ON THE 
 
 MECHANICAL EQUIVALENT OF HEAT, 
 
 WITH SUBSIDIARY RESEARCHES ON 
 
 THE VARIATION OF THE MERCURIAL FROM THE AIR 
 
 THERMOMETER, AND ON THE VARIATION OF 
 
 THE SPECIFIC HEAT OF WATER. 
 
 BY 
 
 HENRY A. ROWLAND, 
 
 PROFESSOR OF PHYSIOS IN THE JOHNS HOPKINS UNIVERSITY. 
 
 Presented June llth, 1879. 
 
 [REPRINTED FROM THE PROCEEDINGS OF THE AMERICAN ACADEMY OF 
 ARTS AND SCIENCES.] 
 
 CAMBRIDGE: 
 
 UNIVERSITY PRESS: JOHN WILSON & SON. 
 
 1880.
 
 OP ARTS AND SCIENCES. 
 
 INVESTIGATIONS ON LIGHT AND HEAT, made and published wholly or in part with 
 appropriation from the KUMFOUD FUND. 
 
 Geol. 
 75 Lib. 
 
 QC 
 
 V. 
 
 ON THE MECHANICAL EQUIVALENT OF HEAT, WITH SUB- 
 SIDIARY RESEARCHES ON THE VARIATION OF THE 
 MERCURIAL FROM THE AIR THERMOMETER, AND ON 
 THE VARIATION OF THE SPECIFIC HEAT OF WATER. 
 
 BY HENRY A. ROWLAND,* 
 
 Professor of Physics in the Johns Hopkins University. 
 Presented June llth, 1879. 
 
 CONTENTS. 
 
 I. Introductory Remarks ... 
 . Thermometry .... 
 (a.) General View of Thermome- 
 try 
 
 (6.) The Mercurial Thermometer 78 
 (c.) Relation of the Mercurial and 
 
 Air Thermometers . . 83 
 
 1. General and Historical 
 
 Remarks . . 88 
 
 2. Description of Apparatus 90 
 
 3. Results of Comparison . 97 
 (d.) Reduction to the Absolute 
 
 Scale .... 112 
 Appendix to Thermometry . . 116 
 
 III. Calorimetry ..... 119 
 
 (a.) Specific Heat of Water . . 119 
 (6.) Heat Capacity of the Calo- 
 
 rimeter .... 131 
 
 IV. Determination of Equivalent . . 137 
 
 (a.) Historical Remarks . . 137 
 
 1. General Review of Meth- 
 
 ods .... 137 
 
 2. Results of Best Deter- 
 
 minations . . .140 
 (6.) Description of Apparatus . 165 
 
 1. Preliminary Remarks . 155 
 
 2. General Description . 157 
 
 3. Details . . . .168 
 (c. ) Theory of the Experiment . 168 
 
 1. Estimation of Work done 163 
 
 2. Radiation . . . .168 
 
 3. Corrections to Thermom- 
 
 eters, etc. ... in 
 (d.) Results 173 
 
 1. Constant Data . . 173 
 
 2. Experimental Data and 
 
 Tables of Results . .174 
 V. Concluding Remarks, and Criticism 
 
 of Results and Methods . . 197 
 
 I. INTRODUCTORY REMARKS. 
 
 AMONG the more important constants of nature, the ratio of the 
 heat unit to the unit of mechanical work stands forth prominent, and 
 is used almost daily by the physicist. Yet, when we come to consider 
 
 * This research was originally to have been performed in connection with 
 Professor Pickering, but the plan was frustrated by the great distance between 
 our residences. An appropriation for this experiment was made by the Ameri- 
 can Academy of Arts and Sciences at Boston, from the fund which was insti- 
 tuted by Count Rumford, and liberal aid was also given by the Trustees of the 
 Johns Hopkins University, who are desirous, as far as they can, to promote 
 original scientific investigation.
 
 76 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 the history of the subject carefully, we find that the only experimenter 
 tho has made the determination with anything like the accuracy 
 demanded by modern science, and by a method capable of giving 
 good results, is Joule, whose determination of thirty years ago, con- 
 firmed by some recent results, to-day stands almost, if not quite, alone 
 among accurate results on the subject. 
 
 But Joule experimented on water of one temperature only, ai 
 did not reduce his results to the air thermometer; so that we are still 
 left in doubt, even to the extent of one per cent, as to the value 
 equivalent on the air thermometer. 
 
 The reduction of the mercurial to the air thermometer, and thence 
 to the absolute scale, has generally been neglected between and 
 100 by most physicists, though it is known that they differ several 
 tenths of a degree at the 45 point. In calorimetric researches this 
 may produce an error of over one, and even approaching two per cent, 
 especially when a Geissler thermometer is used, which is the worst in 
 this respect of any that I have experimented on ; and small intervals 
 on the mercurial thermometers differ among themselves more than 
 one per cent from the difference of the glass used in them. 
 
 Again, as water is necessarily the liquid used in calorimeters, its 
 variation of specific heat with the temperature is a very important 
 factor in the determination of the equivalent. Strange as it may 
 appear, we may be said to know almost nothing about the variation 
 of the specific heat of water with the temperature between and 
 100 C. 
 
 Regnault experimented only above 100 C. The experiments of / 
 Him, and of Jamin and Amaury, are absurd, from the amount of 
 variation which they give. Pfaundler and Plattner confined them- 
 selves to points between and 13. Munchausen seems to have 
 made the best experiments, but they must be rejected because he did 
 not reduce to the air thermometer. 
 
 In the present series of researches, I have sought, firstly, a method 
 of measuring temperatures on the perfect gas thermometer with an 
 accuracy scarcely hitherto attempted, and to this end have made an 
 extended study of the deviation of ordinary thermometers from the 
 air thermometer ; and, secondly, I have sought a method of determin- 
 ing the mechanical equivalent of heat so accurate, and of so extended 
 a range, that the variation of the specific heat of water should follow 
 from the experiments alone. 
 
 As to whether or not these have been accomplished, the following 
 pages will show. The curious result that the specific heat of water
 
 OP ARTS AND SCIENCES. 77 
 
 on the air thermometer decreases from to about 30 or 35, after 
 which it increases, seems to be an entirely unique fact in nature, seeing 
 that there is apparently no other substance hitherto experimented upon 
 whose specific heat decreases on rise of temperature without change of 
 state. From a thermodynamic point of view, however, it is of the 
 Bame nature as the decrease of specific heat which takes place after 
 the vaporization of a liquid. 
 
 The close agreement of my result at 15.7 C. with the old result of 
 Joule, after approximately reducing his to the air thermometer and 
 latitude of Baltimore, and correcting the specific heat of copper, is 
 very satisfactory to us both, as the difference is not greater than 1 in 
 400, and is probably less. 
 
 I hope at some future time to make a comparison with Joule's 
 thermometers, when the difference can be accurately stated. 
 
 II. THERMOMETRY. 
 (a.) General View. 
 
 The science of thermometry, as ordinarily studied, is based upon 
 the changes produced in bodies by heat. Among these we may men- 
 tion change in volume, pressure, state of aggregation, dissociation, 
 amount and color of light reflected, transmitted, or emitted, hardness, 
 pyro-electric and thermo-electric properties, electric conductivity or 
 specific induction capacity, magnetic properties, thermo-dynamic prop- 
 erties, &c. ; and on each of these may be based a system of ther- 
 mometry, each one of which is perfect in itself, but which differs from 
 all the others widely. Indeed, each method may be applied to nearly 
 all the bodies in nature, and hundreds or thousands of thermometric 
 scales may be produced, which may be made to agree at two fixed 
 points, such as the freezing and boiling points of water, but which 
 will in general differ at nearly, if not all, other points. 
 
 But from the way in which the science has advanced, it has come 
 to pass that all methods of thermometry in general use to the present 
 time have been reduced to two or three, based respectively on the 
 apparent expansion of mercury in glass and on the absolute expan- 
 sion of some gas, and more lately on the second law of thermo- 
 dynamics. 
 
 Each of these systems is perfectly correct in itself, and we have no 
 right to designate either of them as incorrect. We must decide 
 a priori on some system, and then express all our results in that 
 system : the accuracy of science demands that there should be no
 
 78 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 ambiguity on that subject. In deciding among the three systems, we 
 should be guided by the following rules : 
 
 1st. The system should be perfectly definite, so that the same 
 temperature should be indicated, whatever the thermometer. 
 2d. The system should lead to the most simple laws in nature. ^ 
 Sir William Thomson's absolute system of thermometry, coinciding 
 with that based on the expansion of a perfect gas, satisfies these most 
 nearly. The mercurial thermometer is not definite unless the kind of 
 glass is given, and even then it may vary according to the way the 
 bulb is blown. The gas thermometer, unless the kind of gas is given, 
 is not definite. And, further, if the temperature. as given by either of 
 these thermometers was introduced into the equations of thermo- 
 dynamics, the simplest of them would immediately become compli- 
 cated. 
 
 Throughout a small range of temperature, these systems agree 
 more or less Completely, and it is the habit even with many eminent 
 physicists to regard them as coincident between the freezing and boil- 
 ing points of water. We shall see, however, that the difference 
 between them is of the highest importance in thermometry, especially 
 where differences of temperature are to be used. 
 
 For these reasons I have reduced all my measures to the absolute 
 system. 
 
 The relation between the absolute system and the system based on 
 the expansion of gases has been determined by Joule and Thomson 
 in their experiments on the flow of gases through porous plugs 
 (Philosophical Transactions for 1862, p. 579). Air was one of the 
 most important substances they experimented upon. 
 
 To measure temperature on the absolute scale, we have thus only 
 to determine the temperature on the air thermometer, and then reduce 
 to the absolute scale. But as the air thermometer is very inconvenient 
 to use, it is generally more convenient to use a mercurial thermometer 
 which has been compared with the air thermometer. Also, for small 
 changes of temperature the air thermometer is not sufficiently sensi- 
 tive, and a mercurial thermometer is necessary for interpolation. 
 1 shall occupy myself first with a careful study of the mercurial 
 thermometer. 
 
 (6.) The Mercurial Thermometer. 
 
 Of the two kinds of mercurial thermometers, the weight ther- 
 mometer is of little importance to our subject. I shall therefore con- 
 fine myself principally to that form having a graduated stem. For
 
 OF ARTS AND SCIENCES. 79 
 
 convenience in use and in calibration, the principal bulb should be 
 elongated, and another small bulb should be blown at the top. This 
 latter is also of the utmost importance to the accuracy of the instru- 
 ment, and is placed there by nearly all makers of standards.* It is 
 used to place some of the mercury in while calibrating, as well as 
 when a high temperature is to be measured ; also, the mercury in the 
 larger bulb can be made free from air-bubbles by its means. 
 
 Most standard thermometers are graduated to degrees; but Reg- 
 nault preferred to have his thermometers graduated to parts of equal 
 capacity whose value was arbitrary, and others have used a single 
 millimeter division. 'As thermometers change with age, the last two 
 methods are the best ; and of the two I prefer the latter where the 
 highest accuracy is desired, seeing that it leaves less to the maker and 
 more to the scientist. The cross-section of the tube changes continu- 
 ously from point to point, and therefore the distribution of marks 
 on the tube should be continuous, which would involve a change of the 
 dividing engine for each division. But as the maker divides his tube, 
 he only changes the length of his divisions every now and then, so as 
 to average his errors. This gives a sufficiently exact graduation for 
 large ranges of temperature ; but for small, great errors may be intro- 
 duced. Where there is an arbitrary scale of millimeters, I believe it 
 is possible to calibrate the tube so that the errors shall be less than 
 can be seen with the naked eye, and that the table found shall repre- 
 sent very exactly the gradual variation of the tube. 
 
 In the calibration of my thermometers with the millimetric scale, I 
 have used several methods, all of which are based upon some graphical 
 method. The first, which gives all the irregularities of the tube with 
 great exactness, is as follows. 
 
 A portion of the mercury having been put in the upper bulb, so as 
 to leave the tube free, a column about l;) mm - long is separated off. 
 This is moved from point to point of the tube, and its length carefully 
 measured on the dividing engine. It is not generally necessary to 
 move the column its own length every time, but it may be moved 
 20"""- or 25 nim- , a record of the position of its centre being kept. To 
 eliminate any errors of division or of the dividing engine, readings 
 were then taken on the scale, and the "lengths reduced to their value 
 in scale divisions. The area of the tube at every point is inversely as 
 the length of the column. We shall thus have a series of figures 
 nearly equal to each other, if the tube is good. By subtracting the 
 
 * Geissler and Casella omit it, which should condemn their thermometers.
 
 80 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 smallest from each of the others, and plotting the results as ordinates, 
 with the thermometer scale as abscissas, and drawing a curve through 
 the points so found, we have means of finding the area at any point. 
 The curve should not he drawn exactly through the points, but rather 
 around them, seeing they are the average areas for some distance each 
 side of the point. With good judgment, the curve can be drawn with 
 great accuracy. I then draw ordinates every 10 mra> , and estimate the 
 average area of the tube for that distance, which I set down in a table. 
 As the lengths are uniform, the volume of the tube to any point is 
 found by adding up the areas to that point. 
 
 But it would be unwise to trust such a method for very long tubes, 
 seeing the mercury column is so short, and the columns are not end to 
 end. Hence I use it only as supplementary to one where the column 
 is about 50 ram - long, and is always moved its own length. This estab- 
 lishes the volumes to a series of points about 50"' m- apart, and the 
 other table is only used to interpolate in this one. There seems to be 
 no practical object in using columns longer than this. 
 
 Having finally constructed the arbitrary table of volumes, I then 
 test it by reading with the eye the length of a long mercury column. 
 No certain error was thus found at any point of any of the ther- 
 mometers which I have used in these experiments. 
 
 While measuring the column, great care must be taken to preserve 
 all parts of the tube at a uniform temperature, and only the extreme 
 ends must be touched with the hands, which should be covered with 
 cloth. 
 
 If V is the volume on this arbitrary scale, the temperature on the 
 mercurial thermometer is found from the formula T =.O V < , 
 where C and f. Q are constants to be determined. If the thermometer 
 contains the and 100 points, we have simply 
 
 
 G _ 100 
 
 - i o v ~ V 
 
 Otherwise C is found by comparison with some other thermometer, 
 which must be of the same kind of glass. 
 
 It is to be carefully noted that the temperature on the mercurial 
 thermometer, as I have defined it, is proportional to the apparent 
 expansion of mercury as measured on the stem. By defining it as 
 proportional to the true volume of mercury in the stem, we have to 
 introduce a correction to ordinary thermometers, as Poggendorf has 
 shown. As I ouly use the mercurial thermometer to compare with 
 the air thermometer, and as either definition is equally correct, I will
 
 OF ARTS AND SCIENCES. 81 
 
 not further discuss the matter, but will use the first definition, as 
 being the simplest. 
 
 In the above formula I have implicitly assumed that the apparent 
 expansion is only a function of the temperature ; but in solid bodies 
 like glass there seems to be a progressive change in the volume as 
 time advances, and especially after it has been heated. And hence in 
 mercurial and alcohol thermometers, and probably in general in all 
 thermometers which depend more or less on the expansion of solid 
 bodies, we firyi that the reading of the thermometer depends, not only 
 on its present temperature, but also on that to which it has been sub- 
 jected within a short time ; so that, on heating a thermometer up to a 
 certain temperature, it does not stand at the same point as if it had 
 been cooled from a higher temperature to the given temperature. As 
 these effects are without doubt due to the glass envelope, we might 
 greatly diminish them by using thermometers filled with liquids which 
 expand more than* mercury : there are many of these which expand 
 six or eight times as much, and so the irregularity might be dimin- 
 ished in this ratio. But in this case we should find that the correction 
 for that part of the stem which was outside the vessel whose tem- 
 perature we were determining would be increased in the same propor- 
 tion ; and besides, as all the liquids are quite volatile, or at least wet 
 the glass, there would be an irregularity introduced on that account. 
 A thermometer with liquid in the bulb and mercury in the stem would 
 obviate these inconveniences ; but even in this case the stem would 
 have to be calibrated before the thermometer was made. By a com- 
 parison with the air-thermometer, a proper formula could be obtained 
 for finding the temperature. 
 
