36 CLARKE 



may have modified the ions in some cases, and this consideration 

 is most important. All thermochemical measurements made upon 

 or in solutions should be reexamined in the light of the modern 

 theory of solutions, in order that the reactions represented may 

 not be misunderstood. The interpretation, by Arrhenius, of the 

 meaning of the neutralization constant, will serve to illustrate 

 the importance of these considerations. I cannot avoid the sus- 

 picion that many of the deductions heretofore drawn from a 

 study of the thermal phenomena of solutions have been al- 

 together fallacious. In brief, it would be wise to concentrate 

 our immediate attention upon the thermal behavior of reactions, 

 rather than upon the heats of formation which may be deduced 

 from them. Reactions should be so classified as to exhibit their 

 common factors in accordance with modern ideas ; when that 

 work has been done a great advance will become possible. 



It would be easy to adduce much more evidence in favor of 

 the principles which I have sought to establish, but I do not 

 wish to overload the argument with details. The definite 

 character of one constant seems to be clear, but the limits of its 

 applicability are yet to be established. A new line of attack 

 upon thermochemical problems has been developed, and it leads 

 to conclusions which are generally novel, and which are some- 

 times antagonistic to the views of other investigators. For 

 example, the thermal equality of single, double and triple 

 carbon unions is directly opposed to the conclusions reached by 

 Thomsen, whose opinions are always entitled to the highest 

 respect. But this equality is implied in the factor 8n of the 

 general formula from which the constant is derived, and with- 

 out it the constancy would vanish. Thomsen holds that the 

 three modes of union have different thermal values, two of them 

 positive and near together, the third being represented by a 

 small negative quantity. I find that the three values are iden- 

 tical ; and the difference between us plainly arises from the 

 different methods of discussion which he and I have employed. 

 This question of method, however, is fundamental, and upon it 

 all further discussion depends. The validity of my formula is 

 the first point at issue, and that must be decided by future re- 

 search and criticism. I fully recognize the fact that the diffi- 



