LIBRARY OF THE UNIVERSITY OF CALIFORNIA. RECEIVED BY EXCHANGE Class The Fractionation of Crude Petroleum by Capillary Filtration. DISSERTATION SUBMITTED TO THE BOARD OF UNIVERSITY STUDIES OF THE JOHNS HOPKINS UNIVERSITY IN CONFORMITY WITH THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY. BY MARSHALL PERLEY CRAM, EASTON, PA. : ESCHENBACH PRINTING COMPANY. 1908 The Fractionation of Crude Petroleum by Capillary Filtration. DISSERTATION SUBMITTED TO THE BOARD OF UNIVERSITY STUDIES OF THE JOHNS HOPKINS UNIVERSITY IN CONFORMITY WITH THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY. BY MARSHALL PERLEY CRAM. 1908 EASTON, PA. : ESCHENBACH PRINTING COMPANY. 1908 CONTENTS. Page. Acknowledgment 4 The Fractionation of Crude Petroleum 5 Water Fractionation 45 Oil Lost in the Earth 47 Fractionating Power of Substances other than Fuller's Earth . . 49 Summary 50 Biographical 51 202452 ACKNOWLEDGMENT. The author takes pleasure in expressing his gratitude to President Remsen, Professor Morse, and Professor Jones for instruction in the lecture room and laboratory. Especial thanks are due to Dr. Gilpin, under whose personal direction this investigation has been pursued, and to Dr. D. T. Day, of the U. S. G. S., who has contributed valuable suggestions and apparatus. The Fractionation of Crude Petroleum by Capillary Filtration. When in process of refinement, black vaseline is filtered through warm dry fuller's earth the first product is an oil perfectly liquid at ordinary temperatures, while the suc- ceeding portions are progressively more viscous until quite hard vaseline is obtained. This observation, that a frac- tional separation of oils in vaseline had been effected, sug- gested to D. T. Day that a like result might be obtained with crude petroleum. He applied this method to a sample of the green crude petroleum from the "third sand," Venango County, Pennsylvania, and found that light products, chiefly gasoline, first appeared when such crude oil was allowed to filter down through a long glass tube filled with granu- lated or powdered fuller's earth. 1 Following this a more elaborate system of specially con- structed funnels similar to those used by the refiners of vaseline in testing the comparative value of various fuller's earths was used. The results from this were briefly sum- marized in a paper upon the ability of petroleum to migrate in the earth. 2 Engler later verified these results and showed that the separation was mechanical and that no oxidation was effected in the process. Day next used a large closed funnel of galvanized iron holding about one hundred pounds of fuller's earth. When crude petroleum was dropped slowly and regularly into this, quite light oils at first, fol- lowed by the usual succession of heavier oils, were obtained. As it was evident from this work that much of the oil passed through crevices without any change he tried the effect of reversing the route of the oil and of allowing it to diffuse 1 Philadelphia Acad. of Sci., 1897. 2 Trans. Petroleum Congress (Paris), 1900. upward through a tube packed tightly with fuller's earth. In such a tube the lighter constituents rose much more rapidly than the more viscous oils so that by separating the fuller's earth from different sections of the tube and displacing the oil by water, quite different oils were obtained from the upper and lower parts of the tube. By using several tubes and uniting oils of the same specific gravity, oil of different grades could be collected in sufficient quantity to be fractionated again, and the process con- tinued until oils result which are not altered by further passage through tubes filled with fuller's earth. At the suggestion and with the cooperation of Day, we have taken up this problem with the results here stated. The tubes used first were three feet long and one and one-eighth inches in diameter. They were closed at the lower end with corks, along the sides of which grooves had been cut, the top of the cork being covered with a bit of cotton cloth to prevent the earth from sifting out of the grooves. Such tubes filled with fuller's earth 1 were placed with their lower ends in an open dish of petroleum and the oil was allowed to rise. At room temperatures (i8-22 C.) and atmospheric pres- sure, the rate of rise of crude petroleum in a tube filled with fuller's earth was very slow. In seven days the oil ascended but 73 cm. in one tube, while ten days in one case and seven in another were required for it to rise 59 cm. To study the effect of heat, a glass tube about three feet long and iV 8 inches in diameter, was filled with earth and placed in a bottle holding about two liters of oil, and the whole heated by an electric stove with which temperatures considerably above those of the room could be maintained day and night. The temperature of the tube was kept between 40 and 70 for three days, in which time the oil rose 54.7 cm. in the tube; in another tube packed in all ways like the former but held at room temperature (about 20), the oil rose 46 cm. in the same length of time. With two tubes in which the earth was packed much less compactly, the time required for the oil to rise 54 cm. was four days for the tube at room 1 The fuller's earth used in this work was kindly furnished by the Atlantic Refining Co., of Philadelphia. temperature and two days for the one at 50 to 80. The rate of rise evidently was but little effected by heat, at least within this range of temperature, and higher tem- peratures could not be used without loss of the more volatile constituents of the oil. The next attempt at increasing the rate of rise of the oil consisted in applying diminished pressure to the top of the tube which reduced the time required for the oil to reach the top of a tube five feet long from several weeks to seventeen hours. If diminished pressure is continued after the oil has reached the top, provided the oil is not exhausted in the reservoir at the bottom, oil will be drawn over from the top of the tube. The specific gravity of the oil thus collected steadily rises as it comes over. The samples so obtained, however, stand under very low pressures for some time, which may cause a loss of their more volatile constituents. This suggested applying increased pressure to the oil in the reservoir rather than diminished pressure to the top of the tube, and an iron bomb, like those used for the transporta- tion of mercury, was fitted with an iron pipe seven feet long to contain earth and a side arm at the bottom of the bomb to which a water column might be attached. The bomb which held about two liters could be partly filled with petroleum and the pipe containing the earth screwed into the top. The side arm which opened into the bottom of the bomb could then be connected with the water pressure so that the lower part of the bomb was filled with water which drove the petroleum upwards. The oil ob- tained at the top, however, was fractionated no further nor in any larger amounts than when the oil was not allowed to emerge from the top of the tube. The difficulty of setting up such a pressure apparatus with tight connections, as well as the range of pressure required a column of water seven feet high being too great when the oil was just started up the tube while a column thirty feet high was insufficient when it was near the