NRLF 5bD LIBRARY UNIVERSITY OF CALIFORNIA, RECEIVED BY EXCHANGE Class The Drop Weights of Twenty Non-associated Liquids and the Molecular Weights Calculated for them DISSERTATION SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE FACULTY OF PURE SCIENCE IN COLUMRIA UNIVERSITY IN THE CITY OF NEW YORK BY GARABED K. DAGHLIAN New York City 1911 GOTCHNAG PUBLISHING CO NEW YORE. N. Y. 1911 The Drop Weights of Twenty Non-associated Liquids and the Molecular Weights Calculated for them DISSERTATION SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FO1 THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE FACULTY OF PURE SCIENCE IN COLUMBIA UNIVERSITY IN THE CITY OF NEW YORK BY GABABED K. DAGHLIAN New York City 1911 GOTCBNAG PUBLISHING CO NEW YORE. N. Y. 1911 ACKNOWLEDGMENT Professor J. Livingston R. Morgan suggested and directed this work. The author begs to tender to Professor Morgan his sincere thanks for the assistance, advice and encouragement accorded him during the work. G. K. D. 226916 CONTENTS Page Object of the Investigation 5 Apparatus and Method 5 Standardization of the Apparatus 6 Twenty New Liquids 9 Discussion of Results 17 Conclusions . 19 OBJECT OF THE INVESTIGATION The object of this work has been to apply the principles arrived at by former workers in drop weights to further non-associated liquids. They proved, first, that the drop weight of a liquid, measured under proper conditions, is proportional to its surface tension ; and consequently can be substituted in place of surface tension in Eotvos' formula as modified by Ramsay and Shields. where y stands for surface tension as measured by capillary rise, M for molec- ular weight, d for density at temperature, t of the liquid, and t c the critical tem- perature. Second, they proved that this method is a great deal more accurate than surface tension method. APPARATUS AND METHOD The apparatus used has been devised and improved into its present form by Professor Morgan. For a detailed description see Jour. Am. Chem. Society, March, 1911. Its essential part consists of a U shaped tube with a capillary bore, one end of which is carefully ground to a diameter of 5.8 mm., and terminates with a smooth circular surface perpendicular to the axis of the tube, and the bore of 0.2 mm. being in the centre. The other leg of this syphon tube dips in a small vessel which contains the liquid worked with, so that by applying gentle suction at the other end the liquid rises in the tube and is syphoned over into another small weighing vessel, which is fitted air-tight at the tip, forming drops. The apparatus is so designed that these two small vessels can be removed and replaced readily; so allows work at different temperatures; and is so arranged that it can be in a water bath at the desired temperature. These baths are heated with the vapor of boiling ether or chloroform or can be cooled by allowing tap water to run through them. The liquid at all times is under perfect control, being regulated by means of a rubber bulb which is connected by capillary rubber tubing to the weighing vessel through the ventilation tubes. The U shaped siphon tube which is called the "tip" is cleaned before starting a new liquid by running through it potassium dichromate solution and sulphuric 6 acid, water, alcohol, ether, and dry air. The supply vessel is then half rilled with the liquid and a drop allowed to form and hang for five minutes, after the constant temperature is attained, to saturate the vessel with the vapor, and then a definite numbe* of drops (15, 25, 30) is run into the weighing vessel. This with its con- tents is then weighed after the apparatus is removed from the bath. A number of such