This section is from the book "Modern Chemistry", by William Ramsay. Also available from Amazon: Modern Chemistry: Theoretical and Modern Chemistry (Volume 2).
In the olden days, no distinction was drawn between a compound and a mixture. Indeed, all " impure 99 substances artificially prepared were termed "mixts.". It was only after the true idea of elements had been arrived at, and indeed not until Dalton had formulated the laws which go by his name, that the distinction was drawn. The ultimate criterion for combination is definiteness of proportion, and this is generally connected with uniformity in properties, or homogeneity. A substance is said to be homogeneous when no one part of it differs from any other part in composition. But this may be predicated of glass, or of air, which are mixtures, and not compounds. A mixture may be homogeneous ; a compound must.
Again, it is usually accepted that the separation of the constituents of a mixture may be effected by mechanical, or at least by physical means ; whereas the separation of the elements from a compound require chemical treatment. Here it is difficult to draw a sharp distinction. The separation of carbon dioxide from soda-water by the application of heat is similar in character to the separation of sugar from water by evaporation of the water ; yet we believe that a solution of carbon dioxide in water constitutes a compound, while that of sugar in water is a mere mixture of the two. It is necessary to be guided by analogy in the former case ; and it is probable that the compound named carbonic acid is really contained in a solution of carbon dioxide in water, on account of the formulae and behaviour of the carbonates.
 
Continue to: