. . . . .... . ,.■: ■ ■-■-.-- ■:■:,- . ■,■■„■ ■ ■- ■... ■ . ■ BOUGHT WITH THE INCOME PROM THE SAGE ENDOWMENT FUND THE GIET OF lUnrij W. Sage 1S91 A.msm :---, Ljtlte Cornell University Library arV17292 Primary batteries. olin.anx The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924031227949 PRIMARY BATTERIES BY HENRY S. CARHART, A.M. Professor of Physics in the University of Michigan SIXTY-SEVEN ILLUSTRATIONS Boston ALLYN AND BACON 1891 -Copyright, 1891, By HENRY S. CARHART. Typography by J. S. Cushing & Co., Boston. Presswork by Berwick & Smith, Boston. PKEFACE. With the exception of a single translation from the French, the material on primary batteries hitherto accessi- ble to English readers has been in detached, portions, partly in books on the general subject of electricity, and partly in scientific journals and technical papers. A thorough knowledge, systematically arranged, of the principles involved in the construction, operation, and theory of primary batteries is of undoubted service to those beginning an extended course of study in the appli- cations and engineering of electricity; while it is indis- pensable to one whose occupation requires familiarity with these most simple and useful means of producing electric currents for practical purposes. This little book has been written with both of these classes of readers in mind. No attempt has been made to compile anything like a complete list and descrip- tion of all the combinations proposed or actually used as primary batteries. A large proportion of them are more curious than" useful, and many have scarcely the merit of novelty. It is hoped that the reader will find a satisfactory account of the theory of a voltaic cell from the point of iv PREFACE. view of the transformation and conservation of energy. In this connection the author desires to acknowledge his obligation to Dr. Lodge's " Modern Views of Electricity." The divisions of the subject are considered to be as logical as the nature of the material permits ; each one is fully illustrated by the most useful types of cells. Some prominence has been given to standards of electromotive force, since they are employed much more commonly than formerly as secondary standards for the measurement of both currents and electromotive forces. Their convenience and, with proper precautions, their accuracy as well com- mend them for general use. It is hoped that the chapter on testing will be of interest to the student, and useful as an outline guide for labora- tory purposes. With scarcely an exception the tests de- scribed have been made either by the author himself or under his immediate supervision. They are believed to be free from bias and to exhibit some facts not heretofore accessible to the public. H. S. C. University or Michigan, June 1, 1891. CONTENTS. CHAPTER I. INTRODUCTION. SECTION PAGE 1. Battery defined 1 2. Batteries : primary and secondary 1 3. Origin of the voltaic cell 2 4. Volta's pile 3 5. The dry pile 4 CHAPTER II. THE SIMPLE VOLTAIC CELL. 6. Fundamental phenomena 7 7. Theory of the voltaic element 8 8. Chemical reaction in the simple voltaic cell 10 9. Inconstancy of the simple voltaic cell 11 10. Experiments on the polarization of a simple cell 12 CHAPTER III. POTENTIAL AND ELECTROMOTIVE FORCE. 11. Electric potential . 15 12. Positive and negative work 16 1 3. Electromotive force 17 14. Relation of electromotive force to difference of potential 18 15. Relation of potential differences to external and internal resistance 20 J6. Volta's contact force 21 17. Explanation of the Volta effect 22 v VI CONTENTS. CHAPTER IV. CLOSED CIRCUIT BATTERIES. SECTION PAGE 18. Distinction between open and closed circuit batteries 27 19. The Daniell battery 28 20. Chemical reactions in the Daniell cell 30 21. Chemical reactions of the cell in relation to energy 32 22. Local action and amalgamation 33 23. The effect of amalgamation 34 24. Relative protection of alloying and amalgamating 35 25. Defects of the Daniell cell 36 26. The effect of temperature changes on a Daniell battery 37 27. The gravity battery 38 28. The Gethins battery 40 29. Delany's modified gravity cell 41 30. Sir William Thomson's tray battery 42 31. Grove's battery 43 32. Bunsen's battery 46 33. The bichromate battery 47 34. Chemical reactions in the bichromate battery 49 35. The advantages of sodium bichromate over potassium bichro- mate 50 36. Directions for setting up a bichromate battery 51 37. The Fuller bichromate cell 53 38. Chromic acid as the depolarizer 53 39. The Partz acid gravity battery 55 40. Taylor's battery 57 41. The copper oxide battery 58 42. The Edison-Lalande battery 60 43. The chloride of silver cell 62 44. Modifications of the silver chloride cell 64 CHAPTER V. OPEN CIRCUIT BATTERIES. 45. The Leclanche' cell 66 46. Chemical reactions in the Leclanchfi cell 67 47. The prism Leclanche battery 69 48. The closed Leclanch6 cell 71 49. LeelanchS cells with carbon cup 73 CONTENTS. Vii SECTION PAGE 50. Leclanchfi cell with agglomerated carbon 74 51. Roberts' peroxide battery 74 52. The sulphate of mercury battery 75 53. The Fitch " chlorine " battery 76 CHAPTER VI. BATTERIES WITHOUT A DEPOLARIZER. 54. The Smee cell 78 55. The sea salt battery 79 56. The Law battery 80 57. The diamond carbon battery 80 58. Cylinder carbon battery 82 59. The Gassner dry battery 83 CHAPTER VII. STANDARDS OP ELECTROMOTIVE FORCE. 60. Latimer Clark's standard cell 86 61. Lord Rayleigh's form of the Clark element 87 62. A standard Clark cell with low temperature coefficient 90 63. The oxide of mercury standard cell 95 64. Sir William Thomson's standard Daniell cell 97 65. Lodge's standard Daniell cell 98 66. Fleming's standard Daniell cell 99 67. The chloride of lead standard cell 102 68. To measure the E.M.F. of a standard cell 103 •CHAPTER VIII. MISCELLANEOUS BATTERIES. 69. Grove's gas battery 106 70. Upward's chlorine battery 109 71. Powell's thermo-electrochemical battery 110 72. A battery absorbing oxygen from the air Ill 73. Minchin's seleno- aluminum cell 112 74. Shelford Bidwell's dry battery 113 75. Jabloehkoff's battery 114 76. Battery with two carbon electrodes 114 viu CONTENTS. CHAPTER IX. BATTERY TESTS. SECTION PAGE 77. What a systematic test includes 115 78. Theory of the method of measuring E.MJF. and internal resistance 116 79. To obtain data for curves of polarization, recovery, internal resistance, and current 119 80. Test of a typical Leclanche cell 121 81. Test of Leclanche 1 cell with depolarizer enclosed in carbon cylinder 124 82. Test of zinc-carbon cell without depolarizer 127 83. Test of a " dry " cell 128 ,84. Test of a silver chloride cell. 130 85. Efficiency test of copper oxide battery 131 86. Testing battery designed for small lamps 134 87. Analysis of the temperature coefficient of a battery 136 88. To determine the thermo-electric power of zinc— zinc sulphate, 136 89. Thermo-electric power of copper— copper sulphate 141 90. Application to aDaniell cell 142 91. Temperature coefficient of a Daniell cell 145 92. Thermo-electric power of mercury— mercurous sulphate 146 93. The experimental cell as a Clark cell 149 94. Electromotive forces of various combinations 151 95. Relative value of oxidants in batteries 153 96. Manganese dioxide in Leclanche' cells 155 CHAPTER X. GROUPING OP CELLS. 97. Activity and efficiency 157 98. Application of Ohm's law to a single cell 157 99. Cells in series 158 100. Grouping in parallel or multiple arc 159 101. Grouping in multiple series 160 102. Arrangement to produce the greatest current 160 103. Grouping of a battery for quickest action 161 104. Grouping together dissimilar cells 164 CONTENTS. ix CHAPTER XI. THERMAL RELATIONS. SECTION PAGE 105. General considerations 166 106. Units of force, work, activity, and heat 168 107. The heat equivalent of a current 169 108. Heat evolved in a circuit with no counter electromotive force, 170 109. Counter electromotive force in a circuit 172 110. Division of the energy in a circuit with counter electromotive force 173 111. Counter electromotive force of electrolysis 173 112. Failure of a cell to effect decomposition 175 113. Calculation of E.M.F. from the heat of combination 176 114. Application to the Smee cell 178 115. Application to the Daniell cell 179 116. Application to the Bunsen cell 179 117. Application to the silver chloride cell 180 118. Helmholtz's formula for electromotive force 180 PRIMARY BATTERIES. oKKo CHAPTER I. EEEATA. Page 17, line 20, for " one ten-millionth of a millimeter," read one twenty -thousandth of a millimeter. Page 28, line 22, for "Edinburgh," read London. Page 58, line 4, for " one of the best substitutes for nitric acid," read nitrate of potassium and sulphuric acid. Page 159, bottom, for "« =^," read n — Page 180, line 4, for "Zn, 2 , S0 4 ," read Zn, 2 , S0 2 . ditions, which determine that the transformed energy- shall be electrical, is called a battery, or voltaic cell. 2. Batteries: Primary and Secondary. — Electric bat- teries may be either primary or secondary. A primary battery is usually understood to be one in which the materials are combined in the cell in such a state as to PRIMARY BATTERIES. CHAPTER I. INTRODUCTION. 1. Battery Defined. — An electric battery, or cell, as a single element is called, is a device for the conversion of the potential energy of chemical separation into the energy of an electric current. Thus the metal zinc and sulphuric acid, which acts chemically on it, represent energy of chemical separa- tion in the potential form. If now the zinc is placed alone in the acid, this energy of chemical separation is converted simply into heat, when the zinc displaces the hydrogen of the acid with the formation of zinc sul- phate. But if the displacement of hydrogen by zinc is made to take place under certain less simple conditions, then a part at least of the kinetic energy developed takes the form of the energy of an electric current. The arrangement of parts necessary to secure these con- ditions, which determine that the transformed energy shall be electrical, is called a battery, or voltaic cell. 2. Batteries : Primary and Secondary. — Electric bat- teries may be either primary or secondary. A primary battery is usually understood to be one in which the materials are combined in the cell in such a state as to 2 PRIMARY BATTERIES. be immediately utilizable in producing an electric cur-, rent ; while, in a secondary battery, the materials or elements of which it is composed need to be modified by electrolysis, due to the passage of a current of electricity from some external source, before the cell is in condi- tion to yield any considerable energy in the form of an electric current. The former possesses a store of poten- tial energy in the materials which admit of chemical reactions ; while the latter is only a reservoir, capable of storing energy by means of the chemical changes produced by electrolysis. Some batteries may combine both characters in one. These are capable of having the chemical changes which take place in them, during the production of a current, reversed wholly or in part upon the passage of a reverse current from some other source ; so that, after they have been exhausted by performing their function as a pri- mary battery, they may again be restored to activity by the passage through them of a current in the opposite direction to the one normally furnished by the cells themselves. This reverse current must be kept flowing for a sufficient time to effect the necessary chemical changes. Such cells are not as efficient in their sec- ondary capacity as storage cells which are designedly such. The energy which they can restore after recharg- ing must always fall far short of the energy expended on them. 3. Origin of the Voltaic Cell. — As early as 1767 Sulzer announced to the Berlin Academy of Science the dis- covery that a peculiar taste is perceived when two different metals are placed together on the tongue and brought into contact at their edges. Such a combina- tion of two metals, as copper and silver, and the saline INTRODUCTION. saliva constitutes, as we now know, a voltaic couple. But the significance of Sulzer's observation was not appre- ciated till more than thirty years later, when Galvani had made his capital discovery (1786) that freshly pre- pared frogs' legs, hung by a copper wire on an iron balcony railing, twitched convulsively whenever the frog touched the iron ; and Volta had demonstrated that the effect was not due to animal electricity, but to the two metals ; and that electricity, identical with that excited by friction, could be produced by means of the metals without the agency of animal tissues, nerves, or muscles. Hence arose Volta's contact theory of electrical ex- citation. This ascribes what is now called the difference of po- tential exhibited by two metals to their mere contact, independ- ently of the medium in which they are immersed. The reader is referred to a later chapter for a discussion of this subject. 4. Volta's Pile. — In pursu- ance of his view of the origin of the electricity producing the muscular contractions of the frog, and in order to increase the electrical action, Volta constructed a chain of ele- ments, to which he gave the name of artificial electric organ, but which has since been known as the Voltaic pile. It consisted of many discs of copper and zinc, or Fig. 1.— Volta's Pile. PRIMARY BATTERIES. preferably silver and zinc, either placed in contact or soldered together in pairs, and piled up with interposed layers of cloth moistened with pure water, or better, with a solution of salt. An essential condition was that the order zinc-copper-cloth, zinc-copper-cloth, must be main- tained from bottom to top. Fig. 1 shows one of the original forms of a voltaic pile. The discs were kept in position by glass rods. The bottom disc of zinc was called the negative pole, and the top one of copper the positive pole. A pile composed of from twenty to forty such pairs of plates produced appreciable physiological effects when the experimenter touched the two poles with moistened hands, or when the positive and negative terminal wires were held in the mouth or touched the eyes. Volta's pile was the immediate forerunner of his " crown of cups," which was the first real voltaic battery. Each element of it was called a galvanic element. Thus the names of both Galvani and Volta became inseparably associated with this earliest device to produce a continuous flow of electricity. 5. The Dry Pile. — Following the principle of Volta, Behrens constructed a pile, in which the moistened cloth was replaced with paper, and which was called, in consequence, a dry pile, though it is inactive unless the paper holds more or less moisture. Zamboni, who interested himself in it and modified it, gave to it the name of Zamboni's pile. It was made of so-called gold and silver paper, the former being coated on one side with copper foil, and the latter with tin. The pairs were made of small discs of the coated paper, from \ to 4 cm. in Figr- 8. Dry Pile. INTRODUCTION. 5 diameter, placed together with their metallic sides out- ward, and then piled up to the number of many hun- dreds in such a way that the copper of every pair was turned in the same direction. The whole column was then firmly pressed into a glass tube, varnished with shel- lac, and finally closed with brass caps, as shown in Fig. 2. Dry piles were made consisting of as many as 20,000 pairs of discs. These were capable of charging a thin Leyden jar of 350 sq. cm. surface, in ten minutes, to such an extent that the discharge melted 2.5 cm. of platinum wire 0.05 mm. in diameter. The dry pile has been applied to the construction of a device for the continuous motion of a light insulated carrier, called an electric pendulum, or perpetual motion. Two columns, 8 and S', Fig. 3, of about 2000 pairs each, are placed so that the positive pole of one and the nega- tive of the other are uppermost. The lower poles are then connected metallically by a wire m, and the whole is placed on an insulating stand. The small metal ring r is attached to a glass rod forming the upper part of the pendulum, which is supported on a knife edge at a, and has a device at b for adjusting the centre of gravity, which is made to assume a position slightly above the point of support. The pendulum, therefore, inclines toward one side, receives a charge from the pole touched, is repelled, and carries its charge over to the opposite pole, by which it is neutralized, and has given to it a charge of the opposite sign. It then reverses its motion toward the pole first approached; and this action is repeated indefinitely. Such a pendulum has been in continuous motion, it is said, in the University at Innsbruck since 1823. 1 The i Miiller's Lehrtrach der Physik, Vol. III. p. 249. 6 PRIMARY BATTERIES. period of oscillation changes within limits with the humidity of the atmosphere. The energy expended by the moving system is exceed- ingly small, and is at the expense of the internal chemical energy of the pile, which is necessarily limited. VHP Fig. 3. — Electric Pendulum. The dry pile has been applied in a similar way to the construction of a delicate electrometer for the detection of minute charges of electricity on a piece of gold leaf suspended between the poles ; or for keeping charged the pairs of quadrants of an electrometer, similar in principle to Sir William Thomson's. THE SIMPLE VOLTAIC CELL. CHAPTER II. THE SIMPLE VOLTAIC CELL. 6. Fundamental Phenomena. — If a strip of pure zinc is placed in sulphuric acid, diluted with from fifteen to twenty times its volume of water, bubbles of hydrogen may be seen to collect on the zinc, but the chemical action soon ceases. If now a strip of copper is placed in the same solution with the zinc, no change is observ- able so long as the two metals are kept out of contact ; but as soon as they are made to touch each other, or are con- nected together by means of a wire or metal strip (Fig. 4), vig- orous chemical action is set up, the zinc is attacked by the acid, and hydrogen gas is liberated in abundance at the surface of the copper plate or strip. Thus, while the chemical action takes place apparently at the zinc, the gas- eous product of the reaction appears only at the copper. As soon as the connection between the two metals is interrupted, the chemical action ceases, and hydrogen is no longer disengaged. If now the proper tests are applied, it will be found that the energetic chemical activity, taking place while the two metals are connected, is accompanied by the Fig. 4. Simple Voltaic Element. 8 PRIMARY BATTERIES. passage of a current of electricity from the copper to the zinc through the metallic connector, and from the zinc to the copper through the liquid in which the plates are immersed. The plates, the liquid, and the connect- ing wire or other conductor constitute the electric cir- cuit. The wire connected with the copper plate is called the positive electrode, and the other the negative. The copper plate itself is called the negative plate, and the zinc the positive plate. This is because it has been demonstrated that zinc in contact with copper in air, either directly or through an intervening metal, assumes a positive charge of electricity, and the cop- per a negative one. Such a system of two different metals, immersed in a liquid which acts chemically on one of them when the circuit is closed, constitutes what is known as a voltaic cell or element. The positive metal is usually zinc ; the negative may be copper,. silver, or platinum; while for the exciting liquid water, salt water, sulphuric acid, hydrochloric acid, or a caustic alkali may be used. 7. Theory of the Voltaic Element. — To make as simple a case as possible, let us suppose that the zinc and copper are immersed in dilute hydrochloric acid, every molecule of which consists of one atom of hydrogen combined with one of chlorine (HC1). Clausius supposed that in a liquid a continual inter- change takes place between like atoms of different mole- cules. Thus the hydrogen of any acid molecule of hydrochloric acid is not permanently attached to the chlorine of the same molecule, but is occasionally sep- arated from it, and then combines with the free chlorine atom of some other molecule. This interchange goes on indifferently in all directions so long as no directive THE SIMPLE VOLTAIC CELL. 9 force is introduced from without. The theory of Clausius is supported by certain facts of double decomposition with strongly combined salts. "When their solutions are mixed, the interchange of atoms allows the formation of weaker compounds ; and that such compounds do form is proved by their appearing as a precipitate, if they are sufficiently insoluble. The chlorine and hydrogen atoms then interchange frequently from molecule, to molecule at random ; and while in the free state between successive pairings, each hydrogen atom carries a charge of positive electricity, and each chlorine atom an equal charge of negative. If now we assume a chemical attraction between the zinc and the chlorine atoms, or imagine with Helmholtz that both zinc and copper have an attraction for the negative charge of the chlorine atoms, the zinc superior to the copper, then it will follow that when the zinc and copper are immersed in the liquid, an extraneous force has been introduced among the chlorine atoms, so that their molecular interchanges are constrained to take place in the direction of the zinc. They unite with the zinc, giving up their negative charge, till this action is arrested by the repulsion between the negative charge accumulated on the zinc and that of the free chlorine. Only incipient chemical action can therefore take place till electrical connection is made between the charged zinc plate and the copper immersed in the liquid with it. Negative electricity then flows toward the copper, through the connecting conductor, and unites with the positive charge of the hydrogen atoms' which move toward the copper plate to meet the negative current. The hydrogen gas thus escapes at the copper plate ; a procession of hydrogen atoms moves steadily in that 10 PRIMARY BATTERIES. direction, either directly or, with greater probability, by successive molecular interchanges ; and the separated electrical charges are reunited through the connecting electrical conductor. When the circuit is interrupted, the charges which quickly accumulate check the move- ment of the disengaged atoms by repulsion of like charges, and all chemical activity ceases. The condition assumed when the circuit is open is one of electrostatic equilibrium. The chlorine atoms continue to unite with the zinc and to deliver to the zinc plate their negative charge, till the repulsion be- tween the negative charges of the zinc and of the momentarily free chlorine atoms equals the chemical attraction between the zinc and chlorine. The two electrodes will then be oppositely charged, and will exhibit a difference of potential dependent upon a number of conditions to be described later. 8. Chemical Reaction in the Simple Voltaic Cell. — If we suppose that the arrangement of metals and acid in the cell is as follows, — Zn | H 2 S0 4 | H 2 S0 4 | Cu , Zinc Sulphuric Acid Sulphuric Acid Copper then the operation which repeats itself over and over when the two metals are electrically connected may be represented thus, — Zn | H 2 S0 4 | H 2 S0 4 | Cu, giving ZnS0 4 | H 2 S0 4 | H 2 | Cu . Zinc Sulphate Sulphuric Acid Hydrogen Copper The arrow represents the direction of the current through the cell. The zinc and hydrogen are both dis- THE SIMPLE VOLTAIC CELL. 11 placed in the direction of the current, while the so-called "sulphion," or S0 4 part of the acid, is displaced in the other direction. All metals and hydrogen are electro- positive, and travel in an electrolyte with the positive current. Zinc sulphate is formed at the expense of zinc and sulphuric acid, and hydrogen gas is set free at the copper plate. The simple chemical action taking place is the displacement of the hydrogen of the acid by zinc, forming zinc sulphate in place of hydrogen sulphate. 9. Inconstancy of the Simple Voltaic Cell. — If the cir- cuit, consisting of zinc, dilute acid, copper, and con- necting wire, is kept closed for some time, the electric current will rapidly decrease in intensity, the chemical action will diminish, and, if the connecting wire offers but little electrical resistance, the action in the cell will shortly cease altogether. This diminution of activity is due to several causes. The chief one is the accumula- tion of hydrogen on the copper plate, causing what is known as the polarization of the cell. The flow of the current is ascribed to what is called the electromotive force (E.M.F.), and by Ohm's law the strength of the current is the quotient of this E.M.F. and the resistance offered by the entire circuit to the flow of electricity. Any condition operating to decrease the E.M.F., to increase the resistance, or to do both, will cause the current to diminish in intensity. Now the hydrogen on the copper plate sets up an inverse E.M.F., so that the effective E.M.F., producing a cur- rent, is diminished by the value of this inverse one. Returning to the theory of the cell, it will be readily seen that both the hydrogen collected on the copper plate and the zinc will attract the free chlorine atoms. 12 PRIMARY BATTERIES. Thus the chlorine atoms will be solicited to cany their negative charge in both directions, and the effective impulse will be the difference of the two. The hydrogen also increases the internal resistance which the cell offers to the passage of electricity, since by its accumulation on the plate a smaller metallic sur- face is actually in contact with the liquid. Independently of the hydrogen, the E.M.F. decreases because of the exhaustion of the acid and the increase in density of the zinc sulphate. Furthermore, when the zinc sulphate in solution reaches the copper plate by diffusion, some of it is liable to be decomposed by the freshly liberated or nascent hydrogen. The zinc is then deposited on the copper, the hydrogen taking its place and forming sulphuric acid. Thus — H 2 + ZnS0 4 = Zn + H 2 S0 4 . When the copper has received a coating of zinc, the two plates are electrically the same, and all action ceases. Because of these faults the simple voltaic cell is of little or no practical value. 