170 RESEARCHES ON THE VOLATILE HYDROCARBONS. 
tainly the appearance of having been performed with great care, conducting to a 
beautiful harmony of results. My confidence in his determinations of boiling-points 
was increased not a little by his alleged discovery in coal-naphtha of xylole, boiling at 
126°.2, indicating a more thorough analysis of this naphtha than those which had been 
previously published. This body, the supposed middle member of the benzole series, 
had up to that time been regarded as wanting in coal-tar naphtha, although all of the 
other members, above and below it, were found to be present,— an anomaly not 
easily reconciled with any plausible theory in regard to the formation of these bodies. 
In view of these circumstances, therefore, I was naturally led, from analogy, to antici- 
pate that the boiling-point difference among the hydrocarbons from petroleum and 
Albert coal would not vary much from 20°. Not being able, however, to reconcile 
with previous facts and theories on this subject, the indications which were being 
gradually unfolded by my seemingly unerring process of separation, I was compelled 
to lay aside all bias, and to regard these indications as pointing unmistakably to a 
much greater difference of boiling-point for the addition of C, H, than had previously 
been supposed to exist in this class of substances. 
Having finally established beyond question the common difference of 30° for the 
addition of C,H, among the hydrocarbons from Albert coal and petroleum (the third 
series from petroleum, with the difference of 20°, had not then been reached), I began 
to surmise that this difference might be found to be common among all other series 
of hydroearbons. In this connection my mind naturally reverted to the earlier deter- 
minations of the boiling-points of the members of the benzole series, some of which, ° 
especially those of benzole and toluole, which had been more studied than the others, 
indicated strongly that 30° might prove to be the true difference for the addition of 
C, H, in this series. My confidence in Church’s determinations thus began to dimin- 
ish, and finally I undertook to make a thorough analysis of coal-tar naphtha, the 
results of which are given in Table 3. As there shown, the boiling-point difference in 
the benzole series is also 30°, and the number of its members is reduced to four, in 
place of five, as alleged by Church. 
This difference of 30°, thus shown to be so common with the hydrocarbons, is so 
much larger than the difference of 19° which Kopp had found so frequent in other 
classes of substances, that the discrepancy cannot be regarded otherwise than as con- 
clusive evidence, if such were wanting, that all liquid bodies do not obey the same law 
in this regard, but that there are unquestionably those series in which the boiling- 
point difference for the elementary difference of C,H, may be greater than 19°, of 
which Kopp has already furnished some examples. 
