cold water, and, in the testing of oils of high flashing point, this water should be heated to about 50° C.

The Parrish Naphthometer.

This instrument is used chiefly in Holland, and differs from those already described, in that the inflammable mixture is carried out of the petroleum-holder to a stationary flame. A represents the tin oil-holder, C the water-bath,

Fig. 52.

Parrish's Naphthometer.

D the support, and E the lamp. The holder is provided with a projecting cover, in which is the cylinder d, having in its axis a small tube, with a wick running into the petroleum. e is a screw, against whose base rests the glass plate f for protecting the thermometer from the heat of the wickflame, and, lastly, B is a chamber communicating with the

air in which are the openings a and b; the former, for the circulation of the air through the petroleum-holder, and the latter, to allow the passage of the oil from B into A. The thermometer c is placed in the vessel B. The bath is filled with cold water, and the oil-holder with the petroleum to be tested, to a point one cm. below the rim. The heating must be slow and is effected by the spirit-lamp, whose flame is only one to one and a half cm. high. The small wick in d is then lighted, care being taken that the flame is not more than six to seven mm. high. The heat of this flame produces a current of air, which, coming in through the opening a, spreads over the surface of the oil and passes out by the tube d, taking with it the vapors evolved from the heated oil. When the oil vapors are sufficient in amount to produce an inflammable mixture they are ig nited by the flame in d, the flame being extinguished by the sudden motion of the air. At this moment the flashing temperature is read.

Engler's Tester.

The apparatus devised by Engler is of the closed form, to which is added an electric mechanism similar to that of the Saybolt tester. It is shown in Figs. 53 and 54. It consists of the copper water-bath A, heated by the spiritlamp B. CC is a glass vessel for water, which has a filling mark etched upon it. mm is the cover, and n the thermometer. In the cover is the glass petroleum-vessel D, also provided with a filling mark, and to which is fitted the brass cover o o. The latter is shown in Fig. 54, in

which will be noticed the following details: 88 are two movable covers, t t the conducting wires insulated by the ebony rings u u, r the thermometer, and q the handle of the stirrer p, seen in Fig. 53. The conducting wires

terminate in platinum points in the vessel D, from one-half to two-thirds cm. above the surface of the oil, and at a distance of one mm. from each other. For the production of the electric spark, a chromate cell is used with an induction apparatus, which gives a spark at least two to three mm. long. The electric apparatus of the Saybolt tester answers very well. In using this tester, the baths A and C are filled with water, and D is filled to the mark with the oil to be tested. When the petroleum vessel is in place, the water in C should stand one cm.

below the rim. The wires are connected with the induction coil and the lamp lighted. As the temperature rises to the testing-point, the spark is passed every degree, care being taken that the spark continues from one-half to one second. After each passage of the spark the oil is gently agitated by the stirrer. The operation is continued in this way until an explosion occurs, by which the covers 8 8 are thrown open.

In the hands of a careful manipulator, reliable results may be obtained by either of the above-described instruments. The difficulty has been that different results have been obtained by different operators in using the same tester. The cancellation of contracts, acrimonious differences, expensive lawsuits have all arisen, simply because legislation of a national or international character has been wanting to regulate so important a feature of a traffic amounting in the aggregate to millions of dollars.

The vast number of experiments which have been made to determine the accuracy of the different kinds of tests has at least pretty conclusively settled the question as to the superiority of the "closed" form of apparatus.

A very exhaustive report on this subject was made "to the Secretary of State for the Home Department" of England by Professor F. A. Abel, Chemist to the War Department, in which, having carefully considered the causes of the variable results obtained in testing the flashing point of different kinds of petroleum, he submitted a form of apparatus for adoption, together with specific directions for its use. This, which we have already described, is, we think,

exclusively employed in England and its colonies, and is also extensively used in this country. It is claimed, however, by those of large experience, that equally reliable results are obtainable by the Saybolt and Engler electrical

testers.

There is considerable difference between the flashing points of samples of oil tested in the "open" and "closed" apparatus. In the open tester, first adopted and legalized in England, 100° F. was chosen as a safe point. In the Abel apparatus now in use, the same point is reached at 73°, a difference of 27°, which Professor Abel says "might without injustice to the trade be accepted as the difference between the results to be furnished by the new test and the present legal test; or, in other words, that 73° might with the new test be accepted as the equivalent for the present legal minimum flashing point of 100°."

There is also a difference of many degrees between the flashing point and the burning point of a given sample of oil. Oils differ greatly in this respect; a very small percentage of light benzene will greatly lower the flashing point; if the great body of the oil is of high test, there will be a large difference between these two tests; but, on the other hand, if the bulk of the oil be of a lower grade, the difference between the flashing and burning points may amount to only a few degrees. The following table prepared by Dr. C. B. White, of New Orleans, being the result of a number of experiments, illustrates the effect of varying quantities of benzene mixed with oil, in changing the flashing and burning points of the same: