Issued March 9, 1907. U. S. DEPARTMENT OF AGRICULTURE, BUREAU OF ANIMAL INDUSTRY. RUI.LF.TIN N. A 1) MKLV1X. Cn IVEST1GATIONS IN THE USE OF THE BOMB CALORIMETER. IN COOPERATION WITH THE PENNSYLVANIA STATE COLLEGE AGRICULTURAL EXPERIMENT STATION. BY J. AUGUST FRIES, M. S., sisduif /.'.r/V/v' /;/ , Inhnal Nutrition . t" o S i i. i CD cr 5. * ft S M So , , WASHINGTON: GOVHRNMHNT I'RINTINC, OFKICH. 1907. Property of the United States Government. Issued March 9, 1907. U. S. DEPARTMENT OF AGRICULTURE, BUREAU OF ANIMAL INDUSTRY. BULLKTIN No. 94. A. D. MELVIN, CHIEF ON BUREAU. IN COOPERATION WITH THE PENNSYLVANIA STATE COLLEGE AGRICULTURAL EXPERIMENT STATION. BY J. AUGUST FRIES, M. S.,- Assistant Expert in Animal Nutrition. WASHINGTON: GOVERNMENT PRINTING OFFICE. 1907. BUREAU OF ANIMAL INDUSTRY. Chief: A. D. MELVIN. Assistant Chief: A. M. FARRINGTON. Chief Clerk: E. B. JONES. Biochemic Division: MARION DORSET, chief; JAMES A. EMERY, assistant chief. Dairy Division: ED. H. WEBSTER, chief; C. B. LANE, assistant chief. Inspection Division: RICE P. STEDDOM, chief; U. G. HOUCK, associate chief; MORRIS WOODEN, assistant chief. I'utliological Division: JOHN R. MOHLER, chief; HENRY J. WASHBURN, assistant chief. Quarantine Division: RICHARD W. HICKMAN, chief. Division of Zoology: BRAYTON H. RANSOM, chief. Experiment Station: E. C. SCHROEDER, superintendent; W. E. COTTON, assistant. Editor: JAMES M. PICKENS. ANIMAL HUSBANDRY OFFICE. .\oV///V//.s.- George R. Samson, herabook assistant. I'lui/trii iiiirxtigntittitx: Rob R. Slocum, poultry assistant. ling investigations: L. R. Davies, scientific assistant. Cooperatirt Stnjj'. Aiiiiiml nutrition iiiri'xtigeititHiK: H. P. Armsby, rx)KTt in charge; J. August Fries, W. \V. Braman, F. \V. Christensen, assistants. /;,,/ jtrmliii-lioii in tin' Xniitli: 1). T. Gray, expert in charge; W. F. Ward, assistant. // t ln-i filing ini-fxtit/utiiHix: W. L. Carlyle, expert in charge of Colorado work; W. F. Hammond, expert in charge of Vermont work. I'nnlii-i/ lii-iiilnii/ iiinxtiiiuiiiHix: Gillx'rt M. (iowell, expi>rt in cliarge. Slieep lirecdiwj nui'slii/ntioiix: George K. Morton, ex[>ert in charge. Turkey breeding investigations: Leon J. Cole, expert in charge; \\ - . F. Kirkpati ick, awistant. 2 LETTER OF TRANSMITTAL. U. S. DEPARTMENT OF AGRICULTURE, . BUREAU OF ANIMAL INDUSTRY, Washington, D. C. , December 15, 1906. SIR: I have the honor to transmit herewith, and to recommend for publication as a bulletin of this Bureau, a manuscript entitled "Investigations in the Use of the Bomb Calorimeter," by J. August Fries, M. S., assistant expert in animal nutrition. The experiments described in this article ae a part of the investigations in animal nutrition conducted at the Pennsylvania State Agricultural Experi- ment Station by cooperation between that station and this Bureau. The particular objects and value of the work reported by Mr. Fries are described in the accompanying letter by Doctor Armsby, the expert in charge of the cooperative investigations. Respectfully, A. D. MELVIN, Chief of bureau. Hon. JAMES WILSON, /Secretary of Agriculture. 3 LETTER OF SUBMITTAL. STATE COLLEGE, PA., October 13, 1906. SIR: The investigations upon the energy values of feeding stuffs which are being carried on, with the aid of the respiration calorimeter, in cooperation with the Pennsylvania Experiment Station, depend in large part for their value upon accurate determinations of the heats of combustion of the feeds and of the visible excreta. The bomb calo- rimeter of Berthelot in its various modifications is recognized as the most convenient and accurate instrument for this purpose which has thus far been devised, and the modification of it known as the Atwater- Hempel bomb has been used in these investigations. Since the values obtained with this instrument are fundamental to the research, it is obvious that critical study should be given to the accuracy of the method employed. I have the honor to submit herewith the results of experiments by Mr. J. August Fries, assistant expert in animal nutrition, upon this subject. These experiments have been made during the past four years in connection with the investigations in animal nutrition reported or in progress, and their results indicate some of the possible sources of error in the method and the 'precautions necessary in its use to secure accuracy, while they also show the need of further experiments on the same subject. It is hoped that these results may be of value to other investigators in the same field. Very respectfully, HENRY PRENTISS ARMSBY, Expert in Animal Nutrition. Dr. A. D. MELVIN, ( 'li'n'f of the Bureau of Animal Industry. 4 CONTENTS. Page. Introductory ,. 7 The apparatus used 7 Manipulation of the apparatus 7 Calculation of results 8 Importance of the apparatus in calorimetric determinations 10 Earlier determinations of heats of combustion 10 Variations in determinations already on record 17 Causes of differences in heat determinations 18 Impure oxygen ;. 19 Testing of oxygen 19 Correction for impurity in the oxygen 22 Rise in temperature corrected for heat due to impurities in the oxygen . 26 Formation of nitric acid in the bomb calorimeter during combustion 27 Oxidation of combined nitrogen to nitric acid 28 Probable error due to disappearance of nitric acid 31 Cause of incomplete combustion 33 Alcohol heat value 35 Alcohol determination used for testing the respiration calorimeter 36 Determination of heat of combustion 36 Conclusion 38 5 INVESTIGATIONS IN THE USE OF THE BOMB CALORIMETER. INTRODUCTORY. In scientific investigations, as well as in technical work, it is often of very great importance to know the exact heat of combustion, or, in other words, the energy value of a substance. To determine the heats of combustion of organic compounds it is necessary to completely oxi- dize them to their most simple or stable decomposition products and measure the heat evolved. Several methods have been used for such determinations, but the latest and best for the greater number of com- pounds is undoubtedly the Berthelot or bomb-calorimeter method, where the substance to be anah T zed is burned in oxygen gas under high pressure. Detailed descriptions of this apparatus and directions for its use are found in various scientific works, hence no full descrip- tion need be given here. THE APPARATUS USED. The apparatus used for the work described in this paper was the Atwater-Hempel bomb calorimeter, which is a modification of the Ber- thelot apparatus." The main features only of its construction and manipulation will be referred to brie% at this time. The apparatus consists of the bomb proper, which is a strong steel cup lined on the inside with platinum, and of a platinum-lined top which rests on a lead washer and is firmly held in place by a threaded collar. A cylindrical vessel of Britannia metal is used for holding the water in which the bomb is immersed. The apparatus is provided with a stirring arrangement operated by means of an electric motor. Outside of this metal vessel are two concentric protecting cylinders made of indurated fiber, each provided with a cover. These fiber cylinders, with the dead air spaces between them, serve to insulate and protect the metal vessel and the water as far as possible from being affected by any change of outside or room temperature or circu- lation of air. A Beckmann mercury thermometer, which can be read to 0.001 C. with accuracy, is used for taking the temperatuie of the water. MANIPULATION OF THE APPARATUS. The substance to be burned, if solid, is pressed into a tablet of con- venient size, placed in a platinum capsule, which is supported on a U. S. Dept. Agr., Office of Experiment Stations Bui. 21. Conn (Storre) Sta. Kept., 1897, p. 199. 8 INVESTIGATIONS IN USE OF BOMB CALOEIMETEK. platinum wire attached to the top or cover of the^bomb, and adjusted so that the substance will come in contact with the fine iron fuse wire by means of which it is ignited. The bomb is then charged with oxygen gas to a pressure of at least 20 atmospheres. It is immersed in the weighed quantity of water, the electric connections are made, and the covers adjusted and clamped so that the stirrer will not rub. After placing the thermometer in position the stirrer is started and the temperature of the water taken, It is intended to have the water as much colder than the surrounding air before the combustion as it will be warmer after the combustion. As soon as a uniform rate of change in the water tempera- ture has been established the substance is ignited. For this purpose an electric current strong enough to fuse instantly the fine iron wire is allowed to pass through, and by the burning of the fuse wire the substance in the capsule is also instantly ignited. The combustion is only of a few seconds' duration, and the heat formed is quickly transmitted through the bomb to the surrounding water. Readings of the thermometer are taken every minute, and from the rise in temperature of the water the heat generated is calculated. CALCULATION OF RESULTS. In the calculation of the results it is necessary to apply several cor- rections. Not only the fixed weighed quantity of water but the whole system bomb, stirrer, metal vessel, etc. is heated up to the same degree, therefore the correct thermoequivalent of it must be deter- mined and used. Thermometer readings must be corrected according to the individual thermometer used. Corrections are also made for the influence of the surrounding air temperature, the heat generated by the burning of the fuse wire, the formation of nitric acid in the bomb, and the lag of the thermometer. The formula which has been used in working out the correction for the outside air influences is the Kegnault-Pfaundler formula, given in Wiley's Principles and Practice of Agricultural Analysis, Volume III, page 572, and the following example will serve to illustrate the main points to be observed in connection with the determination and calcu- lation of the heat of combustion. Formula: nl 6 _ fl 2 O r =Bum of 'i readings minus I increased by an arbitrary factor - 1 r = average rate of radiation during preliminary period. r 1 = average rate of radiation during end period. 6 = number of readings during eombuntion period. / = average 1350^)- 0.0008 = 4- 0.013. 2 4 t= +0.013. Last reading of combustion period, 4- 2.9640 C. First reading of combustion period, - 1.2275 C. Correction for outside air, + 0.0130 C. Correction for thermometer lag, - 0.0006 C. 1.7489 C. Correction for excess oxygen, 4- 0. 0003 C. Correction for impurities in oxygen, 0. 0151 C. Corrected rise in temperature, 4- 1. 7341 C. Water value of bomb system with 20 atmospheres (2, 439. 2) X rise = 4229. 82 calories. Correction tor fuse wire, 16. 00 calories. Correction for HNO 3 , 13. 50 calories. 4229. 32 calories. 4229.32 -s- 1. 0035 4185. 7 calories per gram of cellulose. Should for some reason the current of electricity remain on for a few seconds before the wire fuses, correction should be made for the heat generated by the current. 18399 No. 9407 2 10 INVESTIGATIONS IN USE OF BOMB CALORIMETER. IMPORTANCE OF THE APPARATUS IN CALORIMETRIC DETER- MINATIONS. Not only in physical and chemical laboratories do calorimetric deter- minations find application, but they have acquired special prominence in connection with physiological investigations and research work, and it is in this line of work that at the present time their application in the agricultural science has been found to be of so great importance. Thus, the very elaborate, time-consuming, and expensive nutrition investigation experiments with cattle by the use of the respiration calorimeter are so planned that the successful solving of the question of energy metabolism in the animal body rests almost entirely upon the bomb calorimeter upon its efficiency in quickly and accurately determining the heats of combustion of feed and excreta. Being, then, a most important part of the whole plan, and since so much depends upon the work of this apparatus, one is not justified, because much work has been done with it in the past, in assuming that it is perfect in all respects, or that it is adapted alike to all kinds of organic substances. Instead, it should be severely tested from different points of view, and that was the object of this work. EARLIER DETERMINATIONS OF HEATS OF COMBUSTION. To begin with, it may be well to look a little into the work already done, and note the condition in which our knowledge of the heat values of organic substances is at the present time, and how well different investigators agree in their determinations. For this purpose, as well as to get a general idea of heat values, I have, without going back to the original source of the information, collected and tabulated u large number of results upon the heats of combustion of the. more common organic substance-, as obtained by various investigators. The values are taken from text-books and scientific reference books, some of which are still in use in schools and elsewhere. To make the tables as condensed as possible the works or books referred to are represented in one column, each book by a single num- ber, and the investigators in another column by the first letters of their respective names. EARLIER DETERMINATIONS. 11 TABLE I. Heats of combustion of 1-gram substance, expressed in small calories. Substance. Formula. Calories. ELEMENTS, ETC. Hydrogen Ho ... 34, 462. 634,154.3 c34,800.0 7, 770. 1 8, 080. 7,714.0 8, 039. 8 7, 965. 7, 859. 2, 141. 7 1,600.0 1,352.6 2,260.3 2, 165. 6 2,220.5 2, 162. 5 2, 436. 2, 431. 2, 402. 7 2, 438. 6 2,441.7 1,022.6 652.3 5-, 990. 3 6, 141. 5,961.3 5, 832. 3 5,941.6 6,231.0 5, 917. 8 5, 907. 8 5,910.0 5, 914. 8 5,885.1 5,950.0 5, 949. 5, 626. 4 5, 867. 5, 629. 2 5,717.0 5,785.0 5, 849. 6 Hydrogen (8 investigators) Ho Do Ho 1 .. Carbon 1 f! - Charcoal ( wood) C Charcoal (sugar) C Do C Do C Diamond (to COo) C Diamond (to CO) c Iron (to FeoO 3 ) Fe Iron (to FcO) ... Fe Sulphur So Do So Sulphur (soft) So Do CO to CO.. . . . Do ... Do Do Do Nitrogen to HNO 3 (sol.) No. NO to NO,, PROTEIDS, ETC. (ALBUMINOIDS.) Gluten Do Elastin . Plant fibrin Do Do Serum albumin Syntonin Hemoglobin Do Do ... Do Do Milk casein Do Do Do Do Do Do 5, 858. 5, 855. 8,112.4 5, 840. 9 5, 793. ] Do Yolk of egg Legumin Do 5, 573. 6, 780. 6 5, 745. 1 5. 784. 1 Vitellin Do Do ... Refer- ences. Investiga- tors, a 1,3 F.&S. -10 F.&S. 10 H. 10 F.&S. 1,3,10 F.&S. 10 Gr. 10 F.&S. 10 Sch. 10 B.&P. 10 B. 10 F.&S. 10 F.&S. 10 B. 10 F.&S. 3 F.&S. 3 F.&S. 10 An. 10 F.&S. 10 B. 10 Th. 10 Th. 2 B. 2 D. 1,2,10 St.&L. 2,10 B.&A. 2,10 St. & L. 1,2 D. 1,2,8,10 St.&L. 2,1 St. & L. 2,10 B.&A. 1,3 B.&A. 1,2,3,10 St.&L. 2 Ru. 1,2,3,10 D. 2,10 B.&A. 2, 3, 8, 10 St. & L. 1,3 B.&A. 1,2,3,10 St. 2 D. 2, 3, 10 St. & L. 1 St. & L. 1 I). 2,3. B.&A. 3, 10 St. & L. 1,2,10 St.&L. 2 D. 2,10 B.&A. 1,2,3,10 St.&L. 3 St. & L. 1. 2. 3. 4. 6. 6. 7. 8. 9. 10. A. An B. Bu D. Du F. Ku. Kr. ai. So. Abbreviations: Works referred to Bunge, Physiological Chemistry, fourth German edition, 1902. Armsby, Principles of Animal Nutrition, 1904, and U. S. Department of Agriculture, Office of Experiment Stations, Bulletin 21. Beilage, Chemiker Kalender, 1904. Tollen, Handbuch I der Kohlehydrate, 1898. W. Ostvvald, Grundriss der Allgemeinen Chemie,-2te Auflage, 1890. Tollen, Handbuch II der Kohlehydrate, 1895. Jones, Elements of Physical Chemistry, 1903. Wiley, Principles and Practice of Agricultural Analysis, Vol. Ill, 1897. Journal American Chemical Society, Vol. XXV, 1903. Landolt & Bornstein, Physikalisch-Chemische Tabellen, 1894. Investigators Andre. Gr. Grassi. Andrews. H. Hess. Berthelot. He. Herzberg. Bunson. J. Jahn. Danilewsky. K. Kleber. Dulong. L. Langbein. Favre. LH. Luginni. Fogh. M. Matignon. Frankland. O. Ogier. Gibson. P. Petit. Gottlieb. & Lowest. c Highest. Ro. Ru. Rch. Rechenberg. Re. Recoura. Rodolz. Rubner. Silbermann. Schwackhofer. Stohmann. Tower. Thomsen. Vieille. Sch St. T. Th. V. 12 INVESTIGATIONS IN USE OF BOMB CALORIMETER. TABLE I. Heats of combustion of 1-gram substance, expressed in small calories Cont'd. Substance. Formula. Calories. Refer- ences. Investiga- tors. PROTEIDS, ETC. (ALBUMINOIDS) COU. Egg albumin 5, 687. 4 1,2,10 B.& A. Do . 5, 735. 2 1,2,3,8,10 St. & L. Do 5, 579. 1,2,3,10 St. Do . 5, 690. 6 3 St. Muscle (fat free and water) 5, 728. 4 2 10 B & A Muscle (extracted) . . 5, 720. 5 2, 8, 10 St. & L. Muscle ( fat free ) 5, 662. 6 2,10 St. & L. Do 5, 324. 2,10 St. Muscle (fat free ash free) 5, 656. 2,10 Ru. Do 5, 640. 9 2,10 St.& L. Do 5,345.0 10 Ru. Albumin cryst 5, 672. 1,2,10 St. & L. Do ; 5,598.0 1,2,10 St. Blood fibrin 5, 529. 1 2 10 B. & A. Do 5, 637. 1 2, 3, 10 St. & L. Do 5,511.0 1,2,3,10 St. Do 5, 709. 2 D. Do 5, 772. 1 D. Do 5, 532. 4 3 St. Do 5, 567. 3 3 St. Albumin ( Harnack's) 5, 553. ) St.& L. Wool 5,564.2 2 B.& A. Do 5, 510. 2 2 St.& L. Oongltiten 5, 479. 2,10 St. & L. Do 5, 362. 2,10 St. Fibrin of skin 5, 355. 1 2,10 St. & L. Peptone 5, 298. 8 1,2,3,8,10 St. & L. Do 6, 069. 2 D. Do 4, 876. 1 D. Do 4, 914. 1 D. Fish glue 6, 240. 1 2,10 B. & A. Do 6, 493. 1,2 D. Chondrin 5, 342. 4 1, 2, 3, 10 B. N..H 4 2,537.0 1,2 D. Do CONH 4 2 523 2 10 R I).. OljroOCOll C()N,H 4 CoHrftO, 2,121.0 3, 133. 6 1 2,9,10 F. B. & A. Do (:.n \(>:, 3, 129. 1 2,9,10 St. & L. Do CjH 5 NO 3 050 1 St. Do CfHiNOf 3, 053. 2,10 St. A lun i n C II NO. 4, 370. 7 2,10 B. & A. Do < 1 ! \ i >., 4 355.5 2,10 St. & L. l,i -in -in ', II|,NO . 6,536.5 1,2,10 B. ^ A. Do . 0,,Hi',NOi c. :>:> i 1 2 10 -1 A 1 Sn rki i n C.,H 7 NO 2 4 605 9 1 > 10 SI A L. 11 ip|iiirir iii'id . c II, NO.. 5 659 3 1 2 9 10 H A A Iiii C|HtNO 5 668 2 1 2, 9, 10 St. & L. In, CtHfNO* 6, t>42. 2,5,10 St. ANpartic aci'l C 4 H 7 NO 4 2 911 1 2 10 B. & A. Do O 4 H-NO 4 8 423 1 St. Do i',!l-\(i, J s'.l'.i o 2,10 SI. A; 1, T\ ric-iii . GtH|iNG| 5 915 9 1 '2 10 li A. A. AspiiriiKin !<> IK. ('rriitin rrvat <',HA..O, < < 4 H,N. J (>, ( 4 H,N.,0 S ' II N O..H..O 3, 3%. 8 ::..-,! i.o ::. i-jvii 3 714 1 2,10 2,10 2,10 1 2 10 B. & A. St. & L. SI. -1 A L. f'ri'iitin, HI ih yd n HIM r,H,,N,<>.. 4 275.4 1 '.' 10 St. & L. H (8, 206. 0) 2 |. rri.-urid Do , CiiH 4 N 4 OH 1,754.0 2, 749. 9 2,10 1 , 2, 10 1!. A M. St. .HH.""O 3, 877. 1,2,4,5,10 St. Do CI"HTOH 4, 162. 1 Rch. Do (' oH.O 3,920.0 10 Gi. Milk sugar hydrated (' oHOiiHoO 3, 777. 1 2,10 B. & V Do C oHijoO H~O 3, 736. 8 2,8,6,10 St. & L. Do 3, 663. ~, 1, 10 St. Do 3, 710. 2 Gi. Do CioHOuHoO 3, 945. 1 Rch. Do C ^II^O H~O 3,667.0 1 St. Do . C "HO HoO 3, 772. 2 4 B. Do C oHooO HO 3,721.8 6 Gi. Do C oH~O 3, 949. 3 2,8,6,10 St.& L. Do .' C ."H.")."O 4, 163. 1 Rch. Maltose hydrated C^HooOnH-O 3, 721. 8 2, 6, 10 St. & L. Do 3, 932. 1 Rch. Trehalose C oH^O 3, 947. 2, 10, 6 St & L Trehalose hydrated C jHO H O 3, 550. 3 2, 10, 6 St. & L. Rafflnose (melitriose) C ~H~O 3, 928. 10 St. Do C aHwO 4,020.0 2,6,10 B & M Do C H.iO 4, 020. 8 2,6,10 St. it L. Raffinose (melitriose) hvdrated H- O "(HoO) 3, 400. 2 2,6 St. Do " C H iO s (HoO) 3, 399. 1 10 St. & L Melitose C HO " 3, 880. 4 St. Do c "H jO 4, 122. 2 5 St Melezitose . . .... C^H^'O - 3, 913. 7 2,6,10 St. it L. Glvcogen (C-H O )n 4, 190. 6 2,6 St. & L. Cellulose (CH ' O )n 4,200.0 2,4 B & V. Do . (c'-II "o )n 4, 185. 4 2, (>, 10 St. & L. Do i C"H "o In 4, 146. 1,2,4,10 St Do (c'-H O )n 4, 152. 9 6 Go Do (CHJoOs)n 4, 165. 10 Go. Do (C 6 Hi O5)n 4, 205. 10 B. it V. Dextrin (dextran) (CHi,,Os n . 4, 180. 4 2,10 B. & V. Do cities:.::: 4,112.3 2, 10, 6 St.it L. Starch 4, 228. 2, 4, 10 B & V Do (CoHloOs n . 4,182.5 2,6,10 St. & L. Do (C H 10 O 5 n 4, 123. 2,4,10 St. Do . 4,164.0 2,6,10 GI Do (CgHinOj n 4,479.0 1 Rch. Do 4,116.0 1 St. Inuliii (CcH^O.-)!! . 4, 187. 1 2,10 B & V. Do (C,)H,,,O-,)n 4, 133. 5 2 St. & L. Do (C (1 H 10 Or 1 )n 4 070 2,4,10 St Do 4, 133. 5 6,10 St. ALCOHOIJS. Mcthvl alcohol (liquid) CH 4 O . . . 5 Do CH 4 O . 5, 307. 1 10 F. &S. Do CH 4 O 5,321.5 10 St. & L. Methyl alcohol (gas) CH 4 O . 3, 693. 7 10,5 Th. Kthvl alcohol (liquid) c,H fl o 7,068.0 2,10 H. \ M. bo CjH 6 O 7,183.6 3, 10, 1 I-'. >V s. Do 6,980.0 1 B. Do ( ..H,o 7,095.0 9 A. Do C.,HO 7 044.0 5 St Kthvl alcohol (gas) (H<,O .... 7,402.2 10 Th. Do 7, 321.7 10 B. & M. Do i ' II o 7 394.1 5,7 I'rojivl alcohol (liquid) Do CH0 < H.o 8,010.3 8,005.2 5 10 St. Lu. 1'rojiyl alcohol (gas) OJH B O 8,310.0 10 Th. CjHjO 7 301.7 5,7 A myl alcohol 10 F. ,t S. Do r ll.-.o 9,021.8 in Lu. Do II 9 005.7 5 St. 'ilvccriu < II o (4 112 4) St A I. Do I :UL> I 10 SI A 1. Do ... ' II o I ::17 o _', 10 St. Do 1 II 1 ) 4 305.0 1 St. Do C,H,O, .... 10 1,11. Krvlhrit ' Do ... 4,075.0 1 II" ~> 10 10 St. I.u DM ( 4 ij 'o' 4 117 6 10 It A M Do CJBioOj I i:u :: 10 .-1 ,V I..A K. Do <',.. 4.132.3 10.6 St. & L. EARLIER DETERMINATIONS. TAHLE J. Heats of combustion of J-O 4 678.6 Do C.,H.,O 4 672.5 Benzoic acid C 7 H Oo . 6,360.5 Do C 7 H 6 O 2 6,322 3 Do C 7 H O.. 6,322.1 Do C-H 6 Oo 6, 281. Do 0-HoO." - - -- 6,315.0 "Do C-H 6 O 2 6,345 Do C-H 6 O 2 (7,663.0) Salicylic acid ( Ho . 7,236.0 Do Do C.HO S 5,286.2 C 7 H 8 O 8 . 5, 326. HYDROCARBONS, ETC. Mrtlmne .. CH 4 .. 13,063.0 Do CH 4 13, 218. 9 " Do CH 4 .. 13,243.7 Do ... Ctt... 13.275.0 16 INVESTIGATIONS IN USE OF BOMB CALORIMETER. TABLE I. Heate of combustion of 1-gram substance, expressed in small calories Cont'd. Substance. Formula. Calories. Refer- ences. Investiga- tors. HYDROCARBONS, ETC. continued. Ethane 12,144.2 7,5 Do C~H- 12, 346. 7 10 Th. Do C."H 12, 991. 7 10 B. C- H 12, 027. 3 10 Th. Do C 3 H 8 .. 12, 543. 10 B. Do C S H 8 12, 582. 5,7 \cetylene 11, 919. 7 7 Th. Do CHo 11,923.0 10 Th. Do CH~ 12, 112. 10 B. Kthvlene C."HJ 11,858.0 3,10 F. & S. Do . . CoH* 11,894.4 7,5 Do CH 4 .. 11,883.6 10 Th. Do cX 12, 154. 10 B. 11, 730. 10 Th. Do 0- H- 12,045.0 10 B. Do C- H 11,717.0 5,7 Benzol (benzene). C-H- 9, 949. 10 B. Do C-H- 9,977.5 10 St. & L. Do C-H 9, 997. 10 St.. Ro. 67 Mannite .. . . 5 2 33 Hippuricacid 3 .46 Acetic acid 6 2.11 Asparngin 3 3.31 Palmetic acid 5 1.46 Unc: acid 4 4.83 Stearic acid.. . 5 4.80 Fut of swine 5 3.16 Oxalic acid 5 16.08 Fat of oxen . . 4 3 40 Benzoic acid 6 1.25 Butter fat 6 .57 Methane 4 1.60 Olive oil 7 1 06 3 .48 Arabinose 3 73 Benzol gas 3 2.02 Galactose 2 1.68 Naphthalene 6 1.58 Dextrose 6 1 86 Camphor 3 .12 In chemical-laboratory work, whether it be by advanced students or professional chemists, it is required that the work of the analyst shall be within certain limits of error, varying, naturally, with the nature of the substance anatyzed and the method employed. No fixed rule can be laid down as to the limits of error allowed, but for most of the more common substances such as oxides, carbonates, many ores and clays, etc. some of which have man}' ingredients to be 18399 No. 9407 3 18 INVESTIGATIONS IN USE OF BOMB CALORIMETER. determined, the total of the determinations must come, barring excep- tional cases, within, say, about 99.75 per cent and 100.5 per cent, in order to be accepted. The extent to which the results obtained by different anal3 T sts on a single ingredient are allowed to vary is very variable. In a determination like that of total nitrogen, for example, the substance containing 18 per cent nitrogen, two analysts should come not farther than about 0.05 or 0.06 per cent apart in the results obtained, except for special reasons. Such a variation as, for instance, between 17.97 per cent and 18.03 per cent nitrogen would be equal to 0.33 per cent difference between the analysts. Of the 44 cases just cited in Table II the difference between the results obtained by different investigators is below 1 per cent for only five substances. With most of the substances the analysts are several points apart, and in urea and oxalic acid there are difference's of over 1(> per cent. That two determinations by the same anatyst agree when the same quanta^ of the sample is used, under exactly the same conditions and manipulations, is no guaranty that the method is perfect or the result correct. If, on the other hand, results agree when charges varying in weight have been used under somewhat varying conditions within the working limits of the method, we feel more sure about their being correct. CAUSES OF DIFFERENCES IN HEAT DETERMINATIONS. The method for the determination of heat values of substances can not be considered perfect until work can be done as accurately as is possible in ordinary chemical anal3 r ses at least. From the heat values given we are led to question: What can have mused these differences in the past? What causes differences at present '. Following the directions given in W r iley's Principles and Practice of Agricultural Analysis, Volume III, a number of determinations had been made in this laboratory by means of the apparatus already cribed, using the commercial ox 3* gen gas found on the market, put up in iron cylinders under very high pressure. The water equivalent of the whole bomb S3 r stem had been determined by burning cellulose and cane sugar in oxygen under 'JO atmospheres 1 pressure in the bomb. 4,li>r> calories per gram being the heat value used for the cellulose and 3,065.2 calories per gram for the sugar. The average of a large number of determinat 5on< \\-.i<\ given the value as 43< grains, and since exactly 2,000 grams of water was used in all cases, the factor wa* 2,430. Apparently there was no diflieulty in obtaining concordant results when like charges were burned, but with varying charges the work not -ati-t'actorv. The analvtical error with a small charge mav of CAUSES OF DIFFERENCES TN HEAT DETERMINATIONS. 