 But I hardly believe that any thermometer superior to the mer- 
 curial can at present be made, that is, any thermometer within the 
 same compass as a mercurial thermometer, and I think that the 
 best result for small ranges of temperature can be obtained with it by 
 studying and avoiding all its sources of error. 
 
 To judge somewhat of the laws of the change of zero within the 
 limits of temperature which I wished to use, I took thermometer 
 No. G1G3, which had lain in its case during four months at an average 
 temperature of about 20 or 25 C., and observed the zero point, after 
 heating to various temperatures, with the following result. The time 
 of heating was only a few minutes, and the zero point was taken 
 immediately after ; some fifteen minutes, however, being necessary for 
 the thermometer to entirely cool. 
 VOL. xv. (N. s. vn.) 6
 
 82 
 
 PROCEEDINGS OF THE AMERICAN ACADEMY 
 TABLE I. SHOWING CHANGE OF ZERO POINT. 
 
 Temperature 
 of Bulb 
 before finding 
 the Point. 
 
 Change of 
 Point. 
 
 Temperature 
 of Bulb 
 before finding 
 the Point. 
 
 Change of 
 Point. 
 
 22.5 
 
 
 
 700 
 
 .115 
 
 300 
 
 .016 
 
 81.0 
 
 .170 
 
 405 
 
 .033 
 
 90.0 
 
 .281 
 
 ' 51.0 
 
 .0:59 
 
 100.0 
 
 .313 
 
 CO.O 
 
 .105 
 
 100.0 
 
 .347 
 
 The second 100 reading was taken after boiling for some time. 
 
 It is seen that the zero point is always lower after heating, and that 
 in the limits of the table the lowering of the zero is about propor- 
 tional to the square of the increase of temperature above 25 C. 
 This law is not true much above 100, and above a certain tempera- 
 ture the phenomenon is reversed, and the zero point is higher after 
 heating ; but for the given range it seems quite exact. 
 
 It is not my purpose to make a complete study of this phenomenon 
 with a view to correcting the thermometer, although this has been 
 undertaken by others. But we see from the table that the error can- 
 not exceed certain limits. The range of temperature which I have 
 used in each experiment is from 20 to 30 C., and the temperature 
 rarely rose above 40 C. The change of zero in this range only 
 amounts to 0.03 C. 
 
 The exact distribution of the error from this cause throughout the 
 scale has never been determined, and it affects my results so little that 
 I have not considered it worth investigating. It seems probable, how- 
 ever, that the error is distributed throughout the scale. IE it were 
 uniformly distributed, the value of each division would be less than 
 before by the ratio of the lowering at zero to the temperature to 
 which the thermometer was heated. 
 
 The maximum errors produced in my thermometers by this cause 
 would thus amount to 1 in 1300 nearly for the 40 thermometer, and 
 to about 1 in 2000 for the others. Rather than allow for this, it is 
 better to allow time for the thermometer to resume its original state. 
 
 Only a few observations were made upon the rapidity with which 
 the zero returned to its original position. After heating to 81, the* 
 zero returned from OM70 to 0.148 in two hours and a half. 
 After heating to 100, the zero returned from 0.347 to 110 
 in nine days, and to -0.022 in one month. Reasoning from this, I
 
 OF ARTS AND SCIENCES. 83 
 
 should say that in one week thermometers which had not been heated 
 above 40 should be ready for use again, the error being then supposed 
 to be less than 1 in 4000, and this would be partially eliminated by 
 comparing with the air thermometer at the same intervals as the ther- 
 mometer is used, or at least heating to 40 one week before comparing 
 with the air thermometer. 
 
 As stated before, when a thermometer is heated to a very high 
 point, its zero point is raised instead of lowered, and it seems probable 
 that at some higher point the direction of change is reversed again ; 
 for, after the instrument comes from the maker, the zero point con- 
 stantly rises until it may be 0.6 above the mark on the tube. This 
 gradual change is of no importance in my experiments, as I only 
 use differences of temperature, and also as it Was almost inappreciable 
 in my thermometers. 
 
 Another source of error in thermometers is that due to the pressure 
 on the bulb. In determining the freezing point, large errors may be 
 made, amounting to several hundredths of a degree, by the pressure of 
 pieces of -ice. In my experiments, the zero point was determined in 
 ice, and then the thermometer was immersed in the water of the com- 
 parator at a depth of about G0 cm . The pressure of this water affected 
 the thermometer to the extent of about 0.01, and a correction was 
 accordingly made. As differences of temperature were only needed, 
 no correction was made for variation in pressure of the air. 
 
 It does 'not seem to me well to use thermometers with too small a 
 stem, as I have no doubt that they are subject to much greater irreg- 
 ularities than those with a coarse bore. For the capillary action 
 always exerts a pressure on the bulb. Hence, when the mercury rises, 
 the pressure is due to a rising meniscus which causes greater pressure 
 than the falling meniscus. Hence, an apparent friction of the mercu- 
 rial column. Also, the capillary constant of mercury seems to depend 
 on the electric potential of its surface, which may not be constant, 
 and would thus cause an irregularity. 
 
 My own thermometers did not show any apparent action of this 
 kind, but Pfaundler and Plattner mention such an'action, though they 
 give another reason for it. 
 
 (c.) Relation of the Mercurial and Air Thermometers. 
 1. GENERAL AND HISTORICAL REMARKS. 
 
 Since the time of Dulong and Petit, many experiments have been 
 made on the difference between the mercurial and the air thermometer,
 
 84 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 but unfortunately most of them have been at high temperatures. As 
 weight thermometers have been used by some of the best experi- 
 menters, I shall commence by proving that the weight thermometer 
 and stem thermometer give the same temperature ; at the same time, 
 however, obtaining a convenient formula for the comparison of the air 
 thermometer with the mercurial. 
 
 For the expansion of mercury and of glass the following formulae 
 must hold : 
 
 For mercury, V = V (1 + a t -f J * 2 + &c.) ; 
 glass, V = V' Q (1 + a t 4- (3 1 2 4- &c.). 
 
 In both the weight and stem thermometers we must have V '= V. 
 
 
 
 where V and F are the volumes of the glass and of the mercury 
 reduced to zero, and t is the temperature on the air thermometer. 
 The temperature by the weight thermometer is 
 
 
 P 
 
 
 where P , P t , &c. are the weights of mercury in the bulb at C., 
 t C., &c. 
 
 Now these weights are directly as the volumes of the mercury at 0. 
 
 seeing that V is constant. 
 
 .-. T 100 
 
 In the stem thermometers we have F , the volume of mercury at 0, 
 constant, and the volume of the glass that the mercury fills, reduced 
 to 0, variable. As the volume of the glass V' n is the volume reduced 
 to 0, it will be proportional to the volume of bulb plus the volume 
 of the tube as read off on the scale which should be on the tube.
 
 OF ARTS AND SCIENCES. 85 
 
 At 
 
 .-. r=ioo 
 
 which is the same as for the weight thermometer. 
 
 If the fixed points are and t* instead of and 100, we can 
 write 
 
 T _ t ,At + Bfl^-Cfl + & c . 
 
 r=<{i + (<-<<) [f + fl' + fc+o] +&<>} 
 
 As T and t are nearly equal, and as we shall determine the con- 
 stants experimentally, we may write 
 
 t = T a t (t 1 t) (b t) + &c., 
 
 where t is the temperature on the air thermometer, and T that on the 
 mercurial thermometer, and a and b are constants to be determined for 
 each thermometer. 
 
 The formula might be expanded still further, but I think there are 
 few cases which it will not represent as it is. Considering b as equal 
 to 0, a formula is obtained which has been used by others, and from 
 which some very wrong conclusions have been drawn. In some kinds 
 of glass there are three points which coincide with the air thermome- 
 ter, and it requires at least an equation of the third degree to repre- 
 sent this. 
 
 The three points in which the two thermometers coincide are given 
 by the roots of the equation 
 
 t(*-t) (6-0 = 0, 
 and are, therefore, 
 
 In the following discussion of the historical results, I shall take 
 and 100 as the fixed points. Hence, t' = 100. To obtain a and b, 
 two observations are needed at some points at a distance from and 
 100. That we may get some idea of the values of the constants in 
 the formula for different kinds of glass, I will discuss some of the 
 experimental results of Regnault and others with this in view.
 
 86 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 Renault's results are embodied, for the most part, in tables given on 
 p 239 of the first volume of his Relation des Esper^ces. The 
 Lres given there are obtained from curves drawn to represent he 
 mean of his experiments, and do not contain any theorefcal results. 
 The direct application of my formula to his experiments could hardly 
 be made without immense labor in finding the most probable value 
 the constants. 
 But the following seem to satisfy the experiments quite 
 
 ' Cristal de Choisy-le-Roi i = 0, = -000 000 82. 
 
 Verre Ordinaire 6-245, = .000 000 34. 
 
 VerreVert * = 270, a = .000 000 095. 
 
 Verrede Suede i = +10, a = .000 000 14. 
 
 From these values I have calculated the following : 
 
 TABLE II. REGNAULT'S RESULTS COMPARED WITH THE FORMULA. 
 
 1 
 
 Choisy-le-Roi. 
 
 Verre Ordinaire. 
 
 Verre Vert. 
 
 Verre de Suede. 
 
 1 
 
 | 
 
 1 
 
 1 
 
 1 
 
 1 
 
 1 
 
 1 
 
 *5 
 
 i 
 
 1 
 
 | 
 
 1 
 
 I 
 
 3 
 
 1 
 
 1 
 
 
 
 5 
 
 1 
 
 1 
 
 5 
 
 | .2 
 
 1 o 
 
 5 
 
 
 
 i 
 
 5 
 
 100 
 
 o 
 
 o 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 120 
 140 
 160 
 180 
 200 
 
 120.12 
 140 29 
 160.52 
 180.80 
 201 25 
 
 120.09 
 14025 
 KiO.49 
 
 18083 
 201 .28 
 
 +.03 
 +.04 
 +.03 
 -.03 
 -.03 
 
 119.95 
 139. W 5 
 159.74 
 179.63 
 199.70 
 
 119.90 
 139 80 
 159.72 
 179.68 
 19!l.fi9 
 
 +.05 
 +.05 
 +.02 
 .05 
 + .01 
 
 12007 
 140.21 
 16(1.40 
 180.60 
 20080 
 
 120.09 .01 
 140 22 ! .01 
 160.391 +.01 
 180.62 i .02 
 200.-S9 .09 
 
 120.04 120.04 
 140.11 ! 140.10 
 160 20 i 160.21 
 1MI.33 18034 
 20 ( 50 ! 200 53 
 
 
 +.01 
 .01 
 .01 
 .03 
 
 220 
 
 221 KJ 
 
 221.86 
 
 .04 
 
 2l9.HOl219.78 
 
 + .02 
 
 221.20 221.23 
 
 .03 
 
 220 75 ! 220 78 
 
 .03 
 
 240 
 
 242.55 
 
 242.50 
 
 
 239.90! 239 96 
 
 -.06 
 
 241.60 241.63 
 
 -.03 
 
 241.16 
 
 2*1.08 
 
 + .08 
 
 200 
 
 263 44 
 
 203.4H 
 
 .02 
 
 260.20 
 
 2C021 
 
 -.01 
 
 262.15 262.09 
 
 +.07 
 
 
 
 
 WO 
 
 2H4.48 
 
 2S452 
 
 -.04 
 
 280.58 2SO.O>l 
 
 02 
 
 282.85 262.63 
 
 +.22 
 
 
 
 
 300 
 3?0 
 
 3(15.72 
 
 ::27.25 
 
 3115.70 
 327.20 
 
 -.04 
 
 -.05 
 
 301.08! 301. 12 
 321.80! 321.80 
 
 .04 
 00 
 
 
 
 
 
 
 
 340 
 
 349.30 
 
 348.88 
 
 +.42 
 
 434.00 
 
 34204 
 
 +.36 
 
 
 
 
 
 
 
 The formula, as we see from the table, represents all Regnault's 
 curves with great accuracy, and if we turn to his experimental results 
 we shall find that the deviation is far within the limits of the experi- 
 mental errors. The greatest deviation happens at 340, and may be 
 accounted for by an error in drawing the curve, as there are few ex- 
 perimental results so high as this, and the formula seems to agree 
 with them almost as well as Regnault's own curve. 
 
 The object of comparing the formula with Regnault's results at 
 temperatures so much higher than I need, is simply to test the formula 
 through as great a range of temperatures, and for as many kinds of 
 
 Corrected from 280.52 in Regnault's table.
 
 OF ARTS AND SCIENCES. 
 
 87 
 
 glass, as possible. If it agrees reasonably well throughout a great 
 range, it will probably be very accurate for a small range, provided 
 we obtain the constants to represent that small range the best. 
 
 Having obtained a formula to represent any 'series of experiments, 
 we can hardly expect it to hold for points outside our series, or even 
 for interpolating between experiments too far apart, as, very often, a 
 small change in one of the constants may affect the part we have not 
 experimented on in a very marked manner. ' Thus in applying the 
 formula to points between and 100 the value of b will affect the 
 result very much. In the case of the glass Choisy-le-Roi many 
 values of b will satisfy the observations besides 6 = 0. For the 
 ordinary glass, however, b is well determined, and the formula is of 
 more value between and 100. 
 
 The following table gives the results of the calculation. 
 
 TABLE III. REGNAOI/T'S RESULTS COMPARED WITH THE FORMULA. 
 
 
 Calculated 
 
 Calculated 
 
 
 
 Calculated 
 
 
 . 
 
 ft = .00000032 
 
 a = .00000031 
 
 Observed. 
 
 
 = .00000044 
 
 
 Thermome- 
 
 ft = 0. 
 
 b = 245. 
 
 
 A 
 
 6 = 260. 
 
 J 
 
 ter. 
 
 Clioisy-le-Roi. 
 
 Verre 
 Ordinaire. 
 
 Verre 
 Ordinaire. 
 
 
 Verre 
 Ordinaire. 
 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 10 
 
 10.00 
 
 10.07 
 
 
 .... 
 
 10.10 
 
 
 20 
 
 19.99 
 
 20.12 
 
 .... 
 
 
 20 17 
 
 . . . 
 
 30 
 
 29.98 
 
 30.15 
 
 30.12 
 
 +.03 
 
 30.21 
 
 +.09 
 
 40 
 
 39.97, 
 
 40.17 
 
 40.23 
 
 -.06 
 
 40.23 
 
 
 
 60 
 
 49.96 
 
 60.17 
 
 50.23 
 
 .06 
 
 60.23 
 
 
 
 60 
 
 5:>.'.>5 
 
 60.15 
 
 60.24 
 
 .09 
 
 60.21 
 
 .03 
 
 70 
 
 09.95 
 
 70.12 
 
 70.22 
 
 .10 
 
 70.18 
 
 .04 
 
 80 
 
 79.96 
 
 80.09 
 
 80.10 
 
 .01 
 
 80.11 
 
 +.01 
 
 90 
 
 89.97 
 
 90.05 
 
 
 
 90.07 
 
 
 100 
 
 100 
 
 100 
 
 100 ' 
 
 .... 
 
 100 
 
 
 
 Regnault does not seem to have published any experiments on 
 Choisy-le-Roi glass between and 100, but in the table between 
 pp. 226, 227, there are some results for ordinary glass. The separate 
 observations do not seem to have been very good, but by combining 
 the total number of observations I have found the results given 
 above. The numbers in the fourth column are found by taking the 
 mean of Regnault's results for points as near the given temperature 
 as possible. The agreement is only fair, but we must remember that 
 the same specimens of glass were not used in this experiment as in the 
 others, and that for these specimens the agreement is also poor above 
 100. The values a = .000 000 44 and b = 260 are much better
 
 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 above 100 for the given specimens. 
 
 The table seems to show that between and 100 a thermometer of 
 Choisy-le-Koi almost exactly agrees with the air, thermometer But 
 this is not at all conclusive. Regnault, however, remarks * that be- 
 tween and 100 thermometers of this glass agree more nearly with 
 the air thermometer than those of ordinary glass, though he stat 
 the difference to amount to .1 to .2 of a degree, the mercurial t 
 mometer standing below the air thermometer. With the exception of 
 this remark of Regnault's, no experiments have ever been publi. 
 in which the direction of the deviation was similar to tins. All ( 
 percenters have found the mercurial thermometer to stand above the 
 air thermometer between and 100, and my own expenments agree 
 with this. However, no general rule for all kinds of glass can b< 
 laid down. 
 
 Boscha has given an excellent study of Regnault's results 
 subject, though I cannot agree with all his conclusions on this subject. 
 In discussing the difference between and 100 he uses a formula of 
 the form 
 
 and deduces from it the erroneous conclusion that the difference is 
 greatest at 50 C., instead of between 40 and 50. His results for 
 T t at 50 are 
 
 Choisy-le-Roi ...... .22 
 
 Verre Ordinaire ...... -J-.25 
 
 VerreVert ....... +.14 
 
 Verre de Suede ...... -f- 56 
 
 and these are probably somewhat nearly correct, except the negative 
 value for Choisy-le-Roi. 
 
 With the exception of Regnault, very few observers have taken up 
 this subject. Among these, however, we may mention Recknagel, 
 who has made the determination for common glass between and 
 100. I have found approximately the constants for my formula in 
 this case, and have calculated the values in the fourth column of the 
 following table. 
 
 * Comptes Rendus, Ixix.
 
 OF ARTS AND SCIENCES. 
 
 89 
 
 TABLE IV. RECKNAGEL'S RESULTS COMPARED WITH THE FORMULA. 
 
 Air 
 Thermometer. 
 
 Mercurial Thermometer. 
 
 Difference. 
 
 Observed. 
 
 Calculated. 
 
 
 
 
 
 
 
 
 
 10 
 
 10.08 
 
 10.08 
 
 
 
 20 
 
 20.14 
 
 20.14 
 
 
 
 30 
 
 30.18 
 
 30.18 
 
 
 
 40 
 
 40.20 
 
 40.20 
 
 
 
 60 
 
 50.20 
 
 50.20 
 
 
 
 60 
 
 60.18 
 
 60.18 
 
 
 
 70 
 
 70.14 
 
 70.15 
 
 +.01 
 
 80 
 
 80.10 
 
 80.11 
 
 +.01 
 
 90 
 
 90.05 
 
 90.06 
 
 +.01 
 
 100 
 
 100.00 
 
 
 
 
 
 b = 290 
 
 a = .000 000 33 
 (100 t)(b t) 
 
 It will be seen that the values of the constants are not very different 
 from those which satisfy Regnault's experiments. 
 
 There seems to be no doubt, from all the experiments we have now 
 discussed, that the point of maximum difference is not at 50, but at 
 some less temperature, as 40 to 45, and this agrees with my own 
 experiments, and a recent statement by Ellis in the Philosophical 
 Magazine. And I think the discussion has proved beyond doubt 
 that the formula is sufficiently accurate to express the difference of 
 the mercurial and air thermometers throughout at least a range of 
 200, and hence is probably very accurate for the range of only 
 100 between and 100. 
 
 Hence it is only necessary to find the constants for my thermom- 
 eters. But before doing this it will be well to see how exact the 
 comparison must be. As the thermometers are to be used in a 
 calorimetric research in which differences of temperature enter, the 
 error of the mercurial compared with the air thermometer will be 
 
 which for the constants used in Recknagel's table becomes 
 
 Error = ^ 1 = .000 000 33 j 29000. 780 1 -f- 3 1* I . 
 
 This amounts to nearly one per cent at 0, and thence decreases to 
 45, after which it increases again. As only 0.2 at the 40 point
 
 90 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 produces this large error at 0, it follows that an error of only 0.02 
 at 40 will produce an error of TT ^ T at 0. At other points the 
 errors will be less. 
 
 Hence extreme care must be taken in the comparison and the most 
 accurate apparatus must be constructed for the purpose. 
 
 2. DESCRIPTION or APPARATUS. 
 
 The Air Thermometer. 
 
 In designing the apparatus, I have have had in view the production 
 of a uniform temperature combined with ease of reading the ther- 
 mometers, which must be totally immersed in the water. The uni- 
 formity, however, needed only to apply to the air thermometer and to 
 the bulbs of the mercurial thermometer, as a slight variation in the 
 temperature of the stems is of no consequence. A uniform tempera- 
 ture for the air thermometer is important, because it must take time 
 for a mass of air to heat up to a given temperature within U.01 or 
 less. 
 
 Fig. 1 gives a section of the apparatus. This consists of a large 
 copper vessel, nickel-plated on the outside, with double walls an inch 
 apart, and made in two parts, so that it could be put together Avater- 
 tight along the line a b. As seen from the dimensions, it required 
 about 28 kilogi'ammes v of water to fill it. Inside of this was the vessel 
 m d efg h k I n, which could be separated along the line c/. In the 
 upper part of this vessel, a piston, q, worked, and could draw the 
 water from the vessel. The top was closed by a loose piece of metal, 
 o p, which fell down and acted as a valve. The bottom of this 
 inner vessel had a false bottom, c /, above which was a row of large 
 holes ; above these was a perforated diaphragm, . The bulb of the 
 air thermometer was at <, with the bulbs of the mercurial thermometers 
 almost touching it. The air thermometer bulb was very much elon- 
 gated, being about 18 cm - long and 3 to 5 cra - in diameter. Although 
 the bulbs of the thermometers were in the inner vessel, the stems 
 were in the outer one, and the reading was accomplished through the 
 thick glass window u v. 
 
 The change of the temperature was effected by means of a Bunsen 
 burner under the vessel w. 
 
 The working of the apparatus was as follows. The temperature 
 having been raised to the required point, the piston q was worked to 
 stir up the water ; this it did by drawing the water through the holes 
 at cl and the perforated diaphragm s, and thence up through the
 
 OP ARTS AND SCIENCES. 
 
 91 
 
 apparatus to return on the outside^ When the whole of the water is 
 at a nearly uniform temperature the stirring is stopped, the valve op 
 falls into place, and the connection of the water in the outer and 
 inner vessels is practically closed as far as currents are concerned, and 
 
 Fig.2. 
 
 before the water inside can cool a little the outer water must have 
 cooled considerably. 
 
 So effective was this arrangement that, although some of the ther- 
 mometers read to 0.007 C., yet they would remain perfectly station- 
 ary for several minutes, even when at 40 C. At very high tempera- 
 tures, such as 80 or 90 C., the burner was kept under the vessel w 
 all the time, and supplied the loss of the outer vessel by radiation. 
 The inner vessel would under these circumstances remain at a very
 
 92 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 constant temperature. The water in the outer vessel never.differed 
 by more than a small fraction of a degree from that in the inner one. 
 
 To get the and 100 points the upper parts of the vessel above 
 the line a b were removed, and ice placed around the bulb of the air 
 thermometer, and left for several hours, until no further lowering took 
 place. For the 100 point the copper vessel shown in Fig. 3 was 
 used. The portion y of this vessel fitted directly over the bulb of the 
 air thermometer. On boiling water in x, the steam passed through 
 the tube to the air thermometer. It is with considerable difficulty 
 that the 100 point is accurately reached, and, unless care be taken, 
 the bulb will be at a slightly lower temperature. Not only must the 
 bulb be in the steam, but the walls of the cavity must also be at 100. 
 To accomplish this in this case, a large mass of cloth was heaped over 
 the instrument, and then the water in x vigorously boiled for an hour 
 or so. After fifteen minutes there was generally no perceptible in- 
 crease of temperature, though an hour was allowed so as to make 
 certain. 
 
 The external appearance of the apparatus is seen in Fig. 2. The 
 method of measuring the pressure was in some respects similar to that 
 used in the air thermometer of Jolly, except that the reading was 
 taken by a cathetometer rather than by a scale on a mirror. .The 
 capillary stem of the air thermometer leaves the water vessel at a, and 
 passes to the tube b, which is joined to the three-way cock c. The 
 lower part of the cock is joined by a rubber tube to another glass tube 
 at d, which can be raised and lowered to any extent, and has also a fine 
 adjustment. These tubes were about 1.5 cm - diameter on the inside, 
 so that there should be little or no error from capillarity. Both tubes 
 were exactly of the same size, and for a similar reason. 
 
 The three-way cock is used to fill the apparatus with dry air, and 
 also to determine the capacity of the tube above a given mark. In 
 filling the bulb, the air was pumped out about twenty times, and 
 allowed to enter through tubes containing chloride of calcium, sul- 
 phuric acid, and caustic soda, so as to absorb the water and the car- 
 bonic acid. 
 
 The Cathetometer. 
 
 The cathetometer was one made by Meyerstein, and was selected 
 because of the form of slide used. The support was round, and the 
 telescope was attached to a sleeve which exactly fitted the support. 
 The greatest error of cathetometers arises from the upright support 
 not being exactly true, so that the telescope will not remain in level
 
 OP ARTS AND SCIENCES. 
 
 93 
 
 at all heights. It is true that the level should be constantly adjusted, 
 but it is also true that an instrument can be made where such an ad- 
 justment is not necessary. And where time is an element in the 
 accuracy, such an instrument should be used. In the present case it 
 was absolutely necessary to read as quickly as possible, so as not to 
 leave time for' the column to change. In the first place the round 
 column, when made, was turned in a lathe to nearly its final dimen- 
 sions. The line joining the centres of the sections must then have 
 been very accurately straight. In the subsequent fitting some slight 
 irregularities must have been introduced, but they could not have been 
 great with good workmanship.* The upright column was fixed, and 
 the telescope moved around it by a sleeve on the other sleeve. Where 
 
 the objects to be measured are not situated at a very wide angle from 
 each other, this is a good arrangement, and has the advantage that any 
 side of the column can be turned toward the object, and so, even if it 
 were crooked, we could yet turn it into such a position as to nearly 
 eliminate error. 
 
 It was used at a distance of about HO"" 1 - from the object, and 
 no difficulty was found after practice in setting it on the column to 
 fo mm - at least. The cross hairs made an angle of 45 with the 
 horizontal, as this was found to be the most sensitive arrangement. 
 
 The scale was carefully calibrated, and the relative errors f for the 
 
 * The change of level along the portion generally used did not amount to 
 more than .1 of a division, or about .Ol mm - at the mercury column, as this is 
 about the smallest quantity which could be observed on the level. 
 
 t These amounted to less than .016 mm - at any part.
 
 94 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 denoted by the subfix. Then approximat 
 
 As the height of the barometer varies only very slightly during an 
 experiment, the value of this expression is very nearly 
 
 which does not depend on the absolute value of the scale divisions 
 
 But the best manner of testing a cathetometer is to take readings 
 upon an accurate scale placed near the mercury columns to b 
 measured. I tried this with my instrument, and found that it agree< 
 with the scale to within two or three one-hundredths of a millimeter, 
 which was as near as I could read on such an object. 
 
 In conclusion, every care was taken to eliminate the errors of this 
 instrument, as the possibility of such errors was constantly present in 
 my mind; and it is supposed that the instrumental errors did not 
 amount to more than one or two one-hundredths of a millimeter on the 
 mercury column. The proof of this will be shown in the results 
 obtained. 
 
 The Barometer. 
 
 This was of the form designed by Fortin, and was made by 
 James Green of New York. The tube was 2.0 cm diameter nearly 
 on the outside, and about 1.7 cm - on the inside. The correction for 
 capillarity is therefore almost inappreciable, especially as, when it 
 remains constant, it is exactly eliminated from the equation. The 
 depression for this diameter is about .08 mm -, but depends upon the 
 height of the meniscus. The height of the meniscus was generally 
 about l.S 1 ""' ; but according as it was a rising or falling meniscus, 
 it varied from 1.4 to 1.2 mm -. These are the practical values of the 
 variation, and would have been greater if the barometer had not been 
 attached to the wall a little loosely, so as to have a slight motion when 
 handled. Also in use the instrument was slightly tapped before read-
 
 OF ARTS AND SCIENCES. 95 
 
 ing. The variation of the height of the meniscus from 1.2 to 1.4 mm - 
 would affect the reading only to the extent of .01 to ,02 mm -. 
 
 The only case where any correction for capillarity is needed is in 
 finding the temperatures of the steam at the 100 point, and will then 
 affect that temperature only to the extent of about 0.005. 
 
 The scale of the instrument was very nearly standard at C., and 
 was on brass. 
 
 At the centre of the brass tube which surrounded the barometer, a 
 thermometer was fixed, the bulb being surrounded by brass, and there- 
 fore indicating the temperature of the brass tube. 
 
 In order that it should also indicate the temperature of the barome- 
 ter, the whole tube and thermometer were wrapped in cloth until a 
 thickness of about 5 or 6 cm ' was laid over the tube, a portion being 
 displaced to read the thermometers. This wrapping of the barometer 
 was very important, and only poor results were obtained before its 
 use; and this'is seen from the fact that 1 on the thermometer indi- 
 cates a correction of ,12 mm- on the barometer, and hence makes a 
 difference of 0.04 on the air thermometer. 
 
 As this is one of the most important sources of error, I have now 
 devised means of almost entirely eliminating it, and making continual 
 reading of the barometer unnecessary. This I intend doing by an 
 artificial atmosphere, consisting of a large vessel of air in ice, and 
 attached to the open tube of the manometer of the air thermometer. 
 
 The Thermometers. 
 
 The standard thermometers used in my experiments are given in 
 the following table.
 
 96 
 
 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 
 fl 
 
 >> 
 ffj 
 
 1 -flftrfi 
 
 Owner or Lender. 
 
 ( Physical Labora 
 ( Johns Hopkins Uni 
 
 Prof. Barkei 
 Univ. of Pennsy 
 Chemical Labor 
 Johns Hopkins Un 
 
 n ^o sS ^-s if 
 
 I -siSilill 
 
 ^ >> "^ 
 
 ! 
 
 I 
 
 co 2 : 
 
 1_ 
 
 
 i 
 
 CO 
 
 "3 
 
 CC & Q 
 
 g 
 
 a 
 
 J ^ 
 
 xj JS .S 
 
 a 
 
 ^ = - 
 
 5 .2 
 
 1 5 1 ' 1 : 
 
 
 
 
 ^ 
 
 S o 
 
 4*. 
 