top made its use impracticable. To use diminished pressure, the earth in the tubes must not be packed so hard that the air just above the oil cannot 8 be drawn through the earth above, nor must the earth be packed so loosely that the oil will rise as in a vacuum. The right degree of hardness is obtained by filling about a foot of the tube at a time and packing that much earth as hard as possible with a wooden rod tipped with a rubber stopper. If the tube when pounded upon the floor, rings in the hand, the earth is apt to be packed too closely. Packing tubes may be much facilitated by filling several at once with a separate ramrod for each. By allowing a few minutes to elapse between successive liftings of the ramrod, much labor is avoided. The fuller's earth was first heated in shallow iron pans until it ceased to form geysers when stirred. The earth must be thoroughly cold before it is packed into tubes or the contraction is sufficient to allow the oil to run up the tube immediately when the air is exhausted. The lower end of the tube is best closed with a cork with six or seven grooves cut along the side, and the inner end covered with a bit of cloth to keep the earth from sifting out of the grooves. At the top of the tube a bit of cotton waste below a rubber stopper will prevent any earth from being drawn up when the air is exhausted. The tubes used first were three feet long and one and one- eighth inches in diameter and of glass. Much trouble was experienced on account of their breaking, not when in ser- vice, but soon after they had been used. This was thought to be due to the age of the tubing, but the same happened with new tubes five feet long and one and one-fourth inches in diameter. With the idea that the iron scraper used to remove the earth from the tubes might be the cause, a scraper entirely of wood was tried, but this did not decrease the breakage, it being nothing unusual on going to the laboratory in the morning to find half of the tubes which had been emptied the day before cracked. It had been considered necessary to use tubes of glass in order that the height to which the oil had risen could be seen and that in removing the oil from the middle of the tube it might be scraped out to a sharp dividing line, since the level to which the oil has risen is the point from which all measurements should be made of sections into which the tube is to be divided. Tin tubes were used later to avoid the trouble experienced with glass tubes. These tin tubes were emptied by shaking the earth from the bottom into four thirty centimeter cylinders of the same diameter as the tube, these cylinders being made of two curved pieces of tin held together by a cap at one end and a ring at the other. The cylinders containing the contents of the tube could be opened lengthwise and the earth divided into any desired lengths. Two glass tubes five feet long and one and one-fourth inches in diameter were set up in the same dish of petroleum with ten or twenty tin tubes five and one-half feet long and of the same diameter, and when the oil stood at the top of the glass tubes the tin ones were also opened. Glass tubes, of course, can be emptied as well as tin ones by shaking the contents from the bottom, and no more tubes broke after this method was adopted. The level to which the oil will rise can be regulated by the amount of oil in which the tube is placed, and in the later work the adoption of this method did away with the use of glass tubes entirely. 950 cc. of oil in a tube one and one- fourth inches in diameter and five and one- half feet long will rise to within 20 to 35 cm. of the top. When the oily earth has been removed from the tube, the oil may be separated by adding water. As first prac- tised, enough water was added to form a very thin mud which was thoroughly stirred by a small propeller driven by a water motor. The mixed earth, oil, and water were then poured into a large separating funnel and allowed to stand several minutes until the oil had collected at the top. The earth and water could then be drawn off and the pure oil left. It was found later, however, that if less water is added to the earth as removed from the tubes, after standing a few minutes all the water will pass into the earth and this will be accompanied by the liberation of oil. Oil so liberated can then be poured off directly from the earth without the labor of churning. When water first begins to liberate oil, the 10 earth is granular while when more water has been added and the last of the oil recovered the earth has the consistency of a thin paste which will flow when the dish is inclined, which it will not do when the oil begins to come off. All the oil from the same section of a tube is of the same color irrespective of whether it is the first oil to come off when water is added or whether it does not come until the last. It was assumed at first that all such oil which came from the same section of earth had the same specific gravity irrespective of whether it was the first or last replaced when water was added, but this was found later not to be the case. The first oil to be collected, if taken in sufficiently small volume, is slightly heavier than the next portion. If too much is included in the first sample this will not be the case. Beginning with the second sample the successive portions of oil steadily increase in specific gravity, this gradual addi- tion of water affording another means of fractionation in addition to the separating power of the earth. Both of these methods of separation have been combined in this investiga- tion. The earth must be thoroughly mixed after each addition of water to prevent a layer of water wet earth from isolating earth, which contains oil, from the water added. The petroleum used was a dark green oil from Venango County, Pennsylvania, of specific gravity 0.810. When 950 cc. of this were drawn upwards in a tin tube five and one- half feet long the following separation was obtained as shown in Table I. II TABLE I. SINGLE TUBES IN THE CRUDE PETROLEUM. (0 (2) ( 3 ) Time required 23 . 5 hrs. 17.5 hrs. 17.5 hrs. Distance from top of tube to oil when opened 31 cm. 28 cm. 28 cm. Sp.gr. cc. Sp.gr. cc. Sp.gr. cc. A, 8 cm. at top 0.796 42 0.8012 30 0.8022 18 B, next 8 cm 0.808 45 0.804 37 0.803 35 C, next 18 cm 0.8125 75 0.807 47 0.8075 66 0.8137 24 0.809 22 0.810 25 D, next 30 cm 0.815 130 0.8125 148 0.812 140 E, next 35 cm 0.818 170 0.8185 190 0.8175 145 F, rest 0.8205 125 0.823 100 0.821 105 611 574 534 The oil from section C was collected in two portions, the second being obtained by the addition of more water after the first lot of oil was poured off. Although 950 cc. of crude petroleum were used in each case, it will be noticed that the oil recovered measures much less than that. When several tubes were worked up together in one case 9070 cc. of crude petroleum yielded 5951 cc. of oil, and in another case 8915 cc. gave 5415 cc. To collect a sufficient quantity of oil several tubes were placed in the same container of petroleum, two of the tubes being of glass and the rest of tin. When the oil had reached the top of the glass tubes they and the tin ones were opened, and the earth from the same level in all the tubes mixed in tin pails. The oil was then liberated in several successive fractions by the addition of successive amounts of water. If the earth had been thoroughly mixed after each addition of water the various oils from the same lot of earth would increase regularly in specific gravity, instead of showing the variations which in many cases they do, e. g., the first oil to be displaced by water in D and B of (4), in Table II, with so large volume, would not have been heavier than the succeeding oils if the water and earth had been thoroughly mixed before the oil was poured off. If we were to repeat the work instead of using one common reservoir for all the 12 tubes, we would use a separate reservoir for each tube, and open the tube when the oil in .the reservoir was exhausted. This would do away with the use of glass tubes entirely, besides insuring that the level of the oil in each tube when opened would be practically the same. If a common reser- voir is to be used the tubes should all be packed with prac- tically the same degree of hardness if the oil is to ascend in all with equal rapidity since the ascent in all tubes is checked at the same time, i. e., when the oil in the reservoir is ex- hausted. Diminished pressure was obtained by use of a Chapman water pump which reduced the pressure to from five to twelve cm. Hg when connected with a system of tubes. In the earlier work the pump was not run through the night, which is the reason for the much longer time required for these lots of tubes. The earth from a tube was divided into six sections, the level to which the oil had ascended in the tube being taken as the point to be measured from. A, the top section, in- cludes the eight cm. next the top; B, the next eight cm.; C, the next eighteen; D, the next thirty; B, the next thirty- five; while F includes what earth is left. F varies of course depending upon the height to which the oil has risen. In fractionating the crude petroleum in bulk, F was usually discarded since it was so viscous that it was deemed im- possible to pass it through earth again. Records from several lots of tubes are given in Table II. The specific gravity was measured with a Westphal balance, the oil being in every case at exactly 20 C. While the fourth decimal place is not to be taken as strictly accurate, yet it is con- sidered worth while to record it as giving a nearer approach to the truth than would result from the use of only three decimal places. TABLE II. FIRST FRACTION ATION OF CRUDE (4) (5) Tubes 2 glass, 8 tin. 2 glass, 7 tin. Distance from Glass, o, 15. Glass, 0,6. 5. level of oil to Tin, 40, 15, 24, Tin, 26, 28, 18, top of tube in 22, 28, 19, 24, 15, 28, 12, 24. cm. 1 32. Hrs. required 54 24 PETROLEUM. (6) 2 glass, 8 tin. Glass, o, 12. Tin, 30, 28, 22, 30, 13, 12. 54 Sp. gr. cc. Sp. gr. cc. Sp. gr. cc. A .8015 50 804 IOO O 80 5 65 .8005 350 0, 8055 190 80 5 20O B .807 260 0. 8085 22O 80 7 I4O O .810 190 o ,8ll 1 2O o 8097 125 C .809 IOO 8097 430 810 390 .809 400 o 8l22 300 o 8135 380 o .810 225 o .8115 260 D .815 425 8l 3 530 8133 610 o .8145 625 0, 8135 600 816 325 .8175 460 o 816 200 o 8175 435 o .825 125 E o .8l6 440 o. 8162 480 o ,816 850 o -8I 5 400 o 8162 725 o 8195 260 o .821 830 o 819 390 o .821 330 .827 325 1 The glass tubes are five feet long, the tin tubes five and one-half feet. Both are one and one-quarter inches in diameter. (7) 2 glass, 7 tin. Glass, o, 7. Tin 25. 26. 23, 26, 28, 20, 28, TABLE II. (Continued). (8) (9) 2 glass, 7 tin. 2 glass, 7 tin. Glass, o, 5. Glass, 8, 5, 15. Tin, 16. 34, Tin 58, 31, 28, 15, 18, 23, 10. 60, 40, 16, 53, 16. do) 2 glass, 9 tin. Glass, o, 8. 5. Tin, 40, 32, 36, 26, 27, 27, 20, 25, 18, 84 \ 48 117 84 Ao 800 200 O. 798 130 .8025 175 0. 7995 200 .802 H5 O. 80 1 130 o .8037 I2O O. 8037 160 0. 806 23 Bo .8042 2OO O. 803 155 .8042 180 o. 8085 125 .8048 2OO 0. 8045 230 .8078 215 0. 810 275 o. 8112 23 C o ,808 330 o. 8072 430 o .809 300 0. 810 350 .8078 430 0. 808 275 .8095 440 0. 812 525 .811 95 o. 808 225 .8127 100 o. 8i35 150 Do .812 425 0. 8117 420 .812 390 0. 8148 440 o .812 625 o. 812 580 o 8137 400 0. 8i75 700 .814 360 0. 822 250 1 .8145 300 0. 817 370 0. 8145 300 .8155 200 0. 817 200 0. 8177 42 Bo .8172 240 0. 8135 390 .818 340 0. 8197 315 .816 650 o. 814 560 8197 240 0. 8215 720 o .8162 660 o. 8248 390 1 .818 290 o. 8223 570 .8195 300 0. 817 230 .818 350 o. 821 215 0. 821 42 1 These fractions stood uncovered on top of the earth over night and consequently were exposed to considerable evaporation. ^EHBLAXf^ C\f TH' v a TABLE II. (Continued). (n) 2 glass, 8 tin. Glass, o, 4. Tin, o, o, o, o, 20, 17, 7i 5- ( 2 glass, Glass, Tin, 19, 32, ?5, 12) 6 tin. 7, 13. 32, 48, 29. 2 glass, 9 tin. Glass, 2, 2, 5, Tin, 17, 10, n, 22, 16, 29, 8, 5, 5, 5, 14. Tin (14) 2 glass, 9 tin. Glass, o, o. , 20, 20, 5, 13, 5, , 5, 19, 8, 12, 14. 39 60 20 16.5 Ao 799 80 .798 150 8oi 5 no 798 225 .802 1 80 O .801 140 8055 40 o . 8005 no O .8028 125 O 8015 80 8033 100 Bo :8o7 200 O .8022 1 60 .802 230 0. 80I 5 210 .8115 140 .806 105 .8072 60 O 804 150 .8072 20 .8085 50 0. 8058 1 2O o .8075 75 C o .810 300 o 807 40O o .806 340 O. 8048 500 o .811 490 809 200 o .808 220 0. 806 5 385 o .8145 175 8115 60 .8072 320 0. 8075 225 o .8097 150 D o 8i33 450 0. 811 400 .809 400 0. 810 500 8133 530 o. 8097 42O .8135 250 O. 810 660 0. 816 290 o. 811 290 .8115 680 O. 810 650 0. 8142 250 0. 815 60 0.812 650 0. 812 260 Bo. 8172 400 o. 815 260 o .813 40O 0. 8i35 470 o. 818 520 0. 8148 510 818 285 O. 8i55 530 o. 819 405 0. 815 400 8167 700 0. 8i55 700 0. 8i75 370 0. 817 100 o 8167 805 o. 8162 680 (n) During the night oil was drawn entirely through and out at the top of five tubes which were lost. (12) Three other tubes were set up with these but when opened the oil in these was 81, 81, and 90 cm. respectively from the top, so they were discarded. This unevenness between the tubes was probably caused by using earth in some of them which was not entirely cold. (13) Beginning with lot (13) the pump was run con- tinuously day and night. i6 TABL (15) 2 glass, 9 tin. Glass, o, 10. Tin, 32, 31, 23, 23, 24, 26, 9, 16, 14. II. (Continued). (16) 2 glass, 9 tin. Glass, o, 5. Tin, 24, 19, 1 6, 15, 17, 23, 17, 4, 2. d7) i glass, 9 tin. Glass, 14. Tin, 8, 27, 13, 6, 8,0, 0,0, o. 16.75 16.5 17 A o .800 300 0. 800 125 .804 240 805 95 o. 8025 200 .8055 120 0. 8028 80 o .808 5 23 B 0. 802 300 0. 8042 245 o .812 22O 808 90 0. 8065 100 .812 155 8055 1 20 0. 810 no .8127 45 C o. 8065 800 0. 8085 230 .814 500 o .810 350 0. 8085 600 o .8125 450 0.8115 210 .815 40 D o. 8112 600 o. 8117 435 .814 540 ,8142 525 0. 8i35 650 .814 670 o ,8122 680 o. 8i37 370 .8145 350 8155 200 0. 8148 350 B 0. 8145 740 o .817 740 .820 400 see .8172 750 ,818 4OO below ,817 820 B D TABLE II. (Continued). 08) (19) 2 glass, 9 tin 2 glass, 9 tin. 2 glass, 9 tin. 2 glass, 8 tin Glass, 8, 5. Glass, 14, 17. Glass, o, o. Glass, o, 9. Tin, 17, 2, 15, Tin, 30, 42, 39, Tin, 17, 11, 5 , Tin, 13, 12, 19, 12, 22, 21, o, o. 28, 35, 33, 42, 30, 42. 18, 12, 15, 13, 18, 11. 23, 15, 19, ii, 13, 26. 21 5 23 O .8005 425 .806 200 .8005 225 0.804 300 .8015 190 o .8085 1 80 .803 320 8055 290 o .803 320 .8072 175 .8042 200 .809 175 .8055 75 o .807 200 .8078 475 .809 570 .8078 680 .8097 930 .808 500 8135 140 .8085 200 .813 IOO .8095 170 .8105 150 0. 8117 420 0.8122 500 0.8128 560 0.8128 475 O. 8117 680 0.8125 500 0.813 725 0.813 740 0. 813 660 0.814 350 0.815 400 0.8185 1 80 O. 8155 90 0.817 40 0. 