determinations are made. Then a number of determinations, "blanks," are made with only 5 drops ; but the sixth drop is allowed to hang without falling long enough to make the time of this determination equal to that of the other with more drops. In this way any evaporation through the ventilation tubes, or con- densations, or evaporation from the drops before falling, or evaporation back to the hanging drop, etc., are made equal in the two sets of determinations. By sub- tracting the average weight of one set from that of the other we get the weight of number of drops equal to the difference of those taken in the two. And that gives us the weight of single drop at the temperature of observation. Sometimes, when the liquid has a very high boiling point, at low temperature, it is only necessary to use the weight of the empty vessel as blank. Weight of a single drop is given in milligrams and is indicated by w. w \:\ is called the molecular drop function and we will indicate it by/(M). L a J STANDARDIZATION OF THE APPARATUS We mean by this the testing to see if a tip of this size, 5.8 mm., will give uniform results with the test liquids as tips of different sizes with which work has been done before ; and at the same time to determine the constant of this particular tip. The test liquids are aniline, benzene, pyridene and quinoline. A number of determinations have been made with each one of these at two temperatures. Weight of single drop has been obtained, and k has been calculated for each liquid at both temperatures from equation in which w and t have been measured and observed, and d and / have been taken from former determinations. Table I gives the equations for densities, the critical temperatures (by Higgins), and the molecular weights of the test liquids as has been used in the following calculations ; TABLE I / Mol. Wt. Aniline d t = 1.038797 0.0008605 / 425.8 93.00 Benzene d t = 0.900214 0.0010659 / 288.4 78.00 Pyridene d t = 1.001500 0.0010018 / 347.0 79.00 Quinoline d t 1.109894 0.0008034 / 520.4 129.00 Carbon tetrachloride is one of the usual test liquids. But on account of the great size of this tip and the very smartl volume of the drop that liquid was omitted. It is clear that whenever a drop falls abruptly and not by its own weight, the drop weight is too large. Carbon tetrachloride has a very small surface tension and a large density ; for this reason the drop falls abruptly and so weighs heavier. The tip being very large, too much liquid in weight is contained in the drop volume for the surface tension of the liquid to be able to hold without breaking. As soon as this breaking takes place before the maturity of the drop, the falling part carries with it apparently part of the liquid which would remain on the surface of the tip as remaining-drop if it had not been snatched unduly. On tips still larger carbon tetrachloride gives proportionately larger values. But on smaller tips it is possible to control the drop as the weight of liquid contained in the drop is not too large for the tension of the bounding surface to break it prematurely. The experimental values for the test liquids are given in Table II. TABLE II Aniline Benzene Pyridine Quinoline Temp. 34.30 34.20 34.35 34.30 Wt. vessel -1- liquid 11.0954(10)1 11.0960 11.0960 11.0950 J 36.05 36.05 36.35 11.0871(15)^) 11.0868 11.0868 J 59.575 11.0472 1 59.525 11.0474 59.525 11.0467 J 23.025 22.750 11.2548(15)' 11.2550 23.125 11.2544 22.825 11.2553 58.325 58.425 10.6389(25)-| 10.6384 j- 58.400 10.6385 J 22.375 22,275 10.4846(15) ) 10.484tf ) Average wt. Wt. vessel 4- 5 drops Average Average wt. temp. lin , 71 10.7883) n - 471 10.7882 f 11 2549 10 ' 8484 i U ' 10.8482 J Drop wt. 11.0956 10.6440(0) 10.6440 34.28 45.16 10.7955 36.15 29.14 10.7883 59.64 25.88 10.8483 22.93 40.66 9.9343 58.39 35.215 10.6386 9.9346 9.9340 80.4846 9.7525(0) 9.7525 22.33 48.81 M f w d /(M) k 93 34.3 45.16 1.0093 921.20 2.3898 78 36.15 29.14 0.8617 587.47 2.3857 59.54 25.88 0.8368 532.06 2.3874 79 22.93 40.66 0.9875 759.31 2.3872 58.39 35.215 0.9430 674.23 2.3858 129 22.33 48.725 1.09195 1175.17 2.3882 8 In Table III are given the results of calculation from these experimental values. TABLE III Aniline Benzene Pyridene Quinoline The meaning of k for any one liquid is this: If we work with one of the liquids at different temperatures, and calculate molecular drop function for each temperature, when we plot a curve taking these for ordinates and for abscissas t c -/-6, then we have a straight line for our curve which makes an angle equal to arc tan k with the axis of x. Moreover if we proceed likewise with other liquids, we get a straight line for each liquid equally inclined to the axis of x t only they intercept the axis of x at different distances from the origin. When /(M) plotted in this way does not give a straight line for the curve, that liquid is called to be associated* The average k is called the constant of the tip. For reason to be mentioned later we will adopt as the constant of this tip k value for benzene which is 2.3866. In working with these liquids special care had to be exercised with aniline to have it freshly distilled, and also to see that the capillary bore of the tip is washed with ether and dried, by passing dry air through it for 10-15 minutes, between determinations. Otherwise the successive determinations give larger and larger values. When pure, aniline is almost colorless, and standing for even a day causes it to color a little. In all cases the drops were formed fast, then checked before falling, and allowed finally to fall slowly and of their own weight. Causing a drop to fall rapidly always gives a heavier drop, for it does not fall of its own weight alone, but is. forced out and takes with it more liquid than it should. *See also page (12) 9 TWENTY NEW LIQUIDS The investigation of twenty new liquids by drop weight method forms the main part of this work. These liquids were of greatest attainable purity. Brom- benzene and iodobenzene were specially prepared in the laboratory; diphenyl methane was from Eimer and Amend, redistilled and recrystallized just before the determination was made; isobutylacetate, m-xylene, o-xylene, p-xylene, one sample of mesitylene, bromine and phosphorus trichloride were from Kahlbaum, usually redistilled just before using; while the others, including one sample of mesitylene were especially prepared for this work by the Hoffman and Kropff Chemical Co. (619 Kent Avenue, Brooklyn, N. Y.), and used directly. Experimental results are given below in Table IV. All the individual results are not given, only those where the determination and the blank were at the same temperature being included in the tables. But the mean of all determinations is within a very small error equal to the values given. Brombenzene Temp. Wt. Vessel 4- liquid TABLE IV Wt Vessel Average wt. +5 drops 37.90 11.5863(25) ) 38.40 11.5870 C 11.5867 31.55 11.5867 ) 59.625 11.5210 59.700 11.5207 ..52085 Bromine 11.4376(15)) 11.4380 C 114385 ) 11.43803 Carbon disulphide 20.15 11.1733(15)^ 20.35 11.1718 C 11.17337 19.90 11.1750 ) Cymene 18.95 18.75 18.35 59.45 59.45 11.0995(15)) 11.1015 C 11.1009 11.1018 ) 11.0406 1L0408 10.8345 10.8352 10.8252 10.8247 10.9141 10.9146 10.8190 10.8190 10.7963 10.7968 10.7972 10.7774 Average Average wt. temp. Drop wt. 10.83485 38.3 37.593 10.82495 59.67 34.795 10.91435 0.00 52.368 10.81900 20.13 35.437 10.79677 18.7 30.413 10,7774 59.45 26.33 10 TABLE IV (Cont) Dimethyl aniline Wt. Vessel Temp. -f liquid 21.45 11.2381(15) 21.55 11.2379 59.50 11.1684 59.55 11.1684 59.45 11.1689 59.50 11.1679 Diphenyl methane 59.05 11.2108(15) 59.00 11.2113 58.875 11.2109 Ethyl benzene Average wt. 11.2380 i- 11.1684 t 11.2110 Wt. Vessel -f- 5 drops 10.8419 | 10.8418 | 10.8196 ) 10.8192 10.8194 ) 10.6440(0) 20.90 11 1154^15^ "V 21.10 11.1155 f 11 .11545 10. 