UC-NRLF 35 1DD EXCHANGE ajt : 1 LCHAKGU I. The Detection and Determination of Minute Quantities of Glycerine II. The Volumes of Weight-Normal Cane Sugar Solutions at Different Temperatures 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 FELTON SAMUEL DENGLER, BALTIMORE, 1912 GEORGE W. KING PRINTING Co. BALTIMORE, MD. I. The Detection and Determination of Minute Quantities of Glycerine II. The Volumes of Weight-Normal Cane Sugar Solutions at Different Temperatures 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 FELTON SAMUEL DENGLER, BALTIMORE^ 1912 GEORGE W. KING PRINTING Co. BALTIMORE, MD. CONTENTS. Acknowledgment 5 I. The Detection and Determination of Glycerine. Introduction 7 A. Qualitative Test 7 Description of Method 8 Results 9 Conclusions 10 B. The Determination of Glycerine. Summary of Previous Methods 11 Description of Method 12 Results J4 Results with Electrical Combustion Method 15 Summary and Conclusions 1G II. The Volumes of Weight-Normal Cane Sugar Solutions. Introduction 17 Description of Apparatus 19 Method 20 Results 22 Conclusions 30 Biography 31 253953 ACKNOWLEDGMENT. To Professor Morse, under whose guidance this investigation was pursued, the arthur desires to express his sincere gratitude for kindly assistance and instruction. The author is indebted to President Ira Remsen, Professors Jones and Acree for in- struction and inspiration, also to Professor Whitehead, under whose guidance one of the subordinate subjects was carried out. To Dr. Frazer and Dr. Holland the author's thanks are due for willing assistance. PART 1. The Detection and Determination of Minute Quantities of Glycerine. In the measurement of osmotic pressure in this laboratory, no measurement is regarded as conclusive until it has been shown that the membrane was not broken, thereby allowing some of the solution to escape from the cell. In other words, the solution whose osmotic pressure is being measured must have the same concentration after the cell has been up days, or in some cases weeks, as it had when it was first put up. There are two ways of solving this problem. Either the solu- tion on the inside of the cell must be tested before and after the experiment, to ascertain whether or not any dilution has taken place, or the solution on the outside of the cell must be tested to ascertain whether or not it contains any of the substance whose osmotic pressure is being measured. The polarimeter affords an excellent method for substances which are optically active, but in the case of glycerine, it was thought more practicable to test the water in which the cell is immersed for the presence of this substance. It was necessary to search for some methods for the detection and determination of minute quantities of this substance. Qualitative Test. In the measure of osmotic pressure, it is the practice here to make the water in which the cell is immersed during measure- ment, one-hundredth ion weight normal with copper sulphate; while to the solution on the inside of the cell, there is added, besides the substance whose osmotic pressure is to be deter- mined, a quantity of potassium ferro cyanide, which is osmoti- cally equivalent to the copper sulphate on the outside. The purpose of the addition of these membrane-forming salts, is to 8 repair any break which may occur in the membrane. If there is added to an alkaline solution of glycerine, a small quantity of copper sulphate, the solution becomes deep blue in color, but without precipitation of cupric hydroxide. If the addition of copper sulphate is continued, the precipitate of cupric hydro- xide appears. The color is very similar to the color of solutions containing the Cu (NH 3 ) 4 ion. The cupric hydroxide can be filtered off through an asbestos filter, although some of the colored copper glycerine compound is absorbed by the asbestos in the filtering process. Because this absorbtion takes place, the method cannot be used for the quantitative determination of small quantities of glycerine colorimetrically. It was a question whether or not this color might be due to the alkali dissolving some of the copper, hence the experiments tabulated below were carried out to determine at what concen- tration the alkali dissolves the copper. Column 2 shows the quantities of copper sulphate ; column 3, the quantities of gly- cerine; column 4, the quantities of potassium hydroxide; col- umn 5, the total number of cubic centimeters in the ;final solu- tion; column 6, is the normality of the alkali obtained from columns 4 and 5, and column 7 shows the color of the filtrate. >% - 5 5s j> 2 H o 1 24.78 5 28.05 21 2 24.78 5 280.50 30 3 