8158 1450 .816 930 0. 8158 1000 .816 930 0. 8165 420 .8165 460 O. 8175 350 .8185 400 0. 8187 500 o .8185 520 .820 2OO (18) and (19) Two reservoirs of crude oil were used but the earth from all 22 tubes was worked up together. Grade D from all 22 tubes was first united and then for convenience in working divided into two portions. To study further the fractionation on addition of water, E from lot 1 6 was collected in fourteen fractions. The weight of the earth impregnated with oil before any water had been added was 13.5 pounds, while the weight of the earth containing all the water added, but minus the oil, was 17.5 pounds. The earth was placed in a galvanized iron garbage pail and the water stirred in with an iron paddle. When the first portion of oil was liberated the mass was of i8 the consistency of bran, but as more water was added it turned to a fluid paste. When water was added and the pail inclined, oil would continue to drain for half an hour or longer before the addition of more water became necessary. The oil which was liberated by one lot of water, therefore, could be collected in several portions, and this was done to see whether the oil which comes off immediately after the addition of water is the same as that which drains later. The brackets indicate that the fractions included were liberated by one addition of water. 0.821 25 0.818 0.818 0.8193 70 70 250 0.818 0.818 0.818 0.820 395 350 460 60 0.8208 0.8222 575 55 0.824 0.828 170 16 0.827 95 0.830 45 This occurrence of a. first fraction of higher specific gravity than the ones immediately following occurs regularly if the right amount of water is added to liberate a first fraction of small enough volume. As the volume of the fraction first obtained becomes larger it approaches nearer to the second fraction in specific gravity, and will fall below it if its volume is made too large. The range of specific gravity covered by this first fractiona- tion of the crude petroleum of sp. gr. 0.810 was from 0.800 to 0.830. Oils of the same specific gravity and of the same grade were united and the products chilled and filtered to remove all the dissolved paraffine possible. The oils were 19 chilled and filtered out of doors during the last of December when the thermometer stood at about 4 to 8 C. Lower temperatures as well as throwing paraffine out of solution would cause the whole oil to thicken. The oils were filtered through large plaited filters of drying paper, 24 hours or more often being required for a filter to empty completely. The lighter oils in grades A and B deposited no paraffine. The heavier grades deposited sometimes as much as 10 per cent, of their weight accompanied often by a slight change in specific gravity. When these oils were filtered through earth again they behaved as shown in Table III. 950 cc. were used in each case and the tube divided into five sections. A, as before, is the top 8 cm., B the next 8, C the next 18, D the next 30, and EF the rest. TABLE III. THE SECOND FRACTION ATION. (20) (21) (22) * (23) A 0.8015 A 0.806 B 0.805 B 0.8055 Sp. gr. cc. Sp. gr. cc. Sp. gr. cc. Sp. gr. cc- A . o .8012 36 .8038 45 7997 50 o . 8005 45 B .800 44 8035 48 .802 50 o . 8033 48 C o .8012 68 O 8035 78 -8055 1 08 0.805 115 o .8027 35 o .8052 28 D .8022 170 805 1 60 .8063 175 o . 8063 1 80 EF o .8047 330 .807 320 o .808 260 o . 8085 260 683 679 643 648 (24) (25) (26)* (27)* B 0.8065 B - 8 9 B 0.8105 B 0.812 A 0.8077 38 0.8013 45 0.8075 38 0.8105 42 B 0.807 50 0.805 50 0.8085 50 0.8105 42 C 0.808 80 0.807 75 0.8105 100 0.8085 73 0.810 30 0.810 22 D 0.8092 160 0.8095 180 0.8125 160 0.8115 140 EF 0.8115 300 0.8115 350 0.8135 275 0.8145 250 628 730 5 2 3 569 20 TABUS III. (Continued}. (28) (29) (30) (30* C 0.8095 C 0.810 C 0.811 C 0.811 A .805 52 .8035 40 0, 8005 50 803 50 B .8065 52 o .808 40 o. 809 38 808 55 C . 8085 70 .810 75 ,812 115 .813 105 .811 28 o .8115 30 D .811 160 .8115 140 o 813 175 8135 1 80 o 8135 40 BF .813 350 .813 350 .8145 310 8155 300 712 715 688 690 (32) (33) (34) : * (35)* C 0.8115 C 0.813 C 0.8135 C 0.8135 A .806 45 .8072 20 .803 42 .8025 35 B 807 35 .811 35 ,810 53 o .8077 35 C .810 60 o .812 70 .813 IOO 8135 IOO .812 33 .813 25 D .812 70 .8145 90 .815 150 8145 1 60 813 97 o .815 80 BF .813 103 .8155 200 .817 300 o .817 325 .8135 105 o 8155 125 645 655 813 50 .818 35 813 47 680 Unused, 63 708 (36) (37)* (38)* (39)* C 0.815 C 0.8155 C 0.8155 0.8165 A 805 43 .8053 50 ,808 40 o ,808 50 B .8105 40 .812 45 8095 50 8145 50 C .814 98 .816 103 o. 8i35 IOO 0, .816 82 D 815 155 o .8175 1 60 0. 817 165 0. 8i75 125 BF .817 280 .819 310 0. 819 290 o. 820 310 616 668 645 617 rrv 21 TABLE III. (Continued). (40)* (41) (42) (43) D 0.8135 D 0.814 D 0.814 D 0.814 A .8095 45 o .8045 30 .806 32 .806 25 B .8085 45 .8115 45 .811 45 .8097 30 C .811 95 o .8135 75 .813 92 .814 -50 .8165 28 D .8155 165 o .818 140 .8175 140 .8157 145 BF .817 320 .821 300 o .8195 310 .8175 400 670 618 619 650 (44) (45) (46) (47) D 0.814 D 0.814 D 0.8145 D 0.815 A .8008 45 .800 50 .808 45 .800 42 B .8077 50 .8065 55 o .8115 40 807 47 C .814 103 .8125 100 8135 65 .814 no .8155 30 D o .8175 1 60 .816 160 .817 105 .816 150 .818 58 EF .819 310 .817 300 .8195 1 80 819 330 668 665 .8202 120 679 .821 53 ' 696 (48) (49) (50) (SO D 0.815 D 0.8155 D 0.8155 D 0.8155 A .810 37 .8045 40 .8105 45 . 8058 40 B .805 47 .8n 48 .8148 47 .810 40 C 0.812 105 o .815 98 .810 IOO .8132 60 .8145 50 D .817 1 60 .8185 160 .815 140 8172 75 o .8188 55 o .8188 38 EF .819 300 .820 310 .819 400 .820 38 649 656 732 .819 loo , 0.8208 30 .8208 45 . 8208 30 .8208 95 Unused, 40 736 22 III. (Continued). (52) (53)* (54) (55) D 0.816 D 0.816 D 0.8165 D 0.8165 A .806 38 o .803 43 o .806 47 o . 8095 42 B .8115 42 .8105 30 .811 48 o .8135 40 C o .814 70 o .815 IOO 815 98 o .8145 77 8175 25 D .8185 125 o .8185 175 .8188 150 o .8188 150 BF .821 300 .820 290 .8208 300 .821 295 700 638 643 604 , ,. (56)* (57)* (58)* (59) D 0.8165 D 0.817 D 0.818 D 0.8187 A .806 45 8075 40 ,808 45 .8ll 40 B o .810 45 o .8115 40 o .8135 45 .812 45 C 8145 95 815 IOO .817 105 o .814 92 D .8185 160 o .818 130 ,820 150 .819 150 BF o .821 295 .821 330 .822 300 .823 305 640 640 645 632 (60)* (61) (62) * (63) * D 0.8205 B 0.814 E 0.8163 E 0.817 A 8045 45 .8075 33 o 8i55 42 o .804 45 B o 813 45 o. 810 35 o. 808 50 o. 8075 50 C 8i75 90 o .8125 80 8095 70 8145 102 .812 25 D o .822 170 .818 125 -8i75 105 o .8205 150 o .8182 32 BF .823 270 o .8245 300 823 250 8245 300 Unused, 70 8255 4i 573 647 692 615 23 III. (Continued). (64)* (65)* (66)* (67) E 0.817 E 0.818 E 0.818 E 0.818 A 0. 805 42 0. 8065 90 O. 805 38 0, 805 40 0. 809 20 B 0. 810 42 0. 810 110 0. 811 38 811 45 C 0. 8i45 75 o. 8i55 186 0. 8135 104 o 8145 85 0. 817 50 D o, 820 135 0. 8205 385 o. 819 175 8185 125 0. 8205 75 EF 0. 8255 235 o. 8255 650 0. 8235 240 o 824 325 - 0. 8255 260 529 695 620 1826 3 tubes. (68)* (69) (70) (71) * E 0.8185 E 0.819 E 0.819 E 0.819 A 0. 8205 15 o. 804 24 0. 8115 21 o 8095 23 B o. 8043 35 0. 808 40 0. 814 31 8085 34 C 0. 810 60 o. 8i45 85 0. 815 90 ,814 80 0. 812 30 D 0. 817 160 0. 8i95 140 0. 8165 I2O o .820 1 60 EF o, 8225 300 0. 824 300 0, 824 330 .824 280 590 589 592 577 (72)* (73)* (74) (75) E 0.8195 Eo. 8195 E 0.8195 E 0.8197 A 0. 8i45 30 0. 8055 34 0, ,816 40 .8025 40 B o 8105 42 o. 811 45 8i35 42 o 8105 38 C o, 8135 103 0, 8i55 80 o 809 65 o .816 90 o 814 34 D 0, 8195 160 0. 