19 course magnify the percentage error, but, excluding this possibility, it should be possible with a small charge to obtain results indentical with those obtained from a larger charge of the same material. To say that 4,000 to 5,000 calories should be generated in the bomb in order to give good results, and that with only 2,000 calories the work is not satis- factory, is to say that the entire method is unreliable. IMPURE OXYGEN. Some difficulties which were experienced in elementary carbon and hydrogen determinations in which the same oxygen was used led to the examination of the oxygen for impurities in the form of combus- tible gases. It was found beyond a doubt that such impurities were present in the oxygen in quantity sufficient to be reckoned with, and the magnitude of the error caused by them will be referred to further on. Now arose the questions: To what extent do these combustible gases influence the working of the bomb calorimeter? Is the water value of the bomb, determined by the use of this impure oxygen, correct? Can a correction and the water value be worked out correctly? TESTING OF OXYGEN. In order to give the apparatus a severe test and to try to work out an answer to some of these questions, using the impure oxygen itself for the purpose, the following method was adopted: Cellulose was chosen as the chief substance to be burned, and sugar in the-- form of rock candy was also used. By burning small charges of pure cellulose absorption blocks in the bomb calorimeter, charged with oxygen at various pressures, it was hoped that any increase found with the greater pressure of ox\"gen would represent the total impurity in the additional amount of oxygen equivalent to the increased pressure, and that, having ascertained the correction for impurities, it could be applied and the water value of the bomb worked out without difficulty. Cellulose, although a very hygroscopic substance, and on that account difficult to handle, was chosen for this work because it ignites readily and burns quickly and completely in a comparatively small excess of oxygen. It represents the main class of substances whose heats of combustion are determined in our work, and its heat value is fairly well established. Sugar, on the other hand, is a substance which does not ignite readily and which burns more slowly, often leaving a trace of unburned carbon even with plenty of oxygen. It was soon learned that the combustible gases in the oxygen were not oxidized completely, and perhaps never can be fully burned in a bomb calorimeter, but the tests were carried out as planned. The problem was not so simple as at first thought, but it was hoped that the oxidation of the gases referred to would be in proportion to the density of tho gas mixture or the heat generated in the bomb. This 20 INVESTIGATIONS .IN USE OF BOMB CALORIMETER. is by no means a perfect method, but it may be a means of helping over the difficulty when no pure oxygen or other means for compari- son are at hand. One lot of oxygen (Oxygen I) was tested with both cellulose and sugar in the manner described. A second lot (Ox^ygen II) was tested only with cellulose. In every instance the cellulose was dried at between 103 and 105 C. to constant weight. The effect of the impurities may be seen in Tables III to VII, where the corrected rises in temperature will indicate the differences due to more or less oxygen and to the amount of heat generated in the bomb. But before we discuss these results we may examine Table III, where we find the heats of combustion of the various substances worked out as in the course of ordinary routine work, using the water value 430 for the tomb 83 T stem. This value plus the 2,000 grams of water used makes the factor for the bomb system plus water equal to 2,430 grains. The column headed u Total computed calories" refers to all the heat generated in the bomb, and includes the heat represented by the substance and the iron wire burned, the nitric acid formed and dis- solved, and the heat due to the electric current whenever it failed to ignite the substance instantly. Varying amounts of oxygen will change the water value of the sys- tem : hence in all cases where more or less than 20 atmospheres oxygen was used the temperature rises have been corrected to 20 atmospheres oxygen pressure. For this bomb of about 360 c. c. capacity the correction for 1 atmosphere oxygen is 0.0046 per cent of the observed rise in temperature. Thus, taking for example the first rise referred to in Table III, the rise due to the burning of 0.5092 gram sugar in 10 atmospheres oxygen, corrected to 20 atmospheres oxygen, will be: c. Obwnvd rise in temperatuiv. 0. 8488 Correct inn for 10 atmospheres oxy calorimeter and heat value ii-orkul out l,y iiftiii'/ the inil/'f value 430." Substance. Lot of oxygen used. Oxygen pressure. Weight of substance. Total com- puted. Rise cor- rected for oxygen, 20 atmos- pheres. Bomb water value. Calories per gram. i Sugar 1 AtnioKiiheres. 10 Until!.*. 0.5092 Calories. 2, 050. 26 a 0.8484 416.6 3,976.0 t Cellulose I 10 24 24 24 10 .5019 .5054 .5096 1.0000 .3799 2, 013. 96 2,038.32 2, 046. 35 4, 106. 14 1,617.47 .8352 .8493 .8558 1.7037 .0671 411.3 400.1 390.1 410.0 424.6 3,983.1 4,005.4 4, 020. 2 3, 986. 6 4 194.4 Cell u lost- 11 10 10 15 20 20 20 20 10 .3840 .3711 .3789 .3800 . 3737 .3714 .6800 .4358 1,628.16 1,571.58 1,604.70 1,609.68 1,584.11 1,574.21 2,868.0(5 1,802.34 .6746 .6484 .667H .6734 .6604 .6558 1.1896 . 7381 413. 5 423.8 403.1 390.4 398.7 400.4 410.9 442.0 4,213.8 4,193.3 4, 232. 3 4, 255. 2 4, 240. 2 4,237.2 4,218.3 4 164.2 Naphthalene II 10 24 24 24 24 20 .4257 . 4253 .4276 .6675 1.0035 . 5274 1,803.55 1,824.37 1,815.50 2, 817. 39 4,229.14 6 108 42 .7395 .7489 . 7450 1.1579 1. 7492 2 1111 438.7 435.9 436.8 133. 1 417.7 419 8 4,169.8 4,174.5 4, 173. 4 4,179.5 4,206.4 9 668 9 Camphor II 20 20 .5430 4671 5, 255. 04 4 360 28 2. 1785 1 7985 412. 2 424 4 9,699.3 9 31] 6 Benzoic acid II 20 20 .5034 .6986 4,703.37 4,438.67 1.9409 1.8390 423.3 413 6 9, 315. 8 (> 365 1 20 .7027 4, 473. 37 1.8520 415.4 6,360.4 a No correction for impurities in oxygen is applied. These determinations, too few in number for very good averages, were made now and then as other work permitted; hence the}' cover months of time. A hast}' glance over these figures would be enough to condemn the apparatus as being useless for accurate work or the investigator as being unskilled or careless. Of course the analytical errors of the small charges will, as already said, be magnified; hence the duplicates may not always agree closely. Looking a little closer at the table, we notice that all the different groups that is, those which were treated alike as to quantity of substance and the amount of oxygen agree fairly well, thus proving that the differences between the groups are not due to careless manipulation. The reason for the disagreements must be sought for elsewhere. But how are we to interpret the figures, which, in some cases, appar- ently contradict each other? In the column giving calories per gram we find in the case of cellulose an extreme difference of over 2 per cent in the heat of combustion. Further, we notice that 0.5 gram sugar burned in 10 atmospheres oxygen gave as high results as when 1 gram sugar was burned in 24 atmospheres oxygen. Again, in the case of cellulose in Oxygen I, the smaller charges at 20 atmospheres oxygen gave higher results than the larger charge at the same oxygen pressure, whereas in the case of Oxygen II the opposite is shown. Another very marked peculiarity in the results is that, with the same factor for the bomb, etc., and the same kind of cellulose, in the work with Oxygen I, all the results on both sugar and cellulose are above the accepted values sugar as much as l.fi per cent and cellu- 22 INVESTIGATIONS IN USE OF BOMB CALORIMETER. lose up to 