292 Mr. Griffin on the Constitution of 



fully, and I feel bound to declare that they are at variance. 

 Among a dozen chemical substances, whose solutions I have 

 examined, I find only one that appears to have a fixed atomic 

 volume in solution in water. That substance is ammonia. 

 All the other substances that I have examined vary constantly 

 in their liquid volume, according to the state of dilution, and 

 I am forced to the conclusion, in so far at least as concerns 

 the substances that I shall enumerate, that the doctrine of a 

 fixed atomic volume in solution is fallacious and of no use. 

 While I think it right to state this opinion explicitly, I do so 

 ■with considerable diffidence. It is possible that I may have 

 overlooked some circumstance essential to the full compre- 

 hension of the question. I by no means deny the accuracy 

 of the experiments upon which the opposite opinion has been 

 founded. The difference in our conclusions results from dif- 

 ferent ways of interpreting the experiments. The conclusions 

 are however incompatible; and if the doctrines of Messrs. 

 Playfair and Joule shall prove to be correct, serious fallacies 

 must exist in my calculations or mode of reasoning. 



To place the evidence on this subject before chemists, I 

 proceed to detail my researches into the constitution of solu- 

 tions of acids and alkalies. An investigation of this sort has 

 the unfortunate property, that, dealing with a great variety 

 of separate facts, an account of it must necessarily be long. 

 I shall shorten it to the utmost by throwing as much as pos- 

 sible into the form of Tables ; and in order to render the con- 

 densed account intelligible, I shall begin by describing the 

 plan of the Tables, and explaining such terms and other par- 

 ticulars as seem to require it. 



Of the Measures used. 



I have in all cases examined the constitution of the same 

 volume of solution, namely, the tenth part of an imperial gal- 

 lon, or the bulk of one avoirdupois pound, or 7000 grains of 

 pure water at the temperature of 62° F. I propose to call 

 this measure a Decigallon. 



For the sake of having a manageable term for a small vo- 

 lume of liquid, I divide the decigallon into 1000 parts, each 

 of which contains 7 grains of water, and for which measure I 

 propose the name of Septem. 



A complete scheme of decimal measures thus founded on 

 the imperial gallon is exhibited in the following Table : — 



