6

but is, like anthracene for example, a product of destructive distillation at something like a red heat." "In 1876 Dr. H. W. C. Tweddle exhibited at Philadelphia a greenish substance that he called 'petrocene,' from which he obtained a yellowish-green substance which he called 'thallene.' This was the raw material of this research, the few killograms that were exhibited being obtained from 50,000 barrels of petroleum; the density of petrocene,' that is to say the crude material, is about 1.206. It was separated into lighter paraffines, having a density of about 0.990 and heavier hydro-carbons of about 1.27. Bromine and sulphuric acid separated from 5 to 15 per cent. of paraffine having a very high melting point, 70°, 80°, and 85° C.; ordinary paraffine melting at 65° C. The unsaturated hydrocarbons, anthracene, phenanthrene, chrysene, chrysozene, and pyrene were recognized. Organic analysis showed a hydro-carbon containing from 88 to 96 per cent. of carbon, which is a larger percentage than is found of coal, even anthracite rarely attaining 95 per cent."1 Prof. Samuel P. Sadtler, of the University of Pennsylvania, in a letter dated November 4, 1881, thus details some interesting experiments made by himself:

66

Classifying the subject under the three heads of: 1, Gaseous products accompanying crude petroleum; 2, Crude petroleum; and 3, Solid products accompanying and derived from the petroleum, I started with the first. I made analyses of some ten lots of natural gas' taken from wells in different parts of the oil field, and representing different

1 Prof. S. F. Peckam, U. S. Census Report, 1880.

geological horizons as far as possible.

As there was some

doubt as to whether the results of eudiometric analysis could indicate the presence of the higher members of the paraffine series, I supplemented these analyses by a series of absorption tests made on the spot. Thus I passed a current of natural gas for a time through absolute alcohol, which, while it does not dissolve hydrogen, absorbs marsh gas slightly, ethane, propane, and the higher hydro-carbons in increasing amount.

The hermetically-sealed flasks of the alcohol were then examined in my laboratory, and the gases absorbed driven out by heat, and collected over mercury and analyzed. They proved to be chiefly ethane and propane. I also passed a current of the gas through bromine, both pure and alcoholic, so as to absorb the olefines. On after-examination in my laboratory by neutralizing the free bromine with soda and diluting, I succeeded in separating out colorless oily drops of ethene dibromide, and, presumably, though not certainly, propene dibromide. In the study of the liquid crude oils, after classifying the oils from the different geological horizons (with information given to me by Mr. John F. Carll), and noting gravities, colors, and other physical properties, I proceeded to classify them by filtration (as far as possible in the cold) with animal charcoal and with mineral materials, like clay, alumina, etc. I did this with a view of examining chemically and microscopically the coloring impurities thus withdrawn. My results with these portions withdrawn by filtration are very incomplete; still I think they are largely made up of the members of the higher and more condensed

hydro-carbon series, like anthracene, etc., and not simply amorphous carbon, as supposed by some chemists. In corroboration of this view, I may say that in the crude oils, picric acid will strike a deep blood-red color, like the color of its compound, with anthracene, fluorine, etc. Whereas, in the yellow oil, clarified in the cold by animal charcoal, no such result is obtained. I also verified with a number of crude oils Schorlemmer's observation, that olefines are present, capable of being withdrawn by bromine, and, in small quantities, members of the benzole series, capable of yielding nitro-derivatives like nitro-benzole and nitro-toluole. Indeed, taking several distinct fractions gotten from Bradford oil, I obtained in notable quantities, in the lightest fraction, light-yellow nitro-benzole, and in the higher fractions reddish-yellow nitro-toluole, and probably higher products. I also extracted paraffine from a number of the crude oils, by mixing several volumes of ether with the oil and then chilling, when almost all the dissolved paraffine will separate and can be filtered off.

"Lastly of the solid products which accompany petroleum I examined the paraffine of buttery or firmer consistence, which separates out on the tubing, or derrick-frames, in Bradford oil wells. This was dark in color, looking like the crude ozocerite of Galicia, but not so firm. It had all the characters of a paraffine mixture; I had also collected a whitish, buttery mass from several flowing wells, near Warren, Pennsylvania; this on examination proved to be a perfect emulsion of oil and water, which would stand for months, but separated into distinct layers of oil and water

when warmed. I also took up for examination the solids gotten from Pennsylvania petroleum by pyrogenic formation. Of this character were petrocene and allied products, first mentioned by Dr. Herbert Tweddle, and from which Prof. Henry Morton extracted thallene."

In a subsequent paper published by Professor Sadtler in summing up the results of his own and other observations he says: "What was the material used for these investigations? Were the crude petroleums examined by these different authorities exactly the same, or if, by chance, they ought to have been, are they to be compared with all other petroleums now known? Those familiar with the crude oils as produced in the different sections of Venango, Clarion, and Butler counties, and very recently in Warren and McKean counties also, will know that these oils vary in color from a light amber to a dark black, and in gravity from 30° to 55° Baumé, from thick lubricating oils to nearly pure benzene. Moreover, they come from very different strata or sand rocks' as they are termed. It will thus be seen that if we wish to study the chemical composition of petroleum thoroughly, we have a considerable body of material to choose from. The material must also be carefully assorted before any satisfactory study of petroleum can be made. The great bulk of the crude petroleum that is sent to the refineries, or is exported, is shipped by the pipe-line companies, which have their network of pipes ramifying through whole districts collecting the entire yield of a district and storing it in their immense

tanks. To study such crude petroleums would be like analyzing the sweepings of a mineral cabinet."

When we add to the difficulties, just described, in the way of a correct analysis of petroleum, the different amounts of heat employed in effecting the separation of the compounds found in it, also the impossibility of deciding in many instances whether these compounds are educts or products, we have some conception of the difficulties which beset the path of the investigator. In presenting the present aspect of the chemistry of petroleum, we have endeavored to survey the whole field in order to present in the briefest space possible what is really known respecting this highly complex body.

1 United States Census Report, 1880, Prof. Peckam.