24.78 5 420.75 35 4 24.78 5 504.90 38 5 24.78 5 701.25 45 6 24.78 701.25 45 7 24.78 701.25 45 8 24.78 420.75 35 9 24.78 504.00 38 10 24.78 561.00 40 11 49.56 5 T04.90 38 12 49.5(5 504.90 38 .023 no color .170 slight color .210 slight color .240 good color .280 deep color .280 slight color .280 slight color .210 no color .240 no color very .250 slight color .240 good color .240 no color 9 A study of the table shows that in the first five experiments, as the concentration of the alkali increases the quantities of all the other substances being kept constant from .23 normal to .28 normal, the color of the filtrate increases. Experiments Nos.6 and 7, which contain no glycerine, show by the slight color of the filtrate that some of the copper has been dissolved. In ex- periments 8 and 9, in which the concentration of the alkali was .21 and .24 normal, respectively, no color was obtained in filtrate, while in experiment No. 10, when the alkali was .25 normal, a very slight color was obtained in the filtrate. '_ Some of the cop- per is dissolved then when the alkali is .25 normal, but it is not dissolved if the alkali is kept at .24 normal, or below. In other words, the solution in w r hich this reaction is employed for the detection of glycerine must not be over .24 normal with potas- sium hydroxide. The following experiments were carried out with the view of determining how small quantities of glycerine could be de- tected : w . Sw Otc oo Sg r g X - ) t~ E- 1 'T N 5 123.92 100 1262.25 .23 100 colored X 3 123.92 100 1262.25 .23 100 colored N 2 123.92 100 1262.25 .23 100 colored N 1 123.92 100 12(52.25 .23 100 colored N 123.92 100 1262.25 .23 100 colorless These show that one milligramme of glycerine can be detected in one hundred cubic centimeters of solution. The filtrate of the blank containing no glycerine is colorless, therefore the color in the other experiments can not be due to dissolve) copper. Some idea of the composition of the copper-glycerine com- pound was obtained by mixing copper sulphate and glycerine in different proportions, molecule for molecule. If all' the copper 10 is used up by the glycerine in forming the colored compound, then no precipitate of cupric hydroxide will be formed. A pre- cipitate then indicates that more copper is present than is neces- sary for the formation of the colored compound. ,. - F . ^ . - 2 1 mol. -lye. 10 % mol. (T SO 4 ~ 2 2.71 .23 100 Slight color and no ppt. 1 mol. glye. to % inol. CU SO 4 .'> 4.07 .23 300 s.i:j .2:', 100 Good color and a ppt. 1 mol. glyc. to 1 mol. CU SO 4 100 271.2 .2:5 100 Deep blue color and ppt. 1 inol. glyc. to 1/2 mol. CU SO 4 100 135.62 .23 100 Deep blue color and no ppt. The experiments show that when the ratio is one molecule of glycerine to one-half molecule of copper sulphate, or rather two molecules of glycerine to one molecule of copper sulphate, then all the copper remains in solution. The conclusion to be drawn from these facts is that the -copper atom substitutes two hydro- gen atoms, each in diflfierent molecules of glycerine, and thus serves to hold together two glycerine residues. The logical way to carry out these experiments would be to add the copper sul- phate solution drop by drop to the alkaline glycerine solution until one drop produces a precipitate of cupric hydroxide. This was tried, but a precipitate appeared before the ratio reached iwo molecules of glycerine to one molecule of copper sulphate. It would appear then that some of the copper was used up by the alkali to form cupric hydroxide before all the glycerine had been changed to the colored compound. (_The alkali must then be added last to the solution containing the copper sulphate and the glycerine. The qualitative test can then be used for detecting two or three milligrammes, and with some experience, one milligramme of glycerine in one hundred cubic centimeters of solution. The 11 unknown solution should contain enough copper sulphate to make it one-hundredth normal, and enough alkali is then added to make I lie solution .lio normal. The precipitate of cupric hydroxide is tillered off, and if the filtrate has a blue color, the solution contained glycerine, provided other substances are absent which behave in the same way. Quantitative Detenu inn I ion. The estimation of glycerine can be affected by oxidation with potassium permanganate or potassium diehromate. Hehners 1 dichroiuate method, in which the amount of that salt reduced is determined, has this objection, that since the standard solution is somewhat strong and expands as much as .05 per cent, per degree, great care must be taken to keep the temperature con- stant, and at best, the method does not permit any great re- finement and was wholly inapplicable to our purpose. The permanganate oxidation method was first proposed by Wauklynr'and further worked by Fox, Benedikt and /sigmondy. 