1.04 per cent. In Oxygen II, on the other hand, all the cel- lulose results, except the one where over 1 gram cellulose charge was used, are below the accepted heat value; but the naphthalene, camphor, and benzoic acid are all above naphthalene about 0.7 per cent, cam- phor about 0.3 per cent, and benzoic acid about 0.7 per cent. It should be mentioned here, however, that these substances, while bought as chemically pure, were not tested for impurities before using. Looking at the column giving the water values of the bomb, we lind a corresponding variation, most of the values being below the one used (430), ranging an}- where from 390 to 442. If a general view is taken of the various results and the behavior of the substances under ditfererit conditions, we are convinced that the ox3'gen contains considerable impurity in the form of combustible gases, for which corrections must be applied. This being the case, the old, apparently too high, water value (430) of the bomb will not be high enough. CORRECTION FOK IMPriHTY IN THE OXYGEN. Is the correction a constant quantity? Is it of any value to know the total amount of carbon and h3 7 drogen present as combustible gases in the oxygen ( These are some of the questions which now concern us as we endeavor to work out the water value of the bomb system, applying the corrections as found for the ox3 T gen and for the amount of heat generated in the bomb. We turn our attention to Tables IV to VII and examine first the work with cellulose, Oxygen I. From Table IV we learn that between the averages computed to be due to like charges burned in ten atmos- pheres oxygen pressure and 20 atmospheres there is a slight differ- ence in the rise in temperature in favor of the greater amount of oxy- gen. This difference does not represent the total impurity in that amount of oxygen, but only the amount which will be oxidized under the specific conditions of so much cellulose and so much oxygen. The difference in rise in temperature, when about 1,600 calories were gen- erated in the bomb, was o.ooooo;) C. for 1 atmosphere oxygen. TAHI.K IV. R\y> in tnn/ierdtnn riitixi'il ft;/ InirniiK/ /if.'t' 17 12 1 00 fr"V |l, 0749 6746 0029 10... 3711 16 48 2 00 'f>71 f>H C, |S7 6484 6601 16 1789 16 00 3 00 CiOl 70 6679 665S 20.. . 8NOO 15 f'H ". To I'.ii'i i> 67H4 6734 08M n ::7.;7 15.66 1.50 .584. 11 MM .6604 . r,r,7o ... .3714 it;. (HI 3.90 .6558 .6558 .6665 CORRECTION FOR IMPURITY IN THE OXYGEN. 23 1 ,600 calories in 20 atmospheres oxygen pressure 0. 66763 1 ,600 calories in 1 atmospheres oxygen pressure 66098 Difference 00665 Rise per atmosphere oxygen pressure 000665 Whether this error is uniform throughout that is, whether the eleventh atmosphere oxygen, for instance, would show the same error as the twenty-first atmosphere with the same charge can not be ascer- tained from these figures. At present we can only assume the error to be uniform. Is this then the true correction to apply, no matter how large a charge was burned, or will more of the gases be oxidized in the same quantity of oxygen when more heat is generated? Rock candy, Oxygen I. In Table V we find a difference in the rise of temperature between 10 and 24 atmospheres oxygen pressure equal to 0.00045 C. per atmosphere when about 2,000 calories are generated in the bomb. TABLE Y. Rise In temperature caused by burning like quantities of rock candy in the bomb calorimeter in varying amounts of oxygen (Oxygen 7). Oxygen pressure. Rock candy. Fuse wire. HNO 3 . Total com- puted. Rise in temper- ature. Rise cor- rected for oxvgen, 20 atmos- pheres. Rise due to 2,000 calorics. Atmosphere*. 10 Grams. 0.5092 Calories. 18.08 Calories. 3.25 Calories. 2,050.26 C. 8488 C. 0.8484 C. 0. 8276 10 .5019 16 80 3 15 9 013 96 8356 '8352 8293 24 5054 19 52 4 10 2 038 32 8491 8493 8333 24 50% 13 16 4 50 2*046 35 8556 8558 8364 C. rise. 2,000 calories in 24 atmospheres oxygen pressure 0. 83481 2,000 calories in 10 atmospheres oxygen pressure 82847 Difference 00634 Rise per atmosphere oxygen pressure 00045 Sugar, as has been said, does not always burn completely. Fre- quently a hardly weighable amount of carbon ma} r be seen on the platinum capsule, a very small spot perhaps, yet it is an indication of incomplete combustion of the carbon itself and of the possibility that at least traces of intermediate partially oxidized products may !> found. Cellulose, OxygenIL In Table VI we find the results obtained with cellulose and another lot of oxygen. Here 1,800 calories produced in oxygen at 10 and 24 atmospheres pressure, a difference in temperature of only 0.00007 (\ per atmosphere, which is within the limits of analytical error. 24 INVESTIGATIONS IN USE OF BOMB CALORIMETER. TABLE VI. Rise in temperature caused by burning like quantities of cellulose in the bomb calorimeter in varying amounts of oxygen (Oxygen II). Rise cor- Total Rise in rected for Rise due Oxygen pressure. Cellulose. wire. HN0 3 . com- puted. tempera- ture. oxygen, 20 atmos- to 1,800 calories. .* pheres. Atmospheres. Uranits. Calories. Calories. Calories. C. C. C. 10 0.4258 17.12 3.25 1,802.34 0.7384 0.7381 0. 7371 10 .4257 19.20 2.80 1,803.55 .7404 .7395 .7386 24 .4253 21.12 5.00 1,824.37 .7488 .7489 .7389 24 .4276 21. 44 4.55 1,815.50 .7449 .7450 .7387 C. rise. 1,800 calories in 24 atmospheres oxygen pressure 0. 73881 1 ,800 calories in 10 atmospheres oxygen pressure "... . 73778 Difference 00103 Rise per atmosphere oxygen pressure 00007 We have reason to believe, however, judging from the determina- tions referred to in Table III that this oxygen supply also was not pure, and hence we take up the question asked before: What effect has increased heat production upon the oxidation of these combustible gases ? For some of the following comparisons 1 must depend upon single determinations, which in itself is not very satisfactory; but since no more of the same oxygen is at hand for further work I must take for granted that the determinations are correct. TABLE VII. Comparison of observed ri*e in temperature due to the large and the calculated rise due to the smaller charges of a substance burned in the same quantity of oxygen. Substance. Lot of oxygen used. Oxygen pres- ' sure. Weight of substance. Computed total. EUae. SiiKiir I Miinif ^heres. 24 *>.. so 1.1967 Calculated. 1,278.76 - .0071 Ditlerence. ( 'el 1 11 lorn- 11 21 1 0035 4 229 14 1 7492 21 .4265 1,819.93 1.7359 Calculated. 2, 409. 21 f .0133 Ditlerence. Jl M . tK.7.-. 1966 2,817.39 1,819.98 1.1579 1. 15t>4 Observed. Calculated. 997.46 f- .0015 Difference. In Table VII we find that when I gram of sugar was burned in 24 atmospheira oxygda, 4,106.14 calories of total heat being generated in the bomb, the rise in temperature of the bomb .system was found to be 1.70366 C. The average rise in temperature found when 0.5075 gram of sugar \va- burned, or -JJ>4'J.:'>: J calories of total heat was gene- COMPARISON OXYGEN TESTS. 25 rated in the bomb, in the presence of 24 atmospheres oxygen, was 0.8525 C., and according to that the generation of 4,106.14: calories should have caused a rise of 1.71390 C. instead of 1.70366 C., as found when 1 gram was burned. The difference, therefore, is a minus of 0.0102 C. This indicates that, of the impurities in this particular oxygen, more were oxidized in proportion as the charge of sugar was reduced. Hence the correction referred to in Table V should not be increased in proportion to the increase in the amount of substance used. With cellulose in Oxygen I a charge of 0.6800 gram, which, plus wire, etc., was equivalent to 2,868.06 calories total heat generated, gave a rise of 1.1896 C. instead of 1.1967 C. as it should have done according -to the results upon the smaller charges. Here is also a minus difference of 0.0071 C., showing that this oxygen behaved the same with these two different substances. The combustible gases present in the oxygen must accordingly be of an easily combustible nature, comparatively speaking, when the increase in heat has no effect upon them above that of the lesser heat. Next we examine the results obtained with different amounts of cel- lulose in Oxygen II. Here we see not a decrease, but a very marked increase in the oxidation of combustible gases with increase in heat evolved. A charge of 0.6675 gram cellulose, or a total heat produc- tion of 2,817.39 