10. Experiments on the Polarization of a Simple Cell. — Place enough mercury in a quart jar to cover the bot- tom, and hang near the top of the jar a piece of zinc- Fill up the jar with a nearly saturated solution of salt water, and place the exposed end of a wire, insulated with gutta percha, in the mercury, the upper end form- ing the positive pole of the battery. If now the circuit is closed through some simple current indicator, such as a common telegraph sounder, of a few ohms resistance, the armature will at first be drawn down strongly; but in the course of a few minutes, the time depending upon the total resistance of the circuit, the armature THE SIMPLE. VOLTAIC CELL. 13 will be released, by the magnet, and will be drawn up by the retractile spring. Polarization has then pro- ceeded so far that the current is insufficient to operate the instrument. Next take a small piece of mercuric chloride (HgCl 2 ) no larger than the head of a pin, and drop it in on the surface of the mercury." It will set up a spinning move- ment along the mercurial surface, and the sounder armature will be at once drawn down, indicating that the current has recovered its initial value. The mer- 1,0 Pi 09 $ Ilk, Fig. S. 20 40 - Polarization Curve of Simple Cell. 60 curie chloride furnishes chlorine for the removal of the hydrogen, and so reduces the polarization. In a few minutes the chloride will be exhausted, and polarization will again set in. The introduction of a little mercuric chloride will again restore the cell to activity. A graphical representation of the progress of the polarization in a simple voltaic element is shown by the curve of Fig. 5. A plate of clean zinc and one of 14 PRIMARY BATTERIES. clean copper were immersed in dilute sulphuric acid, specific gravity 1.05. The plates were 5 cms. apart, and 96 sq. cms. surface on each plate were under the liquid. The ordinates of the curve denote the total E.M.F. at intervals of time indicated by the abscissas. The first observations were taken at as short intervals as possible, but after the first few minutes they were less frequent, as the change in the E.M.F. was only slight. The ex- ternal resistance was 20 ohms. With a smaller external resistance the polarization curve is still steeper during the first half-minute, and in the same time the E.M.F. falls to a still lower level. ELECTROMOTIVE FORCE. 15 CHAPTER III. POTENTIAL AND ELECTROMOTIVE FORCE. 11. Electric Potential. — Electric potential is denned in terms of work, and work done is the measure of the energy expended or transformed. It is sufficient for purposes of current electricity to define the difference of po- tential between two points. It is numerically equal to the work done in carrying a unit of electricity in the positive direction from one point to the other. Thus in Fig. 6 the potential differ- ence between the terminals A, B, of Fig. 6. -Simple Battery Circuit. ^ b attely ig the work required to transport a unit quantity of electricity from A round through the external resistance B to the point B. In general it is not the same as the work done in carrying the unit of electricity from B to A through the internal resistance r of the cell, from the negative to the positive terminal. The unit employed in this definition is the " absolute " or centimetre-gramme-second (C.G.S.) unit of quantity, which is ten times the practical unit, called the coulomb. 16 PRIMARY BATTERIES. A point is said to have the practical zero of potential when it is the same as that of the earth. Since difference of potential is the work done on unit quantity, the total work done when any quantity Q is transferred from one point to the other is Q times the potential difference between the points. This remains true whether all the energy expended in the transfer is converted into heat because of the ohmic or frictional resistance M; or whether a portion is converted into mechanical work by means of an appropriate motor device inserted in the external circuit ; or whether the energy is in part stored up by means of electrolysis, as in a secondary battery; or in producing a magnetic field. The work done in one second on any portion of a cir- cuit, included between two points, is the product of the current and the potential difference, both in C.G.S. units. The work is expressed in ergs. It is important to note that the portion of the circuit between the two points considered must not include any source of positive E.M.F. ; that is, an E.M.F. act- ing in the direction of the positive current flow. 12. Positive and Negative Work. — Work done upon the current, or work done in producing a current, is to be considered positive ; while work done by the current is negative. Where the work has the positive sign, energy in some other form is converted into the energy of an electric current ; but when the work is negative, the energy of the electric current is in general expended in heating the circuit, in doing mechanical work, or in effecting chemical dissociation. In the voltaic element the energy of chemical separation is transformed into that of the electric current. The same is true of a sec- ondary battery during its discharge. ELECTROMOTIVE FORCE. 17 In the dynamo-electric machine the power expended in driving the armature is largely reproduced in the energy of the currents traversing it. 13. Electromotive Force. — Electromotive force is the name given to the cause of an electric flow. It is now often called electric pressure from its superficial analogy to water pressure. The origin of the E.M.F. of a vol- taic battery is in the superior affinity of zinc for oxygen as compared with copper. If equivalent weights of zinc and copper are oxidized, the heat of combustion is found to be 85,400 and 37,200 calories respectively. That is, the oxidation of 65 gms. of zinc and 63.4 gms. of copper, requiring equal weights of oxygen, will pro- duce enough heat to raise the temperature of 85,400 and 37,200 gms. of water 1° C. respectively. The strain of the oxygen atoms toward zinc is more than twice as great as toward copper. This strain need not extend to a greater distance in a liquid than the " molecular range," which Quincke has calculated to be about one ten-million th of a millimetre , or one five-hundred-thou- sandth of an inch. As fast as the oxygen is exhausted from the layer of liquid in immediate proximity to the zinc, diffusion supplies the waste. The heat of forma- tion of equivalent weights of zinc and copper with chlo- rine is 97,200 and 51,600 calories respectively. With chlorides, therefore, zinc is still the positive plate, and copper the .negative. If platinum is made to replace copper, the negative strain on the oxygen or chlorine atoms is reduced nearly or quite to zero, and the E.M.F, of the combination is accordingly increased. The E.M.F. of any form of battery depends, therefore, on the materials employed, and is entirely independent of the size and shape of the plates. The condition of 18 PRIMARY BATTERIES. the surface of the plates and the density of the solution or solutions also affect the value of the E.M.F. Thus oxidation of the copper plate increases the E.M.F., while oxidation of the zinc plate decreases it. This result is easily explained in accordance with the theory. The oxygen on the copper plate serves to remove the nascent hydrogen, thus obviating polarization. On open circuit the hydrogen is then attracted toward the copper oxide, and the oxygen toward the zinc. Both operations facilitate the electric separation and transfer of charges in opposite directions. The view here adopted is that the effective E.M.F. of a primary battery is at the contact of the zinc and the exciting liquid rather than at the contact of zinc and copper. 14. Relation of Electromotive Force to Difference of Potential. — The two expressions are not synonymous, neither are they always interchangeable. E.M.F. estab- lishes difference of potential rather than the reverse. This is evident from the fact that there may be a current without any difference of potential between successive points in a circuit, but not without an E.M.F. Such would be the case if a straight bar magnet were thrust through a perfectly uniform circle of wire along the axis of the ring. An induced current would flow along the ring during the motion of the magnet. Every part of the wire would cut equally lines of force, but all points would have precisely the same potential if meas- ured by an electrostatic voltmeter of small capacity. The difference of potential between two points is, however, numerically equal to the effective E.M.F. pro- ducing a current from one point to the other when the circuit between the points contains no source of E.M.F. ELECTROMOTIVE FORCE. 19 In such a case the current flows from the place of higher potential to that of lower, and the loss of potential is proportional to the resistance passed over. Thus in Fig. 6 (p. 15), A has a higher potential than B, and the current flows in the external circuit from the higher potential to the lower. Moreover, the difference of po- tential between A and B is equal to that part of the total E.M.F. of the cell which will produce the given current through the resistance R between the two points. The loss of potential in passing over different portions of this conductor is strictly proportional to the resistance of the several portions. If, however, we direct our attention to the interior of the cell, we find that the current flows across from the zinc to the liquid, or from lower to higher potential. It is so impelled by the vera causa there acting to pro- duce an electric flow. This cause, which is called an electromotive force, may be compared to a pump which lifts water against gravity ; while in the remainder of the closed system of pipes, conveying the water, the liquid flows back again by gravity. It is convenient, therefore, to divide an electric circuit into two regions, one containing the source or sources of E.M.F., and the other containing none. Within the latter region the current flows from higher to lower potential, and the loss of potential is proportional to the resistance passed over. Within the other region, or at some points in that region, the current passes from lower to higher poten- tial, and the change in potential bears no relation to the resistance. In all cases, however, the loss or shrinkage of potential, due to ordinary ohmic resistance, is propor- tional to the resistance passed over. The change in E.M.F. in passing over any resistance is the loss due to 20 PRIMARY BATTERIES. this resistance, added to all the E.M.F.'s encountered, taken with their proper sign. 15. Relation of Potential Differences to External and In- ternal Resistance. — It will be useful to consider atten- tively the distribution of potential throughout the circuit of a simple cell containing no source of coun- ter E.M.F. If the circuit is open so that the external resistance is infinite, then the potential difference between the two electrodes is the total E.M.F. of the cell. Under these conditions the internal resistance of the cell is zero in comparison with the external resistance. Hence the total fall of potential is through the air from one ter- minal to the other. If now the external resistance is gradually diminished, the potential difference between the two poles of the battery becomes less and less, the E.M.F. of the battery remaining constant. If E is the total E.M.F., E the fall of potential between the ter- minals of the cell, and e the loss due to the resistance of the battery itself, then E' = E — e, or E = E' + e ; also E':e::B:r. If now the poles of the cell are connected by a stout conductor of negligible resistance, then E' becomes zero, and e equals E. In other words, the total loss of potential is then entirely internal. If we suppose the seat of the E.M.F. at the surface of the zinc, neglecting the negative E.M.F. at the other plate, then the zinc and connected conductors are at the lowest potential, a sudden rise occurs in passing from the zinc to the liquid, and there is a gradual fall ELECTROMOTIVE FORCE. 21 through the liquid to the negative plate. If the inter- nal resistance is increased, the slope of potential per unit of resistance is diminished, but the total loss through the electrolytic conductor remains the same, and equals the E.M.F. of the cell. It is immaterial whether the two plates with the connecting conductor are partly or wholly immersed in the conducting liquid. 16. Volta's Contact Force. — The muscular convulsions which were observed when the lumbar nerves and the crural muscles of a frog were connected with a bimetal- lic arc of iron and copper, Galvani attributed to a sep- aration of the two electricities at the junction of nerves and muscles. Volta showed that no effect was obtained with a continuous wire of a single metal ; he therefore attributed the effect to the contact of dissimilar metals. After the invention of his pile in 1800 another theory arose, which assigned chemical action as the origin of the E.M.F. In Volta's pile the water moistening the cloth discs was said to be the exciting liquid oxidizing the zinc. Volta assigned to it the function of a con- ductor only. In pursuance of his theory, Volta invented a condensing electroscope with one plate of polished copper and the other of polished zinc. When the zinc plate was placed on the copper and then deftly lifted by means of an insulating handle, the gold leaves of the electroscope diverged with negative electricity. In recent times Sir William Thomson has illustrated the Volta effect, as this has been called, with the appa- ratus shown diagrammatically in Fig. 7. It consists of two half-rings of zinc (Zn) and copper (Cu), placed on insulating supports in the same plane, with a narrow space between their ends. A light aluminum needle is suspended so as to turn freely round the axis of the (7^\ 22 PRIMARY BATTERIES. ring. It is adjusted to hang over one of the spaces between the zinc and copper. If now the needle is charged to a high potential with positive electricity, it will turn toward the copper in the direction of the arrow whenever the two i i i half-rings are metallically con- \y^Z — ' — J nected at AB. If the needle B^-—— — is negatively charged, it turns towards the zinc. This motion • may be interpreted as meaning Fi s- 7 - that the zinc is charged posi- Thomson's Contact Apparatus. , . , ■. , ■. , . ■, tively, and the copper negatively. It also means that there- is a fall of potential in the air from the zinc toward the copper, for the positively charged needle moves in the direction of lower electric potential. It has been supposed by many to demon- strate that the seat of E.M.F. in a voltaic cell is at the contact of the zinc and copper. 17. Explanation of the Volta Effect. — The positive and negative charges exhibited by zinc and copper in con- tact in air may be explained as a simple variation from the ordinary voltaic element. They constitute an air- battery, with the plates immersed in a dielectric or non-conducting fluid ; while the plates of the latter are immersed in an electrolytic conductor. But in each case the fluid bathing the plates acts chemically on both of them. The oxygen is attracted by the zinc and cop- per both, but unequally ; and the effective E.M.F. is a differential result of the two chemical actions. Insu- lated zinc is at a potential of about 1.8 volts lower than the air, while insulated copper is only 0.8 volts lower, ELECTROMOTIVE FORCE. 23 these values being proportional to the heat of formation of ZnO and CuO. When the two metals are brought into contact, their potential becomes the same through- out ; the equalization is brought about by an exchange of electricities, the zinc receiving a positive charge, and the copper a negative one. Their mean potential is then about 1.3 below the average potential of the air. But the normal difference of potential between each metal and the air in the immediate vicinity remains the same as before contact. Hence there is a slope of poten- tial from the air next to the zinc to the air next to the copper of about one volt ; and it is this slope of poten- tial which is indicated by the movement of the needle in the Thomson instrument. The relation of the air voltaic battery to the liquid voltaic battery may be illustrated in a different way. It will be recalled that on open circuit or with infinite external resistance, the potential difference between the zinc and copper is equal to the total E.M.F. of the bat- tery. The copper has then a positive charge, and the zinc an equal negative one, the potential sloping from the positive to the negative. But if the metals are brought into contact, their potential is equalized, and the extreme potential difference is then between the liquid in contact with the zinc and that in contact with the copper, the former being the higher. The plates have no charge, because as fast as oxygen (or chlorine) brings negative to the zinc, and hydrogen brings posi- tive to the copper, both charges are conveyed away by the conductor. This slope of potential in the fluid bathing the plates coexists with their uncharged state only when there is an incessant transfer of electricity throughout the entire circuit. 24 PRIMARY BATTERIES. If now air replaces the liquid, the plates remaining in contact, and hence at the same potential, the internal resistance is infinite, the total E.M.F. is the difference of potential existing in the air surrounding the plates, and the plates acquire a charge, since no current is established. But since in the interior of a battery the current direction is from zinc toward copper, the slope of potential is in this same direction ; therefore the zinc is positively charged, and the copper negatively. To sum up: There are two paths between the zinc and copper plates, the external portion of the circuit and the internal. The plates are charged with electric- ity corresponding to the whole difference of potential of the battery only when one of these resistances or the other is infinite. When the external resistance is infi- nite, and the embracing fluid is an electrolytic conduc- tor, the potential slopes from the copper to the zinc, from the positive charge on the copper to the negative charge on the zinc. When the internal resistance is infinite (air), the plates being directly connected, the slope of potential is from the layer of air in contact with the zinc to the layer in contact With the copper through the non-elec- trolytic medium; while the zinc assumes a positive charge, and the copper a negative one, since in no other way can their, potentials be equalized. With two couples of zinc and copper plates in con- tact, one pair immersed in a conducting liquid and the other in air, the potential in both cases slopes from the zinc toward the copper through the medium ; but in the former there will be a dynamic current, and in the latter only a slight electrostatic displacement sufficient to charge the plates. The displacement in the one is ELECTROMOTIVE FORCE. 25 continuous, in the other momentary. The seat of the electromotive force in either case is at the contact of the metals with the medium, rather than at their con- tact with each other. This is the more apparent from the fact that when zinc and copper in contact are placed in an atmosphere of sulphuretted hydrogen, the zinc acquires a negative charge, and the copper a positive one. In this case the chemical action on the copper is greater than on the zinc, and the electrical conditions are reversed as compared with the same metals in air. Similarly, iron and copper in sulphuric acid give a cur- rent from copper to iron through the external conduc- tor ; but in a solution of potassium sulphide the current is from iron to copper. It is not intended to assert that there is absolutely no true contact force at the junction of two different metals. There is such a contact E.M.F. or potential difference, but it is of very small magnitude, and the evidence of its existence is very different from that furnished by the simple voltaic element. This evidence is furnished by what is known as the Peltier effect. It is a reversible heat phenomenon. The passage of a current through a homogeneous conductor produces heat irrespective of the direction of the current. But when a weak cur- rent is made to pass across a junction from copper to iron, the junction is cooled. This is due to a true con- tact E.M.F. which helps forward the current. Positive work is done at the junction, and energy in the form of heat is absorbed. When the current passes in the opposite direction across the junction, heat is produced additional to that depending upon ordinary ohmic resistance. The same reversible heat production may be observed at the junction of other metals and of dis- 26 PRIMARY BATTERIES. similar substances. But in any case the contact E.M.F., which explains the reversible heat, is at most only a few hundredths of a volt ; it is included in the result- ant electromotive force of a voltaic element, but it is altogether insignificant in comparison with that due to chemical agency. CLOSED CIRCUIT BATTERIES. 27 CHAPTER IV. CLOSED CIRCUIT BATTERIES. 18. Distinction between Open and Closed Circuit Bat- teries. — It has been seen that the inconstancy of the current furnished by a battery through a fixed, resist- ance is largely accounted for by polarization, due to the liberated hydrogen. The agent introduced into the cell to avoid polarization, either by removing the hydro- gen as fast as it is formed or by preventing altogether its disengagement, is called a depolarizer. The distinc- tion between open and closed circuit batteries depends chiefly upon the nature and action of this depolarizer. A battery is entitled to be included in the closed cir- cuit type only when it is capable of working on a closed circuit of moderate resistance for a considerable period, with but slight diminution in the intensity of the cur- rent. It is thus clearly differentiated from those cells that are adapted to stand on open circuit, without wasteful local action, and to furnish current only at intervals, and of a few seconds duration. In a closed circuit cell the depolarizer must act with sufficient promptness and efficiency to prevent polariza- tion quite completely, thus removing this cause of the decrease in the current. In open circuit batteries the depolarizer may indeed be entirely absent, or it may act with so much sluggish- ness that it cannot prevent polarization taking place to 28 PRIMARY BATTERIES. some extent during the action of the cell, but it destroys polarization after the circuit has been again opened. The promptness with which a cell recovers from a depression of its E.M.F. by polarization is a good cri- terion of the efficacy of this class of depolarizers. Batteries provided with such depolarizers occupy an intermediate position between those with a prompt act- ing one and those with none, of which the simple voltaic element is the type. The more efficient depolarizers in general are liquid ; the less efficient or slower acting ones, with only a few exceptions, are solid. The first class must be employed when a continuous current is required, especially if the current is of considerable magnitude. If but a small current is taken from a cell through a high resistance, then a solid depolarizer will suffice. But batteries with no depolarizer for the removal of hydrogen, or an equiv- alent, are adapted only to open circuit use, in which the circuit is to be closed for only a few seconds at a time. 19. The Daniell Battery. — The first constant battery was invented by Professor Daniell, of Edinburgh, in 1836. To prevent the disengagement of hydrogen at the copper plate, it is immersed in a solution of copper sulphate (CuS0 4 ). The nascent hydrogen then decom- poses the CuS0 4 , the result being the formation of sulphuric acid (H 2 S0 4 ) and a deposit of metallic copper on the copper plate. One form of the cell is represented in Fig. 8. is a cleft cylinder of copper, and Z one of zinc. Between the two is a porous cup of unglazed earthenware, so that a continuous liquid circuit is maintained between the zinc and the copper. The zinc is immersed either CLOSED CIRCUIT BATTERIES. 29 in dilute sulphuric acid, or better, in a weak solution of zinc sulphate ; while the copper is surrounded by the solution of copper sulphate contained in the porous cup. Crystals of copper sulphate are shown surrounding the copper cylinder. These are held in a copper wire or perforated basket, and are for the purpose of keeping the solution of the copper salt saturated. The porous Fig. 8. — Daniell Cell. cup serves no purpose except as a partition to separate the liquids surrounding the two plates. Each metal is placed in a salt of itself. The more recent forms of this battery have a zinc prism and the zinc sulphate in the porous cup, while the sheet copper and the copper sulphate solution are outside. The action in either case is the same. 30 PRIMARY BATTERIES. 20. Chemical Reactions in the Daniell Cell. — With acidulated water the chemical action may be represented as follows : — Zn. | H 3 S0 4 | H 2 S0 4 || CuS0 4 | CuS0 4 | Cu„. " — v ' * V ' » > After the first step in the reaction this becomes — Zn z _i | ZnS0 4 | H 2 S0 4 || H 2 S0 4 | CuS0 4 | Cu s+1 . The arrow indicates the direction of the current, and the porous partition is represented by the double verti- cal line. The hydrogen and the metallic elements all migrate in the direction of the current from the zinc toward the copper plate; ZnS0 4 is formed at the ex- pense of CuS0 4 ; metallic zinc disappears, and metallic copper is deposited on the copper plate. The hydrogen is intercepted by the CuS0 4 and never reaches the nega- tive plate. If the zinc is immersed in dilute zinc sulphate instead of acidulated water, the electrolytic circuit, prior to the first step in the chemical reaction, is as follows : — Zn„ | ZnS0 4 | ZnS0 4 || CuS0 4 | CuS0 4 | Cu s . » > After the first step : — Zn,,! | ZnS0 4 | ZnS0 4 || ZnS0 4 | CuS0 4 | Cu, +1 . The action taking place is a very simple one. There is, as before, a decrease of metallic zinc and an increase CLOSED CIRCUIT BATTERIES. 31 of metallic copper, as indicated by the subscripts ; zinc crowds copper out of the copper sulphate, so that there is a continuous transformation of CuS0 4 into ZnS0 4 by this process of replacement. The E.M.F. of a Daniell cell, as ordinarily set up, is about 1.08 volts. The curves in Fig. 9 express the results of a. test made on a Daniell cell set up with < 1.0 n.s m o t> Miu, 20 40 Fig. 9. — Polarization Curves of Daniell Cell. 60 saturated copper sulphate and a 5 per cent zinc sulphate solution. The zinc was amalgamated and the copper carefully cleaned. The external resistance was 5 ohms and the internal 0.85. The upper curve represents the total E.M.F. at small intervals of time, which are laid off horizontally as abscissas, the E.M.F.'s being laid off on the vertical lines as ordinates. The ordinates of the lower curve denote the values of the potential differences at the terminals or electrodes of the cell for the same period of one hour. 32 PRIMARY BATTERIES. This potential difference is the effective E.M.F. pro- ducing the current through the external resistance of 5 ohms. It is then only necessary to divide this terminal E.M.F. by five to obtain the current in amperes. These curves should be compared with the polariza- tion or E.M.F. curve of Fig. 5. They serve to bring out in a forcible manner the contrast between the rapid polarization in a simple voltaic element and the prac- tical freedom from polarization of a well-constructed, clean Daniell cell. The contrast would have been still greater if the voltaic element had been tested with the same external resistance ; but it was not practicable to make a satisfactory time test with an external resist- ance of less than 20 ohms in that case, the polarization being too rapid to follow it with accuracy. 