oil, and air 

 over that of water. One must therefore consider the mixture of 

 oil and water, which is manifest to the senses, since that of air 

 and flame in general escapes the senses. But oil and water mix 

 very imperfectly by composition or stirring, whilst they are 

 exactly and nicely mixed in herbs, blood, and the parts of 

 animals. Something similar, therefore, may take place in the 

 mixture of Hame and air in spirituous substances, not bearing 

 mixture very well by simple collision, whilst they appear, how- 

 ever, to be well mixed in the spirits of plants and animals. 



Again, if the inquiry do not relate to perfect mixtures of 

 spirits, but merely to their composition, as whether they easily 

 incorporate with each other, or there be rather (as an example) 

 certain winds and exhalations, or other spiritual bodies, which do 

 not mix with common air, but only adhere to and float in it in 

 globules and drops, and are rather broken and pounded by tho 

 air, than received into, and incorporated with it ; this cannot be 

 perceived in common air, and other aeriform substances, on 

 account of the rarity of the bodies, but an image, as it were, of 

 this process may be conceived in such liquids as quicksilver, 

 oil, water, and even air, when broken and dissipated it ascends 

 in small portions through water, and also in the thicker kinds of 

 smoke ; lastly, in dust, raised and remaining in the air, in all of 

 which there is no incorporation : and the above representation 

 in this respect is not a bad one, if it be first diligently investi- 

 gated, whether there can be such a difference of nature between 

 spirituous substances, as between liquids, for then these images 

 might conveniently be substituted by analogy. 



And although we have observed of these supplementary in- 

 stances, that information is to be derived from them, when appro- 

 {)riate instances are wanting, by way of refuge, yet we would 

 lave it understood, that they are also of great use, when the 

 appropriate instances are at hand, in order to confirm the in- 

 formation afforded by them ; of which we will speak more at 

 length, when our subject leads us, in due course, to the support 

 of induction, 



XLIII. In the twentieth rank of prerogative instances we will 

 place lancing instances, which we are also wont (but for a different 

 reason) to call twitching instances. We adopt the latter name, 

 because they twitch the understanding, and the former because 

 they pierce nature, whence we style them occasionally the in- 

 stances of Dcmocritus.*^ They are such as warn the understand- 

 ing of the admirable and exquisite subtilty of natiu'c, so that it 

 becomes roused and awakened to attention, observation, and 

 proper inquiry ; as, for instance, that a little drop of ink should 



•^ Alluding to his theory of atones. 



