POLYMORPHISM AT HIGH PRESSURES. 187 



under the more moderate pressures all the liquids tend to a more or 

 less common type of behavior, but that at higher pressures the most 

 varied abnormalities make their appearance, each liquid behaving in 

 its own characteristic waA'. These aljnormalities are doubtless to be 

 explained by the increasing approach to some sort of order brought 

 about when the molecules arc squeezed closer together by high pres- 

 sure. As pressure and temperature shift the predominating type of 

 order in the relation of the molecules may change also, and we get the 

 \ arietl effects which we know occiu- when the arrangement of the atoms 

 is altered as it is in a pt)l\niorphic transition. It seems most natural 

 to suppose that the same sort of considerations and the same mechan- 

 ism that will some day be discovered to account for polymorphic 

 changes will also account for the complicated phenomena in liquids 

 at high pressures. 



Summary. 



New data are presented for the polymorphic transitions of a number 

 of substances between 0° and 200° and to 12000 kgm. All of the 

 substances which I have examined which have shown no polymorphic 

 changes in this same range are enumerated. This enumeration 

 includes also all the substances with polymorphic forms, and classifies 

 them according to chemical structure. The total number of sub- 

 stances examined to date is nearly 150. The discussion deals with all 

 the phase diagrams examined in this series of papers. There is no 

 simple type toward which all polymorphic diagrams tend at high 

 pressures, nor are there a few common types. The complication is 

 very great; probably solid transitions of any type whatever exist, 

 except those involving a critical point. It is suggested that in order 

 to explain these complicated phenomena the effect of the shape of the 

 atoms is the next factor to be considered, and it is shown in detail by 

 considering several artificial examples that assemblages of units of 

 definite shape provided with localized centers of force arc capable of 

 exhibiting the variety of behavior which actual polymorphic sub- 

 stances show. 



Acknowledgment. Grateful acknowledgment is made of liberal 

 appropriations from the Bache Fund of the National Academy of 

 Sciences and from the Rumford Fund of the American Academy of 

 Arts and Sciences, with which expenses for uiaterials and mechanical 

 assistance have been met. 



The Jefferson Physical Labokatory, 



Harvard University, Cambridge, Mass. 