21. Chemical Reactions of the Cell in Relation to Energy. — The question has often arisen why any chemical action should take place upon closing the circuit of a Daniell cell, set up with zinc and copper in their respec- tive sulphates. The answer involves an explanation of the conversion of potential chemical energy into the kinetic energy of dynamic electricity, or at least a statement of the principle upon which this conversion of energy is conditioned. It depends entirely upon whether the heat of formation of the salt that can be formed by the process of replacement is greater than that of the salt or compound decomposed. In the Daniell cell the heat of formation of equivalent weights of ZnS0 4 and CuS0 4 are 242,000 and 191,400 calories respectively. Hence for every 65 grms. of zinc entering into combination as ZnS0 4 , with the reduction of 63.4 grms. of copper from CuS0 4 , the difference CLOSED CIRCUIT BATTERIES. 33 between 242,000 and 191,400, or 50,600, calories of heat, or the equivalent in the kinetic energy of an electric current, must be developed. In the form in which the materials are placed in the cell they represent, therefore, potential energy. Now potential energy always tends to become kinetic whenever the conditions admit of the transformation. The sole condition in the Daniell cell is that the circuit shall be closed. A continuous transformation then goes on, the kinetic energy appearing in the form of an electric current because of the special conditions determining the con- version ; and the process continues so long as there is any available energy left to take part in the opera- tion. 22. Local Action and Amalgamation. — Any chemical action taking place in a cell on open circuit, tending to reduce its available potential energy, or going on when the circuit is closed and not contributing to the produc- tion of the current, is called local action. Local action is always prominent with commercial zinc in an acid solution. The zinc contains foreign particles, such as bits of iron, carbon, or other conducting bodies ; as soon as these are exposed to the liquid, they form closed local circuits, and the zinc is eaten away in patches, or pits. To prevent this wasteful action, the zinc is amal- gamated. Alloys of mercury with other metals are called amalgams. The process of amalgamation consists in forming a zinc-mercury alloy on the surface of the zinc plate or prism. This is best accomplished by first cleaning the zinc by immersion in sufficiently diluted sulphuric acid, and then rubbing mercury over the sur- 34 PRIMARY BATTERIES. face by means of a swab made by tying a piece of cloth round the end of a stick. All excess of mercury should be allowed to drain off. If, however, the plates of zinc stand out of the liquid for some time, the mercury will largely separate, and collect in small globules on the surface. Another method of amalgamating zinc is to dip it in an acid bath containing a mercury salt in solution. This may be prepared by dissolving one part of mercury in three parts by weight of aqua regia (one of nitric to three of hydrochloric acid), and then adding three parts more of hydrochloric acid. There are other forms of local action which amalga- mation does not prevent. Some of these will be more specifically described in connection with the types of batteries most unfavorably affected by them. The zinc of a battery should always be amalgamated when the exciting liquid is acid. 23. The Effect of Amalgamation. — The action of the amalgam appears to be to bring to the surface pure zinc, while foreign materials, especially iron, are left behind. Amalgamated zinc, therefore, acts like pure zinc ; foreign bodies, as soon as they are dislodged, fall to the bottom of the cell ; and wasteful action, due to local currents, is avoided. But amalgamated zinc possesses the singular property of not being attacked when immersed in dilute sulphuric acid. Since this is equally true of pure and commercial zinc, the exemption of amalgamated zinc from attack is not due to the suppression of local cur- rents. The following facts tend to show that the pro- tection of the zinc is to be ascribed to the adhesion of a film of hydrogen to the amalgamated surface. When amalgamated zinc is plunged in water, acidu- CLOSED CIRCUIT BATTERIES. 35 lated with one-twentieth of its volume of sulphuric acid, it is not attacked at ordinary atmospheric pressure. But if a vacuum is produced above the liquid, bubbles of hydrogen are again freely evolved from the zinc surface. Upon readmission of the air, bubbles again adhere to the plate, and the chemical action is arrested. If two plates of ordinary zinc, one amalgamated and the other not, are immersed in dilute acid, the amal- gamated zinc coniports itself as the zinc, and the other as the copper, of a simple voltaic couple. The amalga- mated zinc is thus rather more readily attacked by the acid than the unamalgamated. With pure electrolytic zinc and neutralized sulphate of zinc, there is no potential difference between two plates, one of which is amalgamated and the other not. 24. Relative Protection of Alloying and Amalgamating. — The investigations of Reynier show that the protec- tion secured by mercury is much greater than is gener- ally supposed. In certain liquids the local waste of amalgamated zinc is 50, 100, or even 10,000 times less than that of ordinary zinc. A further question is the relative value of alloying with mercury as compared with amalgamating on the surface. Reynier concludes 1 that zinc alloyed with mercury is, in general, better than zinc amalgamated, especially in experiments of long duration. The first superficial layer of amalgamated zinc is rich in mercury ; but, as the deeper layers are attacked, the proportion of mercury diminishes, and so also the protec- tion obtained. The opposite takes place with the alloy, which is visibly enriched in mercury as its weight i Pile Electrique, p. 21. 36 PRIMARY BATTERIES. diminishes. It is evident that on closed circuit the superiority of the alloy shows itself after a much shorter time. The alloys are more brittle than amalgamated zinc, and they become more so by use, — a fact confirm- ing the preceding observation. The utility of amalgamating the zinc in batteries of the Daniell type has often been contested. Experiment demonstrates that the mercury reduces the loss by one- half in a solution of 15 per cent sulphate of copper. In the alloys referred to the mercury constituted 4 per cent of the entire mass. In a chromic mixture, amalgamated zinc soon loses its brightness, and takes on a dark tint, while the alloy be- comes brighter and brighter up to complete exhaustion. The employment of zinc alloys contributes to the economy of batteries, and increases their constancy. 25. Defects of the Daniell Cell. — The Daniell cell has several rather serious defects. A prominent one is that the copper is sometimes deposited upon the porous cup instead of the copper plate. This deposit grows in the pores, fills them up, and finally cracks the cup and renders it useless. Again, the diffusion of the copper salt through the porous cup, when the battery is not in action, brings it in contact with the zinc ; a spontaneous displacement of copper by zinc then takes place, equivalent to local action. The copper separates in a finely divided state, and is usually oxidized and deposited on the zinc as black cupric oxide (CuO) ; hydrogen is at the same time given off. If the zinc becomes thoroughly black- ened in this way, it should be cleaned. Because of this local action, the Daniell battery should be taken down when not in use. CLOSED CIRCUIT BATTERIES. 37 This reduction of copper and its subsequent oxidation may be illustrated by placing a piece of zinc in a dilute solution of copper sulphate. Immerse a large test-tube rilled with the solution so that its open end shall be over the zinc. As it stands, gas will collect in the tube, displacing the liquid, and the solution will at length lose all its blue color. The black oxide of copper will be found in the vessel, the solution will contain zinc sulphate, and the collected gas will be found, upon test- ing, to be hydrogen. With dense solutions spongy copper will also be found mixed with the oxide. Another objection to the Daniell cell for some pur- poses is its rather large internal resistance, considering its low E.M.F. Only a moderate current, about an ampere, can be taken from a Daniell cell as a maximum. The internal resistance will depend upon the thickness and quality of the porous cup, the size of the plates, and the distance between them. The density of the solu- tions affects the resistance in a minor degree. 26. The Effect of Temperature Changes on a Daniell Battery. — Professor Daniell himself found that his battery yielded a largely increased current when its temperature was raised to 100° C. He attributed this result to increased chemical activity. It is now known that the E.M.F. of this cell changes but slightly with rise of temperature, the decrease per degree Centigrade being less than 0.015 per cent. The most important effect of a rise of temperature of the Daniell cell is the decrease in its internal resistance. It is well known that the resistance of electrolytes diminishes with increase of temperature, and that this inverse relation between resistance and temperature distinguishes electrolytic from metallic conductors, the 38 PRIMARY BATTERIES. temperature coefficient of all metallic conductors being positive, with one exception, — an alloy of ferro-man- ganese and copper. 1 Mr. W. H. Preece found 2 that when a Daniell cell was heated from 0° C. to 100° C, its resistance decreased abruptly at first, and afterwards more gradually, falling from 2.12 to 0.66 ohms. This large decrease of resist- ance accounts for the augmented activity observed by- Daniell, the external resistance in circuit having doubt- less been small. 27. The Gravity Battery. — The gravity battery is a simple modification of the Daniell, designed to avoid the use of a porous cup. It takes its name from the fact that in it the zinc and copper sulphates are sepa- -^v. rated by their difference in den- \«bJL 9 s *ty* form of this battery is """ * shown in Fig. 10. The zinc is suspended, by means of a stout copper wire, from a brass tripod resting on the top of the jar. Thin sheets of copper, riveted together and to the conducting wire, are placed in the bottom of the cell and surrounded with crystals of copper sulphate, known commer- cially as " blue stone " or " blue vitriol." The zinc casting is hung in a weak solution of zinc sulphate from two and a half to three inches above the copper plates. The saturated copper salt has a density greater than the dilute zinc salt. It therefore remains in the bottom 1 American Journal of Science, Vol. XXXIX. p. 471. 2 Proceedings Royal Society, Vol. XXXV. 1883, p. 48. Fig. 10. — Gravity Cell. CLOSED CIRCUIT BATTERIES. 39 of the jar if it is not disturbed, except that it slowly diffuses upward toward the zinc. These cells should be set up with well-diluted zinc sulphate, extending at least an inch below the zinc. If water and crystals of copper sulphate alone are used, the cell will not work at first; and as soon as the copper salt reaches the zinc, either by diffusion or stir- ring, the zinc turns black from the oxidation of the reduced copper, and stalactites will soon be found hang- ing from the zinc. When the cell is properly set up, with copper in copper sulphate and zinc in zinc sulphate, the chemical reactions are the same as in the Daniell cell. If the cell is left standing on open circuit, the copper sulphate diffuses upward, as already explained, and wasteful local action takes place. Besides, the cell becomes foul much more rapidly than if the copper salt were not allowed to reach the zinc. Hence this cell always keeps in better condition if a closed circuit is maintained through a high resistance when the battery is not in use. Zinc then replaces copper in the copper salt as fast as it diffuses upward. The zinc sulphate formed must be occasionally drawn off and replaced with soft water. So, too, crystals of copper sulphate must be added from time to time to keep the solution satu- rated. Care must be taken not to allow these crystals to lodge on the zinc. It is better to add small quantities at frequent intervals than to place too large a supply in the jar at once. When the water evaporates, the zinc sulphate crystal- lizes round the jar, and then creeps up by capillary- action, crystallizing as it ascends, till it finally flows slowly over the top. As a preventive, the tops of the 40 PRIMARY BATTERIES. jars may be dipped in hot paraffin, or a strip of very adhesive tape may be pasted round the rim, inside and out. 28. The Gethins Battery. — The inventor of this form of copper sulphate cell has sought to combine the ad- vantages of a Daniell with those of a simple gravity cell. The cupric sulphate is placed round the sheet copper in the bottom of the jar, as in the gravity form ; while a porous cup, in the shape of a frustum of a cone, is hung in the top of the jar by means of a stout rim, as shown in Fig. 11. The zinc has a broad, heavy foot, and stands in the porous cup. About four pounds of coarse crystals of CuS0 4 are placed in the bottom of the jar, and the jar is about half filled with water. The porous cup with the zinc is then put in position, and a weak zinc sulphate solution is poured in. The battery is then ready for use. Its E.M.F. is slightly over one volt, and its internal resist- ance three ohms. Hence only one-third of an ampere can be taken from it, even on short circuit; and none of this can be utilized, but all is expended in internal Fig. 11. — The Gethins Batte CLQSED CIRCUIT BATTERIES. 41 heat. For energy in the external circuit, there must be external resistance in addition to the internal; and hence the current will be smaller, unless several cells are coupled in parallel. Three of these cells in series will keep a storage battery charged so that it will run a phonograph as much as is required for a private office. The storage cell in the case , tried had thirteen plates, six positive and seven negative, each 60 square inches in area. The primary battery was kept constantly connected with the secondary. The diffusion of the zinc sulphate outward through the porous cup is noticeably greater than that of the copper sulphate inward. The level of the liquid outside the cup rises till the difference in hydrostatic pressure counterbalances the difference in diffusive tendency. 29. Delany's Modified Gravity Cell. — Cells of the Daniell type, in which copper sulphate is the depolar- izer, have been of such great ser- vice when small but constant cur- rents are required, that a brief description of the Delany modi- fication seems desirable. It is shown in Fig. 12. The CuS0 4 is enclosed in a straw- board box, and the zinc in a paper envelope. The box pre- vents the CuS0 4 dust from dis- solving at once, and diffusing so as to reach the zinc. The copper sulphate solution gradually ap- pears by transfusion through the strawboard. The copper of the element consists of heavy wire wound in Tig. 13. Delany's Gravity Cell. 42 PRIMARY BATTERIES. vertical bands about the strawboard box, and an insu- lated wire rises from this to the top of the cell. The paper round the zinc prevents spongy copper or other material falling upon the copper. It is claimed that no stalactites depend from the zinc, and that the deposit on the zinc is easily removed without hacking or scraping. Ordinary gravity cells often need to have this process vigorously applied to them. A band of rubber cloth is attached by a sticky sub- stance to the inside of the rim of the jar to prevent the crystallized salts creeping over. It is said to present a complete mechanical obstruction to the climbing of the zinc sulphate. It may, of course, be applied to any other jar, first making sure that the rim is thoroughly clean ; then after warming the sticky side of the cloth, press firmly all round against the rim. 30. Sir William Thomson's Tray Battery. — Another form of Daniell cell was designed by Sir William Thomson, with a view of diminishing the internal resistance. The cell is made in the form of a large wooden tray, about 20 inches square, lined with lead on which copper has been deposited by electrolysis or during the action of the battery. The lead extends over the outside at the four corners and down under the bot- tom, for the purpose of making contact with the next cell below. The zinc is in the shape of a grate, as shown in Fig. 13, which represents five cells in series. At the corners are feet turned upward. The lead of the cell above rests on the upturned feet of the zinc, making a good electrical connection on account of the weight of the cell. Copper sulphate crystals are spread evenly over the CLOSED CIRCUIT BATTERIES. 43 bottom of the tray, and the zinc is made to rest on four blocks of paraffined wood at the corners. A parchment diaphragm is sometimes placed above the copper sulphate, and a dilute solution of zinc sulphate, density 1.10, is poured on this till it covers the zinc. These cells or trays may be piled up to the extent of ten. The internal resistance may be as low as 0.2 ohm. The circuit must be kept closed to prevent copper sul- phate reaching the zinc. To secure a fairly constant current, the density of the zinc sulphate must not be allowed to greatly exceed 1.1. Some of the liquid at the Fig. 13. — Sir William Thomson's Tray Battery. top must be withdrawn daily, and soft water must be added in its place. Sir William Thomson's cell was originally designed to work the siphon recorder in submarine telegraphy. 31. Grove's Battery. — The Grove battery consists of a cleft cylinder of zinc immersed in dilute sulphuric acid (1 : 12), and a thin plate of platinum in strong 44 PRIMARY BATTERIES. nitric acid (HN0 8 ) contained in a porous cup. The nitric acid is a powerful oxidizing agent ; and, in con- sequence of this property, it acts as an efficient depolar- izer by oxidizing the hydrogen. The nitric acid is easily decomposed, and the nascent hydrogen readily abstracts oxygen from it. The electric chain may then be represented as follows : — Zn x | H 2 S0 4 | H 2 S0 4 || 2HN0 3 | HN0 3 | Pt. » > After the first step in the chemical reaction this becomes — Zn^l ZnS0 4 | H 2 S0 4 || 2HN0 3 | HN0 2 | H 2 | Pt. On one side zinc sulphate is formed as usual at the expense of zinc and sulphuric acid; while on the other a molecule of nitric acid loses one atom of oxy- gen, becoming nitrous acid (HN0 2 ). As the action pro- ceeds, the nitrous acid may lose another atom of oxygen, hyponitrous acid (HNO) remaining. Or further, the nitric acid may break up entirely, according to the fol- lowing reaction : — 3 H + HN0 3 = 2 H 2 + NO. The products are water and nitric oxide. This last is a gas which takes up more oxygen on escaping into the air, forming the red fuming nitrogen peroxide, N0 2 . These fumes are highly corrosive, and are the most objectionable feature of the Grove cell. When a large current is taken from a Grove battery, the nitric acid has the appearance of boiling, on account of the rapid disengagement of the nitric oxide. The acid is carried CLOSED CIRCUIT BATTERIES. 45 off as a spray, corroding the metallic connections and vitiating the air. This battery should therefore be placed in the open air or in a strong draught, and the connectors should be frequently examined and cleaned. The zinc cylinders must be kept well amalgamated, and the platinum plates should be heated to redness occasionally to prevent their becoming brittle from some unexplained cause. These cells must be taken apart and washed with an abundance of water every time they are used. They have the advantages of high E.M.F. and low internal resistance. The former is from 1.8 to 1.9 volts, and the latter is about 0.15 ohm, with a cell 20 cm. high and 9 cm. in diameter. Such a cell is therefore capable of giving 12 amperes on short circuit, or through an external circuit of no appreciable resistance. Before the introduction of dynamo-electric machines and the storage battery, forty Grove cells, requiring only seven or eight pounds of nitric acid, served the writer for many years whenever a brilliant arc light was needed or projection experiments in spectrum analysis were performed. When the nitric acid becomes dilute by the process of decomposition in the porous cup, the reaction may be quite different from that represented above. The acid may give up its oxygen entirely, with formation of nitrate of ammonium. The action may be represented by the following chemical equation : — 2 HN0 3 + 4 H 2 = 3 H 2 + KH 4 N0 3 . The presence of the salt of ammonia in an exhausted Grove cell can be demonstrated by testing the liquid in the porous cup for ammonia in the usual way, by heating 46 PRIMARY BATTERIES. with, powdered lime and water. A saturated solution of ferric chloride, to which 4 per cent of nitric acid has been added, has been recommended as an excellent substitute for nitric acid in a Grove cell. The E.M.F. is then intermediate between that of a Grove and that of a Daniell. 32. Bunsen's Battery. — Soon after the invention of the Grove battery, Bunsen modified it by substituting in Fig. 14. — Buneen Cell. a prism of baked carbon for the platinum. This is an advantage in point of economy. The E.M.F. is slightly less than that of the Grove. The usual construction of the Bunsen cell is shown in Fig. 14. The chemical action in the Bunsen battery is pre- cisely the same as in the Grove. The hydrogen is CLOSED CIRCUIT BATTERIES. 47 intercepted by the nitric acid, and is thus prevented from reaching the carbon prism by oxidation. Another modification of the Grove cell consists in substituting an iron plate for the platinum in strong nitric acid. On account of the passivity of iron in con- centrated nitric acid, it does not dissolve; and it is strongly electro-negative. When the acid becomes weak, however, by the decomposition due to nascent hydrogen, the acid attacks the iron with disengagement of corrosive fumes. On this account iron is not used in practice for the negative plate. 33. The Bichromate Battery. — If the bichromate of potassium or of sodium in solution is treated with sul- phuric acid, chromic acid is formed. This compound (Cr0 3 ) is not only rich in oxygen, but it gives it up readily to nascent hydrogen. Hence the application of bichromates as depolarizers. An ordinary Bunsen cell may be set up as a bichro- mate cell by placing the amalgamated zinc cylinder in dilute sulphuric acid as usual, and filling the porous cup, holding the carbon prism, with a solution of the bi- chromate salt acidulated with sulphuric acid. Or, since both solutions contain sulphuric acid, the porous cup may be dispensed with entirely, both the zinc and the carbon being immersed together in the strongly acid- ulated bichromate solution. In this case the zinc is usually placed between two flat plates of carbon, an arrangement adopted simply to reduce the internal re- sistance of the cell. The E.M.F. does not differ materi- ally from that of the Bunsen. Fig. 15 represents one of the forms of this cell which has been much used, though it is not to be recom- mended. The zinq is attached to a rod, a, by means of 48 PRIMARY BATTERIES. lass. which it can be drawn up out of the liquid when the battery is not in use. The carbon plates are fastened to a metallic clamp, which is attached to the hard rubber top of the cell. The top of the zinc is covered with an insulat- ing strip to prevent direct contact with the carbons. Many forms of " plunge " battery for bichromate solutions have been devised. These are usually arranged as a battery of four or more cells, with the zincs and carbons suspended from a frame, by means of which they may all be lifted out of the liquid together by a wind- Such a battery is shown in Fig. 16. It is a very Fig. 15. Bichromate Cell Fig. 1G. — Plunge Battery. CLOSED CIRCUIT BATTERIES. 49 convenient form for experimental work in physical dem- onstrations. If the current falls off because of the exhaustion of the liquid in contact with the plates, it may be increased again by lifting the plates, by stirring the liquid, or by blowing air through, as is done in the Byrne battery. One inventor gives a slow motion to the carbon plates by means of a small electric motor. Gendron has recently described a bichromate cell, in which the zincs can be easily replaced without interrupting the current. By a system of automatic valves the exhausted liquid is withdrawn at the bottom, while a constant level is main- tained by the supply. The initial E.M.F. of a bichromate battery is a little in excess of two volts per cell. 34. Chemical Reactions in the Bichromate Battery. — When a solution of bichromate of sodium or of potas- sium is treated with sulphuric acid, a purely chemical reaction takes place, resulting in the formation of chromic acid. Thus : — N"a s Cr 2 O r + H 2 S0 4 = Na 2 S0 4 + H 2 + 2 Cr0 3 . The chromic acid, Cr0 3 , is the useful agent to effect depolarization by the oxidation of hydrogen. The pro- cess is supposed to be represented by the following reaction : — 6 H + 2 CrO a + 3 H 2 S0 4 = 6 H 2 + Cr 2 (S0 4 ) 3 . The final result is, therefore, the production of the sulphate of zinc (at the positive plate), the sulphates of sodium and chromium, and water. It will be observed that, while all the oxygen atoms of a bichromate of sodium molecule unite with hydrogen to form water, only three of the seven are concerned with the removal 50 PRIMARY BATTERIES. of the hydrogen displaced by the zinc. The other four oxygen atoms unite with the hydrogen coming from the four molecules of sulphuric acid, which take part in the reactions written above. Only three-sevenths of the oxygen contained in the bichromate salt, therefore, are useful in removing the polarizing hydrogen; and for every three parts of sulphuric acid which are supplied to act on the zinc, four more must be added to decom- pose the bichromate and to release oxygen. When potassium bichromate is used, a double, sulphate of potassium and chromium, K 2 Cr 2 (S0 4 )4, crystallizes out of the liquid as soon as it becomes saturated with these salts. This is known as chrome alum. The crys- tals attach themselves in a compact mass to the bottom of the jar, and are difficult of removal. 