820 120 8185 160 .822 150 BF 0, 824 285 0. 824 290 o ,824 260 8255 300 ,827 30 620 569 618 631 TABLE III. (Continued). (76)*- (77)* (78) (79)* E 0.820 E 0.8205 E 0.8215 E 0.822 A 0. 8045 48 .806 32 0. 8045 28 .8105 32 B 0.812 40 .8125 45 o. 8135 36 .8145 42 C o. 817 91 .8175 90 o. 8185 78 .819 77 D 0. 822 155 .823 100 o. 823 150 8225 155 EF o. 826 260 .8245 330 o. 8275 300 .827 280 594 597 592 586 t (80) (81) (82) (83)* Eo. 822 Eo .822 E 0.824 A-B 0.804 A o. 817 30 .8083 26 o. 8125 48 o .803 32 B 0. 810 40 .814 40 0. 8127 4 8 8035 25 C 0. 8i53 46 8185 92 o. 818 53 8035 63 0. 8163 42 0. 819 50 D 0. 8225 1 60 .824 140 0. 8245 175 . 804 140 EF 0. 8265 295 o .8265 270 0. 828 200 .806 275 o. 830 90 613 568 left ,50 535 7H (84) * (85)* (86)* (87)* A-B 0.8065 A-B 0.808 B-C-D 0.8125 B-C-D 0.8125 A 0. 8035 40 .8085 30 0. 80 5 32 o . 805 80 B o. 8055 40 .807 30 o. 808 42 8085 73 C o. 809 80 .8065 92 o. 813 85 .812 118 813 75 D 0. 8085 155 .8085 "5 o. 815 175 o .815 285 EF 0. 8n 300 .8115 330 o. 820 280 .8175 520 615 597 614 H5I 2 tubes. J A B C D BF 25 III. (Continued). (88)* (89) * (90) * (91)* C-D-E o. 813 C-D-E o. 813 B-C-D 0.8145 D-Eo. 815 8035 40 o 8035 130 o .8065 130 0. 801 165 ,807 20 .808 27 0.8025 176 o 808 40 8077 1 60 .809 137 o. 8075 206 .8115 20 o. 8085 156 8115 73 o 8125 330 o .813 305 0. 812 330 813 30 0, 813 60 .815 40 0. 812 512 0. 8155 160 815 550 o .815 560 0. 817 340 o 8175 90 .8175 75 o. 8i7 800 0. 8i7 150 633 -""/O '-'O'-' '"*! 3 113 .818 425 0.818 410 0.8215 etc. 2595 4 tubes. 4 tubes. 2479 5568 9 tubes. (92)* (93) (94) D-E 0.8163 F 0.822 Fo.822 A 0. 805 84 0. 8107 35 0. 804 28 o. 8075 20 B 0. 807 100 0. 810 43 0. 808 37 0. 8115 20 C o. 815 205 o. 814 70 0. 8165 63 0. 816 45 o. 816 25 o. 817 30 D 0. 820 350 0. 8215 156 0. 8218 146 0. 820 102 BF 0. 8225 500 0. 8285 250 o. 831 255 0. 8225 480 0. 831 60 1906 639 559 3 tubes. * 950 cc. were needed for each tube and for many tubes this amount was available of the same grade (A, B, C, etc.), and of the same specific gravity. In some cases, though, it was necessary to unite oils of the same grade which differed slightly in specific gravity. Such samples differed in no case by more than 0.0015 and are all marked.* To chill and filter these products of two fractionations would have entailed too much loss. As it was, much uniting of samples which differed but silghtly from one another was necessary to obtain sufficient oil for further fractionation. The unions which were made are given in Table IV. 26 5 D -* H s 5 H j 04 to 1 6 p 5 to M M 00 6 pq CO to M 00 6 o j 3 3 to 04 O to ^ ON to O t-^oo 04 VO ON 04 00 HH eg o_ 00 O O 04 v H M 04 M M * CO 00 s to CO to co to to to to O4 a .oo ^ r^ -*N CO CO CO H 10 J*. t4 co to to OO to 04 !> to CO ."tn 10 to to*d ."ti " " 3 CO 00 to tovo OOO 0*0 ON t>. !>. /-\ /-\ s ^ oo ao oo 00 00 00 w w 00 00 OO eg eg 00 oo oo oo o oo oo oo Q o. j o O OOO Q OOO 6 6 6 H 9 5 f-y^ f^y j 2 pq o< H * fr to to 00 /-\ 2 to to M 04 ro to CO ^ ^ to to'co r^- IO tOvO a *> eg 00 U 00 1 00 00 00 00 oo oo oo oo oo oo oo 4 q 6 6 6 o 6 6 6 o 6 66 6666 -t -8 to M- ON i t^ ON 00 ON ON ON 85 01 CO Tf tO 00 vO t^OO ON 0000 1 oo o 00 6 to vg . oooooo 03 ooo o 10 ooooooooooooooooooooooooooooco 606666660000000 O M co iovO ON O M CM CO 28 950 cc. of each of these were fractionated by earth again, with the results as given in Table V. TABLE V. THK THIRD FR ACTION ATION. (95) (96) (97) (98) A-Do .805 A-EF 0.807 A-C 0.8085 A-D 0.8085 A .8065 33 0. 8045 37 o. 806 40 . 8047 38 B .805 33 0, 806 38 0. 8068 40 O .8052 38 C .804 62 o 8065 65 0. 807 58 . 808 70 .805 40 .8083 25 0. 8095 18 . 8093 30 D .8055 150 ,808 142 o. 8093 154 .8095 132 EF o .808 315 o 8095 250 0. 812 295 811 335 633 557 -605 643 (99) (100) (101) (102) A-Co. 8105 EF 0.8115 C-EFo.8i25 C 0.813 A .8105 33 .8065 30 0. 808 33 o . 806 26 B .810 36 o .809 36 8085 34 8105 33 C .8075 7i .810 60 811 54 .8105 50 .8085 17 .812 20 o 8145 22 o .813 17 D o .811 H5 o .812 136 8145 162 o .813 136 EF .814 300 .815 315 .817 295 8i57 365 572 597 600 627 (103) (104) (105) (106) B-Co. 8135 D-EF 0.814 C 0.814 C-D 0.8145 A .8065 28 .8042 35 o .804 40 o 803 33 B .810 33 o .8115 36 .810 40 .810 40 C 8135 60 .8125 60 .8142 58 o 8i45 54 .8165 18 .8147 28 o .816 25 o .8l6 22 D o 815 150 .815 175 .8163 270 o .8l6 150 EF .817 325 o .819 230 8185 280 .8185 260 614 564 713 559 29 TABLE V. (Continued). (107) (108) (109) (no) D 0.815 EF 0.8155 D 0.8163 D 0.817 A 8035 45 .809 52 810 33 O .8065 35 B .8115 47 o .811 47 .8105 45 O .8125 40 C .815 85 o .815 55 o 8132 55 .817 68 .8177 30 .815 40 8i45 33 .818 30 D .8175 156 .8165 170 .8185 140 8195 H5 EF o .8195 300 .8185 255 .8215 275 o .821 290 663 619 58i 608 (ill) (112) (113) (114) EF 0.817 C-D 0.8175 EF 0.8175 EF 0.8187 A .8043 30 8145 35 8065 35 .817 30 B .8105 32 .811 32 .812 48 .8065 25 C .8152 65 o .8165 60 8i55 75 .8122 30 .816 43 0.818 30 8175 18 .813 15 D .8182 160 o .819 150 .8185 150 .8175 150 EF .8205 290 .822 245 .8225 283 .822 340 620 552 609 590 (115) (116) (117) (118) D 0.819 EF 0.819 D 0.819 EF 0.819 A 8055 16 .8032 30 8045 33 .805 30 B .807 lost .8115 36 -813 38 .813 30 C .816 43 .816 52 .8175 60 .8165 60 .817 30 .820 21 8i75 34 10 D .820 130 .8205 1 60 .8215 150 .821 154 EF .8235 300 8235 240 8235 325 8243 295 lost 539 640 579 (119) (120) (121) (122) EF 0.8195 D 0.820 EFo.82i EF 0.82 15 A .805 23 .805 33 o 8075 33 o .803 4i B .814 35 .811 35 .814 41 .811 43 C o .8165 58 .817 53 .8182 63 .8165 60 .8175 32 8175 35 .81,9 18 .8l8 23 D .8205 152 .822 165 .822 150 .8225 182 EF .823 300 -825 310 8245 273 .828 270 600 633 578 629 (123) 30 TABLE V. (Continued). (124) (125) (126) D-EF 0.822 EF 0.8225 EF 0.8235 D-EF o. 824 A .808 34 0.807 31 0.804 35 0.805 35 B .813 35 o . 8095 27 0.813 40 0.814 35 C .8178 52 0.8175 65 0.8185 70 0.820 60 .8192 42 0.8185 17 0.8195 21 0.821 20 D .8233 155 0.8232 150 0.8252 144 0.8253 170 EF o .8265 260 0.8275 300 0.829 280 0.828 300 578 590 590 62O (127) ' ,_- EF 0.8255 A .8055 45 B o .8155 40 C .821 75 .823 25 D o .8275 170 EF .830 280 635 From the products of the third fractionation, oil in sufficient quantity and of five grades was obtained by uniting frac- tions as given in Table VI., the results of the fourth fractiona- tion being given in Table VII. 10 CM co co 10 O o M vo O IO OO 00 00 00 00 00 d d o to w CO VO 00 O O OO CO ON O M o O 10 ^- >o 10 10 10 ^iOrt- ^00 CO OO 00 CO 00 CO CS CO CO cooo vo t^N, to *** 10 10 Q*\ OO OO 00 CO 00 CO 00 00 odd odd WWW o oooo ^O S OO" / 'ON ooo d M C^ CO 000 CO Tj- IOVO 000 ON 00 d M w* 5 * a>5 M ON 8 = = \o O 10 O OO OO OO a d d d WWW CO OO OO o PQP PPQ O M w WWW W g, 0\ ON O O 4 r 1 M M a w OO OO OO OO dodo W CO CO 00 IO O W W CO WWW 00 00 00 odd o w co >o a\ CO w CO w CO 33 TABLE VII. THE FOURTH FRACTIONATION. (128) (129) (130) (131) C-EF 0.815 D-EF 0.8 168 D-EF 0.819 D-EF 0.8205 A o 8135 19 0.812 30 0.8115 24 0.8095 18 B .815 27 0.8122 42 0.8127 35 0.813 26 C .8118 50 0.8165 55 0-8173 60 0.819 45 -813 15 0.818 25 0.8185 22 0.8195 17 D o .8147 140 0.818 1 60 0.820 160 0.8215 130 BF o 8i75 360 0.8195 305 0.8215 310 0.824 340 611 617 611 576 (132) D-EF 0.823 A o .8092 35 B lost 35 C 8195 60 .8213 25 D 8235 150 BF .826 280 585 To better compare the oils of different specific gravities which were obtained by the process just described five samples of 300 cc. each were separated by distillation