Caspian Sea, to which we have alluded in another chapter. We have collected the memoranda and placed them in tabular form. It will be observed that the statements which we have made respecting the superior calorific power of the Russian petroleum are substantiated by these experiments.

There is really no practical difficulty in the way of a much more extensive use of petroleum as fuel, excepting in the matter of cost. For the production of steam, espe cially in the oil regions, it is largely employed. Here a large amount of waste material, such as coke mixed with residuum "settling" from tanks, is employed for this purpose.

Favre and Silberman give as the result of their calculation the following figures respecting the steam-producing power of petroleum: "One pound of carbon combining with 2 pounds of oxygen will evaporate 15 pounds of water at 100° C.; one pound of hydrogen combining with 8 pounds of oxygen will evaporate 64.2 pounds of water; and one pound of petroleum consisting of six parts of carbon and one of hydrogen has for its theoretical evapo

rating power 22.02 pounds. Allowing 20 per cent. of non-combustible matter in anthracite, one pound of it will evaporate theoretically 12 pounds of water; so that one part of petroleum is equal to 1.835 anthracite. A low price for crude petroleum in most markets in America would be four cents per litre, or two cents per pound, or $40 per ton. Average price of anthracite in most markets in America is $8. In some places it is as low as $4, in others $12, but at $8 per ton petroleum would cost for equal heating power about three times as much as anthracite."

Professor S. F. Peckan, in the Tenth Census Report on petroleum, furnishes the following item respecting the steam producing power of petroleum :—

"The oil was burned with a steam jet under four stationary boilers (60 inch shells, 14 feet long with 83.3 inch tubes, and the steam furnished a Worthington compound duplex pump doing an actual work of about 200 horse-power. The indicated horse-power would probably be about 225 or 250). These boilers and this pump use as nearly as possible 4.54 pounds of bituminous coal per horse-power of work done per hour. Using this average, which is pretty well determined as a basis, one ton of 2000 pounds of this coal is equal as fuel to either 3.94 or 4.13 barrels of 42 gallons each of oil. The experiment was not conducted as it should have been, and there is a question as to the pressure against which the pump worked, which accounts for the difference in the estimate. I think it may be stated, however, that four barrels of oil would be re

quired to furnish the equivalent of a ton of good bituminous coal if the oil is burned with a steam jet. With an air jet I look for better results."

It would be quite useless to multiply the results of similar experiments. It could be easily done, but they all point to the same conclusion, and, moreover, they all verify what could so easily have been predicted from an accurate analysis, and the determination of the actual calorific power from such analysis of petroleum.

As we have before observed we must look for the best results from petroleum, both economically and technically, in those uses where the improved product of the manufactured article more than counterbalances the difference in the price of the two kinds of fuel. We shall not have to go far to find such uses. Even in the manufacture of iron, where the cost of fuel enters very largely into the calculations, there are many conditions where the enhanced value of the product more than compensates for this difference.

At Woolwich, England, in the manufacture of armorplate for war vessels, most remarkable results have been obtained from the use of the "liquid fuel."

"Under ordinary circumstances the armor-plate bending furnace is lighted from four to five hours before the plate is placed in it. The time occupied in heating the plate for bending depends upon its thickness, one hour per inch of thickness being allowed. Taking then a six-inch plate, we get from ten to eleven hours from the time of starting before the plate is ready for bending. Let us now see what the liquid fuel will do. The cold furnace is lighted, and

after an hour is deemed sufficiently heated. A 6-inch armor-plate, 7 feet 6 inches long by 3 feet wide, is then consigned to the furnace, and after an hour and a half is drawn out thoroughly heated and ready for bending. Thus in two hours and a half we have the work of ten or eleven hours completely and satisfactorily performed. Nor is this all, the advantages of the system do not stop here. The plate is remarkably free from scale, which can only be accounted for by the absence of the deteriorating influence of the products of combustion in the ordinary furnace. Another valuable result arises from this same cause. Thin plates, when heated by liquid fuel and bent double, show no signs of cracking, as they usually do when heated in the coal furnace. This important feature is said to save ten shillings per ton on the metal, which amount it would lose in value by deterioration under the ordinary method of treatThe fuel, in a vaporized condition, is supplied from the generator to the furnace by six jets, which are led in through small openings, by which means just a sufficient quantity of atmospheric air is admitted to support combustion. This method of supplying the heat also offers another advantage; it can be applied to the whole or any portion of the plate. Thus, if a plate requires to be bent at one end only, then the heat is directed to that part. Further, the rate at which the metal is heated can be regulated to a nicety by increasing or diminishing the number of jets. Close beside the armor-plate furnace is another one for heating thinner ones, which has been regularly at work for some time past. It is heated by four jets, and is

ment.

supplied from the same generator which is placed between the two. The average time occupied in heating is seven minutes; with the ordinary furnace, it takes from twelve to fifteen minutes for each plate. In the working of this furnace we have some remarkable results, which must lead us to expect further and even more important improvements in the application of the system.

We have in the records of the results of the operation of an iron furnace at St. Louis, Missouri, an interesting fact brought out, where the increased yield of the product made it profitable to employ petroleum as fuel.

are the figures:

The following