 
 0> CO CO 
 
 CN ^-i CD (N O5 OS 
 
 
 
 
 
 T 
 
 
 
 
 uation. 
 
 | 
 
 P^ 
 
 |-i . ri d b b 
 
 2 
 
 3 
 
 a, b 
 
 "2 "3 ^ C3 '0 
 
 
 
 
 ^ 3 
 
 
 
 
 CT 1 
 
 
 8 1 1 
 
 co o o 
 
 Tli^ 2 s i 
 
 
 H 
 
 rH i-H 
 
 i-H I-H C^ rH I-H i i 
 
 8> 
 
 
 
 > ^ 2 + 
 
 
 
 
 O O O 
 CN rH . CO 
 
 ^0*^ S= 2s o 
 
 
 1 1 
 
 1 \ \ 
 
 ^_ T I 
 
 ld 
 
 
 
 ~-*~* 
 
 a-| 
 
 
 
 49 
 
 l|| 
 
 
 
 
 58 58 J3 5S 
 
 
 5 CO CO 
 
 'o S 
 
 CO CD O CO CO M* 
 
 1 
 
 S 3 o 
 
 W 
 
 CO CO CSJ CO CO CO 
 
 00 g ^ ^
 
 OF ARTS AND SCIENCES. 97 
 
 The calibration of the first four thermometers has been described. 
 The calibration of the Kew standard was almost perfect, and no cor- 
 rection was thought necessary. The scale divided on the tube was to 
 half-degrees Fahrenheit; but as the 32 and 212 points were not 
 correct, it was in practice used as a thermometer with arbitrary 
 divisions. The interval between the and 100 points, as Welsh 
 found it, was 180.12, using barometer at 30 inches, or 180.05 as cor- 
 rected to 760 mm - of mercury.* At the present time it is 179.68,f 
 showing a change of 1 part in 486 in twenty-five years. This fact 
 shows that the ordinary method of correcting for change of zero is not 
 correct, and that the coefficient of expansion of glass changes with 
 time4 
 
 I have not been able to find any reference to the kind of glass used 
 in this thermometer. But in a report by Mr. Welsh we find a com- 
 parison, made on March 19, 1852, of some of his thermometers with 
 two other thermometers, one by Pastre", examined and approved by 
 Regnault, and the other by Troughton and Simms. The thermometer 
 which I used was made a little more than a year after this ; and it is 
 reasonable to suppose that the glass was from the same source as the 
 standards Nos. 4 and 14 there used. We also know that Regnault 
 was consulted as to the methods, and that the apparatus for calibration 
 was obtained under his direction. 
 
 I reproduce the table here with some alterations, the principal one 
 of which is the correction of the Troughton and Simms thermometers, 
 so as to read correctly at 32 and 212, the calibration being assumed 
 correct, but the divisions arbitrary. 
 
 * Boiling point, Welsh, Aug. 17, 1853, 212.17 ; barometer 30* . 
 Freezing point, " " " 82.05. 
 
 Boiling point, Rowland, June 22, 1878, 2 12.46; barometer 760 mm .. 
 Freezing point, " " " 32.78. 
 
 The freezing point was taken before the boiling point in either case, 
 t 179.70, as determined again in January, 1879. 
 
 t The increase shown here is 1 in 80 nearly ! It is evidently connected with 
 the change of zero; for when glass has been heated to 100, the mean coefficient 
 of expansion between and 100 often changes as much, as 1 in 50. Hence it 
 is not strange that it should change 1 in 80 in twenty-five years. I believe this 
 fact has been noticed in the case of standards of length. 
 
 VOL. xv. (N. s. vn.)
 
 98 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 TABLE VI. COMPARISON BY WELSH, 1852. 
 
 Mean of 
 Kew Standards 
 Nos. 4 and 14. 
 
 Fastre231, 
 Kegnault. 
 
 A 
 
 Kew. 
 
 Troughton and Simms 
 (Royal Society). 
 
 A 
 
 Kew. 
 
 32.00 
 
 32.00 
 
 
 
 32.00 
 
 
 
 38!? 1 
 45.04 
 
 38.72 
 45.03 
 
 +.01 
 -.01 
 
 38.70 
 45.03 
 
 .01 
 .01 
 
 49.96 
 
 49.96 
 
 .00 
 
 49.96 
 
 .00 
 
 55.34 
 
 55.37 
 
 +.03 
 
 55.34 
 
 .00 
 
 60.07 
 
 60.05 
 
 .02 
 
 60.06 
 
 .01 
 
 65.39 
 
 65.41 
 
 +.02 
 
 65.36 
 
 .03 
 
 69.93 
 
 69.95 
 
 +.02 
 
 69.93 
 
 .00 
 
 74.69 
 
 74.69 
 
 .00 
 
 74.72 
 
 +.03 
 
 80.05 
 
 80.06 
 
 +.01 
 
 80.14 
 
 +.09 
 
 85.30 
 
 85.33 
 
 +.03 
 
 85.44 
 
 +.14 
 
 90.60 
 
 90.51 
 
 +.01 
 
 90.56 
 
 +.06 
 
 95.26 
 101.77 
 
 95.24 
 101.77 
 
 .02 
 .00 
 
 95.40 
 101.94 
 
 4-.14 
 
 +.15 
 
 . 109.16 
 
 109.15 
 
 .01 
 
 109.25 
 
 +.08 
 
 212.00 
 
 212.00 
 
 .00 
 
 212.00 
 
 .00 
 
 It is seeruhat the Kew standards and the Fastre agree perfectly, but 
 that the Troughton and Simms standard stands above the Kew ther- 
 mometers at 100 F. 
 
 The Geissler standard was made by Geissler of Bonn, and its scale 
 was on a piece of milk glass, enclosed in a tube with the stem. The 
 calibration was fair, the greatest error being about 0.015 C., at 
 50 C. ; but no correction for calibration was made, as the instrument 
 was only used as a check for the other thermometers. 
 
 3. RESULTS OF COMPARISON. 
 Calculation of Air Thermometer. 
 
 This has already been described, and it only remains to discuss the 
 formula and constants, and the accuracy with which the different 
 quantities must be known. 
 
 The well-known formula for the air thermometer is 
 
 H-h + 
 
 T= 
 
 l + at> _ 
 
 Solving with reference to T, and placing in a more convenient form, 
 we have 
 
 H-h' 
 
 nearly,
 
 OF ARTS AND SCIENCES. 
 
 99 
 
 where 
 
 and 
 
 For the first bulb, 
 
 For the second bulb, 
 
 = .0058. 
 
 To discuss the error of T due to errors in the constants, we must 
 replace a by its experimental value, seeing that it was determined 
 with the same apparatus as that by which T was found. As it does 
 not change very much, we may write approximately 
 
 T= 100 
 
 H h 
 
 V l + yt 
 
 From this formula we can obtain by differentiation the error in 
 each of the quantities, which would make an error of one tenth of 
 one per cent in T. The values are for T = 40 nearly ; t = 20 ; 
 ff lw h = 270""" ; and h = 750 mm -. If x is the variable, 
 
 A x = 
 
 dx 
 dT 
 
 ' *JL _L_ - ru d JL 
 d T looo - u * d T 
 
 TABLE VII. ERRORS PRODUCING AN ERROR IN T OF 1 IN 1000 AT 40 C. 
 
 
 H 
 
 H 100 orA. 
 
 V 
 
 T 
 
 ha 
 
 a 
 
 - constant, 
 a 
 
 b 
 a 
 ^5 constant. 
 
 1,90 
 k^/con*. 
 
 b ino -b 
 a 
 
 -^ constant. 
 
 Absolute 
 ralue, 
 Ax 
 
 n mm. 
 
 27 mm. 
 
 .005 
 
 .00074 
 
 .00087 
 
 .0047 
 
 .00087 
 
 Relative 
 value, 
 Az 
 
 X 
 
 ... 
 
 ... 
 
 0.9 
 
 .10 
 
 .12 
 
 .62 
 
 ... 
 
 From this table it would seem that there should be no difficulty in 
 determining the 40 point on the air thermometer to at least 1 in 2000 ; 
 and experience has justified this result. The principal difficulty is in 
 the determination of H, seeing that this includes errors in reading the 
 barometer as well as the cathetometer. For this reason, as men- 
 tioned before, I have designed another instrument for future use, in 
 which the barometer is nearly dispensed with by use of an artificial 
 atmosphere of constant pressure. 
 
 The value of ^ does not seem to affect the result to any great 
 extent; and if it was omitted altogether, the error would be only
 
 100 
 
 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 about 1 in 1,000, assuming that the temperature t was the same at the 
 determination of the zero point, the 40 point, and the 100 point. 
 It seldom varied much. 
 
 The coefficient of expansion of the glass influences the result very 
 slightly, especially if we know the difference of the mean coefficients 
 between and 100, and say 10 and -f 10. This difference I at 
 first determined from Regnault's tables, but afterwards made a deter- 
 mination of it, and have applied the correction.* 
 
 The table given by Regnault is for one specimen of glass only ; and 
 I sought to better it by taking the expansion at 100 from the mean 
 of the five specimens given by Regnault on p. 231 of the first volume 
 of his Relation des Experiences, and reducing ' the numbers on 
 page 237 in the same proportion. I thus found the values given in 
 the second column of the following table. 
 
 TABLE VIII. COEFFICIENT OF EXPANSION OF THE GLASS OF THE AIR 
 THERMOMETER, ACCORDING TO THE AIR THERMOMETER. 
 
 Tempera- 
 ture ac- 
 cording to 
 Air Ther- 
 mometer. 
 
 Values of 6 
 used for a first 
 Calculation 
 
 b from 
 Regnault's 
 Table, 
 Glass No. 5. 
 
 Experimental Results. 
 
 Apparent 
 Coefficient of 
 Expansion of 
 Mercury. 
 
 b, using 
 Regnault's 
 Value for 
 Mercury .t 
 
 6, using 
 Recknagel'g 
 Value for 
 Mercury 4 
 
 6, using 
 Wullnel's 
 Value for 
 Mercury . 
 
 
 20 
 40 
 60 
 
 80 
 100 
 
 .0000252 
 .0000253 
 .0000256 
 .0000259 
 .0000262 
 .0000264 
 
 .0000263 
 .0000264 
 .0000267 
 .0000270 
 .0000273 
 .0000276 
 
 
 
 
 
 .00015410 
 
 .00015395 
 .00015391 
 
 .00015381 
 
 .0000254 
 .0000258 
 .0000261 
 
 .0000277 
 
 .0000264 
 .0000266 
 .0000267 
 
 .0000277 
 
 .0000273 
 .0000276 
 .0000278 
 
 .0000287 
 
 The second column contains the values which I have used, and one 
 of the last three columns contains my experimental results, the last 
 being probably the best. The errors by the use of the second column 
 compared with the last are as follows : 
 
 TTT ^ from using b m 6 40 = .0000008 instead of .0000011 ; 
 TuW from using b m = .0000264 instead of .0000287 ; 
 
 or, ffa-Q for both together. 
 
 * This was determined by means of a large weight thermometer in which 
 the mercury had been carefully boiled. The glass was from the same tube as 
 that of the air thermometer, and they were cut from it within a few inches of 
 each other. 
 
 t Relation des Experiences, i. 328. 
 
 t Pogg. Ann., cxxiii. 135. Experimental Physik, Wullner, i. 67.
 
 OP ARTS AND SCIENCES. 
 
 101 
 
 As the error is so small, I have not thought it worth while to entirely 
 recalculate the tables, but have calculated a table of corrections as 
 follows, and have so corrected them : 
 
 TABLE IX. TABLE OP CORRECTIONS. 
 
 Tf 
 
 T 
 
 
 Calculated 
 
 Corrected 
 
 Correction. 
 
 Temperature. 
 
 Temperature. 
 
 
 
 
 8 
 
 
 
 10 
 
 9.9971 
 
 .0029 
 
 20 
 
 19.9946 
 
 .0054 
 
 30 
 
 29.9924 
 
 0076 
 
 40 
 
 39.9907 
 
 .0093 
 
 50 
 
 49.9894 
 
 .0106 
 
 60 
 
 69.9865 
 
 .0135 
 
 80 
 
 79.9880 
 
 .0120 
 
 100 
 
 100. 
 
 
 
 T = T' {1 + 373 (b' m - b m ) - (273 + T) (V - &)}. 
 T T 1 {1 .000858 -f (273 -f T'} (b b')}. 
 
 T = .99975 T' approximately between and 40. This last is 
 true within less than T ^ 7 of a degree. 
 
 The two bulbs of the air thermometer used were from the same 
 piece of glass tubing, and consequently had nearly, if not quite, the 
 same coefficient of expansion. 
 
 In the reduction of the barometer and other mercurial columns to 
 zero, the coefficient .000162 was used, seeing that all the scales were 
 of brass. 
 
 In the tables the readings of the thermometers are reduced to 
 volumes of the tube from the tables of calibration, and they are cor- 
 rected for the pressure of water, which increased their reading, 
 except at 0, by about 0.01 C. 
 
 The order of the readings was as follows in each observation : 
 1st, barometer; 2d, cathetometer; 3d, thermometers forward and 
 backward ; 4th, cathetometer ; 5th, barometer, &c., repeating the 
 same once or twice at each temperature. In the later observations, 
 two series like the above were taken, and the water stirred between 
 them. 
 
 The following results were obtained at various times for the value 
 of a with the first bulb :
 
 102 
 
 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 Mean 
 
 .0036664 
 .0036670 
 .0036658 
 .0036664 
 .0036676 
 
 .00366664 
 
 obtained by using the coefficient of expansion of glass .0000264 at 
 100 or a = 0036698, using the coefficient .0000287. 
 
 The thermometers Nos. 6163, 6165, 6166, were always taken out 
 of the bath when the temperature of 40 was reached, except on 
 November 14, when they remained in throughout the whole experi- 
 
 The thermometer readings are reduced to volumes by the tables of 
 calibration. 
 
 TABLE X. IST SERIES, Nov. 14, 1877. 
 
 Relative 
 Weight. 
 
 Air 
 Thermometer. 
 
 V 
 6163. 
 
 V 
 6166. 
 
 7 
 6167. 
 
 Temperature 
 by 6167. 
 
 A 
 
 4 
 
 o 
 
 115.33 
 
 21.25 
 
 6.147 
 
 
 
 
 
 4 
 
 17.1425 
 
 422.84 
 
 255.80 
 
 15.685 
 
 17.661 
 
 .236 
 
 4 
 
 23.793 
 
 634.71 
 
 341.05 
 
 19.157 
 
 24.089 
 
 .296 
 
 5 
 
 30.582 
 
 653.49 
 
 431.71 
 
 22.833 
 
 80.896 
 
 .814 
 
 2 
 
 38.569 
 
 793.18 
 
 
 27.175 
 
 38.935 
 
 .366 
 
 2 
 
 61.040 
 
 
 
 33.864 
 
 61.320 
 
 .280 
 
 4 
 
 59.137 
 
 
 
 38.256 
 
 69.452 
 
 .816 
 
 The first four series, Tables X. to XIII., were made with one bulb 
 to the air thermometer. A new bulb was now made, whose capacity 
 was 192.0 ccm -, that of the old being 201.98 c - cm -. The value of -p for 
 the new bulb was .0058. The values of h' and a were obtained as 
 follows : 
 
 
 a 
 
 hf 
 
 June 8th 
 
 .00366790 
 
 753.876 
 
 June 22d 
 
 .00366977 
 
 753.805 
 
 June 25th 
 
 .00366779 
 
 753.837 
 
 Mean 
 
 .0036685 
 
 753.84 
 
 This value of a is calculated with the old coefficient for glass, 
 new would have given .0036717. 
 
 The
 
 OF ARTS AND SCIENCES. 
 
 103 
 
 W 
 
 53 5 gg 8 
 
 Jf 
 
 fc-|7 '"*'*'*^-* 1 <1 ;i *'*'*T>50Ou3o 
 
 Jih 
 
 ooioo 
 
 s^ i as ^ s s ^ s
 
 104 ' PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 fi- 
 
 J 
 
 ^ 
 
 iii 
 
 s s s s 
 
 T* t-- t 1 * CO 
 OS 00 -H rH 
 
 g 35 3 
 
 f S S5 
 
 a? s 
 
 CO CO ^tl 
 
 00 g O W N 
 
 c$ t- as i-i 
 
 S 8 
 
 
 
 oo ec oj 
 to to to 
 
 S S
 
 OF ARTS AND SCIENCES. 
 
