Januaet 14, 1910] 



SCIENCE 



45 



oughly investigated nothing less than the 

 complete equilibrium curves of the com- 

 pounds in question will suffice; but once 

 the work is carefully done, it is final for all 

 time. This is not the place to dwell upon 

 all the various questions that have been 

 cleared up by the application of the phase 

 rule. It should here be emphasized, how- 

 ever, that the latter deals with the equilib- 

 rium of the various phases whose qualita- 

 tive and quantitative composition is of 

 course ascertained. As to the inner struc- 

 ture of any one of the phases the phase 

 rule is able to tell us nothing. Indeed, in 

 the study of single-phase chemistiy, the 

 phase rule is no help whatever. We may 

 consider the investigation of the constitu- 

 tion of definite chemical compounds a part 

 of single-phase chemistry, and we may 

 similarly consider the question as to the 

 inner nature of a solution {i. e., of a com- 

 pound according to variable proportions) 

 as a problem of single-phase chemistry. 

 In the investigation of the constitution of 

 single phases it is quite impossible to get 

 along without hypotheses. While the 

 phase rule does not involve even the 

 atomic and molecular theories, these are at 

 present indispensable tools in prying into 

 the inner nature of any one phase. But 

 in the study of solutions, interest centers 

 not so much in the equilibrium between 

 phases as in the inner structure of the 

 latter themselves. 



Our methods of ascertaining the struc- 

 ture of chemical compounds are quite 

 numerous, but they readily fall into a few 

 categories. So we argue as to the struc- 

 ture of a compound from its synthesis, 

 from its analysis, from its behavior toward 

 various other chemical agents, from altera- 

 tion by the application of pressure, heat, 

 electricity, light and kindred agencies, and 

 also from its various physical and physio- 

 logical properties. Thus, for example, it 



has always been considered as sound reas- 

 oning that because red precipitate can be 

 formed from mercury and oxygen, these 

 substances are in red precipitate, which 

 conclusion is verified by the fact that the 

 latter compound may be decomposed into 

 oxygen and mercury. There has never 

 been any objection to the argument that 

 if one of the elements actually enters into 

 a compound during the latter 's formation, 

 or can be obtained from the compound 

 either in the free state or in combination 

 with other elements, that element is actu- 

 ally in the compound. So since calcium 

 carbonate may be made from calcium, 

 carbon and oxygen, we argue that these 

 ^ elements and these only are contained in 

 calcium carbonate. Again, when calcium 

 carbonate is heated, calcium oxide and car- 

 bon dioxide, and these only, are obtained; 

 and conversely calcium carbonate may be 

 formed by the union of calcium oxide and 

 carbon dioxide. These facts were duly ex- 

 pressed by the old dualistie formula for 

 calcium carbonate CaCCOj which con- 

 sequently had much to commend it. Yet 

 while we thus hold that the elements cal- 

 citmi, carbon and oxygen are in calcium 

 carbonate, we do not argue that this com- 

 pound contains calcium oxide and carbon 

 dioxide, even though the last two sub- 

 stances will unite and thus form calcium 

 carbonate, or though they may be obtained 

 as decomposition products of the latter 

 compound. We write our formula for 

 calcium carbonate CaCOj because of the 

 precipitation methods by which the com- 

 pound may be prepared, and because of the 

 formulae that we assign to soluble carbon- 

 ates on the basis of the products that they 

 yield by electrolysis. We consequently 

 hold that the carbon dioxide and lime that 

 form when calcium carbonate is heated 

 result from the rearrangement of the atoms 

 and splitting of the compound on account 



