242 



SCIENCE. 



[N. S. Vol. VIII. No. 191. 



which is the more interesting and theoret- 

 ically important. The salts of the first 

 class are comparatively new to chemists, 

 and, in spite of all which has been done, 

 offer a very wide field for investigation. 

 Those of the second class present new 

 species of double and even of triple and 

 quadruple salts. It will first be necessary to 

 study the whole subject by physical as well 

 as by the purely chemical methods which 

 all chemists have hitherto employed, so as to 

 define as distinctly as possible the limit, if 

 there be one, which separates the two classes 

 from each other. I make the point that all 

 known methods of distinguishing double 

 from unitary salts should be employed in 

 each case, as I believe it has not yet been 

 satisfactorily shown that all these methods 

 give coincident results. It seems also pos- 

 sible that there may exist between the two 

 classes given a third class, the salts belong- 

 ing to which exhibit intermediate proper- 

 ties. In chemistry, as in other branches of 

 science, sharply defined boundaries between 

 groups are the exception. I am not aware 

 that the chief methods of distinguishing 

 between single and double salts have been 

 brought together for reference, and perhaps 

 the following list may be convenient : 



1. Purely chemical or reaction tests. 



2. Decomposition by solution alone. 



3. Physical methods. 



a. Determination of the electric conducting 

 power o£ dilute solutions. 



b. Determination of the ions by electrolysis and 

 by spectroscopic observations. 



c. Observation of the change of temperature in 

 the formation of the complex salt. Heat is 

 evolved in the formation of a complex salt 

 from unitary molecules ; none in the formation 

 of a double salt. The formation of a complex 

 salt, therefore, is accompanied by a loss of 

 energy. 



Other methods have been given, but the 

 above mentioned will suffice for most pur- 



In the brief and elementary sketch which 

 I have given to-day of two large and im- 



portant classes of salts I have avoided the 

 use of the so-called structural, or, as they 

 should be termed, valence formulas. That 

 such formulas are often useful, if only as 

 checks upon purelj' speculative work, cannot 

 be doubted. But it seems to me that the 

 time has come for wider and more general 

 conceptions of the chemical action between 

 masses, subject perhaps to the fundamental 

 idea of valence and yet not too dependent 

 upon this idea as at present received. In 

 a paper published long since, an attempt 

 was made to extend the atomistic con- 

 ception and to maintain the theory that the 

 action of one complex mass upon another is 

 determined by the residual unsaturated 

 units of affinity present, without reference 

 to the number of units due to each atomic 

 mass. I venture now to suggest that, in 

 addition to its valence, each atom and each 

 molecule possesses a special chemical poten- 

 tial, not necessarily a function of its valence. 

 The expression ' chemical potential ' is not 

 wholly new, but I think the conception has 

 never been clearly defined. I would now de- 

 fine it as bearing the same relation to chem- 

 ical action which the electric potential bears 

 to electrical action, the two potentials being 

 mutually convertible, all chemical com- 

 pounds having residual affinities or poten- 

 tials besides the valencies. If we suppose 

 that the atoms witliin the molecule are in 

 motion, such motion will be independent 

 of the valences and the molecule will have 

 a certain amount of free kinetic energy con- 

 vertible into chemical electrical energy or 

 into heat. 



In inorganic chemistry four great prob- 

 lems now present themselves for solution. 

 These are : 



The existence and chemical relations of 

 the gaseous elements, of which five are 

 known to exist in the atmosphere. The 

 separation of the elements forming the rare 

 earths, by systematic processes and the de- 

 termination of their positions in the periodic 



