RICHARDS. — SIGNIFICANCE OF CHANGING ATOMIC VOLUME. 595 



Having thus decided that the existence of a polymeric form is prob- 

 able, the next step is a study of the probable mechanism of the act of 

 polymerization. In a simple molecule of water, for reasons given above, 

 it is usually assumed that each hydrogen atom is attached to the oxygen, 

 HOII. When a molecule thus constituted coheres with other molecules 

 to form a liquid, there are two typical situations attainable by the atoms 

 of neighboring molecules. The atoms of each kind may be in juxtapo- 

 sition with the same kind of atoms, or hydrogen of one molecule may 

 be in the neighborhood of oxygen of another. If neither of these two 

 situations involved a great excess of cohesive attraction over the other 

 (the normal condition of an unassociated liquid), both situations would be 

 present to about equal extents. From what has been said before, this 

 may be supposed to be the case in liquid water at high temperatures. . 



What now happens when the water is cooled? Since there is every 

 reason to believe that the act of polymerization evolves considerable 

 heat, this act will tend more and more to occur as the temperature 

 becomes lower, according to the theorem of Le Chatelier. All our ex- 

 perimental evidence points in this direction. But this increasing poly- 

 merization will involve an increasing shift of the cohesive attraction from 

 the condition of equal distribution to that in which oxygen is in juxtapo- 

 sition with oxygen, and hydrogen with hydrogen. 



The effect of such a shift as this upon the volume will depend entirely 

 upon the relative compressibilities of the elements concerned, under the 

 peculiar circumstances to which they are exposed. There is good reason 

 to believe that under conditions as parallel as possible, oxygen is less 

 compressible than hydrogen. For example, pentane (C 5 II 12 , boiling at 

 38°), containing much hydrogen, is half again as compressible as sulphur 

 dioxide (S0 2 , boiling at — 10°) if the figures are compared at corre- 

 sponding temperatures (0° and 50° respectively, each a corresponding 

 number of degrees above the boiling point).* Since sulphur in com- 

 bination is probably at least as compressible as carbon, the difference 

 here is to lie referred primarily to that between hydrogen and oxygen. 

 The study of many other, indeed all the known, compres>ibilities of 

 compounds rich in hydrogen, leads to the conclusion that the compressi- 

 bility of hydrogen is great, and that of oxygen smaller. This is not 

 indeed surprising, since the average density of the oxygen in combination 

 is about eight times as great as that of hydrogen, if Kopp's atomic 



* Landolt ami Burnstein Tabellen, p. 2u\s (1804). The values were found by 

 Grimaldi ami by Colladon and Sturm. 