35. The Advantages of Sodium Bichromate over Potas- sium Bichromate. — The advantages arising from the use of the sodium salt in place of the corresponding one of potassium, appear not to have been appreciated till quite recently. But the sodium salt is to be preferred for the following reasons : — First. It contains a larger percentage of available oxygen. The molecular weight of sodium bichromate is 262.4, and of potassium bichromate 294.6. The two molecules contain the same weight of oxygen. For equal depolarizing capacity, therefore, about 11 per cent less of the sodium salt is required than of the potassium. Unless the cost of the sodium salt is more than 10 per cent higher than that of the potassium salt, the former is the cheaper. Second. It is much more soluble. The potassium bichromate must be dissolved by the aid of heat, and not more than about 100 gms. to the litre will remain in CLOSED CIRCUIT BATTERIES. 51 solution when the liquid cools. The sodium salt dis- solves in the cold, and in any quantity desired. A denser solution can therefore be used with two distinct advantages in this respect alone. The first one very evidently is that the battery does not need to be re- plenished with fresh solutions so frequently. The other advantage is not so obvious, but it becomes apparent when attention is drawn to the fact that there is no liberation of gas in this battery to stir up the liquid ; and the exhausted solution in contact with the carbon plates is replaced by iresh portions only by diffusion, unless the liquid is agitated by lifting the plates or by other mechanical means. The denser sodium bichromate solution is not so soon exhausted of useful oxygen, and will therefore maintain a large current with a smaller rate of enfeeblement. Third. The double sulphates of sodium and chro- mium, if indeed they are formed at all, do not crystallize out as in the case of the potassium chrome alum, but remain in solution. The cells are therefore easily cleaned. 36. Directions for Setting up a Bichromate Battery. — For the solution, Bunsen recommends the following proportions : — Bichromate of potassium . . . 77.5 gms. Sulphuric acid 78.5 c.c. Water 750. c.c. The bichromate must first be dissolved by heating the water to boiling. Time will be saved 'by crushing the crystals in a mortar before putting them into the water. After the solution has cooled, the acid may be slowly added. The acid should be poured into the water, and 52 PRIMARY BATTERIES. not the water into the acid. After cooling" again, the solution is ready for use. Reference to the chemical action of this hattery shows that for every molecule of K 2 Cr 2 7 used, seven mole- cules of H2SO4 are needed, provided the depolarizer is entirely exhausted of its oxygen. The molecular weight of K 2 Cr 2 7 is 294.6, and the seven molecules of H 2 S0 4 weigh 686. Hence, to find the weight of actual acid, corresponding to 100 gins, of the bichromate, write the proportion — 100 : x : : 294.6 : 686. Whence x = 232.8. But sulphuric acid of density 1.8 contains 86 per cent of acid. Hence about 271 gms. of 86 per cent acid are required to furnish the 232.8 gms. of actual acid. This is equivalent to 150 c.c, density 1.8. But since the salt in solution cannot all be utilized to effect depolarization, a residue always being left in the spent liquor, the amount of acid may be reduced. It is better to add a small quantity of fresh acid occasion- ally rather than to supply too much at first. If sodium bichromate is used, 200 gms. may be dis- solved in a litre of water, and to this should be added 150 c.c. strong acid. When the battery begins to show signs of exhaustion, an additional 25 to 50 c.c. per litre may be added. For complete exhaustion of. the oxygen from 200 gms. of sodium bichromate, about 600 gms. of 86' per cent acid would be required. This includes the' quantity necessary to form the chromic acid, and to act on the corresponding weight of zinc. If the sodium salt is powdered, it may be put into the water, and the acid added to the solution at once. Com-: CLOSED CIRCUIT BATTERIES. 53 plete solution will quickly take place, and the mixture is ready for use as soon as it cools. 37. The Fuller Bichromate Cell. — The special object in the design of the Fuller battery is the continuous amalgamation of the zinc. It is shown in section in Fig. 17. The zinc, .to which a brass rod covered with gutta percha is attached, is placed in a porous cup, and an ounce (30 gms.) of mercury is poured in. The cup is then filled with water, and is placed in the -glass or earthen jar containing the solution of bi- chromate and acid and the carbon plate. The acid diffuses through the porous cup fast enough to act continuously on the zinc, which has enough mercury surrounding it to keep it well amalgamated. This insures minimum local action and constancy of current, especially if the current is small. Many thousands of these cells have been in use in the Post-office installation in London, and have given good satisfaction. Each cell is said to serve an entire year by replenishing with acid ten times and potassium bichromate five times. At the end of a year the battery is dismounted, cleaned, and furnished with new zincs. 38. Chromic Acid as the Depolarizer. — Instead of em- ploying either of the preceding bichromates for the sup- ply of chromic acid, the acid may be used directly. It may be obtained in the form of a powder, and is soluble in the acidulated Water. Since one molecule of a bichromate furnishes two of Fig. 17. — Tbe Fuller Bichromate Battery. 54 PRIMARY BATTERIES. chromic acid, it will readily be seen that ten-thirteenths as much powdered chromic acid is required as sodium bichromate. The amount of sulphuric acid is only slightly less. Experiment shows that 150 gms. per litre make a very serviceable solution. The initial E.M.F. is then 2 volts. Another modification, known as the Ward and Sloane battery, employs zinc in caustic soda and carbon in a mixture of chromic acid, nitric acid, and common salt. The proportions are as follows : To one-half gallon of nitric acid add one and a half pounds of chromic acid and one pound of salt. This will make one charge for the porous cup of a cell 12 x 12 x 9 inches. The zincs are the equivalent of twenty-four rods half an inch in diameter, and the carbons are equivalent to fifty electric- light carbons. The initial E.M.F. is 2.9 volts. Such a cell has an internal resistance of one-tenth ohm, and will give a current of 10 amperes for 30 hours; final E.M.F., 2.3 volts. The following solution has been found by Mr. J. W- Swan (British Association, 1889) to give the best results : — Nitric acid (density 1.42) . . 1 part by weight. Chromic acid 3 parts " Sulphuric acid 6 " " Water 5 " " The chromic acid is first dissolved in the water ; the nitric acid is then added, and finally the sulphuric. This solution requires ten parts of acid to five of water. It is scarcely possible to avoid wasteful local action with even well-amalgamated zinc in such a con- centrated acid solution. CLOSED CIRCUIT BATTERIES. 55 The suggestion has recently been made to use with bichromates only enough sulphuric acid to decompose the salts and release chromic acid, and then to add at least as much hydrochloric acid as sulphuric. It is claimed that there is less liability of crystallization and less heat with increased steadiness of current. 39. The Partz Acid Gravity Battery. — This zinc-carbon element possesses several points of novelty and exhibits excellent qualities under ap- propriate tests. It is, in fact, the application of the grav- ity principle to an acid de- polarizer. For this purpose a flat carbon plate, with sur- face increased by means of pointed cones, corrugations, or holes, lies in the bottom of the cell ; and a carbon rod, with the proper taper at the lower end to fit tightly into a hole in the plate made to receive it, leads to the positive terminal on top of the cell. The zinc is either a heavy cylinder where a porous cup is employed (Fig. 18), or a large horizontal plate in the form without porous cup (Fig. 19). In the former case the cup is paraffined to a height of two inches from the bottom to prevent entrance of the acid depolarizer. The depolarizer is a sulpho-chromic salt, in which sulphuric acid has been caused to unite with chromic acid in an amorpho-crystalline state. It is supplied to Vig. 18. Partz Acid Gravity Battery « 56 PRIMARY BATTERIES. the cell when everything else is in place, by filling into the vertical tube shown in the cut to the level of the liquid in the cell. The salt slowly dissolves and dif- fuses over the bottom so as to cover the carbon plate. The excitant is either sulphate of magnesium or com- mon salt. The internal resistance is somewhat lower with the latter. Whenever the cell shows a tendency to weaken and fail, it is necessary only to add one or two tablespoon- Fig. 19. — Partz Acid Gravity Battery. fuls of the sulpho-chromic salt through the tube to restore the current to its normal value. After the spent salts have accumulated to such an extent as to interfere with the working of the cell, it is better to turn out the contents, soak the carbons in warm water, amalgamate the zincs, and set up again with fresh solutions. Since the depolarizer is intended to remain in the bottom of the cell, it is apparent that this battery must be left as much as possible undisturbed. CLOSED CIRCUIT BATTERIES. 57 The form of Fig. 19 is set tip by dissolving 11 oz. of magnesium sulphate in the required amount of water and filling the vertical glass tube to the level of the liquid with the sulpho-chromic salt. One of these cells was tested for E.M.F., internal resistance, current, and polarization. The initial E.M.F. was 2.08 volts ; and in the course of an hour on a closed circuit through one ohm external resistance it fell to 1.85 volts, but recovered to 2 volts again in a few minutes after opening the circuit. The internal resistance was 0.82 of an ohm, and the current about 1.04 of an ampere. 40. Taylor's Battery. — This is a zinc-carbon element capable of maintaining a very large current with small diminution of E.M.F. The carbon rods, eight in number, are attached to a well-shellacked wood cap (Fig. 20) and make contact with the brass plate shown on top. The zinc plate has an active surface of 27.5 square inches, is thoroughly amalgamated, and is wholly im- mersed in the dilute sulphuric acid (1: 15). Contact is made with the zinc plate by means of a heavily amalgamated copper wire, shown in the cut. As the E.M.F. between amalgamated zinc and amalgamated copper is very small, the loss from this cause is inap- preciable. On account of the thorough amalgamation of the zinc, the loss due to local action on open circuit is small. The initial E.M.F. is 1.9 volts, the current on short circuit 10 amperes, and the internal resistance as low as Fig. ZO. Taylor's Battery. 58 PRIMARY BATTERIES. 0.18 ohms. The cell shown weighs, charged, 10.5 lbs. (4765 gms.), and has a capacity, it is claimed, of 70 ampere-hours. The depolarizing solution is one of the best substi- tutes for nitric acid, and is rich in oxygen. 41. The Copper Oxide Battery. — It has been remarked that, in general, the best depolarizers are liquid. There are, however, two exceptions which exhibit notable effi- ciency. They are the oxide of copper and the chloride of silver. Both of these solids readily give up their non- metallic element to nascent hydrogen, and by reduction to the metallic state become excel- lent conductors. The copper oxide cell appears to have been introduced by Lalande and Chaperon, and one of the forms was that shown in Fig. 21. The spiral of zinc is immersed in a solution of caus- tic potash or caustic soda, 30 or 40 parts to 100 of water. The upper vertical part of the zinc Gr, where it passes out of the solution, is covered with a caoutchouc tube to prevent local action at that point. The negative con- sists of a cup of sheet iron containing the copper oxide B. To this cup is riveted an insulated copper wire which passes up through the cover and forms the positive electrode. To prevent action upon the alkaline solution by the carbonic acid gas of the air, it is covered with a thin layer of heavy petroleum oil. The height of the glass jar is 15.5 cm., and the diameter 10.5. Fig. 31. Copper Oxide Battery. CLOSED CIRCUIT BATTERIES. 59 The larger pattern of cell is that of Fig. 22. Here the zinc is a helix of rolled metal suspended from an ebonite cover, which is held in place by means of flanges and nuts. The cell is capable of furnishing 12 amperes, and has a capacity of 540 ampere-hours. The copper oxide battery, invented by Lalande and Chaperon, has a capacity for work per unit weight greater than any other, either primary or second- ary. One kilogramme (2.2 lbs.) is able to furnish 255 xlO 10 ergs, or 188,060 foot-pounds. A disadvan- tage is that only a part of the iron surface, consti- tuting the negative plate, is provided with the cu- pric oxide sufficiently near to be of any service in the removal of the hydrogen, which accumulates on all por- tions of the inner metallic surface. The reduced copper, too, is not in good contact with the surface of the iron cell. The conversion of the alkali into a carbonate, by absorption of carbon dioxide from the air, necessitates the closing of the cell against admission of the air, or else the use of the heavy petroleum oil. The larger cell is closed, and has a relief valve of rubber tubing. The chemical reaction taking place may be written in the form already employed in several cases. Before the first step — Fig. 23. — Copper Oxide Battery. Zn, 2NaOH I 2NaOH ! CuO I CuO I Fe. » -> 60 PRIMARY BATTERIES. After the first step, this becomes — Zvi I Na 2 Zn0 2 | 2NaOH | H 2 | CuO | Fe— Cu. Zinc displaces hydrogen from the caustic alkali, form- ing sodium (or potassium) zincate ; while the ejected hydrogen, travelling with the current, arrives at the cupric oxide, from which it abstracts oxygen, and me- tallic copper is thus reduced at or near the iron of the cell. 42. The Edison-Lalande Battery. — Recognizing the good qualities of the copper oxide as a depolarizer, Edison has devised a form designed to meet the objec- tions noted above. The copper oxide is employed in the form of a compressed slab, which, with its connect- ing copper support, serves also as the negative plate. Two of these plates are enclosed in a copper frame, on the longer arm of which girai^ailllllUIUlllllllUlllUIIH Hill * s ^he binding post. A If |^| I' 1 III ill ' iart * ru bber safety plug in ill IMIIIIIIIDIIIHIUIHIIIII the middle prevents the zinc plate on either side from making contact with the cop- per oxide and copper sup- porting frame. One, two, or three of these copper oxide plates are used, according to the size and capacity of the cells. The weight of the oxide plate for a 15 ampere- hour cell is 2 oz., and for a 300 ampere-hour cell 2 lbs. Fig. 23. — Edison-Lalande Battery. CLOSED CIRCUIT BA TTERIES. 61 Fig. 23 is a 300 ampere-hour cell complete. The cover is porcelain, with small openings for the zinc and copper terminals. Since this cover does not exclude the air, the formation of a carbonate is prevented by pouring on top of the solution of caustic potash (KOH) a small quantity of heavy paraffin oil, so as to form a layer about one-fourth of an inch deep. It is of vital importance that this oil should not be omitted. If it is not used,, the life of the cell is reduced fully two- thirds. If the cell is required to furnish a strong current at once, it should be short-circuited for ten or fifteen min- utes the first time it is used. By this means enough metallic copper is reduced to form a good conducting surface, and the internal resistance of the cell falls to its normal working value. Subsequent short-circuiting should, of course, be avoided, especially because the internal resistance is very low, and the large current flowing causes a great waste of material in the cell. In recent cells the device has been resorted to of reducing a superficial film of copper on the oxide before it is sent from the factory. The 300 ampere-hour cell shown is 11^ inches high and 5f inches in diameter. Its internal resistance is about 0.03 ohm, and its working E.M.F. about 0.7 volt. It is capable of delivering 14 amperes. On open cir- cuit there is practically no local action. The zinc should be well amalgamated. Pressed copper oxide plates have also been used abroad in a cell having the form of Fig. 24, in which the compressed plates, B, are held in contact with the sheet iron, A, by rubber bands. The cell is closed to prevent entrance of air, but has a. relief valve, H, for 62 PRIMARY BATTERIES. the escape of accumulated gas. The small zinc sur- face, V, means relatively large internal resistance. The r ^ plates are made by mixing cop- per oxide with from 5 to 10 per cent of magnesium chloride, and heating the thick mass in an iron mould. 43. The Chloride of Silver Cell. — Marie - Davy appears to have been the first to suggest the use of silver chloride as a depolarizer about 1860 ; but it was brought into prominence by the investigations of War- ren de la Rue, who constructed a battery of this kind contain- ing 15,000 cells. Fig. 84. -Battery with Compressed The e l ements are zm0 an( i Plates of CuO. silver, and on the silver is cast the silver chloride, which is readily reduced to metallic silver by nascent hydrogen. The chloride of silver is easily melted in a porcelain crucible, and may be cast on a silver wire in a hard carbon mould. Silver foil has sometimes been cast in the chloride to give better conductivity. The exciting fluid of De la Rue's battery is ammonium chloride, and contains 23 gms. to one litre of distilled water. A denser solution dis- solves silver chloride. The silver and its chloride are surrounded with a small cylinder of vegetable parch- ment paper (Fig. 25), to prevent short circuits internally, and the zinc rod and silver wire are held in a paraffin stopper. The silver wire of one cell is wedged into the zinc rod of the next. CLOSED CIRCUIT BATTERIES. 63 The following chemical action takes place : — Zn„ 2NH 4 C1 | 2im 4 Cl | 2AgCl | Ag„ = Zn,_i | ZnCl 2 | 2KH 4 C1 | 2NH 4 C1 | Ag, ,»+2- This may be considered the normal action ; but where the cell is worked hard, it may happen that the amnio- nic chloride loses chlorine faster than it recovers it from the silver chloride ; and the ammonium breaks up into ammonia and hydrogen. The ammonic hydrate thus Tig. 85. — Silver Chloride Battery. formed is capable of dissolving silver chloride, with the formation of ammonio-silver chloride. The hydrogen may reduce silver chloride with production of hydro- chloric acid. This acid increases local action. Under such conditions gas may be liberated in the cell, and pro- vision must be made for its escape ; or the cell must be made very strong and must be securely sealed. The initial E.M.F. of a silver chloride cell is about 1.1 volts. Its internal resistance falls very rapidly upon 64 PRIMARY BATTERIES. first closing the circuit, on account of the reduced silver. It polarizes but- slightly, and recovers promptly. It is employed chiefly for testing purposes; sometimes for physicians' use. But it should never be put into service requiring anything more than small currents. Upon standing, the zinc is liable to become coated with a thin, adherent film of the oxychloride of zinc, offering high electrical resistance. 44. Modifications of the Silver Chloride Cell — The mod- ifications thus far introduced consist in the substitution of some other exciting liquid for the ammonic chloride. Thus caustic potash or soda has been used by Scrivanoff . The chemical reaction is then the same as with the copper oxide cell, except that the hydrogen displaced by zinc unites with the chlorine of the depolarizer, form- ing hydrochloric acid. A secondary reaction is thus possible, due to the action of the acid on the zinc. There is then greater liability of local action than if the cell were set up with sal-ammoniac. The excitant may also be zinc sulphate. The dis- placement process taking place is as follows : — Zn. | ZinS0 4 | ZnSO , | 2AgCl | Ag s = Zn,.! | ZnS0 4 | ZnS0 4 | ZnCl 2 | Ag y+2 . In this case zinc chloride is formed at the expense of silver chloride, and the energy appearing in an electrical form may be represented as due to the difference be- tween the heat of combination of zinc chloride and silver chloride. The initial E.M.F. with caustic potash is 1.64 volts ; with zinc sulphate, 1.16 volts ; and with zinc chloride (Gaiffe), 1.01 volts. CLOSED CIRCUIT BATTERIES. 65 It should be remarked that silver chloride is soluble to some extent in the chlorides of the heavy metallic elements. When the liquid contents of the cell contain as much as one part of concentrated zinc chloride in ten parts of water, the silver ehloride is dissolved in quanti- ties which are quite appreciable. Local action then ensues, due to the displacement of silver by zinc, and the zinc rod or plate quickly blackens. The marked efficiency of silver chloride as a depolar- izer is perhaps due to its slow or partial dissolving in the exciting liquid, since liquid depolarizers are, in gen- eral, more effective than solid ones. A weak solution of ammonic chloride may not attack the solid silver chloride. Hence local action does not take place so long as these cells have not been placed in use ; but im- mediately upon closing the circuit through them zinc chloride is formed, and thereafter local action begins to exhaust the silver chloride with blackening of zinc. So that silver chloride cells that have been much used will not stand on open circuit without waste. Moreover, their internal resistance will increase if the zinc becomes encased in the film of oxychloride before mentioned. 66 PRIMARY BATTERIES. CHAPTER V. OPEN CIRCUIT BATTERIES. 45. The Leclanche Cell. — The present chapter will be devoted to open circuit batteries in which a solid depolarizer is used. At the head of this list stands the Leclanche" cell, so called from the name of the inventor. Metallic oxides had been proposed as depolarizers pre- vious to the invention of Leclanche", but without prac- tical results. Thus, with zinc in dilute sulphuric acid and platinum surrounded with the peroxide of lead in a porous cup, Beetz found an E.M.F. of 2.4 volts. Dur- ing 30 minutes short circuit this fell to 1.4, but recov- ered after five minutes rest to 2.16. It is evident that this high E.M.E. is due not only to the oxidation of the zinc, but to that of the hydrogen as well, both chemical processes contributing to the electromotive stress in the same direction. The chief disadvantage in the employment of lead peroxide as a depolarizer lies in the fact that the re- duced lead is converted into lead sulphate. This accu- mulates on the negative plate and has the effect of largely increasing the internal resistance of the cell. It is worthy of note in this connection that one of the more recent forms of storage batteries is composed essentially of the elements used by Beetz ; namely, zinc and lead in an acid solution of zinc sulphate. The depolarizer of the Leclanche - cell is manganese OPEN CIRCUIT BATTERIES. 67 dioxide (Mn0 2 ). It is not used as a powder, but in granules mixed with broken gas carbon to increase the conductivity. The negative plate is baked carbon, and is surrounded with the mixed manganese dioxide and broken carbon, packed in a porous cup, which is finally sealed with pitch, with two small vent tubes inserted. The typical Leclanche" cell, with its porous cup (Fig. 26), has a glass jar moulded with a lip, in which is placed the zinc rod. The carbon plate is usually sur- II m GONtf I Fig. 20. — Leclanche Cell. mounted with a lead cap, cast on the carbon, and hold- ing the binding post of the positive terminal. The cut exhibits a new connection, designed to avoid corrosion of the lead cap. The size of the zinc rod, which never exceeds half an inch (1.25 cm.) in diameter, indicates large internal re- sistance, and shows that this cell is designed to furnish only small currents through considerable external resist- ance. The amount of energy held potentially in the cell is represented approximately by the weight of the zinc. 68 PRIMARY BATTERIES. The exciting liquid is amnionic chloride, the sal- ammoniac of commerce. To set up the cell, five or six ounces of best sal-ammoniac are dissolved in water. Water, or water containing sal-ammoniac, is also poured into the porous cup through one of the vent tubes. If water alone is added, the cell must stand for about 24 hours before use, to permit the diffusion of the ammonium salt through the. porous cup, unless there are holes in it which allow the liquid to pass in rapidly. An incidental advantage of this cell is that the dif- fusion of the liquid through the porous vessel, which serves only to hold the depolarizer and the broken carbon, is of positive utility ; while in two fluid cells the diffusion of the two liquids through the pores is an undesirable feature. The initial E.M.F. of the Leclanche" cell varies from 1.4 to 1.7 volts, and the internal resistance from about 0.4 to 2 ohms. 46. Chemical Reactions in the Leclanch6 Cell. — Theo- retically no chemical reactions take place so long as the circuit remains open, inasmuch as the cell contains neither acid nor an acid salt. But when the circuit is closed, zinc displaces ammonium from the amnionic chloride, and the ammonium breaks up into ammonia gas, which is set free and escapes after the liquid be- comes saturated, and hydrogen which is oxidized by the manganese dioxide. These chemical changes may be represented by the following equation : — Zn x I 2NH 4 C1 | 2NH 4 C1 II 2Mn0 2 1 C = Zn x _! | ZnCl 2 | 2NH.C1 || 2NH S | Mn 2 3 | H 2 | C. OPEN CIRCUIT BATTERIES. 69 If the liquid is allowed to become supersaturated by evaporation, a double salt of the chlorides of zinc and ammonium is liable to crystallize on the zinc. This reduces the E.M.F, and increases the internal resist- ance. A small quantity of hydrochloric acid will usually dissolve these crystals. When a Leclanche" cell is left undisturbed for some time, it will be found that the zinc rod is eaten away at the surface of the liquid, and that it is conical in shape, with the larger end of the cone at the bottom of the zinc. The excessive waste at the surface is doubtless due to oxidation, but the coning is the result of a peculiar local action sometimes seen in other forms of battery. The double chloride of zinc and ammonium gradually settles to the bottom of the cell, becoming progressively denser and denser as the bottom is approached. Now zinc in a solution of amnionic chloride is positive to zinc in zinc chloride; if the latter liquid contains amnio- nic chloride also, the resulting E.M.F. is smaller, but still appreciable. 