calories, gave a rise in temperature of 1.1579 C., which is 0.0015 C. above what it should have been according to cal- culation from the results of the smaller charge of 0.4265 gram cellu- lose, equal to 1,819.93 calories total heat. When 1.0035 grams cellulose was burned there was a difference of 0.0133 C. above the calculated result. This difference in behavior of the two oxygen sup- plies toward larger charges of the substances burned indicated a dif- ference in composition of the combustible gases in the oxygen. In Oxygen I the gases were apparently more readily oxidized than in Oxygen II, and therefore more of them burned with the small charges in proportion to what burned with the larger. The opposite would then be true of Oxygen II. Unfortunately no further and more defi- nite proof can be given, since no qualitative tests were made of the gases. As to the impurities in Oxygen I being different from those in Oxygen II, there can be no doubt. In the first lot a strong, pecu- liar, disagreeable odor was noticed, and when the gases passed through pumice stones saturated with H 2 8O 4 a slight yellow coloration was noticed for a short distance. Oxygen II had no such disagreeable odor and did not color the pumice stone as did Oxygen I. According to anal} r ses made by one of the assistant chemists of the experiment station, Oxygen I contained 0.0355 per cent of hydro- gen and 0.0150 per cent of carbon by weight, while Oxygen II con- tained 0.0335 per cent of hydrogen and 0.0150 per cent of carbon. 26 INVESTIGATIONS IN USE OF BOMB CALORIMETER. These results show that there must have been considerable free Irydrogen present. Not knowing the composition of the gases referred to, we may, for the sake of obtaining at least an approximate calorific value, assume that all the carbon is present as marsh gas, and that the excess of h3 T drogen is free hydrogen. Thus we have in Oxygen I 0.02004 per cent CH 4 and 0.01446 per cent free H 2 , and in Oxygen II 0.02004 per cent CH 4 , and 0.01346 per cent free H 2 . With 20 atmospheres oxygen, or in round numbers 10 grams of oxygen, these gases would represent in Oxygen I 73 calories, equiva- lent to 0.0299 C. rise, and in Oxygen II 69 calories, equivalent to 0.0283 C. rise. In none of the determinations referred to with the above-mentioned oxygen was more than about 45 per cent of these figures reached, which can be seen in Table IV, where the correction for the impurities in 20 atmospheres oxj'gen equals 0.0133 C. This indicates plainly that only a part of the combustible gases mixed with the oxygen can be oxidized in the bomb. RISE IN TEMPERATURE CORRECTED FOR HEAT DUE TO IMPURITIES IN THE OXYGEN. In Table VIII following are given the corrected figures for the rise in temperature, also the values computed for the bomb water value and for the heats of combustion of the substances, respectively. The cor- rected figures for the rise in temperature are obtained by using the results of Tables IV to VI. Thus the corrections used for the sugar and cellulose burned in Oxygen I are the differences found in Tables IV and V. No account has been taken of any increase in total heat formed in the bomb by use of larger amounts of material with Oxygen I. For the smaller charges of cellulose burned in Oxygen II the ditl'er- ences found in Table VI have been used, and for the larger charges the increase in rise of. temperature actually observed (Table VII) has been added to the first-mentioned correction. Thus, where 0.4253 gram cellulose was burned in Oxygen II the correction will be 0.00007 X 24=0.00168 C. For the charge of 0.6675 gram cellulose the correc- tion will be 0.00168+0.0015^ C., and for the largest amount of cellulose burned the correction equals 0.0151 C. To all determinations of the other substances naphthalene, cam- phor, and benzole acid where the total heat was a little higher than it was with the cellulose charge last mentioned, the same correction, 0.0151 C., has been applied. The heats of combustion were computed by using the bomb water Tftloe 489.2. This value is the average of those found for cellulose, naphthalene, camphor, and benzoic acid with Oxygen II. The average of all the results in Table VIII would be437.s, but the values obtained FORMATION OF NITRIC ACID. with Oxygen I were not used on account of the sugar and cellulo.se not agreeing very closely. TABLE VIII. Water value of the bomb calorimeter, and heat value per gram substance, worked out by using the new water value, 439.2, found after applying the corrections for the impurities in the oxygen. Substance. Lot of oxygen used. Oxygen pres- sure. Weight of sub- stance. Total com- puted. Rise cor- rected for oxygen per 20 atmos- pheres. Bomb water value. Calories per gram. Sugar I Atmospheres. 10 Uramx. 5092 Calories. 2, 050. 26 C. 0. 8439 429 5 3 9(19 7 Cellulose I .10 24 24 24 10 .5019 .5054 .5096 1.0000 .3799 2,013.96 2,038.32 2.046.35 4,106.16 1,617.47 .8307 .8385 .8450 1.7031 C605 424.4 430.9 421.7 411.0 449 3, 976. 4 3, 969. 3, 984. 2 4,001.7 4 188 9 Cellulose II 10 10 15 20 20 20 20 10 .3840 .3711 .3799 .3800 .3737 .3714 .6800 .4258 1,628.16 1,571.58 1,604.70 1,609.68 1,584.11 1,574.21 2, 868. 06 1,802.34 .6678 .6418 .6578 .6601 .6471 .6425 1.1834 .7373 438.1 448.7 439.5 438.6 448.0 450.2 423.6 444 5 4,186.9 4, 168. 7 4,184.8 4, 158. 6 4,170.1 4,166.4 4,212.6 4 178 Naphthalene II 10 24 24 24 24 20 .4257 . 4253 .4276 .6675 1.0035 5274 1, 803. 55 1,824.37 1, 815. 50 2, 817. 39 4,229.14 5 108 42 .7388 .7472 .7433 1.1548 1. 7341 2 0960 441.2 441.7 442. 5 439. 7 438.8 437 2 4, 181. 5 4,180.5 4, 179. 4,184.1 4, W6. 7 9 635 fi Camphor II 20 20 .541% 4671 5, 255. 04 4 360 28 2. 1634 1 7834 429. 1 444 9 9,668.3 9 268 1 Benzole acid II ... 20 20 .5034 .6986 4, 703. 37 4,438.67 1.9258 1 S239 442.3 433 6 9, 277. 9 6 336 6 20 .7027 4,473.37 1.8369 435.3 6,332.1 The corrected results in Table VIII agree fairly well among them- selves, and even where the conditions for determinations differed so widely from the ordinary they approach the heat values generally accepted for the substances used. This certainly could not be said when the impurities were not considered, as was seen in Table III, where a higher value was used than the average of the water values given in that table. From all that has been said on this topic we learn that, at best, the determination of the hydrotherrno equivalent, or water value, of the bomb with impure ox\^gen is no easy task, nor can it be very satisfactory. Hence the method explained can be recom- mended for use only where no pure oxygen i. e., oxygen free from combustible gases can be obtained. FORMATION OF NITRIC ACID IN THE BOMB CALORIMETER DURING COMBUSTION. The formation of HNO 3 in the bomb has already been alluded to, also that correction must be made for the heat represented by the HNO 3 solution. In the bomb there is always present more or less free nitrogen, and a portion of it is always oxidized, varying in quantity according to the nature of the substance burned, the total heat generated, and the quantity of nitrogen in the bomb. This quantity of HNO 3 , which 28 INVESTIGATIONS IN USE OF BOMB CALORIMETER. appears to be constant for specific conditions, 1 propose to designate, for convenience, "bomb nitric acid," meaning thereby the HNO 3 , expressed in calories, which would be formed in the bomb from free nitrogen during the combustion of a specified substance under specified conditions. Table IX, following, gives us an idea as to the value of this quantity for the oxygen used. This quantity, however, repre- sents only a very small portion of the free nitrogen present in the bomb. TABLE IX. Formation of HNO 3 in the bontfj calorimeter (hiring combustion of nomnitrog- enous substances. Substance. Determi- nations. Oxygen. Weight of substance. Total heat. HN0 3 . HNO :i per 1,000 calories. Cellulose Number. 6 .{tiiiiinjilieres. 24 Grams. 0.5481 < 'nlorii's. 2, 335. 2 ( 'mir> (trams. 1.0236 .8809 .9133 7. !Mt!2 ti 121C, C..OMI 7.6021 7. 81 34 - HIT r. r. i,* '.'1 59.79 B| 1',-j 1'.' '.M 1.60 i IB i. as 2. 92 A. I'.t 3.36 c. c. 58. HH 58. 27 M M 1.1 3.1 1 'nil,,-!. X. .'.H U88 II 10 3.97 6. H7 4,052.4 1,048.7 1,066.8 1 ,:; :;.; 175.6 4,118.0 4,125.7 4,182.7 '., L"J1 . -1 153.42 176. 73 Per cent. I . is 1.87 l.fiC .79 .04 .07 Do Do Hair Trim- IK. Do... Do Do IK, CAUSE OF INCOMPLETE COMBUSTION. 