2 The following is a brief statement of the same: The glycerine solution is made alkaline with potassium hydroxide and then treated with a saturated solution of permanganate. The solu- tion is boiled one hour and then treated with enough sodium sul- phite to destroy the excess of permanganate. The precipitated manganese dioxide is filtered oil. and known volumes of (lie fil- trate are acidified with acetic acid and then treated with calcium chloride. The precipitated calcium oxalate is either determined gravimetric-ally as the carbonate or the precipitate is rinsed from the filter, acidified with sulphuric acid, heated to 60 de- grees ( \ and titrated with a decinorijnal solution of perman- gamite. The authors of the method obtained satisfactory re- sults with it, but it is long and rather complicated. The method was likewise inapplicable where very minute quanti- ties of glycerine were to be determined. 1. Allan's roininrn-inl (>rir:mir Chemistry. Vol. 2. 316. 2. Allen's ( '<>mni<>rci;il (M-i;;Mii<- ( 'hemisfry. Vol. 2. .",14. 12 The following work was undertaken then to adapt the oxi- dation by permanganate to our conditions. The standard solu- tion of potassium permanganate employed contained from five to six milligrammes of the dissolved salt in each cubic centi- meter. A standard solution of potassium tetroxalate, which was used in determining the strength of the permanganate, was made equivalent, as nearly as possible, to the permanganate solution. Solutions of potassium permanganate are not stable and often deteriorate very rapidly because of the reduction caused by small quantities of the oxides of manganese. v > The best results were obtained by preparing the permanganate solution in the following way: The approximate amount of the salt is dis- solved in water and allowed to stand in the dark for several days. This gives a chance for the oxidation of any oxidizable substance, and also any precipitated oxides are coagulated. The solution is then filtered through two connected asbestos filters. The filtrate is then allowed to stand for several more days and then filtered by the same process into a clean bottle. The necessary amount of water can then be added. Solutions prepared in this way can be kept two weeks or more without any appearance of oxide or any deterioration. The potassium tetroxalate was prepared in the usual way. A saturated solution of oxalic acid was divided into two parts. The smaller part was one-fourth of the whole, less about two or three cubic centimeters. The smaller portion was neutralized with potassium carbonate while boiling. The larger portion was then heated and the potassium oxalate stirred in. The crystals obtained are then recrystalized twice from water and dried on porous- plates. All the oxidation experiments were carried out in alklinc solutions. The solutions also contained enough copper sulphate to make them one-hundredth normal with respect to that sub- 13 stance. Preliminary experiments were necessary to determine the time and temperature necessary for the complete oxidation. The best results were obtained by keeping the solutions for nineteen hours in a constant temperature bath regulated at 50 degrees C. Increasing the time did not increase the amount of permanganate reduced. An experiment was carried out in the following way: To an alkaline solution of copper sulphate and glycerine, a consider- able excess of standard potassium permanganate solution is added. This is then allowed to stand in the 50 degree bath for nineteen hours. To this is then added potassium tetroxalate equivalent to the amount of potassium permanganate used. The solution, after reduction, is then acidified with sulphuric acid, healed to 60 degrees C., and titrated with potassium per- manganate solution until a pin,k color is obtained. The amount of potassium permanganate used in titrating is equivalent to the amount of potassium permanganate reduced by the gly- cerine. The number of atoms of oxygen, equivalent to the amount of potassium permanganate reduced, was then calcu- lated, and from this the number of atoms of oxygen per mole- cule of glycerine determined. Blank experiments were put in every day, which were in every respect identical with the other experiments, except that they contained no glycerine. By means of these blank experiments any reduction outside of that produced by the glycerine itself, could be detected and they also serve as a check on the strength of the potassium permanganate solution. In every instance where a reduction had apparently faken place in these blank experiments, it was found that the potassium permanganate solution had deteriorated. The following oxidation experiments were all kept in a 50 degree constant temperature bath for nineteen hours. The re- sults are calculated in terms of the number of atoms of oxygen per molecule of glycerine. The theoretical number of atoms of oxygen necessary to oxidise a molecule of glycerine to carbon dioxide and water is seven. 