 105 
 
 
 *> it- 
 
 Difference 
 Column 
 reduced 
 
 
 
 ~ 8 S 
 
 i 
 
 s s s 
 
 10 cd ca t^ 
 
 S S 
 
 & & 
 
 If 5 
 
 S 2 -N 
 
 S S 
 
 3 g 
 
 8 S 
 
 5! c 
 
 *Q CO CC 
 
 ^ CO l~ t^ O 
 
 00 O r-! 
 
 2 3 $ 
 
 co rh co 
 S S 58 
 
 !o lo S 
 
 11 
 
 ft 
 
 !J 
 
 S at 
 11 
 
 P CO 
 
 
 cc cc -^ 10
 
 106 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 5 q q 
 i + I 
 
 o o 
 
 5 
 
 
 So o co 3" e 
 S 3 TP eo to 
 
 & 
 
 S 
 
 S 5 
 
 ll 
 
 S S 
 
 r- O 
 
 CO t^- 
 
 Nil- 
 Ili 
 
 ll 2 
 
 8 S I .8 I 
 -f S ^ S S? 
 
 x 
 
 8 9 S 
 
 2 5! S 
 
 S S 8 8 
 
 o> oo 10 -j( 
 
 O CD 
 
 5 
 
 *-l (N (N (M
 
 OP ARTS AND SCIENCES. 
 
 107 
 
 + + + + + + I" I 
 
 t-< (N t-. 
 
 ^ S So 
 
 ? S S 
 
 co Oi I-* I-H o i>- < 100 
 
 T*t-COiOiOOOcO 
 
 S S ^ S S g 
 
 S I S 88 
 
 I S 3 I S g s 
 
 ji 
 
 fe o' 
 
 i 
 
 |U 
 
 I 
 
 J_I_ 
 
 f: S S 
 5 S <p 
 
 1 S; 
 l-sl 
 
 Jt^ O ?O ^O CO CD ^? 
 
 S5 Sg S 9 ^ ri 
 
 I S I 1 I 1
 
 108 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 It now remains to determine from these experiments the most 
 probable values of the constants in the formula, comparing the air 
 with the mercurial thermometer. The formula is, as we have i 
 
 t=T-at(t'-t) (6-0: 
 but I have generally used it in the following form : 
 
 t=CV-t - m (100-0 (1-n (100 + 0), 
 t = C' V t' - m t (40 - n ( 40 + '))' 
 And the following relations hold among the constants : 
 _ Q, (i + m (60 8400 n)), nearly, 
 a = mn, 
 I = JL 100, 
 T GVt M 
 
 In these formulae t is the temperature on the air thermometer ; Vis 
 the volume of the stem of the mercurial thermometer, as determined 
 from the calibration and measured from any arbitrary point; and 
 C", f' , m, and n are constants to be determined. 
 
 The best way of finding these is by the method of least squares. 
 C" must be found very exactly ; t a is only to be eliminated from the 
 equations ; m must be found within say ten per cent, and n need only 
 be determined roughly. To find them only within these limits is a 
 very difficult matter. 
 
 * Determination of n. 
 
 As this constant needs a wide range of temperatures to produce 
 much effect, it can only be determined from thermometer No. 6167, 
 which was of the same glass as 61 63, 61 65, and 6166. It is unfortunate 
 that it was broken on November 2 1 , and so we only have the experi- 
 ments of the first and second series. From these I have found 
 n = .003 nearly. This makes b = 233*, which is not very far from 
 the values found before from experiments above 100 by Regnault 
 on ordinary glass.* 
 
 * Some experiments with Baudin thermometers at high temperatures have 
 given me about 240, a remarkable agreement, as the point must be uncertain 
 to 10 or more.
 
 OP ARTS AND SCIENCES. 109 
 
 Determination of C and m. 
 
 I shall first discuss the determination of these for thermometers 
 Nos. 6163, 6165, and 6166, as these were the principal ones used. 
 
 As No. 61 63 extended from to 40, and the others only from 
 to 30, it was thought best to determine the constants for this one 
 first, and then find those for 6165 and 6166 by comparison. As this 
 comparison is deduced from the same experiments as those from which 
 we determine the constants of 6163, very nearly the same result is 
 found as if we obtained the constants directly by comparison with the 
 air thermometer. 
 
 The constants of 6163 can be found either by comparison with 
 6167, or by direct comparison with the air thermometer. I shall first 
 determine the constants for No. 6167. 
 
 The constants C and t for this thermometer were found directly 
 by observation of the and 100 points; and we might assume these, 
 and so seek only for m. In other words, we might seek only to 
 express the difference of the thermometers from the air thermometer . 
 by a formula. But this is evidently incorrect, seeing that we thus 
 give an infinite weight to the observations at the and 100 points. 
 The true way is obviously to form an equation for each temperature, 
 giving each its proper weight. Thus from the first series we find for 
 No. 6167, 
 
 Weight. Equations of condition . 
 
 4 == 6.147 C 1 , 
 
 4 17.427 = 15.685 C t 930 m, 
 
 4 23.793 = 19.157 0t 1140 m, 
 &c. &c. &c. 
 
 5 100 = 60.156 C t , 
 
 which can be solved by the method of least squares. As t is un- 
 important, we simply eliminate it from the equations. I have thus 
 found, 
 
 Weight. 
 
 1 Nov. 14 (7=1.85171 m = .000217 
 
 2 Nov. 20, 21 G= 1.85127 m = .000172 
 
 Mean 0= 1.85142 m = .000187 
 
 The difference in the values of m is due to the observations not being 
 so good as were afterwards obtained. However, the difference only 
 signifies about 0.03 difference from the mean at the 50 point. After 
 November 20 the errors are seldom half of this, on account of the 
 greater experience gained in observation.
 
 110 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 The ratio of C for 6167 and 6163 is found in the same way. 
 
 Weight. 
 
 1 Nov. 14 .0310091 
 
 2 Nov. 20 .0309846 
 
 Mean .0309928 
 
 Hence for 6163 we have in this way 
 
 G = .057381 C' = .056995 m = .000187. 
 
 By direct comparison of No. 6163 with the air thermometer, we 
 find the following. 
 
 Date. Weight <X. m. 
 
 Nov. 14 1 .056920 .000239 
 
 Nov. 20 2 .056985 .000166 
 
 Jan. 25 3 .056986 .000226 
 
 Feb. 11 4 .056997 .000155 
 
 June 8 3 .056961 .000071 
 
 June 22 2 .056959 .000115 
 
 Mean .056976 .000004 .000154 .000010 
 
 The values of C' agree with each other with great exactness, and 
 the probable error is only O.003 C. at the 40 point. 
 
 The great differences in the values of m, when we estimate exactly 
 what they mean in degrees, also show great exactness in the experi- 
 ments. The mean value of m indicates a difference of only 0.05 
 between the mercurial and air thermometer at the 20 point, the 
 and 40 points coinciding. The probable error of m in degrees is 
 only 0.003 C. 
 
 There is one more method of finding m from these experiments ; 
 and that is by comparing the values of G' with No. 6167, the glass of 
 6167 being supposed to be the same as that of 6163. 
 
 We have the formula 
 
 C= G> (1 +34.8 m). 
 
 Hence m = ~^ . 
 
 We thus obtain the following results : 
 
 Date. Weight. Value of m. 
 
 Nov. 14 1 .000236 
 
 Nov. 20 2 .000218 
 
 Jan. 25 3 .000217 
 
 Feb. 11 4 .000197 
 
 June 8 3 .000215 
 
 June 22 2 .000216 
 
 Mean .000213
 
 OF ARTS AND SCIENCES. Ill 
 
 The results for m are then as follows : 
 
 From direct comparison of No. 6167 with the air thermometer .000187 
 From " " " No. 6163 " " .000154 
 
 From comparison of No. 6163 with No. 6167 .000213 
 
 The first and last are undoubtedly the most exact numerically, but 
 they apply to No. 6167, and are also, especially the first, derived 
 from somewhat higher temperatures than the 20 point, where the 
 correction is the most important. The value of m, as determined in 
 either of these ways, depends upon the determination of a difference of 
 temperature amounting to 0.30, and hence should be quite exact. 
 
 The value of m, as obtained from the direct comparison of No. 6163 
 with the air thermometer, depends upon the determination of a differ- 
 ence of about 0.05 between the mercurial and the air thermometer. 
 At the same time, the comparison is direct, the temperatures are the 
 same as we wish to use, and the glass is the same. I have combined 
 the results as follows : 
 
 m from No. 6167 .000200 
 
 m " 6163 .000154 
 
 Mean .00018* 
 
 It now remains to deduce from the tables the ratios of the constants 
 for the different thermometers. 
 
 The proper method of forming the equations of condition are as 
 follows, applying the method to the first series: 
 
 Weight. 
 
 4 21.25 C tll = 115.33 C, v 
 
 4 255.80 C UI = 422.84 C, v 
 
 4 341.05 C in = 534.71 G t v 
 
 5 431.71 C UI = 653.49 C, v 
 
 where G UI is the constant for No. 6166, C, is that for No. 6163, 
 and v is a constant to be eliminated. Dividing by (7/, the equations 
 
 S~i 
 
 can be solved for --. The following table gives the results. 
 
 * See Appendix to Thermometry, where it is finally thought best to reject 
 the value from No. 6167 altogether.
 
 112 
 
 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 TABLE XVI. RATIOS OF CONSTANTS. 
 
 
 
 6163 
 
 6166 
 
 6166 
 
 6165 
 
 6165 
 
 Date. 
 
 Weight. 
 
 6167 
 
 6167 
 
 6163 
 
 6163 
 
 6166 
 
 Nov. 14 
 
 1 
 
 .031009 
 
 .040058 
 
 1.3111 
 
 
 
 Nov. 20 
 
 2 
 
 .030985 
 
 .040670 
 
 1.3128 
 
 
 
 Jan. 25 
 
 3 
 
 .... 
 
 .... 
 
 1.3122 
 
 
 
 
 Feb. 11 
 
 4 
 
 
 
 1.3115 
 
 8.6588 
 
 6.1449 
 
 June 8 
 
 3 
 
 
 .... 
 
 1.3108 
 
 8.0605 
 
 6.1469 
 
 June 22 
 
 2 
 
 
 
 1.3122 
 
 8.0588 
 
 6.1428 
 
 Mean j 
 
 .030993 
 '.00005 
 
 .040666 
 .000003 
 
 1.31175 
 .0004 
 
 8.0594 
 .0002 
 
 6.1451 
 .0004 
 
 From these we have the following, as the final most probable 
 results : 
 
 ' . C n =8.0601 G t , 
 
 . C UI = 1.31175 (7,, 
 C, = .031003 (7 iT , 
 C,, .24991 C iv , 
 C UI = .040661 <7 iv , 
 
 of which the last three are only used to calculate the temperatures on 
 the mercurial thermometer, and hence are of little importance in the 
 remainder of this paper. 
 
 The value of C" which we have found for the old value of the 
 coefficient of expansion of glass was 
 
 6^ = . 056976; 
 
 and hence, corrected to the new coefficient, it is, as I have shown, 
 C, =.056962. 
 
 Hence, 
 
 <7 ;/ = .45912, 
 <?, = .074720. 
 
 And we have finally the three following equations to reduce the ther- 
 mometers to temperatures on the air thermometer: 
 
 Thermometer No. 6163: 
 T=. 056962 V-t '-.miST(4Q-T) (l - .003 (7*+ 40)). 
 
 Thermometer No. 6165 : 
 ^=.45912 V"- C __. 00018 T(T- 40) (l _ .003 (7+40)).
 
 OF ARTS AND SCIENCES. 113 
 
 Thermometer No. 6166 : 
 
 = . 074720 r'"*,/" .00018 T(T 40) (l . 
 
 where V, V", and V" are the volumes of the tube obtained by 
 calibration ; t ', t ", and t '" are constants depending on the zero point, 
 and of little importance where a difference of temperature is to be 
 measured ; and T is the temperature on the air thermometer. 
 
 On the mercurial thermometer, using the and 100 points as fixed, 
 we have the following by comparison with No. 6167 : 
 
 Thermometer No. 6163 ; t = .057400 V t ; 
 Thermometer No. 6165 ; t = .46265 V t Q ; 
 Thermometer No. 6166 ; t = .4)75281 V t v 
 
 The Kew Standard. 
 
 The Kew standard must be treated separately from the above, as 
 the glass is not the same. This thermometer has been treated as if 
 its scale was arbitrary. 
 
 In order to have variety, I have merely plotted all the results 
 with this thermometer, including those given in the Appendix, and 
 drawn a curve through them. Owing to the thermometer being- only 
 divided to F., the readings could not be taken with great accuracy, 
 and so the results are not very accordant ; but I have done the best I 
 could, and the result probably represents the correction to at least 
 0.02 or 0.03 at every point. 
 
 ('/) Reduction to the Absolute Scale. 
 
 The correction to the air thermometer to reduce to the absolute 
 scale has been given by Joule and Thomson, in the Philosophical 
 Transactions for 1854 ; but as the formula there used is not correct, 
 I have recalculated a table from the new formula used by them in 
 their paper of 1862. 
 
 That equation, which originated with Rankine, can be placed in 
 the form &i/Vl~ ' 
 
 where p, v, and p. are the pressure, volume, and absolute temperature 
 of a given weight of the air ; D is its density referred to air at C. 
 and 760 mm - pressure ; fj, is the absolute temperature of the freezing 
 point; and m is a constant which for air is 0.33 C. 
 VOL. xv. (N. s. vn.) 8
 
 14 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 For the air thermometer with constant volume 
 T = 100 pt ~ Po -, 
 
 PlGO PO 
 
 or, since D = 1, 
 
 ,,- ^ = T- . 00088 T =, 
 
 from which I have calculated the following table of corrections. 
 TABLE XVII. REDUCTION OF AIR THERMOMETER TO ABSOLUTE SCALE. 
 
 T 
 
 Air Thermometer. 
 
 A* Mo 
 Absolute Temperature. 
 
 A 
 or Correction to 
 Air Thermometer. 
 
 
 
 
 
 
 
 10 
 
 9.9972 
 
 .0028 
 
 20 
 
 19.9952 
 
 .0048 
 
 30 
 
 29.9939 
 
 .0061 
 
 40 
 
 39.9933 
 
 .0067 
 
 50 
 
 49.9932 
 
 .0068 
 
 60 
 
 . 59.9937 
 
 .0063 
 
 70 
 
 69.9946 
 
 .0054 
 
 80 
 
 79.9956 
 
 .0044 
 
 90 
 
 89.9978 
 
 .0022 
 
 100 
 
 100.000 
 
 
 
 200 
 
 200.037 
 
 +.037 
 
 300 
 
 300.092 
 
 +.092 
 
 400 
 
 400.157 
 
 +.157 
 
 500 
 
 500.228 
 
 +.228 
 
 It is a curious circumstance, that the point of maximum difference 
 occurs at about the same point as in the comparison of the mercurial 
 and air thermometers. 
 
 From the previous formula, and from this table of corrections, the 
 following tables were constructed. 
 
 
 
 ^ 7 '
 
 OF ARTS AND SCIENCES. 
 
 115 
 
 TABLE XVIII. THERMOMETKB No. 6163. 
 
 = Q 
 11 
 
 i 
 
 li^ 
 
 111 
 
 H 
 
 i! 
 
 II 
 
 If! 
 
 ill! 
 
 ll 
 
 iS 
 
 1 Reading in 
 meters on 
 
 Temperature 
 
 curiiil Them 
 and 100 
 
 Temperature 
 curial Them 
 and 40 fi 
 Air Thermc 
 
 1 
 
 Temperate 
 Absolute Set 
 0C 
 
 1 Reading in 
 meters on 
 
 Temperature 
 curial Them 
 and 100 
 
 Temperature 
 curial Them 
 and 40 fi 
 Air Thermc 
 
 II 
 
 Temperatu 
 Absolute Sea 
 0C. 
 
