522 Dr. D. Mendeleeff [May 13, 



have the one substance, methane, for their origin. The number of 

 substances is so enormous that it is no longer a question of enlarging 

 the possibilities of discovery, but rather of finding some means of 

 testing them, analogous to the well-known two which for a long 

 time have served as gauges for all carbon compounds. 



I refer to the law of even numbers and to that of limits, the 

 first enunciated by Gerhardt forty years ago, with respect to hydro- 

 carbons, namely, that their molecules always contain an even number 

 of atoms of hydrogen. But by the method which I have used of 

 deriving all the hydrocarbons from methane, CH^, this law may be 

 deduced as a direct consequence of the principle of substitutions. 

 Accordingly, in methylation, CH^ takes the jDlace of H, and there- 

 fore CH^ is added. In methylenation the number of atoms of hydrogen 

 remains unchanged, and at each acetylenation it is reduced by two, 

 and in carbonisation by four atoms, that is to say, an even number 

 of atoms of hydrogen is always added or removed. And because 

 the fundamental hydrocarbon, methane, CH"^, contains an even number 

 of atoms of hydrogen, therefore all its derivative hydrocarbons will 

 also contain even numbers of hydrogen, and this constitutes the law 

 of even numbered parts. 



The principle of substitutions explains with equal sim})licity the 

 conception of limiting compositions of hydrocarbons, C'W + -, which 

 I derived, in 1861,* in an empirical manner from accumulated 

 materials available at that time, and on the basis of the limits to 

 combinations w^orked out by Dr. Frankland for other elements. 



Of all the various substitutions the highest proportion of hydi'ogen 

 is yielded by methylation, because in that operation alone does the 

 quantity of hydrogen increase ; therefore, taking metbane as a point 

 of departure, if we imagine methylation effected {n — 1) times we 

 obtain hydrocarbon compounds containing the highest quantities of 

 hydrogen. It is evident that they will contain 



CH^ + (71 - 1) CH^ or C"H2''+2, 



because methylation leads to the addition of CH^ to the compound. 



It will thus be seen that by the principle of substitution — that is 

 to say, by the third law of Newton — we are able to deduce, in the 

 simplest manner, not only the individual composition, the isomerism, 

 and relations of substances, but also the general laws which govern 

 theii' most complex combinations, without having recourse either to 

 statical constructions, to the definition of atomicities, to the exclusion 

 of free affinities, or to the recognition of those single, double, or treble 

 ties which are so indispensable to structurists in the explanation of 

 the composition and construction of hydrocarbon compounds. And 

 yet, by the application of the dynamic principles of Newton, we can 



* ' Essai dune theorie sur les limites des combinaisons organiques,' par 

 D. Mendeleeff, 2/11 aout 1861, ' Bulletin de I'Acadenaie i. d. Sc. de St. Peters- 

 bourf^,' t. V. 