1 Hence local circuits are formed be- tween the upper and lower portions of the zinc rod, the upper portions playing the part of the zinc in a simple voltaic combination. The zinc plates of the copper oxide battery show a similar thickening from the liquid surface downward. The heavy zincate formed can be seen settling toward the bottom of .the cell, and local action sets in, as already explained. 47. The Prism Leclanche Battery. — The prism form of the Leclanche" cell was devised for the purpose of dis- pensing with the porous cup. The carbon plate is sus- 1 See Experiments of Chapter IX. 70 PRIMARY BATTERIES. pended from the cover (Figs. 27, 28), and attached to it by rubber bands are the two agglomerated prisms, containing the depolarizer. They consist of 40 parts granulated manganese dioxide, 52 parts granulated carbon, 5 parts gum shellac, and 3 parts acid potas- Fig. 87. The Prism Leclanche' Battery. Fig. 28. sium sulphate. The mixture is heated to 100° C, and then compressed in moulds under a heavy press- ure. This form of Leclanche' cell has not met the expecta- tions entertained at its first appearance. It appears not to be as efficient and durable as the original form, and has not come into general use in this country. OPEN CIRCUIT BATTERIES. 71 48. The Closed Leclanche Cell. — When an open Le- clanche" cell is kept in a dry place the liquid evapo- rates, and the solution becomes more concentrated, with greater liability of crystallization at the surface and consequent creeping of the salts upward toward the top. To avoid this difficulty, closed cells (of which Figs. 29 and 30 are ex- amples) have been de- vised. In the former, the cover is wood Sat- Fig;. 29. Closed Leclanche Cell. Fig. 30. urated with paraffin and attached to the porous cup, but removable from the outer jar. So also the zinc is held loosely in the cover, and can be taken out. The cover fits down on a shoulder in the top of the jar, and a soft rubber ring makes it tight. In the latter form the porous cup is made with a flange (Fig. 31), which rests ^ipon the top of the jar. 72 PRIMARY BATTERIES. Fig. 31. Closed Leclanche Cell. Both the jar and the flange are paraffined, so that a close joint is made. The zinc passes through an opening in the cell specially provided for it. This is made tight by a piece of soft rubber tubing enclosing the rod at the point where it passes into the jar. Two. or three other modifications of details may be noted in these cells. In the one of Fig. 29, the porous cup, which is unusually large, has in the bottom three large holes covered with burlap. When the cell is set up, the sal-ammoniac solution enters at once, and the cell is ready for use. The porous cup (Fig. 31) has a small hole in the bottom to admit the liquid, and two holes, shown in the cut, on either side of the carbon at the top. The carbon has a special connection by means of a bolt and lock nuts, which serve their purpose satisfac- torily. A stop in the bottom of the glass jar prevents contact be- tween the zinc and the porous cup. The two water marks on the jar serve as a convenient guide in filling. Each cell re- quires 4 oz., or 120 gms., of sal-ammoniac. Both of these types of battery show an unusually Fig. 38. — The Microphone Cell. OPEN CIRCUIT BATTERIES. 73 high E.M.F., and have done excellent service in the hands of the writer. 49. Leclanche Cells with Carbon Cup. — It is entirely practicable to dispense with the unglazed porous cup, and to make a carbon cylinder serve as a receptacle for the manganese dioxide. Two such cells are represented in Figs. 32 and 33. Both of these are loosely covered, to prevent evaporation, and have the depolarizer en- closed by carbon. The zinc of the latter is a cleft Fig. 33. — Samson Battery. Fig. 34. — Ziifo and Carbon. cylinder (Fig. 34), and the carbon cup is corrugated to secure a larger surface. Both the polarization and recovery of these cells are not so rapid as in other forms of Leclanche" cells, but they are more nearly con- tinuous or uniform. A marked feature is the low inter- nal resistance. It is only slightly over 0.3 ohm, and is no lower inthe second than in the first, though the zinc cylinder has so much larger surface than the rod. The intervening distance is greater in No. 33, thus offsetting the larger surface. 74 PRIMARY BATTERIES. With an external resistance of 5 ohms, the loss of potential in the interior of these cells is only 0.07 or 0.08 of a volt, or about 5 per cent of the total E.M.F. of the cell. They show, therefore, high commercial effi- ciency. 50. Leclanche Cell with Agglomerated Carbon. — In the cell shown in Fig. 35 the manganese dioxide appears to be incorporated with the carbon in the paste, and an agglomerate is thus produced by baking. This cell is effectively closed, and the zinc is insulated by a special glass sleeve passing through the carbon cover. BJlliEBifl ^ * u £ on *^ e z * nc roc ^ ^ ts * nto a I It IBM corresponding socket in the glass, I II 11 (HIM I an d serves the double purpose of II HI III homing the zinc up from the bottom iliSlll!!' '''''"HI of the cell and preventing its turn- ing round when the connecting wire is screwed fast to the negative ter- minal. The agglomerated carbon cylinder has a long cleft on either side, and the zinc rod hangs in the centre. The glass insulator holds the zinc somewhat rigidly, and prevents any contact between it and the carbon. This cell exhibits the same peculiarities of moderate but progressive polarization and good recovery as those of the last section. It has a somewhat higher internal resistance, which is, however, less than that of the ordi- nary Leclanche* element. 51. Roberts' Peroxide Battery. — The elements are amalgamated zinc, carbon surrounded with an agglom- erate of peroxide of lead, and a solution of chloride of Fig. 35. — Cell with Ag- gloraerated Carbon. OPEN CIRCUIT BATTERIES. 75 sodium, to which is added a small quantity of bichro- mate of sodium. The E.M.F. is 1.8 volts. • The agglomerate is made by adding minium (red lead) to powdered permanganate of potassium and hydrochloric acid, in quantity sufficient to form a semi- liquid paste. By the combined action of the acid and the permanganate, the Pb 2 3 is converted into lead peroxide (Pb0 2 ). The paste is then introduced into a mould containing a carbon electrode ; and when after a few minutes it has set, it is withdrawn from the mould and dried at the temperature of the air. By this means a mass is obtained as dense as carbon. The bichromate is added to the exciting liquid for the purpose of converting the chloride of lead in the agglom- erate into an insoluble chromate. The partly soluble chloride would form a deposit of lead on the zinc. In the action of the battery, zinc displaces sodium with the production of zinc chloride and sodium hydrate. Hydrogen is released in the formation of the hydrate, and this abstracts oxygen from the lead peroxide. The internal resistance of such cells is large on account of the presence of insoluble lead salts. 52. The Sulphate of Mercury Battery. — Marie" Davy first proposed the use of the sulphates of mercury as the depolarizing agent. For commercial purposes the acid sulphate is used, containing probably both the mer- curic and the mercurous salts. These solids are only slightly soluble, and are therefore slow-acting depolar- izers. The cell has various forms, but always contains zinc as the positive plate, and carbon, surrounded with the mercury salt, as the negative. The form in which it is most used is for medical pur- poses. The carbon is at the bottom of a moulded rubber 76 PRIMARY BATTERIES. case. On this is placed the mercurial salt with a little water. The amalgamated zinc plate is laid on top and is brought into contact with a platinum wire in the body of the rubber cell, and connection is thus made with the electrode. Usually two such cells are mounted together in series. The E.M.F. is about 1.45 volts. 53. The Fitch "Chlorine" Battery. — In Mr. Fitch's original battery the depolarizer was one of the chlorides of mercury ; but in the process of improve- ment the chloride has been replaced by the chlorates of potassium and so- dium. The excitant is composed of the chlorates of potas- sium and sodium and sal - ammoniac, " mixed in their proper combining proportions." Two forms, shown in Figs. 36 and 37, differ only in the extent of carbon surface exposed, and therefore in their internal resist- ance. The internal resistance of the form with carbon cylinder is about 0.35 ohm when the current flowing is 0.2 ampere, or with an external resistance of 5 ohms. Each package of the excitant weighs 145 gms., or 5 oz. About three-quarters of this is amnionic chloride, the remainder being the chlorates. Fig. 36.— The Fitch Battery. OPEN CIRCUIT BATTERIES. 77 The larger cell requires four packages of excitant, each equal to the above. By accident, three of these cells were left on a closed circuit of 75 or 80 ohms for 2375 hours in long-distance tele- phone service. This is about 20 ohms per volt. During this three months ser- vice, their efficiency had not decreased sufficiently to be noticed in using the transmitter. When this cell is exhausted by use, clean thoroughly the jar, the carbon, and the cover ; and after drying, replace the zinc with a new one and supply a fresh solution of the ex- citant. The battery is then again ready for extended service. In case of accidental short-circuiting, extreme cold, or very hard service, crystals of spent residue may form on the zinc and carbon. These may be removed by adding to each cell 1 oz., or 30 gms., of hydro- chloric acid. More than this should never be added at one time, and then only when the accumulation on the plates demands it. Otherwise local action will take place on account of the presence of the acid. Fig. 37. —The Fitch Battery. 78 PRIMARY BATTERIES. CHAPTER VI. BATTERIES 'WITHOUT A DEPOLARIZER. 54. The Smee Cell. — The oldest battery of any prac- tical value without a depolarizer is the Smee (Fig. 38). The positive plates of this cell are zinc, enclosing be- tween them, with proper insulation, a negative of thin silver, corrugated and covered with platinum in a very finely divided state. The excitant or electrolyte is dilute sulphuric acid; and the purpose of the roughened surface of the silver is the mechanical dislodgement of the hydrogen as fast as it is released at the negative plate, since hydrogen is found to be much more easily detached from a rough surface than from a smooth one. The silver plate may be pre- pared as follows: Obtain thick silver foil and roughen the sur- face lightly with fine glass-paper, or by brushing over with strong nitric acid. Unless the surface is rough- ened the platinum black will not adhere. Connect the silver plate, by means of a copper wire, with a small slip of zinc, and insert the silver in a vessel of dilute Fig. 38. — The Smee Cell. BATTERIES WITHOUT A DEPOLARIZER. 79 acid, to which has been added a few drops of platinic chloride. The zinc slip should then be merely touched to the dilute acid at a point remote from the silver. The slight current thus produced will be sufficient to decompose the platinic chloride, and the platinum will gradually deposit on the silver and color it. Then add more of the platinum salt, and insert the zinc deeper into the liquid. Gradually increase the current till the surface of the silver plate is covered with a black coat- ing of finely divided platinum. The platinic chloride may be prepared by dissolving scrap platinum in a mixture of two parts hydrochloric acid to one of nitric acid, and gently warming for some time. For the above use it is not necessary to drive off the acid or to crystallize the salt. 1 A negative plate for the Smee cell has been formed of copper, with the surface roughened by electro-deposi- tion, then plated with silver, and finally platinized. It is said, however, that the silver plating is liable to be porous, and that the acid in time works through to the copper. Also, that the copper dissolves at the edges and is deposited again on the silver. 2 55. The Sea Salt Battery. — A battery which is said to have done good service has been made with sea salt and powdered alum, in the ratio of five parts to two, dis- solved in water, as the excitant. The elements were zinc and carbon, the latter having a very large surface. Zinc chloride and zinc sulphate are formed, and hydro- gen is set free, with formation of sodium and potassium hydrates. Exactly what part the alum takes in the reactions is uncertain and obscure. But such cells are capable of 1 Sprague's Electricity, p. 92. 2 phil. Mag., May, 1840. 80 PRIMARY BATTERIES. intermittent service for certain classes of work requir- ing only small currents. 56. The Law Battery. — In this battery, and in others of similar design, reliance is placed upon a large carbon surface to effect depolarization mechanically. The nega- tive consists of a double cleft cylinder of carbon, with the zinc rod hanging well within the cleft (Fig. 39). The carbon has a surface of about 145 square inches, and the internal resistance is 0.4 ohm when the current flowing is 0.2 ampere. The cell is effectively closed by an insu- lating cover, so made that by a partial turn it locks down tightly against a soft rubber ring. The jar is of flint glass, annealed, and its capacity is one and one-third quarts, or one and a half litres. Sal-ammoniac is the excitant, and each cell takes one litre of the solution containing 150 gms., or 5 oz., of the salt. A renewal of an exhausted cell requires only a new zinc rod and a fresh solution of sal-ammoniac. The spent solution should always be thrown out, and the double carbon cylinder should be thoroughly soaked in water and then exposed to the sun and air, to remove the absorbed salts. This cell is neat, clean, durable, and efficient. For hard work it polarizes more continuously than a Leclanche* cell, but for light currents the polarization is not suffi- cient to be noticeable. The initial E.M.F. is 1.37. 57. The Diamond Carbon Battery. — The negative of this cell is composed of seven rods of soft carbon, 5.5 Fig. 39. —The Law Battery. BATTERIES WITHOUT A DEPOLARIZER. 81 inches long and five-eighths of an inch in diameter, set into a soft metal top and secured by a set screw, in the manner shown in Fig. 40. The metal top is cast round a porcelain insulator through which passes the zinc rod. The zinc is kept from falling too low by an iron cross- pin, and a rubber ring closes the annular opening in the Fig. 40. — "Diamond" Carbon Battery. porcelain round the zinc. Another rubber ring at the bottom of the zinc prevents contact with the carbons. The tops of the cells are covered with paraffin or bees- the inside of the cover and the upper ends of the wax: carbons are also paraffined. Care should be taken not to allow any of the solution of the sal-ammoniac to get on the cover, otherwise the crystallization and creeping 82 PRIMARY BATTERIES. of the salts produce a short circuit, and the cell exhausts itself on apparently open circuit. The internal resistance is only about 0.25 ohm, with 5 ohms external resistance ; the polarization is continu- ous and progressive, as in all cells of this class, but the recovery is very good. The initial E.M.F. is 1.36 to 1.39 volts. 58. Cylinder Carbon Batteries. — In addition to the cylinder carbon battery already described, attention may be drawn to two others (Figs. 41 and 42). In the former, the carbon cylinder and cover of the jar are made in one piece, and the cylinder in both is cleft for free diffusion of the sal-ammoniac solution. The oval form of the Laclede (Fig. 42) has no advantage, except increased carbon surface. The connection with the binding post is made in both cases in such a way as to render corro- sion by capillary ascent of the liquid quite remote. A greater danger in all these cells arises from careless handling after they are set up, during which the liquid splashes up against the top and over the porcelain insulating the zinc. The initial E.M.F. of all carbon cells without depolarizer appears to be about the same,— between 1.3 and 1.4 volts. They quickly drop below this value with a current of two-tenths of an ampere, and subsequently rise but little above a single volt. The ease and cheap- Fig. 41.— The Cylinder Cell. BATTERIES WITHOUT A DEPOLARIZER. 83 ness with which they may be restored to nearly their initial efficiency after exhaustion constitute a strong commendation in their favor. Fig. 48. — The Laclede Battery. 59. The Gassner Dry Battery. — A large part of the most recent batteries appearing as candidates for public favor are of the so-called dry type. They contain the excitant in the form of a paste, the composition of which is in most cases a secret. Their convenience commends them to those having no technical knowledge relating to batteries, and they are very useful in situa- 84 PRIMARY BATTERIES. tions precluding the use of unsealed cells with liquid electrolytes. But their store of available potential energy is, in general, smaller than that of batteries containing a larger quantity of fluid. One of the oldest cells of the dry type is that of Dr. Gassner (Fig. 43). The zinc, composing the positive element, is the containing vessel. It is usually covered with paper, or is enclosed in a paper box. The negative element is carbon, and it occu- pies about one-half the space in the cell. The paste, which is filled in between the zinc and the car- bon in the Gassner cell, has the following composition : "Oxide of zinc, 1 part, by weight ; sal-ammoniac, 1 part, by weight ; plaster, 3 parts, by weight; chloride of zinc, 1 part, by weight; water, 2 parts, by weight. The oxide of zinc in this composition loosens and makes it porous, and the greater porosity thus obtained facilitates the interchange of the gases and diminishes the tendency to the polar- ization of the electrodes." The initial E.M.F. of this cell varies but little from 1.3 volts. It polarizes very rapidly on so low an ex- ternal resistance as 5 ohms; while the internal resist- ance, which is different for cells of different size, is very irregular during the working of the cell, probably on account of the slow and irregular diffusion of the prod- ucts of the chemical action. Fig. 43. — Gaesner Dry Cell. BATTERIES WITHOUT A DEPOLARIZER. 85 Such cells should be employed for intermittent ser- vice, where the circuit is kept closed for short periods only. In such situations they will doubtless prove effi- cient and durable. Their convenience, particularly in the hands of unskilled persons, is much in their favor. Meserole's composition for a dry battery is the fol- lowing : — Charcoal, 3 parts ; mineral carbon or graphite, 1 part ; peroxide of manganese, 3 parts ; white arsenic oxide, I part; a mixture of glucose and dextrine or starch, 1 part ; hydrate of lime, dry, 1 part — all by weight. These are intimately mixed and worked into a paste of proper consistency with a solution composed of equal parts of a saturated solution of chloride of ammonium and common salt, to which are added one-tenth of the volume of a solution of bichloride of mercury and an equal volume of hydrochloric acid. The fluid is added to the dry mixture gradually, and the mass is well worked to insure uniformity. 86 PRIMARY BATTERIES. CHAPTER VII. STANDARDS OP ELECTROMOTIVE FORCE. 60. Latimer Clark's Standard Cell. — The original Latimer Clark normal element was described for the first time in a paper read before the Royal Society, June 19, 1873. 1 The metallic elements are pure zinc in zinc sulphate, and pure mercury in contact with mer- curous sulphate (Hg 2 S0 4 ). The mercury was placed in the bottom of the cell, and contact was made with it, either by passing a platinum wire down through a small glass tube in the cell itself, or else through one blown on the cell near the bottom. The zinc sulphate solution was made by boiling an excess of pure zinc sulphate crystals in distilled water, and decanting the clear solution off from the crystals after cooling. On the mercury was poured a thick paste, made by boiling mercurous sulphate with the solution of zinc sulphate, saturated in the manner just described. Into this paste dipped a zinc rod, or else a plate of pure zinc rested on its surface. Special stress was placed on the boiling of the paste with zinc sulphate solution for the purpose of expelling the air. The cell was imperfectly sealed with a paraffin stopper. 1 Philosophical Transactions, 1874. STANDARDS OF ELECTROMOTIVE FORCE. 87 The normal E.M.F. of this cell, according to Clark, was 1.457 volts. But this was on the basis of the British Association (B.A.) unit of resistance. Now the unit of E.M.F. varies directly as the* unit of resist- ance. If, therefore, the true ohm, which is represented, according to our latest knowledge, by the resistance of a column of pure mercury of one square millimetre cross-section, and 106.3 cm. long at 0° C, is 1.014 times the B.A. unit, then the true volt is also 1.014 times the B.A. volt. Hence, to reduce 1.457 B.A. volts to true volts, divide by the above ratio. The result is 1.437 true volts. This is only 0.002 volt higher than the later value assigned by Lord Rayleigh, as the result of his extended observations. 61. Lord Rayleigh's Form of the Clark Element. — The original Clark cells exhibited certain abnormal and irregular values both of E.M.F. and temperature co- efficient. A thorough investigation of the Clark cell was therefore undertaken by Lord Rayleigh, and the results were published in the "Philosophical Transac- tions of the Royal Society," Part II., 1885, under the title of " The Clark Cell as a Standard of Electromotive Force." This paper was supplementary to one published in the same place in 1884 on " The Electro-Chemical Equivalent of Silver, and the Absolute Electromotive Force of Clark Cells." Only a brief summary of results of this very important investigation can be given here. The E.M.F. of a Clark cell may be too high (1) be- cause the paste is acid ; (2) because the zinc sulphate solution is not saturated. The first fault will cure itself in the course of a month or so. The E.M.F. may be too low (1) because the cell has become dry ; (2) because the solution is supersaturated ; 88 PRIMARY BATTERIES. (3) because the mercury is not pure. The cell loses liquid because of imperfect sealing. Paraffin cracks away from the glass. Lord Rayleigh recommends marine glue. • Supersaturation results from heating the solution or the paste. The strong solution will then cool without any deposit, or will throw down an abnor- mal hydrate. The presence of crystals does not prove that the solution is not in the state of supersaturation, unless it can also be proved that these crystals are those of the normal hepta-hydrated salt. The addition of a few crystals of the normal zinc sulphate will always cause the excess of salt held in solution at a given tem- perature to crystallize out. Respecting the presence of other metals in the mer- cury, it is sufficient to notice only that of zinc. Zinc opposed to pure mercury, without the presence of Hg 2 S0 4 , gives an uncertain E.M.F. of about 1.186 volts. But when the mercury contains one part of zinc in 5,900,000, the E.M.F. falls to 0.513 volt; and with one part zinc in 200,000 it becomes only 0.124 volt. 1 With zinc opposed to pure mercury in a zinc sulphate solution, the E.M.F. is not constant from hour to hour, and is altered by the passage of a minute quantity of electricity which would be insufficient to produce the least effect upon a cell provided with mercurous sul- phate. So marked is the action of the mercurous sulphate in repurifying the mercury, that. Lord Ray- leigh suggests that this may be its principal office in the Clark cell; and he attaches the greatest im- portance to purity of mercury. " It is clear," he says, 1 "On the Electromotive Force of Mercury Alloys," Journal So- ciety Telegraphic Engineers, Vol. VIII, 1879. STANDARDS OF ELECTROMOTIVE FORCE. 89 "that the mercurous sulphate has the property of freeing the mercury from the smallest contamination with zinc." Lord Rayleigh's cell (Fig. 44) is made as follows: A small tube has a platinum wire sealed through the closed end. On this is poured enough pure mercury, dis- tilled in vacuo, to cover the platinum effectively. The paste which covers the mercury is prepared by rubbing together in a mortar 150 gms. mer- curous sulphate, 5 gms. zinc carbonate to neu- tralize acid, and as much zinc sulphate solution, saturated by standing in a warm place, as will i 7i • i , Sealing Wix — - make a thick paste. After the Carbonic acid Fig- ; ^irr.^-.T-. -iPura Bo.-- ■Pt.Wire H 90 PRIMARY BATTERIES. The E.M.F. of a Clark cell, constructed as above, Lord Rayleigh found to be 1.435 true volts. This is equal to 1.438 legal volts, corresponding with the legal ohm, or to 1.455 B.A. volts at 15° C. Using a silver voltameter as a secondary standard, the writer found a Clark cell, made in Berlin after Latimer Clark's directions, to have an E.M.F. of 1.437 legal volts at 15° (strictly 1.434 at 18° C). The value of the temperature coefficient was also investigated by Lord Rayleigh. It was found to vary considerably for different individual cells ; but for cells with saturated solutions the following equation can lead to no appreciable error : — E = 1.435 {1 - 0.00077 (* - 15) } : t is the temperature of the cell. Latimer Clark found a temperature coefficient of 0.06 per cent per degree C. for temperatures within 10° on either side of 15. For higher temperatures he observed a diminution of the coefficient; so that for the whole range of observations, extending up to 100° C, the coefficient was 0.055 per cent per degree C. 62. A Standard Clark Cell with Low Temperature Coefficient. — The objections to Lord Rayleigh's form of the Clark normal element are : (1) the temperature coefficient is high and apparently variable ; (2) it is not constructed in such manner as to keep the zinc and metallic mercury out of contact ; (3) the contact of the zinc and the mercurial salt permits of local action whereby zinc replaces mercury. Respecting the first objection, the method to be pur- sued in reducing the temperature coefficient is suggested by the fact, now well known, that the E.M.F. decreases STANDARDS OF ELECTROMOTIVE FORCE. 