 50 
 
 .923 
 
 .917 
 
 -.911 
 
 .911 
 
 240 
 
 20.557 
 
 20.409 
 
 20.350 
 
 20.345 
 
 58.1 
 
 
 
 
 
 
 
 
 
 250 
 
 21.670 
 
 21.515 
 
 21.457 
 
 21.452 
 
 60 
 
 +.217 
 
 +.215 
 
 +.214 
 
 +.214 
 
 260 
 
 22.776 
 
 22.616 
 
 22.559 
 
 22.564 
 
 70 
 
 1.356 
 
 1.336 
 
 1.328 
 
 1.328 
 
 270 
 
 23.884 
 
 23.713 
 
 23.657 
 
 23.652 
 
 80 
 
 2.494 
 
 2.475 
 
 2.461 
 
 2.460 
 
 280 
 
 24.989 
 
 24.810 
 
 24.755 
 
 24.750 
 
 90 
 
 3.631 
 
 3.604 
 
 3.584 
 
 3.583 
 
 290 
 
 26.093 
 
 25.907 
 
 25.854 
 
 25.848 
 
 100 
 
 4.767 
 
 4.733 
 
 4.707 
 
 4.706 
 
 300 
 
 27.200 
 
 27.006 
 
 26.956 
 
 26.950 
 
 110 
 
 5.903 
 
 5.860 
 
 5.829 
 
 5.827 
 
 310 
 
 28.311 
 
 28.108 
 
 28.060 
 
 28.056 
 
 120 
 
 7.036 
 
 6.986 
 
 6.950 
 
 6.948 
 
 320 
 
 29.425 
 
 29.214 
 
 29.169 
 
 29.163 
 
 130 
 
 8.170 
 
 8.111 
 
 8.071 
 
 8.069 
 
 330 
 
 30.541 
 
 30.324 
 
 30.282 
 
 30.276 
 
 140 
 
 9.304 
 
 9.237 
 
 9.193 
 
 9.190 
 
 340 
 
 31.662 
 
 31.436 
 
 31.398 
 
 31.392 
 
 150 
 
 10.436 
 
 10.361 
 
 10.314 
 
 10.311 
 
 350 
 
 32.782 
 
 32.548 
 
 32.514 
 
 32.508 
 
 160 
 
 11.568 
 
 11.485 
 
 11.435 
 
 11.432 
 
 360 
 
 33.903 
 
 33.660 
 
 33.630 
 
 33.624 
 
 170 
 
 12.700 
 
 12.608 
 
 12.556 
 
 12.553 
 
 370 
 
 35.023 
 
 34.773 
 
 34.748 
 
 34.742 
 
 180 
 
 13.829 
 
 13.730 
 
 13.676 
 
 13.672 
 
 380 
 
 36.143 
 
 35.884 
 
 35.864 
 
 35.857 
 
 190 
 
 14.957 
 
 14.850 
 
 14.794 
 
 14.790 
 
 390 
 
 37.261 
 
 36.994 
 
 36.979 
 
 36.972 
 
 200 
 
 16.081 
 
 15.906 
 
 15.909 
 
 15.905 
 
 400 
 
 38.377 
 
 38.103 
 
 38.094 
 
 38.087 
 
 210 
 
 17.203 
 
 17.080 
 
 17.022 
 
 17.018 
 
 410 
 
 39.492 
 
 39.210 
 
 39.206 
 
 39.199 
 
 220 
 
 18322 
 
 18.191 
 
 18.132 
 
 18.127 
 
 420 
 
 40.604 
 
 40.314 
 
 40.316 
 
 40.309 
 
 230 
 
 19.440 
 
 19.301 
 
 19.242 
 
 19.237 
 
 
 
 
 
 
 TABLE XIX. THERMOMETER No. 6165. 
 
 s 
 1 
 
 !fi 
 
 lilt 
 
 | 
 
 si 
 
 B-3 
 
 21 
 
 ifj 
 
 11! 
 
 | 
 
 si 
 
 EJ 
 
 .S o 
 
 sPa, 
 
 5-S2 
 
 2 
 
 SSd 
 
 a - 
 
 So 
 
 g Q tea 2 
 
 3 
 
 3 d 
 
 ii 
 
 Temperatu 
 curial Thei 
 and 101 
 
 Temperatu 
 curial Thei 
 and 40 
 Air Then 
 
 1 
 
 Tempera 
 Absolute S 
 
 
 bo _ 
 
 - 
 
 Temperatu 
 curial Thei 
 and 1(K 
 
 Temperatu 
 curial Thei 
 and 40 
 Air Then 
 
 II 
 
 - 
 
 po 
 
 '! 
 
 30 
 
 -.464 
 
 -.460 
 
 -.457 
 
 -457 
 
 230 
 
 17.198 
 
 17.067 
 
 17.009 
 
 17.005 
 
 35 
 
 
 
 
 
 
 
 
 
 240 
 
 18.056 
 
 17.920 
 
 17.861 
 
 17.857 
 
 40 
 
 +.463 
 
 +.460 
 
 +.457 
 
 +.457 
 
 250 
 
 18.917 
 
 18.773 
 
 18.714 
 
 18.709 
 
 50 
 
 1.387 
 
 1.376 
 
 1.368 
 
 1.368 
 
 260 
 
 19.771 
 
 19.621 
 
 19.562 
 
 19.557 
 
 60 
 
 2.307 
 
 2.290 
 
 2.276 
 
 2.275 
 
 270 
 
 20.621 
 
 20.465 
 
 20.406 
 
 20.401 
 
 70 
 
 3.216 
 
 3.192 
 
 3.174 
 
 3.173 
 
 280 
 
 21.469 
 
 21.306 
 
 21.247 
 
 21.242 
 
 80 
 
 4.122 
 
 4.092 
 
 4.069 
 
 4.068 
 
 290 
 
 22.308 
 
 22.139 
 
 22.081 
 
 22.076 
 
 90 
 
 5.022 
 
 4.984 
 
 4.957 
 
 4.955 
 
 300 
 
 23.144 
 
 22.969 
 
 22.912 
 
 22.907 
 
 100 
 
 5.916 
 
 5.872 
 
 5.841 
 
 5.839 
 
 310 
 
 23.974 
 
 23.792 
 
 23.736 
 
 23.731 
 
 110 
 
 6.804 
 
 6.753 
 
 6.714 
 
 6.712 
 
 320 
 
 24.796 
 
 24.607 
 
 24.552 
 
 24.547 
 
 120 
 
 7.685 
 
 7.628 
 
 7.590 
 
 7.588 
 
 330 
 
 25.618 
 
 25.424 1 25.370 
 
 25.365 
 
 130 
 
 8.564 
 
 8.500 
 
 8.459 
 
 8.456 
 
 340 
 
 26.433 
 
 26.232 
 
 46.180 
 
 26.174 
 
 140 
 
 9.439 
 
 9.368 
 
 9.324 
 
 9.321 
 
 350 
 
 27.245 
 
 27.038 
 
 16.987 
 
 26.981 
 
 150 
 
 10.309 
 
 10.232 
 
 10.186 
 
 10.183 
 
 360 
 
 28.049 
 
 27.837 
 
 27.788 
 
 27.782 
 
 160 
 
 11.174 
 
 11.091 
 
 11.042 
 
 11.039 
 
 370 
 
 28.856 
 
 28.637 
 
 28.590 
 
 28.584 
 
 170 
 
 12.038 
 
 11.947 
 
 11.896 
 
 11.893 
 
 380 
 
 29.651 
 
 29.426 
 
 29.382 
 
 29.376 
 
 180 
 
 12.900 
 
 12.802 
 
 12.749 
 
 12.746 
 
 390 
 
 30.449 
 
 30.218 
 
 30.176 
 
 30.170 
 
 190 
 
 13.760 
 
 13.655 
 
 13.601 
 
 13.598 
 
 400 
 
 31.249 
 
 31.011 
 
 30.971 
 
 30.965 
 
 200 
 
 14.619 
 
 14.508 
 
 14.453 
 
 14.450 
 
 410 
 
 32.073 
 
 31.829 
 
 31.782 
 
 31.786 
 
 210 
 
 15.479 
 
 15.362 
 
 15.305 
 
 15.302 
 
 420 
 
 32.861 
 
 32.611 
 
 32.577 
 
 32.581 
 
 220 
 
 16.340 
 
 16.215 
 
 16.157 
 
 16.153 
 
 
 
 

 
 116 
 
 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 TABLE XX. THERMOMETER No. 6166. 
 
 1* 
 
 111 
 
 ifi 
 
 1 
 ?. a 
 
 2 
 
 il 
 
 ill 
 
 If* 
 
 13 
 
 1 
 
 11 
 
 *tfl 
 
 H 
 
 Ji 
 
 Temperature < 
 curial Therm( 
 and 100 i 
 
 Temperature < 
 curial Therm < 
 and 40 f 
 
 Temperatui 
 Air Thermoi 
 
 Temperatu 
 Absolute Sea 
 0C. 
 
 2 'ia 
 J?B 
 
 Temperature 
 curial Therm 
 and 100 
 
 Temperature 
 curial Therm 
 and 40 
 
 1 
 
 Temperatu 
 Absolute Sot 
 0C. 
 
 20 
 
 .036 
 
 -.036 
 
 .034 
 
 -.034 
 
 230 
 
 16.478 
 
 16.356 
 
 16.298 
 
 16.294 
 
 30 
 40 
 
 +.770 
 1.574 
 
 +.764 
 1.562 
 
 +.759 
 1.553 
 
 +.759 
 1.553 
 
 240 
 250 
 
 17.259 
 18.042 
 
 17.132 
 17.908 
 
 17.074 
 17.849 
 
 17.070 
 17.845 
 
 50 
 
 2.368 
 
 2.350 
 
 2.336 
 
 2.335 
 
 260 
 
 18.825 
 
 18.686 
 
 18.627 
 
 18.622 
 
 60 
 
 3.156 
 
 3.133 
 
 3.115 
 
 3.114 
 
 270 
 
 19.609 
 
 19.464 
 
 19.405 
 
 19.400 
 
 70 
 
 3.941 
 
 3.911 
 
 3.889 
 
 3.888 
 
 280 
 
 20.392 
 
 20.241 
 
 20.182 
 
 20.177 
 
 80 
 
 4.726 
 
 4.691 
 
 4.665 
 
 4.664 
 
 290 
 
 21.170 
 
 21019 
 
 20.960 
 
 20.955 
 
 90 
 
 5.509 
 
 5.468 
 
 5.438 
 
 5.436 
 
 300 
 
 21.735 
 
 21.793 
 
 21.735 
 
 21.730 
 
 100 
 
 6.293 
 
 6.246 
 
 6.212 
 
 6.210 
 
 310 
 
 22.511 
 
 22.569 
 
 22.511 
 
 22.506 
 
 110 
 
 7.076 
 
 7024 
 
 6.988 
 
 6.986 
 
 320 
 
 23.292 
 
 23.349 
 
 23.292 
 
 23.287 
 
 120 
 
 7.862 
 
 7.804 
 
 7.765 
 
 7.763 
 
 330 
 
 24.075 
 
 24.131 
 
 24.075 
 
 24.070 
 
 130 
 
 8.649 
 
 8.585 
 
 8.544 
 
 8.542 
 
 340 
 
 24.855 
 
 24.910 
 
 24.855 
 
 24.850 
 
 140 
 
 9.437 
 
 9.367 
 
 9.323 
 
 9.321 
 
 350 
 
 25.634 
 
 25.687 
 
 25634 
 
 25.628 
 
 150 
 
 10.228 
 
 10.151 
 
 10.105 
 
 10.102 
 
 360 
 
 26.415 
 
 26.466 
 
 26.412 
 
 26.406 
 
 160 
 
 11017 
 
 10.935 
 
 10.887 
 
 10.884 
 
 370 
 
 27.441 
 
 27.245 
 
 27.195 
 
 27.189 
 
 170 
 
 11.805 
 
 11.717 
 
 11.667 
 
 11.664 
 
 380 
 
 28.240 
 
 28.030 
 
 27.982 
 
 27.976 
 
 180 
 
 12.589 
 
 12.496 
 
 12.444 
 
 12.441 
 
 390 
 
 29.030 
 
 28.814 
 
 28.768 
 
 28.762 
 
 190 
 
 13.370 
 
 13.271 
 
 13.217 
 
 13.214 
 
 400 
 
 29.819 
 
 29.597 
 
 29.550 
 
 29.544 
 
 200 
 
 14.148 
 
 14.043 
 
 13.988 
 
 13.984 
 
 410 
 
 30.608 
 
 30.381 
 
 30.339 
 
 30.333 
 
 210 
 
 14.923 
 
 14.812 
 
 14.756 
 
 14.752 
 
 420 
 
 31.396 
 
 31.162 
 
 31.123 
 
 31.117 
 
 220 
 
 15.699 
 
 15.583 
 
 15.526 
 
 15.522 
 
 430 
 
 32.189 
 
 31.950 
 
 31.914 
 
 31.908 
 
 
 In using these tables a correction is of course to be made should 
 the zero point change. 
 
 TABLE XXI. CORRECTION OF KEW STANDARD TO THE ABSOLUTE SCALE. 
 
 Temperature C. 
 
 Correction in 
 Degrees C. 
 
 
 
 
 
 10 
 
 .03 
 
 20 
 
 .05 
 
 30 
 
 .06 
 
 40 .07 
 
 50 
 
 .07 
 
 60 
 
 .06 
 
 70 
 
 -.04 
 
 80 
 
 .02 
 
 90 
 
 .01 
 
 100 
 

 
 OF ARTS AND SCIENCES. 
 
 117 
 
 Appendix to Thermometry. 
 
 The last of January, 1879, Mr. S. W. Holraan, of the Massachusetts 
 Institute of Technology, came to Baltimore to compare some ther- 
 mometers with the air thermometer ; and by his kindness I will give 
 here the results of the comparison which we then made together. 
 
 As in this comparison some thermometers made by Fastre in 1851 
 were used, the results are of the greatest interest. 
 
 'The tables are calculated with the newest value for the coefficient 
 of expansion of glass. The calibration of all the thermometers, except 
 the two by Casella, has been examined, and found good. The Casella 
 thermometers had no reservoir at the top, and could not thus be readily 
 calibrated after being made. The Geissler also had none, but I suc- 
 ceeded in separating a column. 
 
 The absence of a reservoir at the top should immediately condemn 
 a standard, for there is no certainty in the work done with it. 
 
 TABLE XXII. SEVENTH SERIES. 
 
 
 Original Readings. 
 
 Reduced Readings. 
 
 Air 
 
 
 
 
 
 
 6163 
 
 
 
 
 
 Ther- 
 
 
 
 Kew 
 
 
 
 Reduced 
 
 
 Kew 
 
 
 
 mome- 
 ter. 
 
 6163. 
 
 7334 
 Baudin. 
 
 Stand- 
 ard 
 
 32374 
 Casella. 
 
 Geissler. 
 
 to Air 
 Ther- 
 
 7334 
 Baudin. 
 
 Stand- 
 ard 
 
 32374 
 Casella. 
 
 Geissler. 
 
 
 
 
 No. 104. 
 
 
 
 mome- 
 
 
 No.104. 
 
 
 
 
 
 
 
 
 
 ter. 
 
