Molecular Constitution of Crystals. HI 



several molecules have united and a solid has thus been formed, 

 no ehange will take place in the law of attraction consequent on 

 ?he mutSal action of^ll the accumulated particles; nor can we 

 safelv sunno^e that properties which are themselves transient, 

 Tnd LSy not e xislmg^hen the body is in all its states, may 

 not be alterable after the body has assumed the «o ^^ fo m 



Dana seems to have given "^^^^^^^^^^T cornplication t^a his 

 hypothesis, and by this means to have rendered the explanation 

 of some of the phLomena impossible. It appears to me that 

 all thTconditions would be satisfied by the molecules having s^ 

 poles, all exercising a mutual attraction for each other ; thetoim 

 which would result from the union of such molecules would evi- 

 dentlv be the cube (fis- 5) . But it is natural to suppose in accord- 

 ':: 'I'th'^^uS- ph^noLena m nature, that these po es may be 

 Uableto removal or displacement on account of their ^^'jtua^^attia - 

 tion for each other, and that the attractmg fluid is only held m its 

 place by a certain coercive force of the molecule itself, as is the 

 cJe in the common magnet. As the crystal inci^ases m size 

 those molecules which are at its extremities wdl be n-e and 

 more acted on by this influence, and some may altogethci lose 

 their polarities (k the same time it is by no means a consequence 

 tha" these molecules will drop off, smce the attracting fluid need 

 not be removed from the molecule, but only displaced) ; the 

 cons quenc will be, that the free particles will not be attached 

 to those which have lost their polarities, and modifying planes 



""" Thus Tn" case the molecule at each corner of a cube lose its 

 three unattached poles, a plane will appear at each corner, and 

 thp rictahedron will be thus formed. 



If a row of molecules along each edge lose their unattached 

 poles, planes will appear on each edge, which wi 1 of course belong 

 fo the dodecahedron; but if two rows of mo ecules lose then 

 poles, the resultant form will be a tetrakishexahedron ; it thiee, 

 another tetrakishexahedron. vff„,.o«f 



Let us now consider under what circumstances these diffeient 

 results will take place. If we consider the manner m which he 

 additional particles are laid on ^^ose layers already formed it s 

 evident that those particles nearest the centre will be the first to 

 which new molecules will become attached and that in general 

 the particles will attach themselves as nearly as possible to the 

 centre of the crystal, inasmuch as the attraction is stronges 

 there On the contrary, those molecules which are fuithest 

 from the centre will be the first to lose their attractive force, as 



""""uriutrinolecules in a face already formed be covered except 

 the external row, and if, while the internal molecules were having 



