CITEMICAL CONSTITUTION. The molecular fornuda of a compound represents its composition and its smallest relative reacting-weight. It is not, however, altogether characteristic of the for numerous cases are known in which a number of different compounds are repre sented by one and the same molecular formula. Thus, both ordinary alcohol and di-methyl ether (a substance that may be obtained from wood alcohol) are represented by the formula live compounds, viz. ordinary ether and the four different substances called butyl-alcohols. arc found to have in common the formula etc. Such compounds arc said to be isomeric, or more strictly, met amerir, the term isomeric being often extended to include also the so-called polymeric eompounds, i.e. those which have the same relative composition hut not the same molecular weight. such as acetylene. CM,. and benzene. (Another kind of isomerism may be found discussed in the article STEREO-CHEMIS TRY. ) The isomerism of carbon compounds shows plainly that the composition of a substance does not entirely determine its chemical individuality. For two or more molecules composed of the same kind and number of atoms may represent very different compounds. It is therefore evident
that the character of a compound must depend to a great extent also upon the manner in which the atoms are held in combination within its molecule. This, at least, is the only -possible explanation that presents itself to the mind, as suming that substances are really made up of atoms and molecules: and without this assump tion, i.e. without the hypothesis, isomer ism could not be explained at all. After, there fore. the composition and molecular weight of a newly isolated compound have been determined, a further and much more difficult problem remains to be solved; viz. to determine the chemical con stitution of the compound, i.e. the manner in which the atoms are arranged in its molecules. This problem is solved by combining the results of a careful study of the chemical and physical properties of the compound with a theoretical assumption first induced independently by Ke kule and Couper. The following paragraphs may convey an idea as to how this is done.