into ten fractions. Bach was distilled in the same 500 cc. distil- ling bulb which was heated by an electric stove which en- tirely surrounded the bulb. Bach was first heated to 200 under atmospheric pressure and then to 360 under 50 mm. pressure. The diminished pressure was obtained with a large Chapman water pump and kept constant at 50 mm. by the use of a valve which automatically admitted air to the evacuated system whenever the pressure fell below 50 mm. This valve was constructed from a piece of iron pipe one inch in diameter and five feet long. The lower end was closed with a cap and the pipe filled with mercury to a depth of 76 cm. The upper end of the pipe was closed with a two- hole rubber stopper. In one hole was a long glass tube with the lower end beveled, which reached to the bottom of the mercury and which could be raised or lowered as the 34 barometer varied from day to day. In the other hole of the stopper was a tube which passed just through the stopper and which was connected on the outside with the apparatus to be exhausted. To prevent mercury from being drawn up and over into the apparatus by the air admitted, the end of the tube inside the stopper was drawn out and bent at a right angle and over this was slipped a cap made of larger tubing which was closed at the bottom but which had a fine opening in the side for air. This cap was about six cm. long and extended about three cm. below the end of the tube inside. If any mercury passed through this first fine opening into the cap, it would fall to the bottom without being drawn over into the apparatus or clogging the fine opening in the tube leading thereto. With this valve there was no diffi- culty in keeping a pressure of 50 mm. constant to within one mm. Bach distillate of sufficient volume, which was not tpo viscous or partly solid, was tested as to specific gravity, viscosity, and per cent, absorbed when treated with con- centrated sulphuric" acid (sp. gr. 1.84). Viscosity was measured by taking the time of flow of a measured volume of oil through a capillary, the viscometer used being the one described by Ostwald and Luther as modified by Jones and Veazey. 1 The capacity of the small bulb was 4.5 cc. and the diameter of the capillary such as to require for from five to eight minutes for that amount of oil to flow through it, and one minute 2.6 seconds for the same amount of water. The viscosity as well as specific gravity was always measured at 20. Viscosities have been calculated from the following formula : TS V = *7o ATA in which rj is the coefficient of viscosity for io'wo water, S is the sp. gr. of water, and T the time of flow of water through any given capillary at a given temperature ; rj is the viscosity coefficient of the solution investigated, S is its specific gravity as compared with water as unity at any given temperature, and T is the time of flow of the given solution 1 Z. physik. Chem., 61, 651. 35 at that temperature. The value for pure water at 20 was taken from the work of Thorpe and Rodger. 1 Thirty cc. of each of these distillates, where that much oil was available, all the oil there was where the volume was less than 30 cc., were mixed with an equal volume of con- centrated sulphuric acid (sp. gr. 1.84) and shaken half an hour or longer in a shaking machine. The oil and acid were then poured into a separating funnel and the acid drawn off. The oil was then washed twice with water, once with aqueous NaOH, again with water, and then with this last wash water poured into a burette and allowed to settle. After standing over night the volume was read. - , The oils boiling below 200 (50 mm.) separated clear, but the heavy distillates were milky from water. The volume of these milky oils was read, their specific gravity taken, and then the milkiness was removed by shaking and heating to 60 or so with CaCl 2 . The specific gravity of the clear oil was then taken and the proper correction made to the milky volume. In no case, however, was this correction at all large, and only for the three or four heaviest oils did it ex- ceed one-half of i per cent., the largest correction of all being 2.6 per cent, for the distillate between 23o-26o of the oil of sp. gr. 0.824. An attempt to treat with acid the- oils selected to be distilled, resulted in so much loss from the formation of emulsions that the loss in volume and change in specific gravity could not be determined with any degree of accuracy. 1 Phil. Trans., i8sA, 397 (1894). M O t^* vO IO CM IO O ^t" flj ^J* "^"YS,^ C^ ON t^ ixO CM O -( i^O CO ,-v M ^t-rrt l-l O N M vO 00 O 5 "* O vO t^ co 10 O ^^ t"^* O ^jX l/^ Q^ QQ |4 O^ f^N co !> *-t r^ O ON r^ o o o to o o co o 6 6 ON O O O M o M j^. CM rh to ON r^ Ixp O ^ S O to O \o ^^ o o^ rt- vo o 6 6 M ^ H n O ON ON 10 to 10 oo t^ to vo ^ W IO CO COO-CO Q co oo O t>- O r^ vO f^* CM to O vp O O\ 10 CO CM t> 66 co 6 6 ' 6 6 6 00 O fc H IO to CO ON g gS S?*? M M ^ S" vo r- O O^ TJ- ON ^^00 6 6 4 6 W -L. N W VM rS ^-s. M O O 'OlsP CM PH \^ Q T^- H vO f\j Q i^X f\^ OsK t^ rj- TJ- t~ O N? ON Sj;jt Q 66 6646 i ^ 6 6 CO O $ 1 ; : ^ d -*Q '. o ; 'o 3 TJ : : : rt . ca . oJ M rt> }_i CQ c/J ' QJ fll <1> 3 j -.; ;. (5 ^.g. : . : :U . ^j . -4- 1 !* ! rt hli a t/> ' hi, : : v.- i* . ^ v4 ! V-i On ' *d ^ bjo t/i ^ J^ CJ CO ^ ^ C/2 : ^y* cj t/j *r| ^ CO o CO K^ ^J CO o ' O 0_ III 0? ;, iSi sP ON CO SNO 10 Tj- lOOO x$> O ~~ cN ON O t^ M CX ex oo O r^oo c* co c< ex oo M O^ CO vO 00 r 13 'B 6 6 w 6 6 6 oo 6 6 6 ci 6 M ^4 iO\O to 10 *ON ^2 M O* n O >s^ QO 00 to ><^~^O co O 1-1 ^ t^oo N ON 00 t^ c< COCO O !Ooo ex oo IH vo oo JC o 6 4- o 6 666 M 6 6 I*** O a _ 10 M r>. oo JH >\O iO ex ^O >>5^ *^ CO ^0 Kp CX CX ^ O O^- ON to ex ex oo O ^00 8 CO - 00 M s - CO ON 00 ex" 3 6 6 ON O 6 O 00 O 6 6 00 6 5 M O CO vo .,,., ^ O ON W rO 00 !" O 0^ ON o^ r^** 00 s?s O O 00 O HI eo co HI ex oo HI >^o oo o\ t^ O ?g CO O M CO 00 ONKp CX 100\ M O eooo ex oo M kx * to oo 00 CX CO ex oo H-J 6 6 4 6 6 64-6 6 6 00 O 6 k*! t 1 i r ex ^o iOv> to 10 o til >^^ O CO iO CO to O xP to ON ex w \.^s vO t*"* O ^R ex oo O oo CO CX M 04 00 M 6^ w O co iO ex oo (X H co O q 6 _ 6 6 6 co O 6 6 1 rj-j *& d . -t-> '. 4-> ' +j '- : a :* bo i ^^ ; bio TJ & & : v-i '- : Bto> ^to 8co %"& d 8> o d ^S o ( 3 O CO v2 8, ex ex d CO i i | | o 1 ( 3 O Q _t Q CO vO M es ex es 3 1 o 10 CO oo ON 04 CS 00 O 04 o O ;g 39 "Fluid" means that the oil at 20 was partly solid but would flow when the bottle was inclined; "solid," that the bottle could be turned upside down without the oil changing shape. It was hoped that sulphuric acid of the strength used (sp. gr. 1.84) would dissolve only unsaturated hydrocarbons and leave untouched the paraffines and benzene. By long- continued shaking at ordinary temperature, however, with acid of this strength, benzene is dissolved, provided that the acid is in large excess. One hundred cc. of benzene were completely dissolved in 434 cc. of acid on being shaken four hours. Three of the distillates which had been shaken with acid, however, gave no action when treated with a mixture of equal parts concentrated sulphuric acid and fuming nitric acid, while distillates which had not been previously shaken with sulphuric acid were acted upon by this nitrating mixture. The action of the sulphuric acid, therefore, appears to have been complete. The action of sulphuric acid shows that over 90 per cent, of the oil dealt with consists of paraffine hydrocarbons, and that in the filtration through earth the paraffine hydro- carbons tend to collect at the top of the tube and the un- saturated hydrocarbons at the bottom. The increasing amount dissolved by sulphuric acid in the heavier oils may be seen in the following curves: The abscissas represent temperatures and the ordinates volumes. The same distance upon the X axis is taken to represent a distillate, whatever be the number of degrees over which it may have been collected. The upper curve represents the per cent, of the total volume which distilled between given temperatures; the lower curve, the per cent, of the total volume recovered which was not absorbed by sulphuric acid (i. e., the paraffine hydrocarbons). For the upper curve the ordinates were obtained by dividing the number of cubic centimeters in the distillate by the total volume of oil re- covered. For the lower curve the ordinates were obtained by dividing the number of cubic centimeters in the distillate Temp. o 150 200 140 200 230 260 Normal pressure. 