136 RESEARCHES ON THE VOLATILE HYDROCARBONS. 
has already been described in detail, in a memoir “On Fractional Condensation,” etc. 
(Memoirs of the American Academy, 1864.) 
This new process was first applied, more especially for the purpose of testing its 
efficiency, in the separation of benzole from coal-tar naphtha. This mixture was 
selected for the test on account of the property which benzole possesses, in contradis- 
tinction from its associates, of being crystallizable at a low temperature, thus affording 
an additional test of the purity of the product which might be obtained by the 
process of fractioning. Somewhat to my surprise I found that, after only the fifth 
series of fractionings, I had obtained benzole so nearly pure that the whole of it 
would distil from a tubulated retort between 80° and 81°C.; and that when congealed, 
which was effected by placing the containing bottle in pounded ice, not a drop of 
liquid could be poured from the mass of crystals. From this result, — which, at the 
least, indicated a near approximation to purity,— taken in connection with other 
favorable indications, I felt confident that I had accomplished my first object, and had 
found a process that could, in all probability, be successfully applied in the study of 
the petroleums, which up to that time (1861) had baffled every attempt to resolve 
them into their proximate constituents. 
Being naturally anxious to apply the new process in this seemingly more promising 
field of inquiry, I at once suspended, for the time being, my operations on coal-tar 
naphtha, and commenced simultaneously the investigation of Pennsylvanian petroleum, 
and of the oils distilled from Albert coal (from Hillsboro, New Brunswick) in the 
process of manufacturing illuminating oil. These two substances, neither of which 
had ever been made the subject of special scientific investigation, were selected as 
being fair representative types, on the one hand of the native liquid petroleums, and 
on the other of the artificial coal oils. The comparative study of these two substances 
seemed to promise additional interest on account of the close analogies which they 
present, especially when this circumstance is considered in connection with the fact of 
their great diversity of origin. This is the limit which at that time was assigned for 
these researches ; my intention being, so soon as the separations and analyses should 
be completed, and the boiling-points and some of the other more important physical 
characteristics determined, briefly to publish the results, together with the process of 
fractioning, — preliminary to a complete memoir at a more advanced stage of the 
work. Before this work had been accomplished, however, it became evident that the 
bodies contained in these mixtures could not be studied so satisfactorily by themselves 
as in comparison with other series of hydrocarbons, especially with reference to cer- 
tain important questions of more general interest ; for example, the question in regard 