8022 59.40 59.60 11.0528 11.0519 } 11 .0523 10. 7846 Ethyl aniline 35.80 35.80 35.80 11.2152(15) 11.2150 11.2154 ! 11 .2152 10.8355 10.8355 10.8352 1 58.70 11.1760 ^ 10. 8224 ^ L 1 760 y 58.70 11.1760 > -I- 1 \J \J 10. 8224 > Ethylene chloride 34.20 34.10 11.1483(15) 11.1484 I 11 .14843 10. 8166 1 34.07 11.1486 \ 10. 8167 * 60.00 11.4258(26) 11.4258 10.8404(6) Ethylidene chloride 34.125 34.100 11.0427(15) 11.0422 i 11 .04263 10. 7822 1 34.050 11.0430 i 10. 7822 ' 56.90 11.2097(20) 11 .2097 10. 8683(10) Flourbenzene 9.30 12.8962(20) ) 9.30 12.8965 12 .8962 12. 2747 i (0) 9.30 12.8960 ) 34.50 12.9630(25) ^ 12. 4122 > 34.50 12.9631 j 12 .96316 12.4126 34.50 12.9634 12. 4125 ) Average wt. 10.84185 10.8194 10.8022 10.7846 10.8354 10.8225 10.81665 10.8404 10.7822 10.8683 12.2747 Average temp. 21.5 59.5 10.6440 59.0 21.0 59.5 35.8 58.7 34.1 60.0 34.10 56.9 9.3 12.4124 34,5 Drop wt. 39.15 34.9 37.8 31.325 26.776 37.98 35.35 33.178 29.27 26.043 22.76 31.077 27.537 11 TABLE IV (Con/.) lodobenzene Wt. Vessel Wt. Vessel Average Temp. 4" liquid 23.5 11.9297(50) 59.6 10.9270(30) Isobutyl acetate 23.8 10.5152(30) 59.375 10.2991(25) 59.275 10.4075(30) Mesitylene 23.55 11.0844(15) 23.50 11.0844 57.20 11-0452 57.40 11.0440 Average wt. 11.9297 10.9270 10.5152 i 11.0844 11.0446 + 5 drops 10.1876(10) 10.1441(10) 10.0082(10) 9.8680 (5) 9.9766(10) 10.7812 | 10.7812 ) 10.7787 ) 10.7785 > Average wt. 10.1876 10.1441 10.0082 10.7812 10.7786 temp. 23.5 59.6 23.8 59.3 23.5 57.3 Drop. wt. 43.55 39.085 25.35 ; 21.55 30.32 26.60 Metaxylene 29.90 10.2095(15) 10. 2095 9.9057 9 .9057 24 .9 30.38 58.9 10.1529 10. 1529 9.8888 9.8888 58.9 26.41 Methyl aniline 24.65 24.55 10.8230(25) 10.8232 f 10. 8231 9.9660 9.9659 i 9 .9659 ' 24 .6 42.86 59.7 59.7 11.2219 11.2217 i 11. 2218 10.8378 10.8382 t 10 .8380 59 .7 38.38 O* thoxvlene 23.85 23.85 10.2345(15) 10.2346 i 10.23455 9.9144 9.9144 t 9 .9144 23 .85 32.05 59.30 10.1756 j 10. 1753 9.8965 9 .8965 59 .4 37.88 59.50 10.1750 > Paraxylene 37.9 11.3678(25) 11.3678 10.7921 10 .7921 37 .9 30.785 59.2 59.195 11.3079 11.3083 } 11.3081 10.7820 10.7819 ! 10 .78195 59.2 26.308 Phosphorus trichloride 35.57 11.5517(30) \ 35.60 11.5515 V 11.5514 35.62 11.5510 ) 10.9567(10) 10.9567 1.9567 35.6 29.73 Toluene 36.05 36,00 11.0809(15)^ 11.0808 ] f '" !S!e ! 10.7916 36.0 28.925 59.1 59.3 11.0425 >> ] 11.0421 J f ESS ! 10.78144 59.2 26.075 12 It has been shown by Morgan that the value k ' B found -from the equation for any one tip, with benzene (where 288.5 is the observed critical temperature), can be used as the standard constant of the tip worked with. And that from this B using it in the equation /(M) = B (t c -(-e); for any other liquid, M is shown to be the normal molecular weight of that liquid when the calculated value of t c from w, M, and d for the new liquid at different temperatures, t, is the same; for the constancy of t c independent of temperature shows that B is also the correct temperature coefficient of the molecular drop function of that liquid, and hence it is non-associated, just as the standard liquid benzene is non-associated. Further, it is plain that surface tension in dynes per centimeter of any liquid can be calculated directly from drop weight in milligrams at the same temperature by