14 a 1 1 123.9 140 24X.73 5.29 5.29 7.68 2 1 123.9 140 248.73 5.47 5.47 7.96 3 2 123.9 140 248.75 9.33 4.6(5 6.79 4 2 123.9 140 248.75 9.3:5 4.66 6.79 5 g 123.9 140 248.T) 23.63 4.73 6.87 6 5 123.9 140 248.75 24.06 4.81 7.00 7 10 123.9 140 24S."5 4(5.76 4.66 6.80 8 10 123.9 140 248.75 46.64 4.66 6.79 9 25 123.9 140 20.75 117.17 4.69 6.X 1 10 25 123.9 140 ' 24S."5 1H5.7X 4.157 (5.79 11 30 123.9 140 249.40 141.04 4.70 6.84 12 30 123.9 140 249.40 140.47 4.68 6.82 13 30 123.9 140 249.40 140.85- 4.70 6.84 14 40 123.9 140 311.75 1X5.78 4.64 5.7(5 15 40 123.9 140 311.75 1X5.97 4.65 (5.77 16 40 123.9 140 311.75 187.45 4.69 (5.X2 17 50 123.9 140 446.01 234.79 4.70 <;.X4 18 50 123.9 140 446.01 235.75 4.71 6.86 19 50 123.9 140 446.01 i ) : > ,3,84 4.68 6.83 20 48.5 123.9 140 498.80 224.46 4.63 6.74 Mean 4.69 6.82 Colniuiis 1\ .') and 4 show the (quantities of glycerine, copper sulphate and alkali in each experiment. Column 5 shows the quantity of permanganate added for the oxidation of the gly- cerine. A large excess of permanganate must be added, since in alkaline solution, only one and a half atoms of oxygen per molecule of permanganate are available for oxidizing the gly- cerine. ) Column 6 shows the amount of permanganate reduced by the quantity of glycerine in column 1. Column 7 gives the milligrammes of permanganate reduced by each milligramme of glycerine. Column 8 gives the number of atoms of oxygen for each molecule of glycerine calculated from the quantity of permanganate reduced, and the quantity of glycerine present. The mean values do not include experiments 1 and 2. where only 1 milligramme of glycerine was oxidized. The mean number of atoms per molecule of glycerine is 0.82, while the theoretical number is seven. Two different solutions of gly- cerine were used in these experiments, one containing one milli- IS gramme of glycerine per cubic centimeter of solution, and the other contained five milligrammes of glycerine per cubic centi- meter of solution, so that it was unlikely there was an error in making up the glycerine solution. As a further check on the glycerine solution in experiment No. 20, a weighed amount of glycerine was oxidized directly with practical agreement. In experiments Nos. 1 and -. the results obtained are high com- pared with others. This is probably due to the experimental errors, including the temperature, effects on standard solutions, which are often considerable, accumulating on the small quan- tity of glycerine. In view of the fact that all the results obtained are below in the amount of oxygen necessary to oxidize the glycerine to carbon dioxide and water, it seemed desirable to check this method by some other method. It was first proposed to oxidi/e the glycerine with chromic acid, and collect the carbon dioxide gas formed, and in this way determine the purity of the gly- cerine. This method did not work out, as the chromic acid seemed to absorb the carbon dioxide, and the results were 1 lower than those obtained by the potassium permanganate method. The electrical method for the combustion of organic sub- stances devised by Morse, Taylor and (iray 1 was then used. The glycerine was burned in a current of heated oxygen. The oxy- gen is heated by passing through a porcelain tube around which a platinum wire is coiled, and which wire carries a current of electricity. The results of five combustions are contained in the following table : II 98.0 137.7 :5s.:u8 37.55] 0/767 6.85 81.0 112.4 31.67] :;<).( ;r,t 1.020 0.78 74. ! 103.7 29.286 28.279 1.007 o.76 80.2 112.8 31.358 30.76] .598 6.87 87.] 