time a method of refining petroleum bad been generally introduced, by which a large proportion of the total burning-oil produced consisted of mixed or cracked oils. Such a proportion of high-test oils as were demanded by the market was made, but the great bulk of the distillate had been converted into a cracked or mixed oil. The petroleum was distilled but once, the naphtha was removed, and then the remainder of the oil manipulated to produce such crude burning-oils as were desired, leaving in the still only a small percentage of residuum. These crude burning-oils were treated with sulphuric acid and caustic soda in such a manner as to produce the lightest colored oil possible, and they were further manipulated to bring the test within the legal requirements. As it was much less difficult to bring the mixed or cracked oils within the requirements of a burning rather than a flash test, the burning test has always found strong advocates among a certain class of the manufacturers of petroleum. This method of manufacture was well established, and the markets of the world were well accustomed to handling the various products during that period when the bulk of the crude oil came from the Butler-Clarion district. But gradually, as has been stated, the major portion of the crude oil that flowed into the pipe-lines was no longer from the Butler-Clarion wells, but from those of Bradford. By the end of 1881 more than three quarters of the crude oil was Bradford oil, and the relative proportion has steadily increased. This change in the crude material has been accompanied by a corresponding change in the character of the product. Instead of mixed and cracked oils, consisting largely of normal burning-oil, the products of Bradford crude oil consist largely of the products of destructive distillation, and this is due to the fact that the petroleums of the Butler-Clarion and Bradford districts represent two extremes; the first contains the smallest proportion and the latter the largest proportion of paraffine-oils of any crude petroleums found in large quantities. The proportion of cracked oils in the distillate from the Butler-Clarion petroleum was too small to injure the general quality of the oil. In the Bradford distillates, on the contrary, the products of destructive distillation give character to the whole. And not only is this statement true, but the proportion of high-test normal oils to be obtained at present from the pipe-line crude oil has gradually become so reduced that the best brands of oil on the market have deteriorated, until it is very difficult, if not impossible, to purchase an article of burning-oil equal in quality to the best offered for sale a few years since. All this time the requirements of law in regard to test have been met, perhaps it may be said, with increased faithfulness.
From the foregoing pages it must be manifest that any improvement in the increasingly bad quality of kerosene can be looked for only from one of two directions. Either it must come from the development of a new field for crude oil of superior quality, or from the in-