aid of the proportion 7 : w = ' B : B where H is the corresponding value for the surface tension as found from and so by using these two values ' B and B comparisons at different tempera- tures can be made of drop weight and surface tension. We can look at this also from the graphical side. It was pointed out that graphical meaning of k is that it is the slope of a certain curve, and that this slope changes in value according to the size of the tip. If we strike a tip of just proper size so that its k is the same as the k obtained from surface tension (which is about 2.12) then drop weights of a liquid from that tip in milligrams is equal to the surface tension of that liquid in dynes per centimeter at the same temperature. It must be remembered here that drop weight in essence is a more directly experimental value than surface tension by capillary rise, for the former is simply a singly determined weight, while the latter is equal to^2 r h (d-d') where r stands for the radius of a small capillary bore, assumed to be constant, h for the height of the column of liquid rising into the capillary bore, which is burdened with an error of 6-9% due to the volume of the meniscus, and d-d' for the difference of the density of the liquid and density as vapor at that temperature. This, perhaps, will serve to account for the variable results below of surface tension as obtained from capillary rise. Below in Table V are given the values of the function /(M) and \ c as cal- culated from drop weights, together with the same values, as far as they can be found in the literature, calculated from capillary rise. Each liquid is taken as a subdivision of Table V, and the molecular weight and the formula for the density as used in the following calculations are given; also the observed t c . Of the columns the first is the temperature, and the others are in order drop-weight or surface-tension, density, molecular function, and t c calculated. 13 TABLE V 1. Brombenze, 38.3 M=156.96, 37.593 4=1.52030.001282, . , 1.4713 845.6 ^f 59.67 34.794 1.4438 792.557 397.8 2. Bromine, 0.0 52.368 ,4=3.18718, /=:302.2 3.1872 ' 712.17 304.4 10.6 Ramsay and Aston, B =2.12112 40.27 3.152 552.08 276.9 46.0 34.68 3.031 487.98 282.1 78.1 29.51 2.917 426.09 285.0 3. Carbon disulphide, M=76, 20.13 35.440 ,14, 4=1.29215 0.0013025/, /=275(*) 1.26594 544.05 254.1 4. Cymene, 18.69 M=134.112, 4=0.8620.0008044 Gf-11.9), 30.414 0.85554 883.52 394.9 59.45 26.330 0.82376 785.20 394.5 Renard and Guye, p =2.1108 11.9 27.98 0.862 809.38 401.3 31.7 26.19 0.846 767.13 401.1 54.8 23.95 0.828 711.65 397.95 74.5 22.16 0.812 667.08 396.5 91.8 20.60 0.798 627.35 395.0 108.9 19.18 0.784 591.04 394.9 117.0 18.45 0.778 571.57 393.7 134.9 16.97 0.762 532.96 393.4 146.5 15.94 0.751 505.48 392.0 163.4 14.60 0.735 469.68 391.9 172.8 13.92 0.726 451.50 392.7 5. Dimethyl aniline, M=121 21.5 39.615 .098, 4=0.95890.000825 (/f-16.7), / =:414.45 0.95494 999.94 446.5 59.5 34.90 0.92359 900.74 442.9 22.7 Dutoit and Friderich, B =2.10124 35.31 0.9540 891.5 453.0 43.5 32.81 0.9368 838.6 448.6 76.7 29.24 0.9086 762.5 445.6 99.0 26.80 0.8895 708.3 442.1 (*) also 272.96, 277.63, 271.8, 279.6, 278,05 14 TABLE V (Con/.) Renard and Guye, B =2.1108 10.9 36.27 0.964 909.75 447.9 41.0 33.12 0.939 845.41 447.5 55.0 31.53 0.927 811.76 445.6 78.9 28.84 0.907 753.38 441.8 96.0 27.04 0.892 714.26 440.4 108.8 25.71 0.882 684.25 438.9 126.7 23.91 0.866 644.16 437.9 134.8 23.12 0.858 626.26 337.5 154.0 21.20 0.840 582.87 436.2 165.0 20.18 0.828 560.61 436.6 175.5 19.19 0.822 535.28 435.1 6. Dephenyl methane, M=16 1.1, 4=1.01260.0007914 (, f-11), / c =497 59.0 37.80 0.9745 1171.26 555.8 Dutoit and Friderich, B =2.10124 108.3 27.86 0.9209 