 
 
 
 
 
 
 *58.83 
 
 .11 
 
 32.68 
 
 +.20 
 
 +.69 
 
 
 
 6 
 
 a 
 
 
 
 
 
 t.43 
 
 63.5 
 
 
 3360 
 
 .71 
 
 
 .52 
 
 
 .52 
 
 .51 
 
 
 6.08 
 
 113.0 
 
 
 43.65 
 
 6.33 
 
 
 6.08 
 
 
 6.11 
 
 6.13 
 
 
 12.68 
 
 171.55 
 
 12.59 
 
 55.47 
 
 12.91 
 
 13.42 
 
 12.65 
 
 12.73 
 
 12.68 
 
 1270 
 
 12.82 
 
 20.49 
 
 242.0 
 
 20.48 
 
 69.56 
 
 20.77 
 
 21.29 
 
 20.49 
 
 20.63 
 
 20.57 
 
 20.56 
 
 20.74 
 
 24.55 
 
 278.8 
 
 24.50 
 
 76.90 
 
 24.80 
 
 25.33 
 
 24.54 
 
 24.66 
 
 24.61 
 
 24.59 
 
 24.81 
 
 29.51 
 
 323.9 
 
 29.49 
 
 85.88 
 
 29.80 
 
 30.32 
 
 29.52 
 
 29.66 
 
 29.61 
 
 29.58 
 
 29.83 
 
 39.45 
 
 4131 
 
 39.43 
 
 103.72 
 
 39.76 
 
 40.22 
 
 39.47 
 
 39.62 
 
 39.53 
 
 39.54 
 
 39.80 
 
 39.15 
 
 4107 
 
 39.15 
 
 103.23 
 
 39.48 
 
 39.08 
 
 39.20 
 
 39.34 
 
 39.26 
 
 39.26 
 
 39.56 
 
 51.17 
 
 
 51.10 
 
 124.84 
 
 51.49 
 
 51.83 
 
 .... 
 
 51:32 
 
 51.29 
 
 51.26 
 
 51.49 
 
 61.12 
 
 
 61.05 
 
 142.73 
 
 61.47 
 
 61.69 
 
 
 61.29 
 
 61.24 
 
 61.23 
 
 61.41 
 
 70.74 
 
 ' | 
 
 70.57 
 
 159.87 
 
 71.00 
 
 71.14 
 
 ..... 
 
 70.83 
 
 70.78 
 
 70.76 
 
 70.92 
 
 80.09 
 
 
 79.74 
 
 176.50 
 
 80.31 
 
 80.25 
 
 
 80.02 
 
 80.04 
 
 80.06 
 
 80.10 
 
 80.39 
 
 
 80.15 
 
 177.23 
 
 80.74 
 
 .80.66 
 
 . 
 
 80.43 
 
 80.44 
 
 80.49 
 
 80.51 
 
 89.95 
 
 
 89.63 
 
 194.35 
 
 90.22 
 
 90.11 
 
 
 89.93 
 
 89.97 
 
 89.97 
 
 90.03 
 
 89.92 
 
 
 89.59 
 
 194.22 
 
 90.18 
 
 90.06 
 
 . 
 
 89.89 
 
 89.90 
 
 89.93 
 
 89.98 
 
 100.00 
 
 
 99.69 
 
 212.37 
 
 100.06 
 
 99.32 
 
 . . . 
 
 100.00 
 
 100.00 
 
 100.00 
 
 100.00 
 
 * The original readings in ice were 58.68 and 58 45, to which .15 was added 
 to allow for the pressure of water in the comparator. This, of course, gives 
 the same final result as if .15 were subtracted from eacli of the other tempera- 
 tures. No correction was made to the others. 
 
 t Probably some error of reading.
 
 118 
 
 PROCEEDINGS OP THE AMERICAN- ACADEMY 
 
 TABLE XXIII EIGHTH SERIES. 
 
 
 Original Readings. 
 
 Reduced Readings. 
 
 Air 
 Ther- 
 mome- 
 
 6163 
 
 376 
 
 7316 
 
 Baudin. 
 
 368 
 Fastr<5. 
 
 3236 
 
 Casella. 
 
 6163 
 Reduced 
 to Air 
 Ther- 
 
 376 
 Fastrd. 
 
 7316 
 Baudin 
 
 868. 
 
 Fastr<5. 
 
 3235 
 Casella. 
 
 ter. 
 
 
 Fastre. 
 
 
 
 
 mome- 
 
 
 
 
 
 
 
 
 
 
 
 ter. 
 
 
 
 
 
 
 
 *5860 
 
 111.3 
 
 .23 
 
 87.6 
 
 32.80 
 
 5 
 
 
 
 ft 
 
 
 
 
 
 3.67 
 
 90.7 
 
 130.0 
 
 
 106.25 
 
 39.35 
 
 3.61 
 
 3.64 
 
 
 3.64 
 
 3.65 
 
 11.55 
 
 161.6 
 
 170.9 
 
 11.40 
 
 147.2 
 
 53.70 
 
 11.56 
 
 11.60 
 
 il.64 
 
 11.62 
 
 11. Q3 
 
 20.72 
 
 243.7 
 
 217.9 
 
 20.59 
 
 194.2 
 
 70.15 
 
 20.70 
 
 20.75 
 
 20.84 
 
 20.80 
 
 20.79 
 
 32.19 
 
 347.4 
 
 276.9 
 
 32.09 
 
 253.2 
 
 90.80 
 
 32.17 
 
 32.24 
 
 32.34 
 
 32.28 
 
 32.29 
 
 39.36 
 
 411.85 
 
 313.85 
 
 39.26 
 
 290.1 
 
 103.68 
 
 39.36 
 
 39.43 
 
 39.52 
 
 39.48 
 
 39.45 
 
 50.71 
 
 
 372.0 
 
 50.57 
 
 248.2 
 
 123.65 
 
 i 
 
 50.75 
 
 50.84 
 
 50.80 
 
 50.57 
 
 60.10 
 
 . '. 
 
 420.0 
 
 59.92 
 
 39(5.45 
 
 140.80 
 
 . 
 
 60.10 
 
 60.19 
 
 60.21 
 
 60.12 
 
 73.82 
 
 
 490.6 
 
 73.59 
 
 466.85 
 
 165.68 
 
 
 73.84 
 
 73.87 
 
 73.93 
 
 73.97 
 
 86.50 
 
 
 655.25 
 
 86.16 
 
 531.22 
 
 188.20 
 
 . 
 
 86.48 
 
 86.51 
 
 86.56 
 
 86.56 
 
 
 
 550.2 
 
 85.21 
 
 525.95 
 
 186.42 
 
 
 86.45 
 
 85.60 
 
 85.45 
 
 85.51 
 
 100.66 
 
 
 624.93 
 
 99.70 
 
 600.58 
 
 212.45 
 
 . . . 
 
 100.00 
 
 100.00 
 
 100.00 
 
 100.00 
 
 From these tables we would draw the inference that No. 6163 
 represents the air thermometer with considerable accuracy. At the 
 same time, both tables would give a smaller value of m than I have 
 used, and not very far from the value found before by direct compari- 
 son, namely, .00015. 
 
 The difference from using m = .00018 would be a little over 
 0.01 C. at the 20 point. 
 
 All the other thermometers stand above the air thermometer, 
 between and 100, by amounts ranging between about 0.05 and 
 0.35 C., none standing below. Indeed, no table has ever been 
 published showing any thermometer standing below the air ther- 
 mometer between and 100. By inference from experiments above 
 100 on crystal glass by Regnault, thermometers of this glass should 
 stand below, but it never seems to have been proved by direct experi- 
 ment. The Fastre thermometers are probably made of this glass, 
 and my Baudins certainly contain lead ; and yet these stand above, 
 though only to a small amount, in the case of the Fastre's. 
 
 The Geissler still seems to retain its pre-eminence as having the 
 greatest error of the lot. 
 
 The Baudin thermometers agree well together, but are evidently 
 made from another lot of glass from the No. 6 1 67 used before. These 
 last two depart less from the air thermometer. The explanation is 
 plain, as Baudin had manufactured more than one thousand ther- 
 
 * See note on preceding page.
 
 OF ARTS AND SCIENCES. 
 
 119 
 
 mometers between the two, and so had probably used up the first 
 stock of glass. And even glass of the same lot differs, especially as 
 Regnault has shown that the method of working it before the blow- 
 pipe affects it very greatly. 
 
 It is very easy to test whether the calorimeter thermometers are of 
 the same glass as any of the others, by testing whether they agree 
 with No. 6163 throughout the whole range of 40. The difference 
 in the values of m for the two kinds of glass will then be about 
 .003 of the difference between them at 20, the and 40 points 
 agreeing. The only difficulty is in calibrating or reading the 100 
 thermometers accurately enough. 
 
 The Baudin thermometers were very well calibrated, and were 
 graduated to -fa C., and so were best adapted to this kind of work. 
 Hence I have constructed the following tables, making the and 40 
 points agree. 
 
 TABLE XXIV. COMPARISON OF 6163 AND THE BAUDIN STANDARDS. 
 
 6163 
 Mercurial 
 and 40 fixed. 
 
 7334* 
 
 Difference. 
 
 6163 
 Mercurial 
 and 40 fixed. 
 
 7316.* 
 
 Difference. 
 
 
 12.699 
 20.547 
 24.604 
 29.564 
 39.337 
 
 
 12.673 
 20.553 
 24.567 
 29.550 
 39.337 
 
 
 
 +.026 
 .006 
 +.037 
 + 014 
 
 
 
 11.609 
 20.762 
 32.203 
 39.358 
 
 
 11.584 
 20.746 
 32.211 
 39.358 
 
 
 +.025 
 +.016 
 .008 
 
 
 Taking the average of the two, it would seem that No. 6163 stood 
 about .015 higher than the mean of 7334 and 7316 at the 20 point, 
 or 6163 has a higher value of m by .000045 than the others. 
 
 These differ about .17 from the air thermometer at 40, which gives 
 the value of m about .000104. Whence m for 6163 is .00015, as we 
 have found before by direct comparison with the air thermometer. 
 
 I am inclined to think that the former value, .00018, is too large, 
 and to take .00015, which is the value found by direct comparison, as 
 the true value. As the change, however, only makes at most a differ- 
 ence of 0.01 at any one point, and as I have already used the previous 
 value in all calculations, I have not thought it worth while to go over 
 all my work again, but will refer to the matter again in the final 
 results, and then reduce the final results to this value. 
 
 * A correction of 0.01 was made to the zero points of these thermometers 
 on account of the pressure of the water.
 
 120 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 III. CALORIMETRY. 
 (a) Specific Heat of Water. 
 
 The first observers on the specific heat of water, such as De Luc, 
 completed the experiment with a view of testing the thermometer ; 
 and it is curious to note that both De Luc and Flaugergues found the 
 temperature of the mixture less than the mean of the two equal por- 
 tions of which it was composed, and hence the specific heat of cold 
 water higher than that of warm. 
 
 The experiments of Flaugergues were apparently the best, and he 
 found as follows : * 
 
 3 parts of water at and 1 part at 80 R. gave 19.86 R. 
 2 parts of " " 2 parts " " 39.81 R. 
 
 1 part of " " 3 parts " " 59.87 R. 
 
 But it is not at all certain that any correction was made for the 
 specific heat of the vessel, or whether the loss by evaporation or 
 radiation was guarded against. 
 
 The first experiments of any accuracy on .this subject seem to have 
 been made by F. E. Newmann in 1831.f He finds that the specific 
 heat of water at the boiling point is 1.0127 times that at about 28 C. 
 (22 R.). 
 
 The next observer seems to have been Regnault,t who, in 1840, 
 found the mean specific heat between 100 C. and 16 C. to be 
 1.00709 and 1.00890 times that at about 14. 
 
 But the principal experiments on the subject were published by 
 Regnault in 1850, and these have been accepted to the present time. 
 It is unfortunate that these experiments were all made by mixing 
 water above 100 with water at ordinary temperatures, it being 
 assumed that water at ordinary temperatures changed little, if any. 
 An interpolation formula was then found to represent the results; 
 and it was assumed that the same formula held at ordinary tempera- 
 ture, or even as low as C. It is true that Regnault experimented 
 on the subject at points around 4 C. by determining the specific heat 
 of lead in water at various temperatures ; but the results were not of 
 sufficient accuracy to warrant any conclusions except that the variation 
 was not great. 
 
 * Gehler, Phys. Worterbuch, i. 641. 
 
 t Pogg. Ann., xxiii. 40. j Ibid., li. 72. 
 
 Pogg. Ann., Ixxix. 241 ; also, Rel. d. Exp., i. 729.
 
 OF ARTS AND SCIENCES. 121 
 
 Boscha has attempted to correct Regnault's results so as to reduce 
 them to the air thermometer ; but Regnault, in Comptes Rendus, has 
 not accepted the correction, as the results were already reduced to the 
 air thermometer. 
 
 Him (Comptes Rendus, Ixx. 592, 831) has given the results of 
 some experiments on the specific heat of water at low temperatures, 
 which give the absurd result that the specific heat of water increases 
 about six or seven per cent between zero and 13! The method of 
 experiment was to immerse the bulb of a water thermometer in the 
 water of the calorimeter, until the water had contracted just so much, 
 when it was withdrawn. The idea of thus giviug equal quanti- 
 ties of heat to the water was excellent, but could not be carried 
 into execution without a great amount of error. Indeed, experi- 
 ments so full of error only confuse the physicist, and are worse than 
 useless. 
 
 The experiments of Jamin and Amaury, by the heating of water by 
 electricity, were better in principle, and, if carried out with care, 
 would doubtless give good results. But no particular care seems to 
 have been taken to determine the variatiorfof the resistance of the 
 wire with accuracy, and the measurement of the temperature is 
 passed over as if it were a very simple, instead of an immensely diffi- 
 cult matter. Their results are thus to be rejected ; and, indeed, 
 Regnault does not accept them, but believes there is very little change 
 between 5 and 25. 
 
 In PoggendorfFs Annalen for 1870 a paper by Pfaundler and Plat- 
 ter appeared, giving the results of experiments around 4 C., and 
 deducing the remarkable result that water from to 10 C. varied as 
 much as twenty per cent, in specific heat, and in a very irregular man- 
 ner, first decreasing, then increasing, and again decreasing. But 
 soon after another paper appeared, showing that the results of the 
 previous experiments were entirely erroneous. 
 
 The new experiments, which extended up to 13 C., seemed to give 
 an increase of specific heat up to about 6, after which there was 
 apparently a decrease. It is to be noted that Geissler's thermometers 
 were used, which I have found to depart more than any other from 
 the air thermometer. 
 
 But as the range of temperature is very small, the reduction to the 
 air thermometer will not affect the results very much, though it will 
 somewhat decrease the apparent change of specific heat. 
 
 In the Journal de Physique for November, 1878, there is a notice of 
 some experiments of M. von Munchausen on the specific heat of
 
 122 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 water. The method was that of mixture in an open vessel, where 
 evaporation might interfere very much with the experiment. No 
 reference is made to the thermometer, but it seems not improbable 
 that it was one from Geissler ; in which case the error would be very 
 great, as the range was large, and reached even up to 70 C. The 
 error of the Geissler would be in the direction of making the specific 
 heat increase more rapidly than it should. The formula he gives for 
 the specific heat of water at the temperature t is 
 1 _|_ .000302 t. 
 
 Assuming that the thermometer was from Geissler, the formula, 
 reduced to the air thermometer, would become approximately 
 
 1 _ .00009 t + .0000015 t 2 . 
 
 Had the thermometer been similar to that of Recknagel, it would 
 have been 1 -f .000045 t + .000001 t\ 
 
 It is to be noted that the first formula would actually give a decrease 
 of specific heat at first, and then an increase. 
 
 As all these results vary so very much from each other, we can 
 hardly say that we know anything about the specific heat of water 
 between and 100, though Regnault's results above that temperature 
 are probably very nearly correct. 
 
