ON THE ESSENTIAL PROPERTIES OF MATTER. 467 



to the divisibility of matter with respect to coloured substances ; for the 

 colours of thin transparent substances, which he considers as resembling those 

 of most other substances, are no longer observable, in any known medium, 

 when the thickness is less than about ^mAnnr of an inch. But we have 

 positive evidence that coloured substances may be reduced to dimensions 

 far below this limit ; besides the instance of the gilt wire, which has already 

 been mentioned, a particle of carmine may still retain its colour, when its 

 thickness is no more than one thirty millionth of an inch, or one sixtieth 

 part of the limit deduced from the supposition of Newton ; and it is there- 

 fore scarcely possible that the colours of such substances can precisely 

 resemble those of thin plates, although they may perhaps still be in some 

 measure analogous v to them. 



Impenetrability is usually attributed to matter, from the common obser- 

 vation that two bodies cannot occupy the same place at once. And it is 

 thus that we distinguish matter from space ; for example, when we dip 

 an inverted jar into mercury, the air contained in the jar depresses the 

 surface of the mercury, and prevents its occupying the space within the 

 jar : but if the jar had been void of matter, like the space above the mer- 

 cury of a barometer, nothing would have prevented its being filled by the 

 mercury, as soon as either its weight or the pressure of the atmosphere, 

 urged it to enter the jar. 



But it does not appear that our senses are fully competent to extend this 

 proposition to all substances, whether material or not. We cannot prove 

 experimentally that the influence of gravitation is incapable of pervading 

 even the ultimate particles of solid matter, for this power appears to suffer 

 no diminution nor modification, when a third body is interposed between 

 the two gravitating masses. In the same manner, a magnet operates as 

 rapidly on a needle, through a plate of glass or of gold, whatever its thick- 

 ness may be, as if a vacuum only intervened. It may, however, be 

 inquired if the gold or the glass has not certain passages or pores, through 

 which the influence may be transmitted : and it may be shown, in many 

 instances, that substances, apparently solid, have abundant orifices into 

 which other substances may enter ; thus mercury may easily be made to 

 pass through leather, or through wood, by the pressure of the atmosphere, 

 or by any other equal force : and, however great we may suppose the pro- 

 portion of the pores to the solid matter, it may be observed, that it requires 

 only a more or less minute division of the matter, to reduce the magnitude 

 of the interstices between the neighbouring particles within any given 

 dimensions. Thus platina contains, in a cubic inch, above 200 thousand 

 times as many gravitating atoms as pure hydrogen gas, yet both of these 

 mediums are free from sensible interstices, and appear to be equally con- 

 tinuous ; and there may possibly be other substances in nature that contain 

 in a given space 200 thousand times as many atoms as platina ; although 

 this supposition is not positively probable in all its extent ; for the earth is 

 the densest of any of the celestial bodies with which we are fully ac- 

 quainted, and the earth is only one fourth as dense as if it were composed 

 entirely of platina ; so that we have no reason to believe that there exists 



2n2 



