ON A PROCESS OF FRACTIONAL CONDENSATION. 123 
tion of the petroleum from Sehnde, near Hannover, also with the use of Wurtz’s bulbs, 
both assert in the most positive manner the impossibility of separating from petro- 
leum, by fractional distillation, products of constant boiling-point. 
Such is the general character of the results obtained in the attempts which 
have been made to separate the constituents of such mixtures by fractional distil- 
lation. 
The treatment with strong acids, etc.,as an auxiliary to the common method of 
fractional distillation, which is claimed to have given good results in some cases, is 
open to serious objections in its application to mixtures of unknown substances, as 
must be readily apparent. The further consideration of this subject is reserved for 
another occasion, when I shall submit the results which I have obtained by my 
process in the study of mixtures almost identical with some of those in the investiga- 
tion of which the acid process has been employed. I shall then be able to show that 
the results obtained by that process are, to a considerable extent, inaccurate and by 
no means exhaustive; and that it is still of the highest importance to have a process 
which shall be generally applicable in all such cases, without resort to any harsh and 
uncertain treatment. 
With regard to the value of constancy of boiling-point above referred to, as a test 
of purity of a liquid substance, I may here say that, without scarcely lessening the 
importance of obtaining constancy of boiling-point, before resorting to harsher treat- 
ment, in the study of mixtures of unknown substances, I think I shall be able to 
show, on another occasion, that this property is not necessarily indicative of so high a 
degree of purity as has generally been supposed; and that a body may have a con- 
stant boiling-point, and yet contain enough of a foreign substance to appreciably — 
and, in delicate cases, seriously — affect the determination of its constitution and of 
some of its other properties. But in no such case have I yet found that the removal 
of the impurity by chemical means has essentially changed the boiling-point, —z. e., 
never to the extent of 1° C. of temperature. I propose, at a future time, to study this 
question synthetically, operating with pure liquid substances, with the view to deter- 
mine, in a few cases, how much of a foreign substance may be present, — which would 
probably be variable in different cases, — without sensibly affecting the boiling-point. 
A solution of this question would, I think, be of considerable practical value in some 
instances.* 
* Since this was prepared for the press I notice that late experiments by Berthelot go to show the correct- 
ness of my conception of the value of constancy of boiling-point, as above stated. 