91 with an increase in the density of the zinc sulphate solution. Hence, if the solution is saturated at 30° or 40°, upon a lowering of temperature the excess crystal- lizes out with a decrease of density. The reverse pro- cess takes place with rise of temperature, with the additional disadvantage that time is required for the diffusion of the redissolved salt. The temperature coefficient in such a cell is therefore made up of two parts : one a real temperature effect, the other a second- ary change resulting from a variability in the density of the zinc sulphate solution. A rise of temperature lowers the E.M.F. by increasing the density of the solution in addition to the direct primary effect of the temperature change. The slowness of diffusion when the temperature rises makes the coefficient for a rapid rise of temperature smaller than for a slow one. Thus Professor Threlfall, 1 investigating Clark cells made in accordance with Lord Rayleigh's directions, found the coefficient to be 0.000402 for a rapid rise of temperature from 21° to 34° C. This is less than half the value found by Lord Rayleigh between the same temperatures. The magnitude of the temperature coefficient depends, then, upon the temperature at which the zinc salt is saturated ; and, because of diffusion, upon the rapidity of the temperature change. To obviate these difficulties the zinc sulphate should be saturated at some definite temperature lower than any at which the cell is to be used. The temperature selected by the writer is that of melting ice. The following table exhibits the observed and calcu- 1 Philosophical Magazine, November, 1889. 92 PRIMARY BATTERIES. lated values of the E.M.F. of a cell, set up with such a solution, in terms of a Rayleigh cell at 20° C. : — Temperature C. Observed. Calculated. o 8.3 1.0108 1.0106 8.5 1.0103 1.0105 9.3 1.0104 1.0102 11.8 1.0093 1.0092 13.8 1.0084 1.0085 15.0 1.0080 1.0080 18.1 1.0069 1.0068 19.4 1.0064 1.0063 19.9 1.0062 1.0061 20.3 1.0060 1.0059 20.8 1.0054 1.0057 21.1 1.0057 1.0056 21.6 1.0054 1.0055 - 22.4 1.0050 1.0052 23.3 1.0048 1.0048 25.1 1.0044 1.0041 26.4 1.0035 1.0036 30.2 1.0019 1.0022 33.1 1.0014 1.0013 39.1 0.9991 0.9989 41.7 0.9980 0.9979 50.4 0.9949 0.9947 52.7 0.9939 0.9940 The Rayleigh cell was always very near 20° C, and the reduction to that temperature was made by means of the coefficient 0.00077. The equation for the E.M.F., derived from the above observations, is — E t = E lb \1 - 0.000387 {t - 15) + 0.0000005 (t - 15)*}. STANDARDS OF ELECTROMOTIVE FORCE. 93 The calculated values of the second column were all obtained by this formula. The change for one degree C. is the following linear function of the temperature : — - 0.000386 + 0.000001 (t - 15). The coefficient ranges from 0.00040 at 0° to 0.000376 at 25°, and 0.000361 at 40° C. At the highest observed temperature of the table it was only 0.000348. The 1.010 \ \ s s S u \ S 1.005 \ s \ s ri N S \ H \ \ \ 1.000 N .995 T« TO >t1 n* 1 ut es < 10 20 30 40 50 Fig. 45. — Relation between B.M.F. and Temperature. curve of E.M.F. with temperatures as abscissas is clearly concave upward (Fig. 45), indicating a fall in the tem- perature coefficient with rise of temperature. Lord Rayleigh's cell showed a considerable increase in the coefficient with rise of temperature, the sign of the second term in his equation expressing the relation between E.M.F. and temperature being negative. The other two objections urged against the usual form of Clark cell are founded chiefly on the local 94 PRIMARY BATTERIES. action taking place when the zinc and mercurial salt are in contact. Zinc replaces mercury to some extent when in contact with a salt of mercury. With the oxide of mercury this action is very marked, resulting in reduction of the mercury and oxidation of zinc. The same replacement process goes on with mercurous sul- phate, zinc sulphate being formed at the expense of zinc and mercury sulphate, while the zinc is amalga- mated with the reduced mercury. A progressive change in the density of the solution ensues, entailing perhaps a rise in the value of the temperature coefficient. It may be noted, further, that if the cell contains crystals of zinc sulphate, the liquid at the surface of the mercury salt in an undisturbed cell is likely to be denser than it is even a few millimetres higher up, because the zinc sulphate crystals form at the bottom of the liquid. Bearing in mind that zinc in dilute zinc sulphate is positive to zinc in a relatively denser solution, it is easy to see that a voltaic couple is thus formed of one metal and two solutions of different densities. That this is actually the case is proved both by experiment 1 and by the deposit of zinc on the zinc rod just at the surface of the mercurous sulphate. Upon dismounting and opening one cell, which was perhaps a year old, it was found that zinc had been removed from the rod at the surface of the liquid, and some of it had been deposited again upon the rod at the surface of the mercury salt in a solid frill, which was not easily detached. This action is analogous to the transfer of copper from one plate to another in electrical connection with it, when the two are immersed in a solution of copper sulphate, and the tem- perature at one plate is kept higher than at the other. 1 See Chapter IX. STANDARDS OF ELECTROMOTIVE FORCE. 95 The obvious remedy is to insert a porous partition between the mercurous sulphate paste and the zinc in zinc sulphate solution. For cells not intended for transportation, plaster of paris, mixed up with a somewhat dilute solution of zinc sulphate, answers perfectly. Its effect on the E.M.F. appears to be nil. But if much disturbed it is liable to break up after a few months. A slip of cork is better if the cell is to be roughly shaken, as in trans- portation. The separation of the zinc from the mercury salt increases the E.M.F. about 0.4 per cent, or from 1.435 to 1.440 true volts at 15° C. Since mercurous sulphate is almost insoluble in concentrated zinc sulphate, the separation of the zinc from the mercury salt appears to present a complete mechanical obstacle to local action. This view is confirmed by observations on cells two years old. To prevent accidental short cir- cuits, it is desirable to mount a standard cell with a high resistance in series with it. This resistance of about 10,000 ohms, consisting of plumbago on glass, is mounted in the case (Fig. 46), and is, therefore, always in circuit with the cell. It can give rise to no error so long as zero or condenser methods are employed. 63. The Oxide of Mercury Standard Cell, — This normal element was described by M. Gouy in the " Journal de Physique," Tom. VII., 1888, p. 532. M. Gouy employs mg. 46. Caihart-Clark Standard Cell. 96 PRIMARY BATTERIES. the oxide of mercury instead of the sulphate as a de- polarizer. He further makes use of a 10 per cent solu- tion of crystallized zinc sulphate, of density 1.06, in place of a saturated one. M. Gouy finds that the negative polarization of his cells, due to closing the circuit, does not amount to one one-thousandth of the E.M.F. after the cell has been agitated and left standing for a short time. On the other hand, the positive polarization, arising from a reverse or charging current, persists longer than the negative. It can be gotten rid of by closing the circuit for a short time to produce negative polarization, from which the cell rapidly recovers. The reverse cur- rent undoubtedly forms some mercurous sulphate, which gives a higher E.M.F. as a depolarizer than the oxide ; and, while it lasts, produces an apparent polarization in the positive sense. The E.M.F. of this cell is 1.390 legal volts at 12° C, and the change due to temperature is 0.0002 volt per degree. The formula for the E.M.F. is then E t = 1.390 - 0.0002 (t - 12) . This is equivalent to a temperature coefficient of 0.000104, or only about 0.01 per cent per degree C. The E.M.F. of this cell is said to increase with in- crease of density of the zinc sulphate solution. To prevent local action, the zinc is not allowed to come in contact with the mercuric oxide. For use in which high internal resistance is of no consequence, the zinc rod is placed in a glass tube having in it a small hole near the lower end. If it is necessary to decrease the internal resistance, the zinc is enclosed in a linen bag. Detailed directions are given for the preparation of STANDARDS OF ELECTROMOTIVE FORCE. 97 zinc sulphate and mercuric oxide ; also for the purifica- tion of zinc and mercury. 64. Sir William Thomson's Standard Daniell Cell. — Some form of Daniell cell has long been used as a standard of E.M.F., partly because its polarization is small, and partly because its E.M.F. is near unity. To insure constancy, some provision must be made to pre- vent, or at least to greatly retard, the mingling of the two sulphates. Thus Raoult's cell consists of two glass vessels, one containing zinc in zinc sulphate, and the Pig. 47. — Thomson Standard Daniell Cell. other copper in copper sulphate. When in use the two vessels are connected by an inverted U-tube, filled with zinc sulphate solution, and closed at both ends with a piece of thin bladder. The normal Daniell element of Sir William Thomson (Fig. 47) consists of a rather low glass jar, with a plate of zinc in saturated zinc sulphate solution at the bottom. Above is suspended the copper plate ; and the copper sulphate, which is a half-saturated solution, is introduced through the funnel, connecting by a rubber tube to a 98 PRIMARY BATTERIES. siphon which terminates in a pointed horizontal tube at the surface of the zinc sulphate. By filling the funnel and gently raising it, the copper sulphate will flow over the surface of the saturated zinc sulphate, so that the surface of separation between the two liquids will be clearly defined. Upon the termination of the experi- ment the funnel is lowered and the solution is run out. It should be used but once. Just before making quan- titative use of the cell a feeble current should be sent through for a short time to freshly coat the copper plate. The E.M.F. of a cell thus set up has been found to be 1.072 true volts at about 15° C. The temperature coefficient is small, but appears not to have been care- fully determined. Dr. Fleming found it to be about one-fifth of the variation of the Rayleigh-Clark cell between 0° C. and 20° C. ; a but Mr. Preece found a greater variation, amounting to 9 parts in 1000, for one-half the range of temperature, or between 17° and 28°. 2 If Mr. Preece is correct, the temperature co- efficient of the normal Daniell cell within the above range is quite as high as that of the Rayleigh form of Clark element. Mr. Preece's method was scarcely sensi- tive enough to admit of a good determination of the variation of E.M.F. with temperature. 65. Lodge's Standard Daniell Cell. — A wide-mouthed bottle (Fig. 48) is provided with a cork, through which passes a large test-tube B with a small opening at the bottom. The zinc rod Z is held in this tube by a cork. A small test-tube c is fastened to R by an elastic band. This tube contains the solution of copper sulphate, and into it dips a gutta-percha covered copper wire, bared at 1 Philosophical Magazine, August, 1885, p. 136. 2 Proceedings Royal Society, Vol. XXXV. 1883, p. 48. STANDARDS OF ELECTROMOTIVE FORCE. the lower end and furnished with a fresh deposit of electrolytic copper. The insulated wire passes through a cork in the small tube. This tube is immersed in the zinc sulphate solution contained in the bottle Gr up to a point near its -top. Fig. 48. Lodge's Standard Daniell Cell. Fig. 49. Fleming's Standard Daniell Cell. The two sulphates are by this device kept entirely separate, and the electric connection between them is established by means of the moisture covering the glass. The internal resistance of the element is enormously high, and the cell is applicable only to zero methods or comparisons by means of a condenser. 66. Fleming's Standard Daniell Cell. — The form of Daniell cell shown in Fig. 49 was specially designed 100 PRIMARY BATTERIES. by Dr. Fleming as a standard of E.M.F. 1 It consists of a U-tube 8 inches long and J inch in diameter, provided with side tubes, glass taps, and reservoirs as shown. To fill the cell, the tap A is opened, and the tube is filled with the denser zinc sulphate solution. A is then closed, and the zinc rod is secured in the left-hand branch by means of an air-tight rubber stopper P. The tap C is now opened, and the liquid falls in the right- hand branch only; and if the tap B is opened at the same time, the copper sulphate solution will flow in gently as the level of the zinc solution sinks in this branch. The operation may be so conducted that the surface of separation between the two solutions will remain quite sharp, and will gradually sink to the level of the tap 0. All the taps are then closed, and the copper rod is inserted in the right-hand limb. When the surface of contact ceases to be sharply defined by reason of diffusion, it is only necessary to draw off the mixed liquid at the level of the tap C, and to supply fresh solutions from the reservoirs above. The extra tubes, L and M, are for the purpose of holding the electrodes when not in use, each in its own solution. The exact value of the E.M.F. of a Daniell cell is dependent upon the density of the solutions and the condition of the zinc and copper surfaces. Thus ' Increase in density of the CuS0 4 solution increases E.M.F. Increase in density of ZnS0 4 solution decreases E.M.F. Oxidation of the copper surface increases E.M.F. Oxidation of the zinc surface decreases E.M.F. Moreover, for an equal increment or decrement of density of both solutions the increment and decrement 1 Philosophical Magazine, 5 S., Vol. XX. p. 126. STANDARDS OF ELECTROMOTIVE FORCE. 101 of the E.M.F. are so nearly equal, that for equi-dense solutions, within limits, the E.M.F. is independent of the absolute density of either. It is of the utmost importance that oxidation of the copper surface should be carefully guarded against. Even slight oxidation, indicated by brown spots, raises the E.M.F. by as much as 4 parts in 1000, while a film of dark brown oxide may affect the E.M.F. as much as 2 per cent. Since rolled copper sheets or drawn wire probably enclose more or less oxide mechanically, it has been found necessary to freshly electroplate the copper surface immediately before use. Raoult found that copper foil gave a higher E.M.F. than electro-deposited copper by about one two-hundredth ; and he attributed it to the oxides of copper enclosed in it. If a newly electroplated copper rod is left in the copper sulphate solution, it is gradually oxidized; and the oxidation is more rapid if the rod is exposed to the air and contains even a trace of the copper sulphate. The rod should be electroplated with a thin film of copper immediately before it is transferred to the standard cell for use. If a chemically pure zinc rod is used, it is immaterial whether it is amalgamated with pure mercury, or is freed from oxide on the surface by slight rinsing in dilute sulphuric acid before placing it in the sulphate of zinc. For general use Fleming recommends two standard solutions of each salt. First, a solution of copper sul- phate, saturated at 15° C, and of density 1.2, and a solution of zinc sulphate of the same density. Second, a solution of the copper salt, of density 1.1 at 15°, and one of the zinc salt, of density 1.4 at the same tempera- ture. 102 PRIMARY BATTERIES. If equi-dense solutions are used, with the precau- tions already described respecting the surfaces of the zinc and copper rods, the E.M.F. is very close to 1.102 true volts. If, however, copper sulphate of density 1.1 and zinc sulphate of density 1.4 are used, then the E.M.F. of the cell is 1.072 volts. These last solutions corre- spond with those employed by Sir William Thomson in his standard form of gravity cell. If the cell is allowed to stand an hour or so after the freshly electroplated copper pole is introduced into it before measuring the E.M.F., then its value will be about 0.003 volt higher than the above, provided the zinc retains a bright untarnished appearance. But the smallest deposit of copper on the zinc, due to the diffu- sion of the copper salt into the zinc sulphate, lowers the E.M.F. 2 or 3 per cent. The many precautions required to insure a normal E.M.F. in a standard Daniell cell, on every occasion of its use, are more than an offset to a negligible tempera- ture coefficient in comparison with that of a Clark cell, particularly if the latter is reduced to 0.038 or 0.039 per cent. 67. The Chloride of Lead Standard Cell. — MM. Bailie and Fe"ry have proposed 1 the use of a salt of lead as a depolarizer. The best results were obtained with the chloride. It has one of the disadvantages of the Daniell, but in an inferior degree ; that is, the deposition on the zinc of the metal contained in the depolarizer. But with proper precautions, the formation of this metallic deposit may be greatly retarded. The cell is mounted as follows: Powdered lead 1 Journal de. Physique, Tome IX. p. 234. STANDARDS OF ELECTROMOTIVE FORCE. 103 chloride, precipitated from a warm solution and of crystalline texture, is introduced into the tube A (Fig. 50) which encloses a lead wire, forming the negative of the element. The positive is a plate of zinc, amalga- mated and immersed in a solution of chloride of zinc, of density 1.157. When the circuit is closed zinc is dissolved, and chloride of lead is reduced. The E.M.F. decreases with the concentration of the zinc chloride solution. With the above density, made by dissolving 17.2 gms. pure zinc chloride in 100 c.c. distilled water, the E.M.F. is exactly one- half a volt. Dr. Fleming's standard Daniell cell was taken for com- parison. The variation of E.M.F. with temperature was found to be almost negligible, amounting to only 0.005 volt in 46° C: The solution of should be made neutral by agitation with zinc oxide, since the presence of free acid aug- ments the electromotive force. The polarization, though greater than in the Daniell cell, is still very small, and the cell recovers promptly and exactly its normal value. 68. To Measure the E.M.F. of a Standard Cell. — In the absence of means of making an absolute determination of the E.M.F. of a standard cell, the silver voltam- eter may be resorted to as a secondary standard. Of this method, Lord Rayleigh remarks : " It will be seen ■Li 'J Fig. SO. — Chloride of Lead zinc chloride B standard Cell . 104 PRIMARY BATTERIES. that in this way any one may determine the E.M.F. of his standard battery with a very moderate expenditure of trouble, and without the need of any special apparatus." * The method of making the determination is shown in Fig. 51. The main battery B is a storage cell, and in series with it is a carefully adjusted resistance SB of 10 legal ohms at 14° C, made of platinoid wire im- mersed in kerosene; also a silver voltameter V, and a second resistance \® B of heavy iron wire for the pur- pose of adjusting the current to the proper value. The standard cell B 1 is placed in a derived circuit at the ter- minals A O of the 10-ohm coil. In circuit with it is a sensitive "long coil " galvanome- ter Cr, and a carbon resistance, SB, of 100,000 ohms. A balance is effected between the E.M.F. of the standard and the fall of potential over the 10-ohm coil by vary- ing the auxiliary iron resistance, and by greater or less immersion of the vertical silver plates of the voltameter in the silver nitrate solution. If any small change occurs in the current during the deposition of the silver, the balance may be maintained perfectly by changing slightly the depth of immersion of the silver 1 Philosophical Transactions, Part II. 1884, p. 453. Fig. 51. — E.M.F. Measured by Silver Voltameter. STANDARDS OF ELECTROMOTIVE FORCE. 1.05 plates. For this purpose the voltameter is provided with a rack-and-pinion movement for the plates. All the conditions for a balance being ascertained, the gain plate is carefully washed, dried, and weighed. It is then replaced, and the circuit is kept closed for a sufficient time to secure enough gain in the kathode plate to weigh accurately, the balance being carefully maintained as described during the entire time. The gain plate is again removed, and the amount of silver deposited is determined. This gives the value of the mean current through the 10-ohm coil. Then by Ohm's law, E= OR ; and since both current and . resistance R are known, JE is in this manner determined. Example. Temperature of standard cell, |(15°.5 + 15°.7) = 15°.6 C. Temperature of 10-ohm coil, i(16°.6 + 16°.7) = 16°.65 C. Resistance of 10-ohm coil at 16°.65 = 10.00583 legal ohms. Weight of silver plate after deposit, . . . 29.99292 gms. Weight of silver plate before deposit, . . 29.79942 " Weight of silver deposited, 0.1935 " Time of deposition, 20 minutes. 1 ampere deposits 4.0246 gms. per hour. Hence the current equals 0.1935 -=- £(4.0246) = 0.14424 amperes, and E = 0.14424 x 10.00583 = 1.44324 legal volts at 15°.6 C. Reducing to 15° by the formula 1.4432 = E £1 - 0.000386 (t - 15)], the E.M.F. of the standard equals 1.4435 legal volts. 106 PRIMARY BATTERIES. CHAPTER VIII. MISCELLANEOUS BATTERIES. 69. Grove's Gas Battery. — The polarization current obtained from a water voltameter, and due to the oxygen and hydrogen clinging to the two platinum plates, sug- gested to Grove the possibility of prolonging this cur- rent by supplying a suf- ficient quantity of the two gases in contact with platinum. The po- larization current soon exhausts the films of oxygen and hydrogen on the two respective plates. By extending the strips of platinum so that they are partly in the liquid and partly in the gas of each tube of the voltameter sup- plied with water acidu- lated with sulphuric acid, density 1.2, Grove succeeded in producing continuous currents of sufficient intensity to decompose water and to produce a brilliant spark in broad daylight between two carbon Fig. 52. — Grove's Gas Battery. MISCELLANEOUS BATTERIES. 107 points. For this latter purpose he employed fifty- pairs. The figure exhibits the form of gas battery preferred by Grove. Fis a three-necked Woulff's bottle. In the two outer holes are fitted two glass tubes by means of rubber stoppers. Each tube is open below and contains a piece of platinum foil ending above in a platinum wire, which is sealed into the top of the tube. The entire apparatus is filled with acidulated water through the middle opening B, and a current is then passed through till one tube S is filled with hydrogen, and the other half-filled with oxygen. If now the battery is removed, and the terminals at PandiVare connected by a conducting circuit, a current flows from the oxygen tube to the hydrogen through the external circuit. In order to increase the surface of the liquid in con- tact with the platinum and exposed to the gas, Grove covered the foil with pulverulent platinum by Smee's method of electrolytic deposit. The liquid then rises along the roughened surface by capillary action. The hydrogen in this cell plays the part of the zinc in a voltaic element. The current through the cell is from the hydrogen to the oxygen — the reverse of the decom- posing or charging current. In modern nomenclature this is a storage battery. The effect of the charging current is to decompose sul- phuric acid primarily and water as a secondary reaction ; and the accumulation of the products of the electrolysis in the two tubes is a storage of potential energy. When this potential energy is converted into the kinetic energy of a current, all the processes are reversed, the current with the others. In the same way, when energy is stored in the potential form by lifting a weight from the 108 PRIMARY BATTERIES. earth, the running down of this energy by conversion into the kinetic variety involves a reversal of the mo- tion of the weight. In the electrolytic process the chain of molecules may be represented as follows : — H 2 | H 2 S0 4 | H 2 S()TpHA < « After the first step in the electrolysis this becomes — H 2 | H 2 | H 2 S0 4 | H 2 S0 4 | 0. The oxygen and hydrogen are now at the two ends of the chain; and, leaving out the water as unessential, the chain of the gas battery may be written — H 2 | H 2 S0 4 | H 2 SQ 4 | 0; » > and this becomes, after the first exchange of atoms among the molecules — H 2 S0 4 | H 2 S0 4 | H 2 0. Hydrogen is in both cases transferred in the direction of the current, which is shown by the arrow. In the discharge process the oxygen may equally well be sup- posed to suffer a transfer in the opposite direction, though it is simpler to conceive of the motion of the hydrogen only. The operations of the electrolytic process are then strictly reversed in the recombining process. The tubes of the gas battery may be filled with the two gases obtained by any other method than elec- trolysis, with no difference in the result. . MISCELLANEOUS BATTERIES. 109 If one tube is filled with hydrogen and the other with acidulated water, a current is still obtained, and hydro- gen gradually disappears on closed circuit. Grove showed that the current in this case was due to the oxygen absorbed from the atmosphere. Similar results were obtained with other gases, not- ably hydrogen and chlorine ; also with one gas and a liquid whenever chemical reaction was possible between the two. 70. Upward's Chlorine Battery. — The electrodes are zinc and carbon, the former immersed in water contained in a porous cup, and the latter in water saturated with chlorine gas. The space between the porous cup and the carbon is filled with broken retort carbon. Each cell contains several zincs and carbons joined together in multiple. Since the chlorine is both the active exciting agent and the depolarizer, the liquid about the carbon is kept saturated with the gas, which passes into the porous cup by diffusion, while the zinc chloride formed diffuses outward. The cell must be closed air-tight to prevent the escape of chlorine. Each cell consists of a glazed vessel, with an inlet tube near the bottom and an outlet near the top. A glazed cover, with the requisite provision for the passage through of the two electrodes, closes the cell tightly. The chlorine, made from chloride of lime (CaOCL.), is stored in a glazed earthenware cylinder provided with inlet and exit tubulures. The cells and the reservoirs are connected together in series, the top of the reservoir to the bottom of the first cell; the top of this cell to the bottom of the second; and the top of the second back again to the. reservoir. Each cell is further pro- 110 PRIMARY BATTERIES. vided with a draw-off stone tap for removal of the zinc chloride formed in the action of the cell. The E.M.F. is 2.1 volts and very constant. Large cells have been built by Woodhouse & Rawson for charging storage batteries, and they are said to furnish a current of 150 amperes on short circuit. 