ADVANCEMENT OF LEARNING 



141 



sun's rays are so contracted by a burning-glass, and heightened 

 to such a degree as to fire any combustible matter, may not the 

 rays of the moon, by the same means, be actuated to some small 

 degree of warmth, so as to show whether all the heavenly 

 bodies are potentially hot ? and as luminous heats are thus in- 

 creased by glasses, may not opaque heats, as of stones and 

 metals, before ignition, be increased likewise, or is there not 

 some proportion of light here also ? Amber and jet, chafed, at- 

 tract straws, whence query, if they will not do the same when 

 warmed at the fire? 



An experiment may be varied in quantity, wherein very 

 great care is required, as being subject to various errors. For 

 men imagine, that upon increasing the quantity the virtue 

 should increase proportionably ; and this they commonly pos- 

 tulate as a mathematical certainty, and yet it is utterly false. 

 Suppose a leaden ball of a pound weight, let fall from a steeple, 

 reaches the earth in ten seconds, will a ball of two pounds, 

 where the power of natural motion, as they call it, should be 

 double, reach it in five ? No, they will fall almost in equal 

 times, and not be accelerated according to quantity. Suppose 

 a dram of sulphur would liquefy half a pound of steel, will, 

 therefore, an ounce of sulphur liquefy four pounds of steel ? 

 It does not follow ; for the stubbornness of the matter in the 

 patient is more increased by quantity than the activity of the 

 agent. Besides, too much, as well as too little, may frustrate 

 the effect thus, in smelting and refining of metals, it is a com- 

 mon error to increase the heat of the furnace or the quantity 

 of the flux ; but if these exceed a due proportion, they prejudice 

 the operation, because by their force and corrosiveness they 

 turn much of the pure metal into fumes, and carry it off, 

 whence there ensues, not only a loss in the metal, but the re- 

 maining mass becomes more sluggish and intractable. Men 

 should therefore remember how <sop's housewife was de- 

 ceived, who expected that by doubling her feed her hen should 

 lay two eggs a day ; but the hen grew fat, and laid none. It 

 is absolutely unsafe to rely upon any natural experiment before 

 proof be made of it, both in a less and a larger quantity. 



An experiment is produced two ways ; viz., by repetition 

 and extension, the experiment being either repeated or urged 

 to a more subtile thing. It may serve for an example of repe- 

 tition, that spirit of wine is made of wine by one distillation, and 



