602 PROCEEDINGS OF THE AMERICAN ACADEMY. 



of formation of these compounds. Dimethyl ethyl carbinol has a dis- 

 tinctly greater heat of formation than isoamyl alcohol, while in the case 

 of the propyl alcohols there is no great difference between the heats 

 of formation of the isomers. Since a greater heat of formation usually 

 signifies greater affinity, since greater affinity must cause greater pres- 

 sure, and since greater pressure must diminish the volume, it appears 

 highly probable that in the case of the amyl alcohols a greater stress 

 within the molecule in one case may cause an increase in the density 

 which is enough to compensate for the decrease of density on the 

 molecular surface due to a smaller pressure of cohesion between dif- 

 ferent molecules. Thus, the apparently anomalous behavior of the alco- 

 hols not only offers no argument against the j^rinciples involved, but 

 rather supports the other arguments in favor of those principles. 



These cases may be talien as typical ; and i^robably most of the varia- 

 tions in density, comioressibility, and boiling point may be explained in 

 this way. As before, however, it is not to be expected that every case 

 could exactly correspond, for the densities, compressibilities, and heats of 

 formation of complex compounds of this kind represent average values, 

 and much would depend upon the constitution of the body, and the 

 arrangement of the atoms in space. As in the case of water and tin, 

 cited above, it might easily happen that a considerable excess of heat 

 of formation in a given case might be evolved by a shift from a more 

 compressible union to a less compressible one, involving the expenditure 

 of more chemical energy without causing a decrease in volume. 



It is evident upon consideration that all these typical cases are really 

 dependent for their interpretation upon the following very simple con- 

 ceptions: first, that internal pressure, due to both cohesive attraction and 

 chemical affinity, must tend to diminish the volume of condensed sys- 

 tems (i. e. solids and liquids) ; secondly, that the greater the pressure, the 

 greater is this tendency ; and tliirdl}', that, otlier conditions being equal, 

 a substance already much compressed by internal pressure suffi^rs less 

 contraction upon subjection to additional pressure than one but little 

 compressed. The circumstance that the facts above cited are all consist- 

 ent with this obvious logic afibrds a strong presumption that it is correct:ly 

 applied. 



The existence of these different intensities of internal pressure affords 

 a basis also for explaining the partial parallelism and frequent apparent 

 irregularity of the various physical properties of material, such as tenacity, 

 hardness, ductility, malleability, etc., and the relation of these properties 

 to boiling points. For example, in order that a liquid may vaporize 