71. Powell's Thermo-Electro-Chemical Battery. 1 — Differ- ences of potential have often been observed between two plates of the same metal in a solution of a salt of the same, when one plate is at a higher temperature than the other. Thus two zinc rods in sulphate of zinc are at a different potential if their temperatures are different, the one of higher temperature constituting the positive electrode (negative plate) of a voltaic couple. This property has been applied to the construction of a thermo-chemical couple with copper plates in copper sulphate solution. A horizontal plate is placed in the bottom of the cell, and a well insulated wire leads out, preferably through a glass tube. Another copper plate, with a copper tube attached to its centre, is suspended so that its under surface touches the surface of the solu- tion. Half-a-dozen small openings at the bottom of the copper tube convert it into a rose burner. Gas is con- ducted in through the tube, lighted at the openings, and the small flames heat copper wires riveted to the copper plate. The transfer of heat to the plate, and so to the liquid, is thus increased. Under these conditions, a current flows from the warm to the cold plate through the external circuit, and copper is transferred from the cold plate to the warm one through the solution. In other words, the cold plate performs the same function as the zinc in a simple 1 London Electrical Review, "Vol. XX. p. 2. MISCELLANEOUS BATTERIES. Ill voltaic element. 1 The energy concerned in the transfer comes from the heat applied. The combination is thus both a primary (heat) battery and an electrolytic cell. The potential energy transformed is in this case repre- sented by the illuminating gas. The E.M.F. is small, only about 0.035 of a volt with a difference of temperature of 50° C. between the upper and the lower plates. A small addition of sulphuric acid, which is of utility in an electrolytic cell for copper sulphate, reduces the E.M.F. of the thermo-chemical battery to zero. Copper nitrate may be used in place of the sulphate. Note. — The inventor of this battery describes it with the current flowing through the cell from the warm plate to the cold one, and says expressly that copper is transferred from the top to the bottom (Electrical Review, Vol. XX. p. 2, London). But if the reader will consult the next chapter, he will find an account of tests on this point, with a table of E.M.F. 's at different temperature differences. 72. A Battery Absorbing Oxygen from the Air. 2 — When copper is alternately exposed to the air and immersed in an aqueous solution of ammonia, it oxidizes, and the oxide dissolves as a blue solution of ammoniacal cupric oxide. If the copper remains immersed in the solution at a considerable depth, the supply of oxygen that can reach the copper plate is very limited, and cuprous oxide is formed and dissolved. If now an aerating plate of platinum foil or platinum sponge is supported on the liquid surface, and connected by a wire with the copper, a current flows through the liquid and the wire, and the process of oxidation and solution is greatly hastened. The platinum plate or i See Chapter IX. 2 Proceedings Royal Society, Vol. XLIV. p. 182. 112 PRIMARY BATTERIES. sponge condenses oxygen, which is gradually transferred to the copper. The current rapidly runs down if its density is more than one micro-ampere (millionth of an ampere) per square centimetre of the aerating plate. The E.M.F. may be from 0.5 to 0.6 of a volt. The addition of common salt or of sal-ammoniac reduces the internal resistance and increases the E.M.F. With a thin iayer of spongy platinum as the aerating plate the E.M.F. may be as high as 0.8 of a volt. Similarly, if a platinum plate is immersed in a solu- tion of ferrous sulphate or sulphurous acid, and an aerating plate is placed on the surface of some dilute sulphuric acid in another vessel ; and if the two vessels are connected with a siphon or a piece of moistened candle wick, and the two plates are joined by an electric conductor, the oxygen condensed by the aerating plate will be transferred to the oxidizable solution in the other vessel, with the formation of ferric sulphate or sulphuric acid, and at the same time a current will flow through the circuit. 73. Minehin's Seleno-Aluminum Cell. — Professor Min- chin 1 constructs a cell sensitive to light in the following manner:. Two small clean plates of aluminum are taken, and a thin layer of sensitive selenium is spread over one of them. Fine platinum wires are then attached to both plates, and they are immersed in presence of each other in a small glass cell containing acetone. Alcohol — preferably methylic — answers very well, except that in a few days the plates become covered with a gelati- nous deposit of aluminate of alcohol. The selenium must be treated by heating and care- 1 Philosophical Magazine, Vol. XXXI. p. 207. MISCELLANEOUS BATTERIES. 113 fully keeping it near the melting-point for some time, till it assumes a very dark brown color. It has then its most sensitive surface. When a cell, constructed as described, is exposed to light, an E.M.F. is at once developed, and the sensitive seleno-aluminum plate is negative towards the insensi- tive one, i.e. as copper to zinc. This photo-electric cell is sensitive to all parts of the spectrum, with a maximum in the yellow near the bor- der of the green. The variation in sensitiveness through- out the entire visible spectrum is about 30 per cent. 74. Shelford Bidwell's Dry Battery. — This cell, which grew out of an investigation into the sensitiveness of selenium to light, has thus far only a scientific interest. On a plate of clean copper is spread a layer of copper sulphide. The sulphide is then compressed in a vise between the copper plate and one of polished steel. The steel plate is next carefully removed, and a thin layer of silver sulphide is spread over the compressed copper sulphide. Finally, a plate of silver is pressed down upon the sulphide and the cell is complete. The copper plate constitutes the positive electrode, the current flowing through the cell from the silver to the copper. The chemical action consists in the reduc- tion of the sulphide of copper with deposition of copper on the copper plate, and the simultaneous formation of an equivalent amount of the sulphide of silver. The cell is entirely analogous to the Daniell, with copper and silver in their sulphides in place of copper and zinc in their sulphates. With copper and silver separated by copper sulphide only no current was obtained; but when free sulphur was mixed- with the sulphide, the cell became active. 114 PRIMARY BATTERIES. 75. Jablochkoff's Battery. — Carbon is attacked by nitrates in a state of igneous fusion, while iron is not. Hence a vessel of cast iron, cylindrical in form and filled with fused nitrate of potassium or sodium, serves at the same time as a receptacle and as an unattacked electrode. An iron wire helix serves to hold the coke and to con- duct to the external circuit. If the nitrate is maintained in a state of fusion, the cell will have an E.M.F. of from one to two volts. It has been observed that if an aqueous solution of the salt is used instead of the fused nitrate, the poles are reversed, or the iron is the negative electrode and acts like zinc in a simple cell. 76. Battery with Two Carbon Electrodes. — This was devised by Tommasi and Radiguet in 1884. At the centre of a cylindrical glass jar is placed a carbon rod, covered with a thick layer of peroxide of lead, the whole enclosed in a linen bag. This enclosed electrode is placed in a carbon tube pierced with holes ; the two electrodes are then put into the glass jar and filled around with fragments of retort carbon, and a concentrated solution of chloride of sodium added to chloride of calcium. This latter salt serves to retard very much the evaporation of the water. The carbon rod with the coating of lead peroxide is the positive electrode. The E.M.F. is from 0.6 to 0.7 of a volt. No action takes place on open circuit, but since polarization sets in rapidly on closed circuit, the cell can be used only for applications requiring an intermittent current. For such purposes it has a very long life. Some of these cells, after remaining in service for several years, operate absolutely as well as the first day they were set up. BATTERY TESTS. 115 CHAPTER IX. BATTERY TESTS. 77. What a Systematic Test Includes. — The most ob- vious quantities to be measured are the E.M.F. and internal resistance. While a high E.M.F. is desirable for most purposes, a low E.M.F. is no indication that a battery may not be admirably adapted to its intended work. So low internal resistance is a commendable feature, because, caeteris paribus, low internal resistance means high efficiency ; but if a battery is to be used on a circuit of high resistance, its own resistance is rela- tively of less account. For large currents, low internal resistance is a necessity. It is further very desirable to know the rate, progress, and total amount of polarization that takes place when a cell is kept on a closed circuit of known resistance for a definite period. The results of a test to determine such data respecting polarization can all be expressed graphically in the form of a curve. So also the promptness and extent of the recovery from polarization are equally essential objects of investi- gation, and the results can be expressed in the same manner as the polarization. These data, together with the potential difference at the terminals or electrodes, when the battery is on closed circuit, furnish all that is needed to compute the internal resistance and the current. 116 PRIMARY BATTERIES. An efficiency test can be made only by working a battery to exhaustion. This is not practicable for one of relatively large internal resistance and rapid polariza- tion. For open circuit cells many plans have been devised to secure continuous intermittent test service extending over long periods. But none of these is so satisfactory as to place a battery in actual service and wait for results. Another important object of inquiry is the amount of depreciation and local action taking place on prolonged standing on open circuit. This is applicable strictly to open circuit cells only. The practised eye of the observer with experience will not overlook many details of mechanical construc- tion, which are as important to the satisfactory working of a battery as its electrical features. 78. Theory of the Method of Measuring E.M.F. and Internal Resistance. — The E.M.F. is measured by com- paring it with that of some standard which is known. The standard employed in the following tests Was the author's form of the Latimer Clark cell, having an E.M.F. of 1.44 true volts, or 1.444 legal volts, at 15° C. For ordinary battery tests a rapid method of comparison, accurate to one-half per cent, is all that is required. The condenser method is the only one that admits of sufficient rapidity, and it possesses the required accu- racy. For this purpose, a standard mica condenser, divided into fractions so as to admit of using from 0.05 to one microfarad, and a sensitive reflecting galva- nometer of from 5000 to 7000 ohms resistance, are required. Also the proper charge and discharge keys, and an ordinary circuit-closer. The condenser is then charged with the standard cell BATTERY TESTS. 117 and discharged through the galvanometer, and the deflec- tion noted. The same process is repeated with the cell to be tested. The ratio of the deflections produced is the ratio of the electromotive forces of the. standard and the cell in question ; for the deflections are at least approximately proportional to the quantities of elec- tricity discharged through the galvanometer, so long as. those deflections are not large and not widely different; and the quantities are proportional to the electromotive forces charging the condenser, the capacity of which remains constant. To obtain the internal resistance, we must know the total E.M.F. of the cell, and the difference of potential between its terminals when the circuit is closed through a known external resistance. If, now, it is assumed that the potential difference at the terminals can be measured so soon after closing the circuit that no polarization has set in, then the total E.M.F., previously measured, is the whole fall of potential over the resistance of the entire circuit, while the difference of potential at the battery terminals represents the fall over the known external resistance, which must contain no source of E.M.F. If, therefore, E and E' represent total E.M.F. and terminal potential difference, r and R the internal and external resistance respectively, then — E:E'::\r + ifyjl. Hence E — E' :E' ::r: B, and r=R E ~ W - E' Since R is known, and E and E' have been measured, r is also known for the given conditions of external resistance and current. 118 PRIMARY BATTERIES. There is reason to believe that the resistance, and probably the electromotive force, of a battery depends to a certain extent upon the current flowing through the battery, and upon the rate of diffusion of the products of the chemical changes taking place. The resistance, and generally the electromotive force, varies also with the temperature of the battery. All that can be positively affirmed of the value r, obtained as described, is that it satisfies the equation Fig. 53. — Diagram, of Battery Tests. expressing the relation between R, U, and E'. Still, it is true that for widely different values of M, the value of r ascertained by this process will enable us to compute with considerable accuracy the potential difference E' at the terminals available to produce a current through a known external resistance M. In Fig. 53 are shown diagrammatically the connec- tions of the apparatus for making the measurements BATTERY TESTS. 119 described. The condenser is at 0, the galvanometer at o IS) (M o | © o> 1 © 1 © | OJ | © 1 © T-H T-H o ' o o o o © o T-H t-l -CM CM •o © © CO N to i-t 1 © o © l-~ CD -* CO 1 CN 1 o ■a IO o CO 1 CO 1 IN 1 CN 1 CO CN 1 CN 1 CN 1 3 1 CO CN IO •e ■* 1 CO 1 o 1 IO 1 o 1 lO 1 ■* 1 tH 1 - O CN -* CO 00 o t-H CN -* CO CO o tN CM -* CO 00 © CO BATTERY TESTS. 123 have the significations already given to them ; -while and r are current in amperes and internal resistance in ohms' respectively. The E.M.F. of the standard was 1.443 legal volts. The same results are expressed graphically in the curves of Fig. 54, all of them being drawn to the same scale, except the internal resistance as indicated. 3.0 2.0 to I 1-5 1.0 OS I hkv V ^ \ _PQ LAE IZA Tior 1 ) 1 E RMI NAL P.D Hb ■bl.b TAN CE s. cu RRE NT Min. 20 40 Fig. 54.— Tent of Leclanch(5 Cell. 60 The polarization curve shows a very rapid fall of electromotive force for the first four minutes, and quite a steady decrease up to three-quarters of an hour. The recovery curve shows an almost equally rapid rise of 124 PRIMARY BATTERIES. electromotive force for the first four minutes after open- ing the circuit. It continues to increase up to the end of the hour, when it is still a quarter of a volt helow its initial value. The recovery curve is plotted back from the end of the polarization curve toward the left, so as to exhibit more plainly the depression of the voltage at the end of the two hours test. The terminal difference of potential is more nearly constant in value after the first steep incline than the total E.M.F. ; and the shortening of the intercepts be- tween the two curves shows the decrease in the internal resistance during the hour. The current after the first two minutes exhibits great steadiness for an open circuit battery. The fall to the end of the hour is only 0.016 ampere, or a little less than 8 per cent. The initial resistance of this particular cell is high, but it falls more than 50 per cent during the hour. Other individual cells made at the same factory show an internal resistance as low initially as 0.8 of an ohm. 81. Test of Leclanche Cell with Depolarizer Enclosed in Carbon Cylinder. — In cells of this character the depolar- izer is not favorably located to accomplish its purpose, since the current leaves the outside of the carbon cylin- der rather than the inside where the manganese dioxide is placed. It is exceedingly doubtful if the depolarizer is of much value in this relation to the carbon surface unless it is a soluble salt and diffuses through the liquid. The large area of carbon surface is an offset, however, to the unfavorable location of the manganese dioxide. A large carbon surface diminishes polarization. It has been found as a result of many experiments that reduc- BATTERY TESTS. 125 tion of zinc surface does not exercise so notable an effect on the current strength as the reduction of carbon sur- face. Hence the practice of employing zinc rods of small surface area, and carbon plates, rods, and cylin- ders of much greater superficies. Attention is called to the slower rate of polarization of this cell, Fig. 55, as compared with Fig. 54, on first !.Q 0.5 \ x RE cov fffy Pot AS; tAl 10M Te 1 R«i fVAL ~Pn ' — Int. RE bi5_ £q/y ce CU P3F?C MT •-»-.. / ) Min. 20 40 Fig. 55. — Test of Cell with Carbon Cylinder. 60 closing the circuit. The polarization is more continu- ous, but not so precipitate. The recovery is of the same character. This feature in the polarization curve, which may be called the " characteristic " of a battery, is advantageous in cells which are designed for service requiring ordinarily the closing of the circuit for only 126 PRIMARY BATTERIES. a few seconds intermittently. The polarization is less for short intervals than with cells having a steeper polarization curve at the beginning. The terminal potential curve runs nearly parallel with the total E.M.F. curve, and the vertical intercepts be- tween the two are short. With 5 ohms external 1-6 to OS s _PO Lar •Za r '0/Y TE RMi NAl .fio \nT RE. SI&T ANC E CU RRE (NT Min, Fig. 56. 20 40 -Test of Another Carbon Cylinder Cell. 60 resistance, the uniform value of R for all these curves, unless another value is given, the internal loss of energy in this cell is only 6.2 per cent, the internal resistance averaging about 0.33 of an ohm. The cur- rent fell from 0.26 to 0.2 of an ampere during the entire hour. Fig. 56 illustrates another cell of the same general BATTERY TESTS. 127 characteristics, but of a different manufacture and smaller zinc surface. Both have the black oxide of manganese enclosed in a carbon cylinder, and both show polariza- tion and recovery curves of the same character, though the recovery of the latter is less marked. It has a higher E.M.F. and a slightly smaller internal resistance. The energy wasted internally averages about 5.7 per cent. 1.0 O.S ; Pn L ■M^l {AT_ lOrv RE COV E(?Y ' ^ »L *?0. INT RE SIS TAN CE_ cu R(?E hT , Min. 20 40 60 Tig. 57. — Zinc-Carbon Cell without Depolarizer. It is to be borne in mind that these particular values are derived from individual cells, and do not represent the average obtained from a number of the same type. 82. Test of Zinc-Carbon Cell without Depolarizer. — The curves of Fig. 57 are derived from an investigation of a well-known type of battery employing amnionic chloride, but no depolarizer whatever. The polarization is some- 128 PRIMARY BATTERIES. what more pronounced at the start, but has the same progressive character as in the two preceding cases. The internal resistance exhibits marked irregularities, and is higher than would be anticipated, considering the extent of carbon surface. The current is nevertheless quite regular and has a mean value somewhat above 0.2 of an ampere. 1.0 OS . 1 \ v s (-OV fey £(JL ARJ_ Z_AT_ ION rt F?MI NAL PC II-1T -££ 5'5T <*fvr F cu Rf?E NT 1 Min, 20 40 Fig. 58. — Curves from a Dry Cell. 60 83. Test of a " Dry " Cell. — A dry cell has the excitant in the form of a semi-fluid or porous, pasty mass. In so far as polarization depends upon diffusion the dry cell may be expected to show a more marked and persistent depression of voltage when placed on an external resist- ance of no more than 5 ohms. Such anticipations are abundantly justified by the curves of Fig. 58, derived from a test of one of the best known cells of this class. The E.M.F. fell to less than one-half its initial value in BATTERY TESTS. 129 the hour, and its recovery during the following hour was quite leisurely. The semi-liquid electrolyte admits of only slow diffusion, even though ingredients may be added to make the mass porous. The internal resist- ance of this cell was not large, but was irregular, and the current fell during the test to less than half its initial value, because of the great drop in potential. 1.0 03 PO(. 4Ri ZAT iO* F?F rov FPY re RM/ NAL PD. K INT. RE 515 TAN CE cv RRE NT Min, 20 40 Fig. 59. — Second Dry Cell. 60 Another type of dry cell, not described in the preced- ing pages, furnished the curves of Fig. 59. It must be admitted that this cell makes a showing comparing favorably with zinc-carbon cells set up with a liquid electrolyte. The polarization is leisurely, the internal resistance only three-tenths of an ohm, and the current averages fully 0.2 of an ampere. 130 PRIMARY BATTERIES. Two of these same cells were subjected to a test in which for four months continuously they actuated a relay-sounder of about 100 ohms resistance every second, by means of a seconds-pendulum. They showed no perceptible deterioration in that time, though on two or three occasions the clock was stopped for several hours, during which time the circuit remained closed. On ' >., RE cov ERY PO LAR IZA TIO M P. D. jjj erfr i / y \ / / / \ V f INT RES IST/> NCE Mill. 20 40 Fig. 60. — Curves from Chloride of Silver Cell. 60 starting the clock, the relay-sounder again operated without readjustment. After standing idle for seven months these same cells were again put to the same service in time measurements, and they are still as efficient as ever after three months' additional use. 84. Test of a Silver Chloride Cell. — The curves (Fig. 60) obtained from a small silver chloride cell, made by BATTERY TESTS. 131 the author, serve to illustrate a number of points. The chloride was cast in a carbon mould on a silver wire, leaving a very small surface of silver exposed to the exciting liquid, which was dilute ammonic chloride. The internal resistance on first closing the circuit, before any metallic silver had been reduced from the chloride, was 79.7 ohms. It fell during the hour to one ohm, the scale of the resistance curve being one-tenth as large as that of the others. This precipitate fall of resistance is due to the reduction of silver from the chloride, which converts a poor conductor into the best one known. Coincident with this fall of internal resistance is the rise of potential difference at the terminals and of the current. The latter does not rise above 0.17 of an ampere. The cell was a small one, with about two square inches of zinc surface. The polarization of the silver chloride cell is so slight as to justify its inclusion in the list of constant current batteries ; for the E.M.F. is nearly constant, and the drop in internal resistance causes the current to increase in intensity instead of the reverse. The recovery from polarization is extremely prompt, and occurs within the first two or three min- utes after opening the circuit. The initial value of the E.M.F. is not regained, but the final loss is less than 0.05 of a volt. 85. Efficiency Test of Copper Oxide Battery. — An effi- ciency test consists of two parts : — First, the determination of the total quantity of elec- tricity obtained by the consumption of a definite weight of zinc, compared with the quantity of electricity required to deposit the same weight in electrolysis. Second, a comparison of the useful energy in the 132 PRIMARY BATTERIES. external circuit with the internal energy as heat waste in the cell itself. -" // \ 1 \ 9 / / // \ \ 1 / /, \ / 1 / / i ' / / | w z s * h ■i - I u z < -S IU u. s > 55 & » g " t w z -J ■.-. £E 1- 3 - T 5 z ^ 1 4 £1 ' / is! / a A-l 1 z l \ "1 A \ \ / ' ! ! ! J ! site i -i 18 tiun 8 ' • NH a C2 i 8 a - LtXV For the first, the zinc must be weighed at the beginning and end of the test, and the whole number of ampere- BATTERY TESTS. 133 hours must be determined. This gives the quantity of electricity obtained by the consumption of a known weight of zinc. The quantity required to deposit the same weight of the metal can be calculated from the electrochemical equivalent of zinc. For the second part of the test, the internal resistance must also be measured at intervals during the run. Then the energy lost as heat in any circuit of resistance r is C\ ; for energy is the product of current and elec- tromotive force, and by Ohm's law electromotive force is Or. Hence energy is C 2 r. The external resistance being known also, the external energy is calculated in the same way. The curves of Fig. 61 express graphically the results of such a test made with great care by Mr. A. E. Ken- nelly in the Edison laboratory. Four 300-ampere-hour cells were taken at random from the stock. They were joined in series in a circuit of 0.8 of an ohm external resistance. The total run was 108 hours. The external energy increased quite up to the middle of the time, because of the continuous reduction in internal resist- ance. The following are details of the computation : — Weight of zinc before test .... 10,017 gms. " " " after " .... 8,567 " Total loss 1,450 " Loss calculated from output . . . 1,444 " Loss by local action 6 " Mean current 2.76 amperes " E.M.F 2.8 volts Total quantity in ampere-hours . . 298 134 PRIMARY BATTERIES. The ampere-hours are the product of the mean current and the time, or 2.76 x 108 = 298. Taking the electrochemical equivalent of zinc as 0.0003367, the calculated loss is as follows : — 298 x 3600 X 0.0003367 = 361 gins, per cell. 361 X 4 = 1444 gms. for 4 cells. The quantity 0.0003367 is the weight in grammes deposited by one coulomb, — an ampere for one second. Hence ampere-hours must be multiplied by 3600 to reduce to ampere-seconds or coulombs. In reading the figures at the left of the diagram, all except those relating to current must be divided by 4 to reduce to the values for a single cell. An efficiency test of this same type of cell, conducted by the author, showed curves approximating much more closely to straight lines than those of the diagram. The E.M.F., current, and internal resistance were even more constant after the first few hours than those represented above. The total output for a single cell was 390 ampere-hours. A 15-ampere-hour cell tested to exhaustion gave 10.1 ampere-hours and 7.5 per cent loss of zinc by local action. This cell had been standing a long time with the undissolved alkali exposed to the air. 86. Testing Battery Designed for Small Lamps. — The following method for testing primary batteries, designed for lighting small incandescent lamps, is recommended by Mr. I. Probert. It may be impracticable to measure exactly the cur- rent flowing by the direct use of the ammeter, as the resistance of the latter, though low, is usually sufficient to materially reduce the current when the instrument is BATTERY TESTS. 