931.4 557.5 210.2 19.11 0.8438 677.1 538.4 7. Ethyl benzene, M=106.08, ^=0.883160.0008333^ , f =346.4 c 21.9 31.325 0.86566 772.83 350.8 59.5 26.775 0.83358 681.13 351.3 8. Ethyl aniline, M=121.098, 4=0.9796 0.000831/, /=:425.4 35.8 37.89 0.9488 862.06 444.90 58.7 35.35 0.9310 907.51 444.96 Dutoit and Friderich, k B =2.10124 7.4 37.26 0.9738 927.9 455.0 107.8 22.89 0.8886 698.4 446.2 210.0 16.76 0.7996 477.6 443.3 9. Ethylene chloride, M=98.95, 4=1.280149 0.0015277/ ( 288.4 , f= ] 289.3 ( 283.3 34.1 33.178 1.22805 618.79 299.4 60.0 29.27 1.18850 558.09 299.8 10. ( 250.0 Ethylidene chloride, M=98.95, / =1.206951 0.0015992/, t -} 254.5 ( 260.0 34.065 26.042 1.1525 506.87 252.5 56.8 22.76 1.1160 452.56 252.5 15 11. 12. 13. 14. TABLE V (Cont.) Flourbenzene, M=96.04, 4=1.04655-0.001208*, 9.3 31.077 1.03435 636.80 34.5 27.537 1.00537 574.62 282.1 281.7 lodobenzene, M=203.96, 4=1.8606-001535*, ^=448.0 23.5 43.55 1.82445 1010.62 453.0 59.6 39.08 1.7691 925.73 453.5 ( 288.3 Isobutyl acetate, M=H6.096, ^=0.8802-0.001065 (-10), * tf =| 2 95.8 23.8 25.35 0.86565 664.23 308.1 59.3 21.55 0.82820 581.54 309.0 15. Mesitylene, M=120.1, d =0.8746 0.00081/, ^36=7.7 23.5 30.32 0.8656 812.61 370.0 57.3 26.60 0.8282 734.21 370.9 Dutoit and Friderich, B =2.10124 7.4 27.92 0.8686 746.1 368.5 108.4 18.47 0.7846 528.9 360.1 Renard and Guye, k B =2.1108 11.4 28.3 0.866 758.23 376.6 25.1 26.7 0.054 722.05 373.2 36.3 25.84 0.845 703.75 375.7 55.4 23.99 0.829 661.74 374.9 64.8 23.29 0.821 646.60 377.1 74.3 22.20 0.814 619.39 373.8 92.2 20.57 0.798 582.01 373.9 108.9 19.03 0.784 544.83 373.0 127.0 17.43 0.769 509.38 374.3 146.6 15.83 0.752 465.98 373.4 156.2 15.02 0.744 445.30 373.2 Metaxylene, M=106.08, 4=0.8740 .000944 (/-10) , *,=346.l 24.9 30.38 0.8599 752.85 346.4 58.9 26.41 0.8271 671.67 346.3 Dutoit and Friderich, 15.7 28.97 0.869 74.9 22.71 0.814 136.7 16.56 0.759 714.2 583.4 445.8 361.6 358.5 354.8 16. Methyl aniline, M=107.08, 24.6 42.86 59.7 38.38 TABLE V (Con/.) 351.4 352.7 351.6 350.7 350.4 850.2 3a0.2 351.5 351.7 352.1 4=0.9944 0.000801(^-10), / f =428.6 0.9827 977.78 440.^3 0.9546 892.68 440.3 Renard and Guye, k B =2.1108 10.0 28.88 0.874 707.95 38.0 26.06 0.849 651.57 49.0 24.75 0.837 624.47 63.9 23.25 0.824 592.78 76.8 21.94 0.812 564.88 88.0 20.85 0.803 540.82 99.4 19.74 0.792 516.76 109.0 18.94 0.784 499.18 128.3 17.16 0.767 458.93 136.5 16.43 0.759 442.49 Dutoit and Friderich, B =2.10124 9.9 39.19 0.9947 886.5 437.8 108.5 26.37 0.9128 690.2 443.0 210.8 18.59 0.8227 477.7 444.1 17. Orthoxylene 23.85 , M=106.08, 32.015 4=0.8932 0.0008425/, 0.8731 785.36 / f =358.3 358.9 59.40 27.88 0.8432 701.07 359.15 18. Paraxylene, 37.9 M=106.08, 28.785 4=0.8801 0.0008468/, 0.8480 720.0 / =344.4 345.6 59.2 26.308 0.8300 667.51 344.9 19. Phosphorus 35.6 trichloride, M 29.73 1.5649 4=1.612750 587.24 .0013463/, 287.66 16.4 28.71 46.2 24.91 20. Toluene, M=92.064, 36.0 28.925 59.2 26.075 Ramsay and Shields 1.582 562.3 1.527 499.8 4=0.86820.0009526(1-15.2), ^ 0.8484 0.8263 Ramsay and Aston, 15.2 28.18 0.8682 46.6 24.60 0.8380 78.4 20.93 0.8080 132.5 15.53 0.7535 658.57 603.75 k B =2.12112 631.0 563.9 490.8 382.2 287.50 287.80 ( 320.8 1 320.6 318.04 318.02 318.7 318.0 315.8 319.5 11 TABLE V (Con/.) Renard and Guye, B =2.1108 13.1 28.21 0.871 630.65 317.9 29.1 26.33 0.853 596.87 317.9 48.0 24.15 0.835 555.29 317.1 59.0 23.10 0.827 532.93 317.5 79.0 20.92 0.807 490.01 318.1 91.5 19.55 0.795 464.48 317.5 108.9 17.89 0.778 431.21 319.2 DISCUSSION OF RESULTS It will be noticed that four liquids have been studied only at one temperature ; viz. : bromine, phosphorus trichloride, diphenyl methane, and carbon disulphide. In the case of bromine, one temperature was considered sufficient