122.3 34056 :;: 1.351 .70.". r,.sc, Mcan=s= (5.82 1. Morse, "Exercises in Quantitive Chemistry." ]>;iir> r.:;7 16 The results obtained then by the potassium permanganate and electrical combustion methods agree, and the glycerine used has a purity of 97.29%. The glycerine oxidized by the per- manganate was therefore only 97.29 per cent pure. If we apply this correction to 6.S2, the mean number of atoms of oxygen which appear to be used by a molecule of glycerine, we get 7.01 atoms of oxygen per molecule of glycerine, while the theoreti- cal number is seven. Glycerine is very hydroscopic, and in all operations care was taken so that the substance was exposed as little as possible to the moisture of the air. The potassium permanganate oxidation method then can be used for determining quantities of glycerine as small as two milligrammes in alkaline copper sulphate solutions. The solutions inside and outside the cell are in ordinary practice, made one-thousandth normal with thymol to prevent the growth of penicillium in the sugar solutions. If the potas- sium permanganate method is used for the quantitative deter- mination of the glycerine, some other way of destroying the penicillium will have to be devised. The thymol is readily oxi- dized by the permanganate, and the amount of permanganate reduced by the thymol would be large compared to the amount reduced by glycerine, so that all the experimental errors would accumulate on the relatively small quantity of that substance. PART 2. of Wcif/Jit Xornuil Cam' fruyar Solutions at Different Temperatures. 1 When a solid substance is dissolved in a liquid, the volume of the solution is not equal to the sum of the volumes of the solute and solvent, but is usually smaller. This shrinkage is often quite large, and in the case of weight normal solution of- glucose, it was found by Morse, Frazer and Dunbar 2 to be 6.03 cubic cen- timeters, when 178.74 grams of glucose are dissolved in 1,000 grams of water at degrees. The exact nature of the cause of this contraction is not known. This investigation was undertaken with the view of deter- mining the contraction in cane sugar solutions at different tem- peratures. It was proposed to measure the increase in volume directly in going from a lower to a higher temperature. In other words, the apparatus was to be of the dilatometer rather than the pycnometer type. With this in view, an apparatus illus- trated in Figure 1 was devised. It consists essentially of a bulb and a calibrated tube "ab," in which the increase in volume is read with the cathetometer. The stop cock is placed at the bottom for convenience in clean- ing and drying the apparatus. It also has the advantage that no small gas bubbles can collect there when the temperature is 1. This work was clone in collaboration with Mr. Eyssell, and all the data on the odd concentrations will be found in his dissertation. 2. Am. Chem. Journal, 38, 222. 1 100 cc F ic i 19 raised, which might happen were it placed on the side or near the top. It is, of course, necessary that the stop cock fit per- fectly, and considerable grinding was necessary in order to ob- tain the desired result. The bulb part and the capillary tube "ab" are prepared sepa- rately and then sealed at "c." The bulb is weighed empty. Tt is then filled with air-free water of known temperature to the mark on Ihe tube and weighed. It is then filled with water of a known temperature to the 100 mm. mark, and again weighed. From the weights and temperatures of the water, the capacity of the bulb to mark and of the tube from to 100mm. can be calculated. The small tube "ab" .has an internal diameter of 2-2.25 nun., and the distance between the scratches "a" and "b" is approximately 350 mm. The part between the scratches was carefully calibrated by means of a short thread of mercury, and a curve drawn for the corrections which must be applied on account of the inequalities of the bore. The calibrated tube was then sealed on to the graduated tube at "c." The capacity between the 100 mm. mark and the lower scratch "a"was deter- mined by means of a mercury thread which rested on the 100 inni. mark and extended above scratch "a" into the cali- brated part of the tube. The thread was weighed and its volume calculated at the temperature. The known volume of the portion of the tube above the scratch "a" which is filled by the mercury, is then deducted from the total volume. The dif- ference, after applying the meniscus correction, is the capacitv of the apparatus between "a" and the upper limit of gradu- ation. All weighings were made with a tare of the same volume and form, so that the weighings were not influenced by tempera- ture, moisture or air displaced. In all, eleven such pieces of apparatus were prepared so that ten could be used for the fe ; weight normal cane sugar solutions, and the eleventh for air free wafer. The water used in making up the cane sugar solutions was boiled to free if from air. Tn making up a solution, the quantity of water was taken, which in a vacuum, would weigh 150 grams. 