135 inserted in the circuit. The present method is said to overcome this difficulty entirely. The illustration shows the connections. The battery B to be tested is joined up to the lamp (which has a voltmeter V across its terminals), the switch S being turned to the position shown. Under these conditions the battery works directly on the lamp, and the voltmeter V gives the voltage between the lamp terminals. In order to determine the current, the switch S is turned to the position shown by the dotted Fig. 68. — Testing Battery for Current. lines; this brings into circuit the auxiliary battery (preferably small portable accumulators, as they have a low resistance), the ammeter A, and the electrolytic resistance E. The current from B, though reduced by the resistance of the ammeter, is reinforced by the auxiliary battery B' ; and by adjusting the distance be- tween the plates of the electrolytic resistance the current can be adjusted to the greatest nicety, until the deflec- tion of V is the same as it was previous to the turning of the switch 8. Hence the ammeter A now shows the 136 PRIMARY BATTERIES. current which, under the former conditions, was flowing through B. The observation being taken on A, the . switch 8 is turned back to the position shown in the figure, and the battery B continues to work under the practical conditions. 87. Analysis of the Temperature Coefficient of a Battery. — If the temperature coefficient is a purely thermo- electric effect, then it should be susceptible of analysis by a measurement of the thermo-electric power of the two metal-liquid pairs. If, for example, the thermo- electric power of zinc— zinc sulphate and copper— cop- per sulphate can be measured separately, then their algebraic difference should represent the temperature coefficient of the Daniell cell, except so far as it may depend upon the thermo-electric power of the liquid pair, zinc sulphate— copper sulphate, which is the only other contact of dissimilar substances in the cell. So, also, if we combine the results obtained by measur- ing the thermo-electromotive force of zinc— zinc sulphate and mercury— mercurous sulphate in zinc sulphate, the result should be the temperature coefficient of a Clark standard cell. The meaning of thermo-electric power may perhaps be explained with advantage. If two junctions are at two temperatures f 1 and i° 2 , of which fO __ *i ~r tg 2 is the mean ; and if E is the E.M.F. of the pair under these conditions, then The thermo-electric power at f = - t 2 — tx 88. To Determine the Thermo-Electric Power of Zinc- Zinc Sulphate. — For this determination it is necessary BATTERY TESTS. 137 to have two contacts of zinc and a solution of its sulphate so related that one can be kept at a constant temperature, while the other is brought to successive different temperatures. Two stout glass tubes, about four inches (10 cm.) long and three-quarters of an inch in diameter, were connected near the tops by a narrow glass tube 10 inches (25 cm.) long. This will be called the " experimental cell." It was filled with zinc sulphate solution saturated at zero, and two zinc wires about a foot in length were suspended so as to dip half Fig. 63. — Diagram Showing Method of Measuring Thermo-Electric Power. or three-quarters of an inch into the liquid. The immersed ends were slightly amalgamated. Two ther- mometers were hung from a convenient support so that their bulbs dipped into the solution at the same depth as the zinc wires. The liquid filling the small tube served to make the electrical connection between the two limbs. The electromotive force was measured by the follow- ing method: R and R' in Fig. 63 are two resistance boxes of 10,000 ohms each. For most purposes R' may be less than 10,000. They must be of the most exact adjustment, or the errors, if any, must be known. They are connected in series with a good Leclanche" cell of 138 PRIMARY BATTERIES. higher E.M.F. than a standard Clark cell. Two Daniell cells would perhaps answer as well, but they are not so convenient. The total resistance in the circuit must be kept at 10,000 ohms, partly in R, and the remainder in R'. In a derived circuit from AB, the terminals of R, are placed in series a Clark cell SO, the experimental cell HO, and the long coil galvanometer Gr. It is better to include a resistance of 10,000 or 20,000 ohms besides in this circuit. The standard cell must be so connected that its positive is joined to the same terminal as the positive of the main circuit Leclanche" cell LO. A key must be placed in both circuits, preferably a double successive contact key of the style used with a Wheat- stone's bridge. The first points coming in contact close the main circuit ; increased pressure brings the second pair of contact points together, closing the derived circuit. When the pressure is relieved, the derived circuit opens first, and finally the main circuit. The adjustment consists in changing the resistance in the two boxes, keeping their sum 10,000, till the closing of the circuit does not cause the galvanometer needle to swing. A balance then subsists between the E.M.F. of SO and the fall of potential in the main circuit over the resistance between A and B. The cell EO is not in- cluded in the derived circuit in this first balance. The E.M.F. of the standard cell being known, the fall of potential over a single ohm in the main circuit is then known. The galvanometer employed was a Thomson reflecting instrument, astatic, and having a resistance of 7000 ohms. A change of a single ohm from R to R', or the reverse, when the balance is nearly effected, is perfectly BATTERY TESTS. 139 evident in the swing of the mirror. In fact, when a balance has been secured, if the key is kept closed for two or three seconds, the polarization of the main circuit Leclanche" cell is always evident in the overthrow of the balance. The next step is to include the experimental cell in the circuit as shown in the figure. Both limbs are sur- rounded with broken ice, and their temperature is nearly or quite the same. It is usually necessary to change the resistance R by a small number of units, perhaps two or three, in order to restore the balance. One of the limbs is then heated by successive stages, using a bath of warm water. The temperature is allowed to become as nearly stationary as possible, and a balance is again brought about as before. If the resistance JR must be increased to bring the galva- nometer needle to zero, then the E.M.F. of the experi- mental cell is so directed as to place the cell in series with the standard cell. If R must be diminished to secure a balance, the experimental cell JEO is in opposi- tion to the standard. The closing of the key therefore indicates at a glance which pole of HO is positive. For if HO is in series with SO, the galvanometer needle will swing in one direction ; if in opposition to SO, it will swing in the other direction ; and the direction of the swing always indicates to the operator whether R must be increased or diminished to effect a balance. With zinc in zinc sulphate the heating of one limb always produces an E.M.F. tending to make the zinc in the cold the positive plate, or to produce a current from cold to hot through the cell. The zinc in the cold limb acts like the zinc of a simple voltaic couple. The table gives the data of one series of experiments. 140 PRIMARY BATTERIES. TABLE I. Temp. C. Left Limb. Temp. C. Bight Limb. Temp. Difference (corrected). Resistance in It to Balance. Change in E.M.F. in Legal Volts. E.M.F. per Degree C. 0.6 9.8 14.4 19.0 27.8 37.6 47.3 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.2 9.2 13.8 18.4 27.2 37.0 46.7 9141 9183 9199 9225 9269 9320 9377 0.00660 0.00911 •>. 0.01319 0.02011 0.02812 0.03707 0.00072 0.00065 ? 0.00072 0.00074 0.00076 0.00079 The observation marked doubtful I had reason to think included an error in making the balance. 0° 10° 20° 80° 40° 60° Fig. 64. — Thermo-EIectric Power of Zn-ZnS0 4 and Cu-CuS0 4 . The mean thermo-electric power for a temperature of 23°.85 C. is therefore 0.00079. BATTERY TESTS. 141 89. Thermo-Electric Power of Copper— Copper Sulphate. — The apparatus was set up in precisely the same man- ner as before, with a solution of chemically pure copper sulphate of density 1.11. Two freshly electroplated copper wires were used as electrodes to dip with the thermometers into the solution. The current produced on heating one limb was found to have the same direc- tion as in the case of zinc sulphate, viz. from the cold limb to the warm through the liquid. The copper in the cold acts like the zinc of a simple voltaic cell. The table following contains all the data. TABLE II. Temp. C. Left Limb. Temp. C. Right Limb. Temp. Difference (corrected) . Resistance in JR to Balance. Change in B.M.F. in Legal Volts. E.M.F. per Degree C. 0.6 0.4 0.2 9129 5.1 0.4 4.5 9145 0.00252 0.00056 9.6 0.4 9.0 9166 0.00582 0.00065 16.1 0.2 15.7 9189 0.00944 0.00060 21.7 0.6 20.9 9216 0.01369 0.00066 32.1 0.5 31.4 9260 0.02061 0.00066 38.4 0.5 37.7 9295 0.02611 0.00069 44.7 0.6 43.9 9330 0.03162 0.00072 49.4 0.6 48.6 9353 0.03524 0.00073 The mean thermo-electric power for the copper- copper sulphate couple is therefore 0.00073 for the mean temperature of 25° C. The results for both zinc and copper immersed in their sulphates are plotted in the curves of Fig. 64, in which curve A refers to zinc and zinc sulphate, and curve B to copper and copper sul- phate. The total E.M.F.'s in legal volts, due to heating one limb, are plotted as ordinates, and the differences of 142 PRIMARY BATTER TES. temperature as abscissas. It will be noticed that both are slightly concave upward, indicating an increase of the thermo-electric power with rise of temperature. 1 90. Application to a Daniell Cell. — Since both zinc and copper, each in a solution of its sulphate, tend to become negative when heated, or to play the role of copper in a simple voltaic element, it is evident that they will exhibit the same phenomenon when set up together as a Daniell cell. When the entire cell is heated, the E.M.F. tends to rise because of the effect at the copper side of the couple, while the heating of the zinc and its sulphate gives to the zinc the power of gen- erating a counter E.M.F. Whether or not the E.M.F. of the cell as a whole will rise or fall with rise of temperature depends upon the relative thermo-electric power at the two sides. The thermo-electric power of Zn— ZnS0 4 is a little greater than that of Cu— CuS0 4 , so that the voltage of the cell falls by a very small coefficient per degree rise of temperature. The near equality of the two thermo-electric powers explains the small temperature coefficient of the Daniell cell. A little consideration will show that if the Cu— CuS0 4 side of a Daniell cell alone is heated, the E.M.F. of the cell will increase, while heating of the Zn— ZnS0 4 side 1 Since these experiments were made, the author has ascertained that similar ones were made by Bouty in 1880 (Journal de Physique, 1880, p. 229) . Bouty made use of a similar method, but employed a Lippmann voltmeter for electromotive forces. His results are 0.0006947 and 0.0006885 for zinc and copper in their own salts respectively, expressed as fractions of the E.M.F. of a Daniell. If it is assumed that the Daniell had an E.M.F. of 1.08 volts, the results of M. Bouty are 0.00075 for Zn-ZnS0 4 , and 0.00074 for Cu-CuS0 4 . Considering the small electromotive forces to be measured and the many disturbing causes, such as oxidation of the surfaces and convec- tion currents, the results are in very good agreement. BATTERY TESl'S. 143 alone will cause a somewhat greater decrease of E.M.F. The relative coefficients in the two cases were measured by setting up the experimental cell as a Daniell, making use of the same solutions that were used in the preced- ing determinations, and inserting in the small connect- ing tube between the two limbs a plug of purified asbestos to prevent intermixture of the two sulphates. After a balance had been obtained with a Rayleigh standard cell the experimental Daniell was substituted for it. A comparison of the E.M.F. of the two was thus made, and data secured to calculate the changes in the voltage of the Daniell by the subsequent heating. The resistance to balance the Rayleigh cell at 18°.7 C. was 9134 ohms. The E.M.F. of the cell at this tem- perature is 1.434 legal volts. Hence the fall of poten- tial over one ohm is 0.000157 legal volt. This constant is used to calculate changes in E.M.F. of the Daniell under test. TABLE III. Temp. 0. Zn-ZnS0 4 Limb. Temp. C. Cu-CuS0 4 Liml). Corrected Temp. Dif- ference. Resistance in E to Balance. Change in E.M.F. in Legal Volts. E.M.F. per Degree C. 1.0 0.9 0.1 6935 10.8 0.8 9.9 6896 0.00612 0.00062 18.8 0.9 17.8 6864 0.01115 0.00063 29.6 1.2 28.3 6815 0.01884 0.00067 45.9 1.3 47.5 6709 0.03548 0.00075 1.5 1.8 0.3 6948 1.4 15.2 13.5 6994 0.00722 0.00053 1.4 26.4 24.7 7036 0.01382 0.00056 2.0 38.8 36.5 7087 0.02182 0.00060 1.4 40.0 38.3 7093 0.02277 0.00059 1.3 48.5 4 f '.9 7129 0.02842 0.00061 144 PRIMARY BATTERIES. The foregoing table of results justifies the anticipa- tion respecting the changes in E.M.F. ; for it will be observed that heating the zinc end of the experimental cell causes a marked diminution of the E.M.F., while the opposite result follows the heating of the copper end. The coefficients in this case are both smaller than when each metal in its sulphate was used in both limbs 2 ^ ^ L — ,^ &■ + ^ ■*' 7 S * ~* 4^ S 7??'* / / L y^L^ h- -**& -P *£' U 117. Application to the Silver Chloride Cell. — If we assume the cell set up with a dilute solution of zinc sulphate, then the result of the action taking place when the cell is in operation is the formation of zinc chloride and the decomposition of silver chloride. Hence we have only to find the difference between the heats of formation of the two chlorides. From Thomsen 's investigations these are — Heat of formation of Zn, Cl 2 . . . 56,420 calories. Ag 2 , Cl 2 . . . 29,380 " .-. SH = 27,040 Therefore E = 0.000043 x 27,040 = 1.16 volts. 118. Helmholtz's Formula for Electromotive Force. — The direct measurement of the E.M.F. of a battery rarely gives a result agreeing exactly with the value calculated from the thermo-chemical data of the reaction accompanying the work of the battery. Helmholtz has THERMAL RELATIONS. 181 accordingly modified the formula from thermo-dynamic considerations so as to express the E.M.F. by the equation, E = 0.000043 C ± T-, dT in which O equals the heat of the reactions, E the electromotive force, and T absolute temperature, or temperature reckoned from a zero equal to — 273 of the Centigrade scale. The last term of the equation ex- presses a general relation which may admit of different interpretations. We may suppose that the chemical energy can be only partially transformed into electric energy, while the rest is directly converted into heat. Or an explanation of the discrepancy may be sought for in phenomena that tend to prevent the integral trans- formation of the chemical energy. An examination of this problem has been undertaken by Chronstchoff and Sitnikoff - 1 They have applied to the solution of the problem the thermo-electromotive force produced by the passage of a current at the con- tact surfaces of liquids and metals in a battery. This is known as the Peltier phenomenon. The expression for this E.M.F. of thermal origin is identical with the final term of the Helmholtz equation, which represents the difference between the chemical heat and the voltaic heat of a battery ; and the question arises whether they are equivalent expressions for the same identical quantity. The results of their experimental investigation of the problem raise a strong probability at least that this explanation is the correct one. One or two examples must suffice to illustrate the application of this method i Comptes Eendus, Tom. 108, 1889. 182 PRIMARY BATTERIES. to the explanation of the discrepancy existing between the observed value of the E.M.F. and that calculated from thermo-chemical data. The thermo-electromotive forces of the metal-liquid contacts were carefully measured by the experimenters in each case. 1. Case in which the E.M.F. observed is greater than the calculated value. Pb | PbS0 4 | ZnS0 4 | CuS0 4 | Cu. » > E = 0.61 volts at 20° C. E calculated from thermal values of CuS0 4 and PbS0 4 is 0.383. The thermo-electromotive force for the system Cu | CuS0 4 | Cu, between 0° to 50°, was found to be 0.00066 =^- dT For the system Pb | PbS0 4 | ZnS0 4 | PbS0 4 | Pb, — =-0.00011 volt. dT The value of T— is therefore 293x0.00077=0.225 volt. dT Then 0.383 + 0.225 = 0.608 volt. This is almost exactly identical with the observed value. 2. Case in which the observed E.M.F. is less than the calculated value. Zn | ZnS0 4 | PbS0 4 | Pb. E = 0.500 volt at 20° C. THERMAL RELATIONS. 183 E calculated from thermo-chemical data of ZnS0 4 and PbS0 4 is 0.697 volt. The thermo-electromotive force for the system Zn | ZnS0 4 | Zn was found to be 0.00076 volt per degree. For the system Pb | PbS0 4 | ZnS0 4 | PbS0 4 | Pb, *? = - 0.00011 volt as before. dT Hence T— = 293 x 0.00065 = 0.190 volt, dT and E = 0.697 - 0.190 = 0.507 volt. In this last example the authors of the paper appear dE to have made an error in respect to the sign of —— for Pb - PbS0 4 . The corrected value gives E= 0.473 volt. The conclusion derived from all the experiments is that the Peltier effect is of a nature to make up for the discrepancy between the electromotive force observed directly, and that calculated from the thermal values of the chemical actions. The Peltier effect gives a value of the same sign and of the same order as this differ- ence. INDEX. Activity and efficiency 157 " unit of 168 Advantages of sodium over potassium bichromate 50 Agglomerated carbon, Leclanche" cell with 74 Air voltaic battery 23 " battery absorbing oxygen from Ill Alloying, relative protection of 35 Amalgamation and local action 33 " effect of 34 Analysis of the temperature coefficient of a battery 136 Application to a Daniell cell 142, 179 " of Ohm's law to a single cell 157 " to a Smee cell 178 " to a Bunsen cell 179 " to a silver chloride cell 180 Arrangement to produce greatest current 160 Artificial electric organ 3 Jjaked carbon 46 Batteries without a depolarizer 78 " miscellaneous 106 Battery defined 1 primary and secondary 1 gravity 38 the Gethius 40 Sir William Thomson's tray 42 Grove's 43 Bnnsen's 46 bichromate 47 plunge 48 Ward and Sloane 54 Partz acid gravity 55 Taylor's 57 copper oxide 58 Edison-Lalande 60 185 186 INDEX. PASS Battery, chloride of silver 63 " open circuit 66 " prism Leclanche 69 " Samson 73 " Roberts' peroxide 74 " sulphate of mercury 75 " Fitch " chlorine " 76 " sea salt 79 Law 80 " diamond carbon 80 " closed carbon 82 " Laclede 83 " Grove's gas 106 " Upward's chlorine 109 " Powell's thermo-electro-chemical 110 " absorbing oxygen Ill " Jablochkoff 's 114 " with two carbon electrodes 114 " tests 115 Beetz 66 Behrens 4 Berlin Academy of Science 2- Bichromate battery 47 " chemical reactions in 49 " " directions for setting up 51 " Fuller cell 53 Bidwell's dry battery 113 Blue vitriol 38 Bunsen's battery 46 (calculation of E.M.F. from heat of combustion 176 Calorie 17, 169 Carbon cup, Leclanche' cells with 73 Carhart-Clark standard cell 95 Cells in series 158 Change in potential 19 " E.M.F 19 Chemical changes 2 " reaction in the simple voltaic cell 10 " reaction in the Daniell cell 30 " reactions in relation to energy 32 " " in the bichromate battery 49 " " in the Leclanche' cell 68 Chloride of lead standard cell 102 " silver cell 62 INDEX. 187 Chromic acid 49 " " as the depolarizer 53 Circuit, simple hattery 15 " electrolytic 30 Clausius, theory of 9, 175 Closed circuit batteries 27 " Leclanche' cells 71 " carbon batteries 82 Compressed plates of CuO 62 Condensing electroscope 21 Conductor, electrolytic 21 Contact force 21 Copper oxide battery 58 Coulomb 15 Counter electromotive force in a circuit 172 " " of electrolysis 173 Coupling together dissimilar cells 161 Daniell, Professor 28, 37 battery 28, 29 cell, E.M.F. of 31 ' ' polarization curves of 31 " defects of 36 " effect of temperature changes on 37 " temperature coefficient of 145 Data for polarization curves 119 Davy, Marie 75 Defects of the Daniell cell 36 Delany's modified gravity cell •■ 41 Depolarizer 27 " efficient 28 solid 28 " batteries without 78 Diagram of battery tests 118 Diamond carbon battery 80 Difference of potential 15 " " relation of E.M.F. to 18 " " between two points 18 Diffusion through porous cup 36, 68 " of zinc sulphate 41 " of the redissolved salt 91 " slowness of 91 Directions for setting up bichromate battery 51 Dissimilar cells in parallel 165 Dissipation of energy 16° 188 INDEX. PAGE Distinction between open and closed circuit batteries 27 Division of energy in a circuit with counter E.M.F 173 Double sulphate of potassium and chromium 50 " of sodium and chromium 51 Dry pile 4 " battery, Gassner 83 " " Meserole's composition for 85 Shelford Bidwell's 113 Ji/dison-Lalande battery 60 Effect, Volta 21, 22 " of amalgamation 34 " Peltier 25, 183 Efficiency test of a copper-oxide battery 131 Electric pendulum 6 " potential 15 " pressure 17 Electrode, positive 8 " negative 8 Electrolytic conductor 21 " circuit 30 " zinc 35 " process 108 Electrometer, quadrants of 6 Electromotive force 15, 17 " " relation of, to difference of potential 18 " " positive 16 " " standards of 86 " " depends on materials 17 " " effective 18 " " seat of 20 " " of the Daniell cell 31 " " of the Clark cell 90 " equation for 90,92,96 " " and temperature, relation between 93 " " of standard Daniell cell 98 " measured by silver voltameter 104 " " of various combinations 151 Electroscope, condensing 21 Element, simple voltaic 7 Energy of chemical separation 1 " expended g " chemical reactions in relation to 32 " conservation of Igg " dissipation of ^gg Equation for electromotive force 90 92 96 INDEX. 189 PAGE JB ailure of a cell to effect decomposition 175 Favre 171 Fitch " chlorine " battery 76 Formula for electromotive force 96 Fuller bichromate cell 53 Fundamental phenomena 7 " units 168 Galvani 3, 4, 21 Gassner dry battery 83 General considerations 166 Gravity battery 38 " cell, Delany's modified 41 " battery, Partz acid 55 Grouping of cells 157 " in parallel or multiple arc 159 " in multiple series 160 " of a battery for quickest action 161 Grove's battery 43 " gas battery 106 Heat equivalent of a current 169 " evolved in a circuit with no counter E.M.F 170 " of formation 23, 32 " of combination 17, 176 " reversible 26 " mechanical equivalent of 169 Helmholtz's equation 162 formula forelectromotive force 180 Hydrogen, accumulation of 11, 12 ' ' nascent 12 " sulphuretted 25 " plays the part of zinc 107 Internal resistance, to obtain 117 J oule's equivalent 169 " law 170 L/aclede battery 83 Latimer Clark's standard cell 86 Law battery 80 Leclanche cell 66 " " chemical reactions in 68 190 INDEX. FAQB Leclanche cell with carbon cup 73 " " with agglomerate-carbon 74 Leyden jar - 5 Local action 65 " " and amalgamation 33 Lord Eayleigh 87 form of Clark element 87 Loss of potential 19 Manganese dioxide 66, 67 " "in Leclanche cells 155 Mechanical equivalent 169 Microphone cell 72 Minchin's seleno-alumiuum cell 112 Miscellaneous batteries 106 Modifications of the silver chloride cell 64 N eedle, aluminum 21 Negative pole 4 " electrode 8 Nitrate of ammonium 45 Nitric oxide 44 Open circuit batteries 66 Origin of the voltaic cell 2 Oscillation, period of 6 Oxide of mercury standard cell 95 Jr artz acid gravity battery 55 Peltier effect 25, 183 " phenomenon 181 Pendulum, electric 5 Period of oscillation 6 Platinum black 78 " pulverulent 107 " finely divided 7U Polarization of a simple voltaic cell 12 ' ' curve of a simple cell 13 " progress of 13 " curves of Daniell cell 31 " progressive 74 Positive pole 4 " electrode 8 INDEX. 191 PAGE Potassium bichromate 47, 50, 52 " sulphide 25 Potential, electric 15 " difference of 15, 18, 20, 23, 25, 100 " practical zero of 1U " loss of 1!) " fall of 20 " slope of 21, 23 Practical unit 15 " zero of potential 16 Preece, W. H 38 Prism Leclanche' battery 69 Rack-and-pinion movement for plates 105 Reaction, chemical, in the simple voltaic cell 10 " " " Daniellcell 30 " " " bichromate battery 49 " " " Leclanche cell 68 " in relation to energy 32 Reduction of copper 37 Relation of potential differences to external and internal resistance 20 " between E.M.F. and temperature 93 Relative protection of alloying and amalgamating 35 " value of oxidants in batteries : 153 Removal of crystals of spent residue 77 Resistance, internal 20, 24 " external 14, 20, 24 Reversible heat 26 Roberts' peroxide battery 74 Oamson battery 73 Sea salt battery 79 Seleno-aluminum cell 112 Simple voltaic cell 7 Smee cell 78, 171 Sodium bichromate, advantages over potassium bichromate 50 Standards of electromotive force 86 Standard cell, Latimer Clark's 86 " " with low temperature coefficient 90 " " Carhart-Clark 95 " " oxide of mercury 95 " " chloride of lead 102 " " to measure the E.M.F. of 103 " Daniell cell, Sir William Thomson's 97 " Lodge's 98 192 INDEX. PAGE Standard Daniell cell, Fleming's 99 " solutions 101 Sulphate of mercury battery 75 Sulpho-chromic salt 55 Systematic test, what it includes 115 1 aylor's battery 57 Temperature coefficient, analysis of 136 " "of the Daniell cell 145 " " of the Clark cell 147,149 Test of typical Leclanche' cell 121 " of Leclanche' cell with depolarizer in carbon cylinder 124 " of zinc-carbon cell without depolarizer 127 " of a " dry " cell 128 " of a silver chloride cell 130 " of battery for small lamps 134 Thermal relations 166 Thermo-electric power of zinc — zinc sulphate 136 " of copper — copper sulphate 141 " " of mercury — mercurous sulphate 146 Thermo-electro-chemical battery 110 Theory of the voltaic element 8 " of Clausius 9, 175 Thomson, Sir William 6, 21 Thomson's contact apparatus 22 " tray battery 42 Time-constant 162 Typical Leclanche cell 67 " " " test of 121 Units of force, work, activity, heat 168 Upward's chlorine battery 109 Volte 3, 4, 21 " effect 21, 22 Voltaic cell, inconstancy of 11 " " origin of 2 " " simple 7 ' ' element 8 Volta's pile 3 Voltameter 104 Voltmeter 135 " electrostatic 18 " Lippmann 142 INDEX. 193 PAGE Water marks in jar ... ,. 72 What a systematic test includes 115 Work done 16 " positive and negative 16 Woulff 's bottle 107 Zamboni i Zamboni's pile 4 .';■ . MMWSj' v" vXv^SviJ'VV '■;■