33 These few determinationvS will suffice to show that a very large error may be introduced by the H 2 SO 4 . In the case of urine the error intro- duced by reckoning sulphuric acid as nitric acid is small, but with the hair it reached about 0.8 per cent, and in the case of the epidermic tissue, dandruff, etc., which is richer in sulphur, the error was in one instance about 1.9 per cent. CAUSE OF INCOMPLETE COMBUSTION. The shape and size of crucible, or capsule, in which the substance is burned may influence the combustion. Owing to inability to watch the process of the combustion in the bomb, the general opinion, gained perhaps from the observations of explosives, is that the combustion in the bomb calorimeter is very sudden and violent, in nature like an explosion in the free air. This, however, is not correct. The time of combustion varies with the nature of the substance, but, with the materials mentioned in the previous tables, only by a very few seconds. It is of a very short duration, but by no means an instantaneous flash. The supposed vio- lence, implying that the flash fills the bomb, has absolutely no foundation and would not happen except when the oxygen contains a considerable percentage of highty combustible gases. To prove the correctness of this statement direct tests were made as follows: Small pieces of filter paper, about one-fourth inch square,, were fastened by one edge to the sides, top, and bottom of the bomb by means of paste, and allowed to dry. In one trial the burning of a small charge of cellulose did not affect the few pieces of filter paper which were fastened on the top and on the bottom of the bomb. After that a more complete test was made by fastening three pieces to the top, one at the center of the bottom, and ten distributed on the sides of the bomb; four of them about opposite the top of the capsule, four nearer the top, and two nearer the bottom. The charge was 1 gram powdered rock candy plus a little naphthalene, burned in 20 atmospheres oxy- gen. The test resulted in two of the three pieces on the top being burned, the third one showing in one corner a trace of brown as when paper has been held near strong heat. All the rest were absolutely untouched. Hence, the flames may at times reach the top, and per- haps more seldom the sides, though the distance is much shorter. AVith more oxygen the flame very likely would not even have reached the top of the bomb. It is a known fact that tire will put itself out in a closed room, even before the oxygen is fully consumed. If this is true in a room, might it not also be true in a deep capsule surrounded by plenty of ox}*gen ? To have some light shed upon this question, and to be able to decide in favor of some one of the different shaped capsules used, the follow- 34 INVESTIGATIONS IN USE OF BOMB CALORIMETER. ing tests were made: A wide-mouthed bottle of 1,500 c. c. capacity was filled with oxygen, and the crucible, attached to a wire fastened in the cork, was let down in the bottle, and the substance ignited by means of an iron fuse wire as in the case of the bomb combustion, and the burning watched. The substances burned were filter paper (cellu- lose), cut in disks and laid on the bottom of the crucible, and powdered sugar (rock cand}'). After combustion the ordinary test was made for the oxygen by means of a glowing splinter of wood. In all cases there was a large excess of oxygen, since the large cork could be removed and the splinter plunged in and ignited several times in suc- cession. In the following table, which gives the results of the tests, crucible No. 1 was a large nickel crucible, If inches in diameter at the top and 1 T \ inches deep. Crucible No. 2 was a small nickel crucible, inch in diameter by \ g-inch deep. Crucible No. 3 was the same as No. 2, per- forated, and crucible No. -i was the same as No. 3 with the holes enlarged. A small piece of cellulose, not weighed, was put under the sugar in the crucible in one case to facilitate the combustion. The time which the substance stood in the bottle before ignition was noted, as well as the length of time it continued to burn. The last column in the following table gives the unburned residue: TAI-LK XIII". Powdered sugar and cellulose burned indifferent shaped crucibles in oxygen under atmospheric pressure. Substance. Cruci- ble. Weight of substance. Time be- fore igni- tion. Time of combus- tion. l"nl)iiriu'l residue. Cellulose 1 Gram. 0. 2190 Minutes. 80 Seconds. 60 Oram. 0.0000 Sugar. . . 1 .2675 5 75 .0077 Do 2 . 2454 5 380 .0081 Cellulose 2 .1890 5 165 .002(5 Sugar . .... 3 2582 8 195 .0122 Do 4 .2132 90 75 icon Sugar HI xl cell ii lose 4 .2037 5 00 .0060 Cell ulose 4 ! :,s;, 15 40 .0002 In only one instance did the substance burn without leaving any unburned residue of carbon. There was a marked difference between the two substances in the way they burned, the sugar taking longer time and leaving a great deal of unburned matter. A number of holes were made near (he bottom of the small crucible. During the burning some of the holes became coated over with a film of charred material, tin- flame remained smaller, and the circulation of the leases, i. e., of oxygen, to the burning substance was lessened. When the holes wen- enlarged these last-mentioned hindrances were magnified and the sugar burned only in part, most of it was left untouched by the small flame, and in one spot it was not even fully melted. It is possible that in this ca-e -nine of the oxygen might have leaked out on standing, I nit after ALCOHOL HEAT VALUE. 35 the combustion there was enough left in the bottle to allow the glowing splinter to burst into flame several times. When the sugar rested on a small piece of filter paper the combustion of the sugar was more nearly complete, and filter paper alone was practical^ completely burned in crucible No. 4. The figures of the table, as well as the actual behavior of the com- bustion, indicate the shape of the larger crucible, which was very much broader at the top than at the bottom, as being the best. The combustion itself, or we may say the supply of oxygen to the burning points, was more uniform with the larger and more intermittent with the smaller crucible. A good start, and much heat quickly generated, are factors which undoubtedly contribute much to make a combustion complete, but these may not in the least change the conclusion reached as to the shape of the crucible or capsule best suited for the work. ALCOHOL HEAT VALUE. Time did not permit me to take up the question of alcohol heat value as fully as I desired, hence at this time 1 shall make mention of only a few points which have come up in connection with the determi- nation of heat of combustion, and make some suggestions which may at least be of practical value. Alcohol being a volatile liquid it can not be satisfactorily burned in an open dish, but must be inclosed in a receptacle from which it can not evaporate, and gelatin capsules of known heat value are generally used for this purpose. The alcohol charge should never be weighed out by difference i. e., by weighing the bottle but the actual amount placed in the capsule should be weighed. The reason for this is to make sure that there is no evaporation. Unless the capsule with the alcohol is weighed a little evaporation may not be noticed, and if the capsule is not tight enough to hold the alcohol while it is being weighed it is not fit for the work. A little evaporation of alcohol from the capsule in the bomb will always cause the results to come too low, since alcohol vapors are not fully burned. To insure complete combustion of the alcohol and the gelatin the capsule may be filled with clean, ignited asbestos to absorb the alcohol. Another way to obtain good combustions of the alcohol and gelatin, though not so simple as the gelatin-capsule method, is to make use of a tubular platinum capsule, about one-fourth of an inch in diameter, having perforated sides and bottom. This platinum capsule is coated with a mixture of the best gelatin with about 10 per cent of glycerin and water, all gently heated. The coating is allowed to dry, and the weight of the gelatin mixture is ascertained. Separate portions of the mixture are dried and analyzed for the heat value. This gelatin- coated capsule can be used with or without asbestos. Caps can be made of the same material to fit tightly over the end, or inverted, used 36 INVESTIGATIONS IN USE OF BOMB CALORIMETER. as plugs. It requires less gelatin to coat such a capsule and make it alcohol tight than is found in the ordinary gelatin capsules used. The weight of all gelatin capsules, etc., should be taken when the the material is dry, and this weight should be used in correcting for the heat of combustion of gelatin. But before using they should have been exposed to room conditions for a considerable time, else there is no constancy about them when the}' have to be reweighed and filled with alcohol. ALCOHOL DETERMINATION USED FOR TESTING THE RESPIRA- TION CALORIMETER. This was absolute alcohol diluted with distilled water, and hence the first step was to obtain its specific gravity. Two pycnometers were used and the specific gravity was taken at 15.6 C., as follows: No. 1. No. 2. Pycnometer plus distilled water Grains. 