824 TRANSACTIONS OF THE AMERICAN INSTITUTE. 



ever, to suppose th;it in the molecule composed of six atoms, the 

 centre of motion Avill be the centre of gravity of the system, and 

 that the eliipticity of the orbits will result entirely from the 

 inequality of the difierent pairs and not from excentricity of 

 motion. 



If, therefore, we suppose the molecules of different chemical 

 compounds to be made up of atoms, or clusters of atoms, of differ- 

 ent weights, we have an adequate cause for the ellipsoidal forms 

 of such molecules and all the consequent varieties of crystalline 

 form and structure. If this view be correct it may not be impos- 

 sible for the mathematician to determine the chemical constitution 

 of a crystal by its geom.etrical properties. 



The interlinking of the solid molecules will be seen in Fig. 10, 

 which is a section through a single layer of molecules in the plane 

 of two axial directions. According to the idea of impenetra1>ility 

 here advanced, it seems to be immaterial whether each couple of 

 ai)proacliing atoms actually collide at the crystalline poles, or pass 

 around them as foci. 



The most formidable argument, perhaps, which may be brought 

 qo-ainst this view of cohesion or crystalline attraction, is the dis- 

 crepancy between terrestrial and atomic gravitation, when we con- 

 sider the enormous disproportion between the masses. A sus- 

 pended body, for example, is sustained by 'its own cohesion or 

 atomic o-ravitation against the gravitating force of so immense a 

 body as the entire earth. A thread or wire, moreover, will sus- 

 tain many times its own w^eight, and yet the attraction between 

 two consideral)le bodies of heavy material, even when delicately 

 suspended at an extremely minute distance from each other, is so 

 small as to be practically insensible. 



The explanation of this apparent anomaly may be found in the 

 infinitessimal distances at Avhich the atomic gravitation acts, when 

 compared with the gravitation of masses. The powerful resistance 

 to compression in solids and liquids, affords an indication of the 

 entire force associated with atoms at minute distances, to which 

 the resultant virtual attraction in solids, pr()l)ably bears no incon- 

 siderable proportion, even at the maximum distance between the 

 adjacent atoms of a crystal, which is that of a quarter of an orlntal 

 revolution. This distance may be supposed sutEciently minute 

 to produce, in atoms, at the first order of distances, a very power- 

 ful attraction, when multiplied by the number of interlinking 

 molecules iu the cross section of a solid body. When once scpa- 