50 mm. pressure. Temp. o 150 200 Normal pressure. mm. pressure. Temp. o 150 200 140 200 230 260 Normal pressure. 50 mm. pressure. 43 not absorbed by sulphuric acid, by the total volume of oil recovered. The area between the two curves represents the proportion of hydrocarbons soluble in sulphuric acid, which it will be seen is greatest for the oils of highest specific gravity. On referring to Tables III. and V. it will be noticed that there are several tubes given where the specific gravity of the oil of grade A is heavier than grade B and sometimes than grade C. Tubes where this irregularity is marked are 4$, 62, 68, 72, 74, and 80 of Table III., and 112 and 114 of Table V. A slight irregularity appears in 20, 21, 24, 27, 40, 71, and 85 of Table III. and in 99 of Table V. If the oils in these cases are not colorless, the color is strongest where the specific gravity is greatest, so that although oil of the grade A has passed through the most earth it is yet more strongly colored than oil of grade B or C. No reason for this variation has been established. It should be remembered, however, that the different oils rise in the earth with differing velocities, not because they differ from one another in specific gravity but because they differ in surface tension. A rough attempt was made to measure relative surface tensions by measuring the height to which different oils rise in the same capillary tube, but although a kathotometer was used and the level of the oil in the capillary brought to the same spot each time, the work sufficed only to show that the difference between the surface tension of the oils obtained was so slight as to require very careful measure- ment for the results to be of any value. That viscosity shows the same irregularity in these oils as color and specific gravity, appears from the following measurements. (62) (68) Sp. gr. Vis. Sp. gr. Vis. A 0.8155 0-0539 0.8205 0.0626 B 0.808 0.0469 0.8043 0.0469 C 0.8095 0.0509 0.810 0.0520 0.812 0.0555 0.812 0.0554 D 0.8175 0.0525 0.817 0.0524 0.8182 0.0535 BF 0.823 0.0612 0.8225 0.0606 44 In (50) of Table III. an irregularity appears in grade B which is also found in the viscosity. (5o) Sp. gr. Vis. A 0.8105 0.0532 B 0.8148 0.0559 C 0.810 0.0526 D 0.815 0.0526 EF '..... 0.819 0.0552 The oils obtained by one fractionation leum have the following viscosities, the being those previously given in Table I. of the crude petro- fractions measured (I) (2) (3) Sp. gr. Vis. Sp. gr. Vis. Sp. gr. Vis. A 0. 796 o 0376 0.8012 .0408 0. 8022 .0401 B 808 0529 0.804 0485 803 .0470 C o 8125 0501 0.807 0443 8075 o 0453 o 8137 .0529 0.809 .0476 810 .0471 D 815 .0504 0.8125 .0460 ,812 .0472 E o ,818 0521 0.8l85 0537 O 8175 .0529 F o ,8205 . . 0.823 . . o .821 That the viscosity does not increase with the specific gravity, particularly with the higher fractions, is apparent in two of the three series just given. The same is also shown in the following four tubes taken from Table III. A B C D EF (21) Sp. gr. 0.8038 o 0.8035 o 0.8035 o 0.8052 o o . 805 o 0.807 o Vis. 0465 0456 0456 0485 0479 .0480 (22) Sp. gr. 0.7997 0.802 0.8055 0.8063 0.808 Vis. o . 042 i 0.0485 0.0502 o . 0496 0.0510 (47) A B C D EF Sp. gr. O.SOO 0.807 0.814 0.816 0.819 Vis. 0.0453 0.0538 0.0542 0.0528 0.0556 (53) Sp. gr. 0.803 0.8105 0.815 o 0.8l85 0.820 Vis. 0.0515 0.0563 0684 0.0570 0-0559 45 This drop in viscosity in the oils which occurs at the bottom of the tube, appears to be a regular occurrence in the dozen or so oils which have been tested. Further investigation of this point is intended. WATER FRACTIONATION. To test the effectiveness of water fractionation alone, 1000 cc. of crude petroleum, previously chilled and filtered, of specific gravity 0.807, were mixed with 1000 gms. of earth and allowed to stand 24 hours. Water was then added in small amounts and the oil collected. Sp. gr. Volume of oil. Total water present, cc. cc. A 0.8148 44 500 B 0.8139 278 650 C 0.816 211 800 D 0.820 84 950 E 0.8225 28 1400 F 0.8245 28 2750 673 The fractions of large enough volume were then mixed with earth again and the oil replaced with water. One gram of earth was used for each cc. of oil, the earth having been heated first and allowed to cool. B. o 8139 278 cc. stood 1.5 hours. C. 0.816 211 cc. stood 6 hours. D. 0.820 84 cc. stood 2.5 hours. Sp. gr. Oil. Water. Sp. gr. Oil. Water. Sp. gr. Oil. Water. cc. cc. cc. 0. 8185 IO 70 O .820 10 80 0.822 32 7 6 O. 818 10 1 10 .820 20 125 0.823 20 207 0. 818 21 164 .8195 72 250 0. 818 20 .820 30 4IO 52 0. 817 42 . . O .820 10 588 0. 819 10 216 0. 820 44 277 142 820 16 428 8215 20 686 193 4 6 It is apparent that while petroleum is fractionated by simply mixing the oil with fuller's earth and then displacing the oil from the earth with water, the fractionation is much less complete than when tubes are used as previously de- scribed. It will be noticed that although fractions C and D in the table last given are separated hardly at all by further treat- ment with earth and water, yet the specific gravity of all the oil recovered is higher than that of the oil used, e. g., from C of sp. gr. 0.816 is obtained nothing lighter than 0.8 195, and from D of sp. gr. 0.820 is obtained nothing lighter than 0.822. To determine whether the specific gravity of the oil re- covered will continue to rise after the oil is fractionated no further by repeated treatment, 330 cc. of sp. gr. .819, ob- tained by uniting several products of one fractionation of the crude petroleum, were mixed with 330 grams of earth and water was added. Sp. gr. Vol. oil. Total water present. ' cc. A 6 64 B 0.8215 50 C . . 