20 To this was added the necessary weight of cane sugar, also cor- rected for air displaced. The rotation of the solution was taken with a polarimeter as a check on the concentration. The rotation was also taken after the experiment, in order to deter- mine whether any increase or loss of concentration had taken place. The cane sugar solution was cooled below the tempera- ture of the bath in order that there might be an increase rather than a decrease in volume in reaching the temperature of the bath. A decrease in volume would mean the leaving of a film of solution on the walls of the calibrated tube, and the exact volume would necessarily be diminished by that amount. The top of the calibrated tube was closed by means of a rubber cap used in fountain pen fillers. This cap effectually prevented any evaporation, and at the same time allowed for any ex- pansion. The temperature of the hydrant water during April, made it necessary to give up all the temperatures below 15 degrees. The pieces of apparatus were weighed empty, filled with the solutions, and placed in a constant temperature bath regulated automatically to keep the desired temperature with a maximum variation of about .01 of a degree. A description of the bath used will be found in Vol. 45 of the American Chemical Journal, page 381. Several days were required for the solutions and glass to come to temperature, after which constant readings were obtained on the height of the liquid in the calibrated tube. After the volumes of the solutions had remained constant for several days and the necessary readings secured, the bath was regulated to the next desired higher temperature. In this work, the volumes were determined at 15, 20, 25, and 30 degrees. The pieces of apparatus were then taken down and weighed with the contained solution. From the weight of the apparatus, empty and full, the weight of the solution is obtained. This weight must be corrected for the weight of the solution contained in the bore of the stop cock, since this solution in the stop cock did not enter into the volume changes. The correc- tion was applied as follows: The volume of the bore of the stop cock was determined by means of mercury. This volume 21 . was then added to the observed volume of the whole solution at '20 degrees. The weight of the solution, divided by this cor- rected volume, gives the density of the solution. The density times the volume of the bore of the stop cock, gives the weight of the solution in the stop cock, and this weight, subtracted from the original weight, gives the corrected weight of the solution in the tube. In the following table 1, column 1, con- tains the number of the pieces of apparatus; colume 2, the weight normal concentration of the solution; column 3, the volume of the bore of the stop cock ; column 4, the density of the solution ; column 5, the weight of the solution in the bore, and column 6, the corrected weight of the solution in the apparatus. TABLE 1. 1 n 3 4 5 6 | J. _. 6 6 ^ J b Hj j II fill E ^ 2 3 - x J -~ >^~ o "3 3 ^ ^ r"*" ^ v. r. w x rt 7 0.2 .0510 1.0228 .0522 102.6412 g. 13 0.4 .0576 1.0458 .0602 105.1192 g. 6 0.6 .0393 1.0663 .0419 107.2322 g. 10 0.8 .0472 1.0869 .0513 109.2404 g. 15 1.0 .0461 1.1048 .0509 110.9894 g. Water .0452 0.9972 .0451 100.2209 g. In order to calculate the volume of the sugar used in making up the solutions, the specified gravity of solid sugar must be known. On looking this up, a wide variation was found in the results obtained by different investigators. It was difficult to decide on which one was correct, so it was decided to work out two tables for each temperature, one using the value 1.5813, ob- tained independently by Kopp and Gerlack, and the other table based on the value 1.5860 obtained by Schroeder. Joule and Playfair give .0001116 per degree as the cubical expension of solid sugar between and 100 degrees. The volume of 1,000 grams of water at the desired tempera- ture was calculated from the values found in Landolt-Boern- stein-'s Physikalisch-Chemiscbe Tabellen. 22 The actual volume of a sugar solution, containing 1,000 grams of water, was calculated from the observed volume and the per- centage of water by weight in that volume. The following table 2 gives the percentage by weight of sugar and water in the different concentrations of weight normal sugar solutions. TABLE -2. ^ c f- ? ? ^ V. i. v. 0.2 tooo 0:5.65 (57.8784 6.85 0.4 1000 88.05 135.7568 1 1 .05 0.6 1000 83.08 203.6352 1(5.02 0.8 1000 78.65 271.5136 21.35 1.0 1000 74.66 339.3020 25.34 The results obtained at 15 degrees are given in the two fol- lowing tables. For calculating the volumes of the sugar 1.5813 was used as the specific gravity of solid sugar in table 3 and 1.5860 was used in table 4. TABLE :;. Temperature 15 degrees. Sp. (Jr. of solid sugar = 1.5X13. o * S . I * o e.. a - ;-" 0-43 & CS V- 5. 