owing to the fact that practically the calculated t c agreed with observed one, and the Ramsay and Aston figures by aid of capillary rise, which disagreed with this, were stated by the authors to be inaccurate. With phosphorus trichloride at 60 difficulties were encountered, possibly due to the presence of water vapor in the air, although it was passed first through drying tubes, a yellow substance forming in the ventila- tion tube, and giving a slightly variable result, although the mean value agreed practically with that at the lower temperature which leads to practically the observed t c and agrees with that found from capillary rise. With diphenylme- thane at 23 it was impossible to get satisfactory results owing to the fact that it was below its melting point, and crystallization in the tube was hard to avoid, so that checking results at this temperature were difficult to obtain. Since the value of tc at 59 does not differ greatly from the value found by Dutoit & Friderich at 108, it was thought more important to consider further liquids rather than continue work with this. Carbon disulphide was used here only to see what effect would be produced by the known purity of this sample, as compared to that purified in this laboratory. Assuming, since in the case of bromine and phosphorus trichloride, the t c calculated agrees with the critical temperature, and, in the case of diphenyl- methane it is in agreement with that found at 108 by Dutoit & Friderich, that these liquids are non-associated, we can conclude that all of the liquids examined above, with the exception of dimethyl aniline, are non-associated, for all, with this exception, lead to the same calculated value of t c (within a very small dif- ference) at both temperatures of observation. In the case of dimethylaniline the same downward trend is shown by drop weight as is to be noticed from the results of capillary rise. It is probable however, that this discrepancy of 0.8% between 59.5 and 21 is due to the decomposition of the liquid, for it darkens 18 very rapidly on exposure to light, and the low temperature observation was made first when the liquid was perfectly fresh. Possibly a reversed trend would be the case if observation at high temperature was made while the liquid was fresh. The agreement of the calculated t c with the observed critical temperature was found to be very satisfactory for the following liquids: brombenzene, bromine, ethylidene chloride, toluene, phosphorus trichloride, paraxylene, orthoxylene, metaxylene, while in mesitylene, ethylbenzene, and iodobenzene the discrep- ancies are below \% for the former, and slightly above that for the two latter. For the other liquids the calculated t i. e., the fictitious critical temperature, viz. the point 6 below which, by the formula /(M)= B (t-t-), /(M) would be zero, is larger than critical temperature for cymene, dimethyeaniline, ethylene chloride, methylaniline, ethylaniline, isobutylacetate and diphenylmethane. It is possible to compare with these values of the calculated \ c , the corre- sponding values found from capillary rise, and consequently to find proof for the relationship for only the following liquids: carbon disulphide, cymene, dimethylaniline, diphenylmethane, ethylaniline, toluene, phosphorus trechloride, metaxylene, methylaniline, and mesitylene. In the cases of toluene and phosphorus trichloride the agreement is practically perfect between the critical temperatures from drop weight and capillary rise. With diphenylmethane, ethylaniline, cymene, and methylaniline, a value of tc at some temperature in capillary rise values is always to be found which is equal to the one constant at both temperatures