A MODEL OF NATTTRE. 185 



similar l)ricks. but the (li.scoverv In' LeiKU'd that the case with which 

 the eorpuscK's peneti'ate different bodies depends only on the (U'nsity 

 of the <)l)staeles, and not on tlieir eheniieal constitution, is held ))y 

 Professor Thomson to he "a stronu- contirniation of tlu^ view that the 

 atoms of the elementary substances are made up of simpler parts, all 

 of which are alike.""'' On the present occasion, however, we are occu- 

 pied rather with the foundations than with these ultimate ramifications 

 of the atomic theory: and having- shown how wide its rano-e is, I must, 

 to a certain extent, retrace my steps and return to the main line of 

 m}' argument. 



THE PROPERTIES OF ATOMS AND MOLECULES. 



For if it be granted that the evidence that matter is coarse grained 

 and is formed of separate atoms and molecules is too strong to be 

 resisted, it may still be contended that we can know little or nothing 

 of the sizes and properties of the molecules. 



It must be admitted that though the fundamental postulates are 

 alwa3's the same, dilierent aspects of the theorv, which have not in all 

 eases been successfully combined, have to be developed when it is 

 applied to different plol)lems: ))ut in spite of this there is little doubt 

 V>ut that we have some fairly accurate knowledge of molecular motions 

 and magnitudes. 



If a liquid is stretched into a veiy thin film, such as a soap l)ul)ble, 

 we should expect indications of a change in its properties when the 

 thickness of the film is not a very large multiple of the average 

 distance between two neighboring molecules. In 1890, Sohncke'' 

 detected evidence of such a change in films of average thickness 

 of 106 millionths of a millimeter (/^yw), and quite recently Rudolph 

 Weber found it in an oil film when the thickness was 11.5////. " 



Taking the mean of these numbers and combining the results of 

 different variants of the theor}", we may conclude that a film should 

 become ;mstable and tend to rupture spontaneously somewhere be- 

 tween the thicknesses of 110 and 55 ///<, and Professor Reinold and I 

 found b}" experiment that this instability is actuallj' exhibited between 

 the thickness of 90 and 45 ////.'' There can therefore be little doubt 

 that the first approach to molecular magnitude is signaled when the 

 thickness of a film is somewhat less than 100 /<yu, or four millionths of 

 an inch. 



Thirteen 3'ears ago 1 had the honor of laying before the Chemical 



"For the most recent account of this subject, see an article on "Bodies smaller than 

 atoms," by Prof. J. J. Thomson, in the Popular Science Monthly (The Science Pre&s), 

 August, 1901. [Reprinted in the present Smithsonian Report.] 



"Wied. Ann., 1890, XL, pp. 345-355. 



■= Annalen der Physik, 1901, IV, pp. 706-721. 



iPhil. Trans., 1893, 184, pp. 505-529. 