1 0.2 0.4 0.6 0.8 1.0 1000.857 1000.857 1000.857 1000.857 1000.857 42.024 85.848 12X.772 171.696 214.620 1043.781 1086.705 11 29.62! > 1172.553 1215.477 1042.940 1084.756 1127.430 1168.358 1210.672 0.832 1.949 2.190 4.195 4.805 TABLE 4. Temperature 15 decrees. Sp. Gr. of solid sugar = 1.5860. f| "OCJ - II a*TJ 1.2 kj| ^ c '^ -J * V. ac-'S c > > 0.2 1000.857 42,788 1048.645 1042.040 0.60(5 0.4 1000.857 X5.576 1086.438 1084.75<; 1 .677 0.6 1000.857 128.364 1129.221 1127.430 1.782 0.8 1000.857 171.152 1172.009 1168.358 3.651 1.0 1000.857 213.940 1214.70 -t 1210.672 4.125 23 Tables .") and (i contain I lie results obtained at -0 degrees. TABLE 5. Temperature 20 delves. Sp. (Jr. of solid sugar = 1.5X13. i. c g -r -r I* Q 88 ~=L~ ~ ._. | . -gg ^S : ~~ it Z E ^T Z- C ^ ^ in /- - :r ~ :: "3 ^f 3 *4 C . "E ^r "5 ^ '^> > '^ 7 */: c S - > . ^ 0.2 1001.7.",! 42.94s 1044.699 1044.007 0.692 0.4 1001.751 sr,.896 1087.647 1065.1*74 1.673 0.6 1001.7."! 128.844 1130.595 112S.806 1.789 0.8 1001.751 171.792 1173.543 1 ir.ii.S77 3.666 1.0 1001.751 214.740 1216.41)1 1212.343 4.148 TABLE 6. Temperature 20 degrees. Sp. Gr. of solid sugar = 1.5860. w ; c e c; ?. r 5 r &.' 2 - ts S ^ 5 ?" r F ~ -r ?r i ^~ I fc^ > 3 '^ i; ^ 7 00 ^S ~ >~ > 0.2 1001.751 42.812 1044.563 1044.007 0.556 0.4 1001.751 85.624 1087.375 1085.974 1.401 0.6 100li751 128.436 1130.187 1128.806 1.381 0.8 1001.751 171.248 1172.999 1169.877 3.122 1.0 1001.751 214.060 1215.811 1212.343 3.468 Tables 7 and 8 contain the results obtained at 25 degrees. TABLE 7. Temperature 25 degrees. Sp. Gr. of solid sugar = 1.5813. c X "o *-4 C *G c "* e t3 V u xj'S a * 3 r >g t. ji 3 0.2 1002.911 42.972 1045.883 1045.307 0.57(5 0.4 1002.91 1 85.944 1088.805 1087.417 1.438 0.6 1002.911 128.9 1C 1131.827 1130.412 1.415 0.8 1002.K11 171.888 1174.799 1171.596 3.203 1.0 1002.911 214.860 1217.771 1214.209 3.562 24 TABLE 8. V Temperature 25 degrees. Sp. Gr. of solid sugar = 1.5800. "^ s. c I "S y o 2 c is ^ ri C 1 ^ CD tJ3 C^ re a os Is I*; O+J tt Q '~ f ~ S z K a^-S ll 14,2 ^: o ._-, p ^ a -**" ? --- 7 'A o d C > 0.2 1002.911 42.836 1045.747 3045.307 0.440 0.4 1002.911 85.672 1088.583 1087.417 1.116 0.6 1002.911 128.508 1131.409 1130.412 .997 0.8 1002.911 171.344 1174.255 1171.596 2.659 1.0 1002.911 214.180 1217.090 1214.209 2.881 Tables 9 and 10 contain the results obtained at 30 degrees. TABLE 9. Temperature 30 degrees. Sp. Gr. of solid sugar = 1.5813. 11 pa ^ ||l II' !/ >c || "3 p 1 5w c g"e SO & C ^ ^ x - 3 > 0.2 1004.314 42.994 1047.308 1046.801 0.507 0.4 1004.314 85.988 1090.302 1089.055 1.247 0.6 1004.314 128.982 1132.296 1132.182 1.114 0.8 1004.314 171.976 1176.290 1173.479 2.811 1.0 1004.314 214.970 1219.284 1216.235 3.049 TABLE 10. Temperature 30 degrees. Sp. Gr. of solid sugar = 1.5860 V m g g 3 So o o o > __ ^ M a S'S IP o Is ^ll 1 S.S .5? a 3 _2 ! J5 ^ c ll SI ^w C 9Q o ti 1 0.2 1004.314 42.860 1047.174 1046.801 0.373 0.4 1004.314 85.720 1090.034 1089.055 0.979 0.6 1004.314 128.580 1132.894 1132.182 0.712 0.8 1004.314 171.440 1175.754 1173.479 2.275 1.0 1004.314 214.300 1218.614 1216.235 2.379 The values of the contraction obtained at the different tem- peratures are all brought together in Tables 11 and 12, in order to observe what takes place as the temperature is raised. 25 TABLE 11. Sp. (Jr. of sol i - w ej u Q u 0.2 0.832 0.692 0.57; 0.507 .325 0.4 1.9*9 1.673 1.438 1.247 .702 0.6 2.190 1.789 1.415 1.114 1.079 0.8 4.195 3.666 3.203 2.811 1.384 1.0 4.805 4.148 3.562 3.049 1.756 TABLE 12. Sp. Gr. of solid sugar 1.5860. 0.2 0.696 0.556 0.440 (}.:\i:\ .323 0.4 1.677 1.401 1.116 0.979 .698 0.6 1.782 L.381 .997 0.712 1.060 0.8 3.651 3.122 2.659 2.275 1.376 1.0 4.12." 3.468 2.881 2.379 1.756 A stud}' of the table shows that as the temperature is raised, the contraction, or the difference between the sum of the volumes of the solute and solvent on the one side, and the ob- served volume of the solution on the other side, diminishes. This decrease is not directly proportional to the rise in tempera- ture, but is greater between 15 and 20 degrees than it is be- luccn 20 and 25 or 25 and 30 degrees. At each temperature, the contraction is proportional to the concentration, as is also the decrease in contraction. Sufficient data is not at hand to pass final judgment on these facts, but it appears that as the temperature is raised, the observed volumes approaches that of the sum of the volumes of the solvent and solute. Whether or not they ever become equal remains for future investi