77.2663 Grams. 87 2591 Pvcnometer, emptv . 27.2385 37. 2259 50 cc. air 50.0278 .06894 50.0332 B6M Water 50 08674 50 09214 I'ycnorneter, plus alcohol 68. 4121 78.4056 PvcnoineUT, empty 27 2385 37 2259 , r )0 cc. air 41.1736 05894 41. 1797 05894 Alcohol '. 41 23254 41 23864 No. 1. 41.23254--50.08674=.823222 specific gravity. No. 2. 41. 23864--50.09214=. 823255 specific gravity. Average, =.823238 specific gravity. According to Squibb, at 15.6 C. 0.82755 specific gravity =88.0 per cent pure alcohol. .81684 specific gravity=92.0 per cent pure alcohol. Difference, .01071 specific gravity= 4.0 per cent pure alcohol. 0.82755 =88.0 per cent alcohol. .823238= unknown alcohol. .004312 0.01071 : 0.00431 2 ::4.0 :X X = 1.6105 per cent. Hence 88.0 1.6105=89.611 per cent ethyl alo.hol. DETERMINATION OF HEAT OF COMBUSTION. I -ing the old water value of the bomb and without any correction for oxygen, two determinations of the capsule gelatin gave an aver- age of 4,871 calories per gram, and the average of three determina- tions of the alcohol gave <,4.">,s.7 calories per gram, or 7,207 calories per gram purr, alcohol. Applying the correction found for impurities in oxygen by burning 1 gram cellulose, and using the corresponding water value of the bomb, the average for the gelatin was 4,S(K) calo- DETERMINATION OF HEAT OF COMBUSTION. 37 ries per gram, and for the alcohol 6,447 calories per gram, or 7,194 calories per gram pure alcohol. A sudden jar of the bomb caused one capsule containing alcohol to fall to the bottom of the bomb. It was ignited, and the determina- tion was carried through as usual, but only 6,949.8 calories per gram pure alcohol was measured, that is, 3.4 per cent less than the above average. Upon opening the bomb there was found a black spot and some 3^ellowish oily liquid at the bottom, and a strong, very peculiar odor mixed with odor of alcohol was noticed, thus showing incomplete combustion very decidedly. Another charge of alcohol which had remained in the capsule in the bomb for several hours before ignition gave 7,091 calories per gram pure alcohol. Here, where there was opportunity for evaporation, the heat obtained was less than the aver- age of the three other determinations. These observations are sufficient to show that to obtain correct values great care must be taken when volatile substances are to be burned in the bomb. One or two questions which may possibly prove to be of some im- portance in connection with this work will be considered briefty. These questions concern the contents of the bomb before and after combustion, and their influence upon the measurement of the heat, namely, the influence which the changes in the contents have upon the water value of the bomb, and the quantity of heat held by these vari- ous compounds formed and not accounted for. In the following calculations I shall use the values for specific heats, etc., as given below: Per unit weight: Cane sugar 0. 301 specific heat. Water =1. 000 specific heat. O 2 = .2175 specific heat. CO 2 = . 1875 specific heat. 2 1 liter =1. 4298 gram. To illustrate these problems and for the sake of simplicity we shafl consider one example, leaving out all of the minor things which would complicate the calculations. Thus, we have a bomb of 360 c. c. capacity. In the bomb, before combustion, are 1 gram cane sugar and 20 atmos- pheres pure oxygen, all at 20 C. What changes take place when the sugar is burned, and what are their significance? According to the weight and specific heat of the substances in the bomb, they represent a certain mass of water at the same tempera- ture; also, a definite quantity of heat is held by them, which may be expressed in calories. Twenty atmospheres oxygen would then be 20X360 c. c. =7,200 c. c. oxygen. 7,200X1.4298 (weight of 1 liter oxygen) =10. 294 56 grams oxygen. 10.29456x0.2175 (specific heat)<>2.239 grams water. 1 gram cane sugar X 0.301 specific heat<>0-301 gram water. 88 INVESTIGATIONS IN USE OF BOMB CALORIMETER. Hence the substances in the bomb represent Oxygen 2. 239 grams H 2 O Sugar . 30] grams H 2 O Total 2. 540 grams H 2 O The sugar burns to CO 2 and H 2 O, and after the combustion of 1 gram sugar we find 0.5793 gram water and 1.5426 gram CO 2 . 1.5426 grams CO 2 X0.1870 specific heat=0.288 gram water. 10.294561.1219=9.1727 grams O 2 , and X0.2175=1.933 grams water. Thus in the bomb after combustion we have Oxygen equivalent to 1. 933 grams H 2 O Carbon dioxide equivalent to . 288 gram H 2 O Water . 5793 gram H 2 O Total 2. 8003 grams H 2 O After combustion 2. 8003 grams water Before combustion 2. 5400 grams water Difference (increase) .26,03 gram water The whole bomb system therefore has changed to an extent equiva- lent to 0.2603 gram water during the combustion. This is equal to 0.011 per cent of the total bomb water value plus water used, or an error of about one-half calory in the ordinary determination. If, for instance, 1 gram butyl alcohol should be burned instead of sugar, the change in the system would be equal to 0.4685 gram water an error of 0.019 per cent. This is a very small error, and, at the present, is perhaps altogether negligible. The next question is, To what extent does this change of contents affect the heat evolved during the combustion '* This question refers to the same changes in the contents of the bomb as did the preceding one, but instead of studying their effect upon the water value of the bomb we want to see what the effect will be upon the determination of heat of combustion when the factor for the bomb and the water used remains the same. In the case of the sugar there was an increase of 0.2603 gram water in the bomb, which, raised to 1.7 C., would equal about one-half calory held by it. The correction will, in other words, in the case of sugar, be tin- >ame whichever way we apply it. lnt this is not necessarily true of all -iibstances burned. As said before, the correction is so small that it need seldom be considered. CONCLUSION. From what has been said. then, we learn that there are many possi bilitie- for error in the work with the bomb calorimeter. I 'ndoubtedly inanx invent jn-aiors j n the past have worked with impure oxygen and CONCLUSION. 39 never questioned its purity. In the light of our present experience it is questionable whether Stohmann himself, by the use of a heated copper tube, could have removed the last traces of combustible gases from his oxygen. The disappearance of nitric acid formed and its relation to the ash has not been taken into consideration, and it is only within a couple of years that the thermometer lag has come to be applied in the calcu- lations of the results. These overlooked or at times unknown difficulties, which have been referred to throughout this paper, may be the cause of some of the disagreements of results as experienced by different investigators and referred to in the earlier part of this paper. From what has been done and said I believe we also have learned that much work is yet needed in the different lines indicated before the method for determinations of heat of combustion by means of the bomb calorimeter can be called perfect. o University of California SOUTHERN REGIONAL LIBRARY FACILITY 405 Hilgard Avenue, Los Angeles, CA 90024-1388 Return this material to the library from which it was borrowed. SRLF 8CT 1 6 1995 SRLF QUARTER LO/ RECEIVED EMS LIBRARY r