208 ANNUAL EEPORT SMITHSONIAN INSTITUTION,' 1916. 



To avoid the force of this objection, it would be necessary to elimi- 

 nate the cyclising action of heat during the decomposition of coal. 

 This is what I have attempted to do with the assistance of my two 

 expert collaborators. In operating the distillation of coal in vacuo, 

 so as not to admit of an increase in temperature above 450°, we ob- 

 tained a special coal tar and a new kind of coke. But in studying 

 this vacuum coal tar and coke we have assured ourselves that each 

 of them, like ordinary coal tar and coke, are exclusively of cyclic 

 nature. We conclude from this that the cyclic compounds pre- 

 exist in coal and certainly form its major part. From these ex- 

 perimental results there follow, in our opinion, the three following 

 conclusions : 



(1) Berthelot's theory of the formation of coal tar can not be con- 

 sidered as accurately interpreting the facts. All the derivatives of 

 coal tar which chemical industry has utilized in such a brilliant man- 

 ner, are no longer believed, as formerly, products of heat action. It 

 is not at all to the heat of the gas jets that is due their well-lmown 

 aromatic radical so rich in valuable properties. This radical al- 

 ready existed though in a more hydrogenated condition, in the 

 plants of the carboniferous age. All chemistry of the aromatic com- 

 pounds thus owes a dependence on plant chemistry. 



(2) Vacuum coal tar is in reality nothing more than petroleum, 

 having its odor, density, fluorescence, and weak rotatory power. All 

 the definite compounds that we have derived from it are found to be 

 identical, with those other compounds isolated from the petroleums 

 of Canada, California, and Galicia. We therefore verify for the first 

 time, a relation of a chemical order between these two natural i^rod- 

 ucts of such high importance, coal and petroleum. Does this rela- 

 tion imply a common origin, and can it serve as an argument for those 

 who claim that petroleum, like coal, is of plant origin ? For my part 

 I believe so, but to enter into a discussion of that point would be too 

 far from my subject. 



(3) If coal, as we believe we have demonstrated, is formed of a 

 mixture of cyclic substances, one could hardly fail to attribute the 

 same structure to cellulose, which, of all the substances contained in 

 plants, is the one that plays the greatest part in the formation of 

 coal. The objection that my opponents would make in this respect 

 therefore falls and my hypothesis conversely finds a new example for 

 its support. 



With one span we will now bridge the entire distance separating 

 the first products of plant assimilation from its final product, 

 namely, living matter. And it should be understood at the outset 

 that I employ this term "living matter" only as an abbreviation, 

 and to avoid long circumlocution. You should not, in reality, attribute 