cations to show. The two most concentrated so- 26 hitions were found to have lost in rotation while they were in the bath, but lack of time prevented repetition of the ex- periments. The exact rotations in degrees are given in Table 13. The first column gives the rotation when the solution was made up, the second column, the rotation when the apparatus was taken down after the completion of the experiment, and the last column, the loss in rotation. TABLE 13. o c c % & % r< S-3 3 s .2 s ce i -! S^'S -5 K % ^ SE t-i 3 |i ^ ^ ~ '_ fi 'o *i ^ ;l U* "-- ^. ~ - ^- ^ Water 100.4096 100.4976 .0880 .000175 0.2 100.2520 100.3536 .1017 .000203 0.4 100.4023 100.5150 .1128 .000224 0.6 100.4419 100.5637 .1218 .000243 0.8 100.3826 100.5130 .1305 .000260 1.0 100.3220 100.4605 .1385 .000276 Table 15 contains the expansion coefficients of .the sugar solu- tions, also air-free water between 20 and 25 degrees, based on the volumes at 20 degrees as unity. TABLE lo. 3c K a; II I II g| 11 Water 100.4976 100.6132 .1156 .000230 0.2 100.3536 100.4786 .1249 .000249 0.4 100.5150 100.6486 .1335 .000266 0.6 100.5637 100.7067 .1430 .000284 0.8 100.5130 100.6607 .1477 .000294 1.0 100.4605 100.6151 .1546 .000308 Table 10 contains the expansion coefficients of the sugar solutions, also air-free water between 25 and 30 degrees, based on the volumes at 25 degrees as unity. TABLE 36. ! 9) . c If 2*3 ~ 5 C 'o ^'\ O ^j i"i ti 11 ^ ^- 7; '^ r. >-H > w Water 100.6132 100.7.107 .1375 .000274 0.2 100.4786 100.6222 .1436 .OOOL'SC, 0.4 100.6486 100.8002 .1516 .000301 0.6 100.7067 100.8644 .1577 .000313 0.8 100.6607 100.8225 .1618 .000322 1.0 100.6151 100.7829 .1678 .000334 28 All the coefficients are brought together in Table 17 for the sake of comparison. TABLE 17. 1 K MS Eg Water 0.2 0.4 0.6 0.8 1.0 III Hi .000175 .000203 .000224 .000243 .000260 .000276 .000230 .000249 .000266 .000284 .000294 .000308 .000274 .000286 .000301 .000313 .000322 .000334 In the preceeding expansion coefficients, the ones obtained be- tween 15 and 20 degrees are referred to the volume of the solu- tion at 15 degrees as unity, those between 20 and 25 degrees are referred to the volume at 20 degrees as unity, and those be- tween 25 and 30 degrees are referred to the volume at 25 de- grees as unity. It seemed desirable to calculate the expansion coefficients basing them all on the same unit. In the following, then, the unit employed is the volumes of the solutions at 15 degrees, and in the case of air-free water, the volume of the water at 15 degrees. Table 18 contains the expansion coefficients between 15 and 20 degrees. TABLE 18. *ji| gtt 58 = 3e II 3 c > K c e %^ ? Water 100.4096 100.4976 .0880 .000175 0.2 100.2520 100.3536 .1017 .000203 0.4 100.4023 100.5150 .1128 .000224 0.6 100.4419 100.5637 .1218 .000243 0.8 100.3826 100.5130 .1305 .000260 1.0 100.3220 100.4605 .1385 .000276 29 Table 1!) contains the expansion coefficients between 20 and > dfgm's. based on the volume at 15 degrees as unity. TABLE 19. .2 e u E = r i If If | >** >* c c 5 Water ]00.4!>7; 100.6132 .1156 .000230 0.2 100.3536 100.4786 .1249 .000249 0.4 100.5150 100.6486 .1335 .000266 0.6 100.5< ;:57 100.7067 .1430 .000285 0.8 100.5130 100.6607 .1477 .000295 1.0 100.4605 100.6151 .1546 .000308 Table 20 contains the expansion coefficients between 25 and :',< de^rese, based on the volume at 15 degrees as unity. TABLE 20. Water 100.6132 100.7507 .1375 .000274 0.2 100.4786 100.6222 .1436 .000286 0.4 100.6486 100.8002 .1516 .000301 0.6 100.7067 100.8644 .1577 .000315 0.8 100.6607 100.8225 .1618 .000322 1.0 100.6151 100.7829 .1678 .000344 All the expansion coefficients, based on the volumes at 15 de- grees, are brought together in Table 21. TABLE 21. > W.-itcr 0.2 0.4 0.6 0.8 1.0 .000175 .000203 .000224 .000243 .000260 .000276 I .000230 .000249 .000266 .000285 .000295 .' .000274 .000286 .000301 .000315 .000322 .000344 30 The expansion coefficients increase with the temperature, as shown by the table. They are also roughly proportional to the concentration, but the results can not be regarded as final since the 0.8 and 0.9 weight normal solutions lost in rotation. This portion of the work will therefore have to be repeated. The expansion coefficients obtained for water show practical agreement with those given in Landolt-Boernstein's Physik- alisch-Chemishe Tabellen. so that the apparatus may be re- garded as accurate. With apparatus on hand, then, the vol- umes of various solutions can be determined over a consider- able range of temperature, and some light may be thrown on facts whose relation at present is not clearly seen. BIOGRAPHY. Felton Samuel Denser was born .May 111, lSS