12 214 D 0.821 60' 270 E 0.821 82 413 F 0.8225 26 613 236 Seventy-five cc. of E of sp. gr. 0.821 were next mixed with 75 grams of earth and 150 cc. of water were added. Fifty- one cc. of oil whose specific gravity was unchanged, but whose color was reduced, were obtained. Fifty cc. of this when treated with earth and water returned 34 cc. of oil with the color considerably lighter, but the specific gravity still 0.821. Although only two-thirds of the oil used are recovered whenever oil is mixed with earth and then displaced with water, yet this loss does not seem to affect the specific gravity of the oil obtained for longer than one or two treatments after the oil ceases to be fractionated. After this the oil recovered has the same specific gravity as the -oil used. 47 THE OIL LOST IN THE EARTH. The sum of the fractions of oil displaced from the earth is usually about two-thirds of the volume of the oil used. A pressure of approximately 200 tons per square inch upon the earth from which water has displaced all oil that it will, results in the liberation of considerable water but very little oil. When earth which has been pressed is heated to 165 for three hours, considerable water distills over but much less oil than would be expected, e. g., from 75 grams of earth which should contain 25 cc. of oil, but 4 cc. of oil were ob- tained. The earth was removed once from the flask and pulverized, and when the heat was discontinued the, earth was thoroughly dry. On extraction with ether the earth gave a solution having the color of the original petroleum. The extraction was made with a Soxhlet extractor and con- tinued until the extract was colorless. On evaporation of the ether there remained about 8 cc. of a heavy oil with the color of the natural petroleum. Pressure, heat, and ex- traction with ether together gave about half the amount of oil which the earth must have contained. Earth which had been used once was allowed to dry for several weeks at room temperatures until it had lost all appearance of containing moisture. It was then pulverized, sifted, and used in a tube with the crude petroleum of sp. gr. 0.810 with the following results. 8 cm. at top . . o Next 8 cm 0.8284 10 Next 18 cm 0.8225 45 Next 30 cm 0.8143 60 0.8155 80 Rest 0.8175 83 0.819 114 392 cc. Earth used, 720 gms. Crude petroleum used, 740 cc. 4 8 The first oil up the tube evidently is absorbed by heavy material in the earth, while the first oil recovered dissolves material from the earth which increases its specific gravity beyond that of the next fraction. To see how much of the weight of the earth just used was due to material which it had retained from its first use, 300 grams of earth were mixed with 300 cc. of crude petroleum and the oil displaced by water. The oil recovered measured 205 cc., and the weight of the earth after drying for several weeks at room temperature was 347.5 grams. Fully 15 per cent., therefore, of the weight of the earth used in the tube just mentioned was solid matter which it had retained from its first use. In all cases the earth was heated before it was used be- cause it was believed that heating decreased the amount of oil lost in the earth. The earth was heated usually in iron pans on a gas stove until it ceased to form geysers when stirred. A tube packed with earth which had not been heated gaye results as follows with crude petroleum of sp. gr. 0.810. Sp. gr. Vol. cc. Top 8 cm o . 803 30 Next 8 cm o . 8045 38 Next 18 cm 0.8103 &5 Rest . . 440 593 Crude oil used 930 Earth used, 948 gms. Tube 5 feet long, i */4 inches diameter. 20 hours required at diminished pressure. In a case of water fractionation alone with unheated earth but 242 cc. of oil were recovered from 500 cc. of crude petroleum. Results obtained toward the close of our work indicate that the loss of oil when unheated earth is used is much less than we had supposed it to be. The gain from heating the earth may not pay for the trouble of heating it, and this point should be investigated before any very extensive in- vestigation is again undertaken. 49 Earth after heating must become thoroughly cold before it is used to pack tubes. The earth holds its heat for several hours, and if it is used the same day upon which it is heated, there is apt to be contraction in a tube so packed sufficient to allow the oil to run up the side of the tube as it would in a vacuum. The length of the tubes used was five and one-half feet. A tube nine feet long was held for two days with a constant diminished pressure of about 10 cm. Hg, and connected to the same vacuum pump with several five and one-half foot tubes. The oil was drawn to the top of the latter, and these removed and a second lot substituted which were likewise fully impregnated with oil before the long tube was opened. When it was opened the oil had ascended but 45 cm., showing that the diminished pressure had not penetrated that length of earth and reached the bottom of the tube. A shorter tube in which the earth was packed very much harder, so that the tube filled with earth rang like an iron rod when pounded upon the floor, when connected with a vacuum pump at one end and a manometer at the other, showed diminished pressure on the manometer when the column of earth was two feet long but not when two feet, four inches. THE; FRACTIONATING POWER OF SUBSTANCES OTHER THAN FULLER'S EARTH. A clay from Topsham, Maine, was found which in tubes showed a power of fractionating as well as decolorizing the higher fractions. Compared with fuller's earth, the action was: Sp. gr. Clay. Fuller's earth. 8 cm. at top 0.799 0.793 Next 8 cm o . 804 o . 800 Next 8cm 0.810 0.806 Next 10 cm. 0.810 0.807 Next 30 cm 0.812 0.8092 Next 45 cm 0.812 0.8112 Time required 69 hours, 76 hours. Sp. gr. petroleum used, 0.806. Tubes 5 feet long, iV 4 inches in diameter. 50 Neither powdered brick made from the same clay nor powdered feldspar showed any power of fractionation. Another similar clay (from Mere Point, Brunswick, Maine) showed a power of water fractionation, but its behavior in a tube was not tested. 400 grams of this clay, previously sifted and heated, were mixed with 170 cc. of crude petro- leum of sp. gr. 0.806 and allowed to stand 14 hours. Water was then added and the following fractions obtained: Sp. gr. Vol. in cc. Total water present. 0.8l65 24 104 0.817 60 133 0.8188 20 234 6 374 no The color was scarcely changed at all. SUMMARY. (1) When petroleum is allowed to rise in a tube packed with fuller's earth, there is a decided fractionation of the oil, the fraction at the top of the tube being of lower specific gravity than that at the bottom. (2) When water is added to fuller's earth which contains petroleum,- the oil which is displaced first differs in specific gravity from that which is displaced afterwards when more water is added. (3) When petroleum is allowed to rise in a tube packed with fuller's earth, the paraffine hydrocarbons tend to collect in the lightest fraction at the top of the tube and the unsaturated hydrocarbons at the bottom. (4) Whenever oil is mixed with fuller's earth and then displaced with water, about one-third of the oil remains in the earth. BIOGRAPHY. Marshall Perley Cram was born in Brunswick, Maine, on January i, 1882. He was prepared for college at the Bruns- wick High School, and entered Bowdoin College in 1900, from which he was graduated in 1904 with the degree of A.B. He was assistant in chemistry in the same college for the year 1904-5 and received the degree of A.M. in 1905. Since October, 1905, he has been a graduate student in chemistry in the Johns Hopkins University, his subordinate subjects being physical chemistry and geology. OF THE UNIVERSITY OF THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW AN INITIAL FINE OF 25 CENTS WILL BE ASSESSED FOR FAILURE TO RETURN THIS BOOK ON THE DATE DUE. THE PENALTY WILL INCREASE TO SO CENTS ON THE FOURTH DAY AND TO $1.OO ON THE SEVENTH DAY OVERDUE. 27 1 MO LD 21-100m-7,'89(402s)