by drop weight method. In the case of methylaniline, the large change from 9.09 to 108.5 when compared with the smaller one between 108.5 and 210.8 would seem to indicate error somewhere, which of course may be due to decomposition. With dimethyl- aniline the trend is the same in all results, although it can hardly be said that the agreement in the capillary rise values is satisfactory all this, including the trend, may well be due to the decomposition of this liquid, which seems to be very un- stable in pure condition. For cymene the mean of all values from capillary rise is 395.5, which agrees fairly well with the constant drop weight value of 394.7. For metaxylene the drop weight values are constant but lower than either set found from capillary rise, but the fact that capillary rise values do not even agree well enough with one another, indicates that there is a source of error in capillary rise in case of this liquid. 1 Q . * *.,. * CONCLUSIONS ' The results of this research may be summarized as follows : I. According to the new definition of the normal molecular weight, i. e., that the normal (benzene) constant B gives for the liquid a constant value of t independent of the temperature of observation, in the relationship /(M) = B (^-/-6) the following liquids are to be regarded as perfectly non-associated as is benzene itself : brombenzene, bromine, ethylidene chloride, toluene phosphorus trichloride, para-, ortho-, and meta-xylenes, mesitylene, ethylbenzene, iodobenzene, flour- benzene, cymene, ethylenechloride methylaniline ethylaniline, isobutylacetate, carbon disulphide, and diphenylmethane. In the case of dimethylaniline the discrepancy in the values of t g is 0.8%, which is much too great for an experi- mental error, and can only show that the molecular weight is slightly abnormal, which could be due to the fact that the liquid is readily decomposed. II. Of the values of t e calculated for the above liquids, those for brom- benzene, bromine, ethylidenechloride, toluene, phosphorus trichloride, ortho-, meta-, and para-xylenes agree excellently with the observed critical temperature; while the disagreement with this for ethylbenzene and iodobenzene is around 1%. III. The agreement between the calculated values of tc from drop weight and those for capillary rise, with the 11 liquids which have been studied by that method is exceedingly good for toluene and phosphorus trichloride. In other cases the mean of the t e from capillary rise agrees well with that from drop weight, and it is only for mesitylene and metaxylene that the values in mean from capillary rise disagree with those from drop weight. In the latter case, however, this seems to mean little, as the values by two observers, although higher through- out, do not agree even fairly with one another. BIOGRAPHY Garabed K. Daghlian was born January 11, 1882, in Aintab, Turkey. He graduated from Central Turkey College in 1902, spent the year 1906-7 in graduate study in Syrian Protestant College, Beiruit, Syria, and attended Columbia Uni- versity as a graduate student in the faculty of Pure Science, i909-19il. He received an M.A. from Columbia University in 1910, and was appointed Uni- versity Scholar in Physical Chemistry for 1910-11, He was art assistant and instructor in Central Turkey College, 1902-1909. UNIVEESITY OF CALIFORNIA LIBRARY, BERKELEY THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW Books not returned on time are subject to a fine of 50c per volume after the third day overdue, increasing to $1.00 per volume after the sixth day. Books not in demand may be renewed if application is made before expiration of loan period. FEB 29193;> 757n-8,'31 FES 29 1932 22691 6