 It seems to me probable that my results with the mechanical 
 equivalent apparatus give the variation of the specific heat of water 
 with considerable accuracy ; indeed, far surpassing any results which 
 we can obtain by the method of mixture. It is a curious result of 
 those experiments, that at low temperatures, or up to about 30 C., 
 the specific heat of water is about constant on the mercurial ther- 
 mometer made by Baudin, but decreases to a minimum at about 30 
 when the reduction is mvde to the air thermometer or the absolute scale, 
 or, indeed, the Kew standard. 
 
 ' As this curious and interesting result depends upon the accurate 
 comparison of the mercurial with the air thermometer, I have spent 
 the greater part of a year in the study of the comparison, but have 
 not been able to find any error, and am now thoroughly convinced of 
 the truth"of this decrease of the specific heat.' But to make certain, 
 I have instituted the following independent series of investigations 
 on the specific heat of water, using, however, the same thermome- 
 ters. 
 
 The apparatus is shown in Fig. 4. A copper vessel, A, about 20 cm -
 
 OP ARTS AND SCIENCES. 
 
 123 
 
 in diameter and 23 cm high, rests upon a tripod. In its interior is a 
 three-way stopcock, communicating with the small interior vessel , 
 the vessel A, and the vulcanite spout G. By turning it, the vessel B 
 could be filled with water, and its temperature measured by the ther- 
 mometer D, after which it could be delivered through the spout into 
 the calorimeter. As the.vessel B, the stopcock, and most of the spout, 
 were within the vessel A, and thus surrounded by water, and as the 
 vulcanite tube was very thin, the water could be delivered into the 
 
 calorimeter without appreciable change of temperature. The proof 
 of this will follow later. 
 
 The calorimeter, JE, was of very thin copper, nickel-plated very 
 thinly. A hole in the back at F allowed the delivery spout to enter, 
 and two openings on top admitted the thermometers. A wire attached 
 to a stirrer also passed through the top. The calorimeter had a 
 capacity of about three litres, and weighed complete about 388.3 
 grammes. Its calorific capacity was estimated at 35.4 grammes. It 
 rested on three vulcanite pieces, to prevent conduction to the jacket.
 
 124 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 Around the calorimeter on all sides was a water-jacket, nickel- 
 plated on its interior, to make the radiation perfectly definite. 
 
 The calorific capacity of the thermometers, including the immersed 
 stem and the mercury of the bulb, was estimated as follows: 14 cm - of 
 stem weighed about 3.8*% and had a capacity of .8^; 10"- of 
 mercury had a capacity of .3**- ; total, I.!* 1 -. 
 
 Often the vessel B was removed, and the water allowed to flow 
 directly into the calorimeter. 
 
 The following is the process followed during one experiment at low 
 temperatures. The vessel A was filled with clean broken ice, the open- 
 ing into the stopcock being covered with fine gauze to prevent any 
 small particles of ice from flowing out. The whole was then covered 
 with cloth, to prevent melting. The vessel was then filled with water, 
 and the two thermometers immersed to get the zero points. The 
 calorimeter being about two thirds filled with water, and having been 
 weighed, was then put in position, the holes corked up, and one ther- 
 mometer placed in it, the other being in the melting ice. An obser- 
 vation of its temperature was then taken every minute, it being 
 frequently stirred. 
 
 When enough observations had been obtained in this way, the cork 
 was taken out of the aperture .Fand the spout inserted, and the water 
 allowed to run for a given time, or until the calorimeter was full. It 
 was then removed, the cork replaced, and the second thermometer 
 removed from the ice to the calorimeter. Observations were then 
 taken as before, and the vessel again weighed. 
 
 Two thermometers were used in the way specified, so that one 
 might approach the final temperature from above and the other from 
 below. But no regular difference was ever observed, and so some 
 experiments were made with both thermometers in the calorimeter 
 during the whole experiment. 
 
 The principal sources of error are as follows : 
 
 1st. Thermometers lag behind their true reading. This was not 
 noticed, and would probably be greater in thermometers with very 
 fine stems like Geissler's. At any rate, it was almost eliminated in 
 the experiment by using two thermometers. 
 
 2d. The water may be changed in temperature in passing through 
 the spout. This was eliminated by allowing the water to run some 
 time before it went into the calorimeter. The spout being very thin, 
 and made of vulcanite, covered on the outside with cloth, it is not 
 thought that there was any appreciable error. It will be discussed 
 more at length below, and an experiment given to prove this.
 
 OP ARTS AND SCIENCES. 
 
 125 
 
 3d. The top of the calorimeter not being in contact with the water 
 rt. temperature may be uncertain. To eliminate this, the 
 was often at the temperature of the air to commence with. Also the 
 water was sometimes violently agitated just before taking the final 
 leading previous to letting in the cold water. Even if the tempera 
 ture of this part was take* as that of the air, the error would scarcely 
 ever be of sufficient importance to vitiate the conclusions 
 
 5th Some water might remain in the spout whose temperature 
 might be different from the rest. This was guarded against. 
 
 6 h. Evaporation. Impossible, as the calorimeter was closed. 
 
 7th. The introduction of cold water may cause dew to be depos- 
 octrred rimeter ' The ex P eri <*^ were rejected where this 
 
 The corrections for the protruding thermometer stem, for radiation 
 &c, were made as usual, the radiation being estimated by a series of 
 observations before and after the experiment, as is usual in determin- 
 ing the specific heat of solids. 
 
 June 14, 1878. Wr* Experiment. 
 
 Time. 
 
 41 
 42 
 43 
 44 
 
 Ther. 6168. 
 
 296.75 
 296.7 
 296.7 
 296.65 
 
 Points. 
 
 6163, 57.9 Air, 21 C. 
 
 6165, 34.8 Jacket about 25 C. 
 
 6166, 20.5 
 
 44-44f Water running. 
 *6i 218.7 251.7 
 4 H 218.8 251.8 
 *8 218.9 252.0 
 
 Calorimeter before 
 after 
 
 Water at added 
 Thermometer 
 
 2043.0 
 2853.3 
 
 810.3 
 1.1 
 
 
 Total' at 
 
 811.4 
 
 Temperature before 296.6 
 Correction for -}- .2 
 
 Calorimeter before 
 Weight of vessel 
 
 Water 
 Capacity of calorim. 
 thermom 
 
 2043.0 
 388.3 
 
 TesI? 
 
 35.4 
 1.1 
 
 296.8 = 26.597 
 Correction for stem -{- .019 
 
 Initial temp, of calorimeter 26.616 
 
 
 Total capacity 
 
 1691.2
 
 126 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 218 . 6 + .2 = 218.8 = 17.994 251.6 - 1 = 251.5 = 17.962 
 Correction for stem 006 Correction for stem 006 
 
 1^988 17 ' 956 
 
 Mean temperature of mixture, 17.972. 
 
 Mean specific heat 18 1691.2 X 8.644 __ l m5 
 Me^n specific heat 18 27 811.4 X 17-972 
 
 June 14. Second Experiment. 
 
 Calorimeter before 2016.3; temperature 361.4 by No. 6163. 
 after 3047.0; " 244.5 and 288.7. 
 
 Air, 21 C.; jacket about 27. 
 
 361.4 + .2 = 361.6 = 33.803, or 33.863 when' corrected for stem. 
 
 244.5 -j- .2 = 244.7 = 20.865 ; no correction for stem. 
 
 288.7 .1 = 288.6 = 20.846 ; 
 
 Mean, 20.855. 
 
 Mean specific heat between and 21 ^ ^^gg 
 
 Mean specific heat between 21 and 34 
 
 June 14. Third Experiment. 
 
 Calorimeter before 1961.8; temperature 293.6 by No. 6166. 
 " ' after 3044.6; " 243.7 and 213.0. 
 
 Air and jacket, about 18 C. 
 
 393.6 .1 = 393.5 = 29.036, or 29.077 when corrected for stem. 
 
 243.7 .1 = 243.6 = 17.349 ; no correction for stem. 
 213.0 4- .2 = 213.2 = 17.374 ; 
 
 Mean, 17.361. 
 
 Mean specific heat between and 17 _ QQOI 
 Mean specific heat between 17 and 29 ~ 
 
 It is to be observed that thermometer No. 6166 in all cases gave 
 temperatures about 0.02 or 0.03 below No. 6163. This difference 
 is undoubtedly in the determination of the zero points, as on June 15 
 the zero points were found to be 20.4 and 58.0. As one has gone up 
 and the other down, the mean of the temperatures needs no cor- 
 rection.
 
 OP ARTS AND SCIENCES. 127 
 
 i 
 
 June 15. 
 
 Calorimeter before 2068.2 ; temperature 364.6 by No. 6166. 
 after 2929.2 ; " 249.7 and 217.7. 
 
 Air and jacket at about 22 C. 
 
 264.6 == 26.766, or 26.782 when Corrected for stem. 
 
 249.7 = 17.822, or 17.812 " 
 
 217.7 + .1 =217.8 = 17.884, or 17.874 when corrected for stem. 
 
 Rejected on account of great difference in final temperatures by the 
 two thermometers, which was probably due to some error in reading. 
 
 June 21. 
 
 Calorimeter before 2002.7 ; temperature 330.3 by No. 6163. 
 
 " after 3075.2; " 221.9 and 256.6. 
 
 .- 
 Air and jacket, 21 C. 
 
 330.3 -f- .1 = 330.4 = 30.321, or 30.359 when corrected for stem. 
 
 221.9 + - 1 == 222 - = 18.349, or 18.343 
 
 256.6 -j- .0 = 256.6 = 18.358, or 18.352 " " 
 
 Mean, 18.347. 
 
 Specific heat between and 18 
 Specific heat between 18 and 30 
 
 June 21. 
 
 Calorimeter before 2073.8 ; temperature 347.8 by No. 6166. 
 " after 2986.8 ; " 234.5 and 206.6. 
 
 Air and jacket, about 21 C. 
 
 347.8 -|-.0 = 347.8 = 25.457, or 25.471 when corrected for stem. 
 234.5 -f -0 = 234.5 = 16.643, or 16.636 " " 
 206.6 -j-.l = 206.7 = 16.651, or 16.644 
 
 Mean, 16.640. 
 
 Specific heat between and 17 99971 
 
 Specific heat between 17. and 25 
 
 Rejected because dew was formed on the calorimeter. 
 
 A series was now tried with both thermometers in the calorimeter 
 from the beginning.
 
 128 ' PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 June 25. 
 
 Calor. before 2220.3 ; temperat. 325.6 by No. 6166 ; 309.9 by No. 6165. 
 after 3031.4; 233.4 224.6 
 
 Air, 24.2 C. ; jacket, 23.5. 
 
 325 6 4- .0 = 325.6'= 23.725, or 23.72'6 when corrected for stem. 
 309'.9 + '.2 = 310.1 = 23.739, or 23.740 
 
 233.4 -f -0 = 233.4 = 16.558, or 16.545 
 224.6 + .2 = 224.8 = 16.562, or 16.549 
 
 Means, 28.733 and 16.547. 
 Specific heat between and 16 _ 1 QQ ^ 
 Specific heat between 16 and 24 
 
 June 25. 
 
 Calor. before 2278.6; temperat. 340.35 by No. 6166; 324.1 by No. 6165. 
 after 3130.2; " 242.5 " 232.8 " " 
 
 Air, 23.5 C. ; jacket, 22.5. 
 
 340.35 + .0 = 340.35 = 24 .877, or 24.881 when corrected for stem. 
 324.1 +.2=324.3 = 24.899, or 24.903 " " 
 
 242.5 -f .0 = 242.5 = 17.264, or 17.253 " " 
 232.8 +.2 = 233.0 = 17.261, or 17.250 " " 
 
 Specific heat between and 17 ^7 
 
 Specific heat between 17 and 25 
 
 , June 25. 
 
 Calor. before 2316.35 ; temperat. 386.1 by No. 6166 ; 368.4 by No. 61 65. 
 
 " after 2966.90; " 295.4 " " 281.7 " " 
 
 Air, 23.5 C.; -jacket, 22.5. 
 
 386.1 -f .0 = 386.1 = 28.455, or 28.465 when corrected for stem. 
 368.4 -f- .2 = 368.6 = 28.472, or 28.482 " 
 
 295.4 -f -0 = 295.4 = 21.374, or 21.368 
 
 281.7 -j- '-2 = 281.9 = 21.400, or 21.394 " " 
 
 Means, 28.473 and 21.381. 
 
 Specific heat between and 21 
 
 Specific heat between 21 and 28 * 
 
 Two experiments were made on June 23 with warm water in 
 vessel A, readings being taken of the temperature of the water, as it
 
 OF ARTS AND SCIENCES. 129 
 
 flowed out, by one thermojneter, which was then transferred to the 
 calorimeter as before. 
 
 June 23. 
 Water in A while running, 314.15 by No. 6163'. 
 
 Calor. before 1530.9 ; temperat. 281.1 by No. 6166. 
 
 " after 2996.3; % " 328.4byNo.6166; 272.7byNo.6163. 
 
 314.15 + .1 = 314.25 = 28.526, or 28.552 when corrected for stem. 
 281.1 -{-.0 = 281.1 = 20.262, or 20.258 " " " 
 
 328.4 4- -0 = 328.4 = 23.945, or 23.950 
 272.7 -f- 1 = 272.8 = 23.960, or 23.966 
 
 Specific heat between 20 and 24 _ 
 Specific heat between 24 and 29 ~ 
 
 June 23. 
 Water in A while running, 383.9 by No. 6163. 
 
 Calor. before 1624.9 ; temperat. 286.75 by 6166. 
 after 3048.2 ; 392.45 by 6166, and 318.1 by 6163. 
 
 383.9 -f- .1 = 384.0 = 36.303, or 36.357 when corrected for stem. 
 286.75 -j- .0 = 286.75 = 20.702, or 20.700 
 392.45 -f -0 = 392.45 = 28.954, or 28.980 
 318.1 -f .1 = 318.2 = 28.9.64, or 28.992 
 
 Specific heat between 21 and 29 
 
 Specific heat between 29 and 36 
 
 To test the apparatus, and also to check the estimated specific heat 
 of the calorimeter, the water was almost entirely poured out of the 
 calorimeter, and warm water placed in the vessel A, which was then 
 allowed to flow into the calorimeter. 
 
 Water in A while running, 309.0 by No. 6163. 
 
 Calor. before 391.3; temperat. 314.5 by 6166. 
 
 " after 3129.0; " 308.3 by 6166, and 378.5 by 6163. 
 
 Air about 21 C. 
 
 Therefore, water lost 0.078, and calorimeter gained 5. Hence 
 the capacity of the calorimeter is 39. 
 
 Another experiment, more carefully made, in which the range was 
 greater, gave 35. 
 
 VOL. xv. (N. s. vn.) 9
 
 130 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 The close agreement of these with the estimated amount is, of 
 course, only accidental, for they depend upon an estimation of only 
 0.08 and 0.12 respectively. But they at least show that the 
 water is delivered into the calorimeter without much change of 
 temperature. 
 
 A few experiments were made as follows between ordinary tempera- 
 tures and 100, seeing that this has already been determined by Reg- 
 nault. 
 
 Two thermometers were placed in the calorimeter, the temperature 
 of which was about 5 below that of the atmosphere. The vessel B 
 was then filled, and the water let into the calorimeter, by which the 
 temperature was nearly brought to that of the atmosphere ; the opera- 
 tion was then immediately repeated, by which the temperature rose 
 about 5 above the atmosphere. The temperature of the boiling 
 water was given by a thermometer whose 100 was taken several 
 times. 
 
 As only the rise of temperature is needed, the zero points of the 
 thermometers in the calorimeter are unnecessary, except to know that 
 they are within 0.02 of correct. 
 
 Jujie 18. 
 
 Temperature of boiling water, 99.9. 
 
 Calor. before 2684.7; temperat. 259.2 by 6166, and 248.3 by 6165. 
 " after 2993.2; " 381.0 363.4 " 
 
 259.3 = 18.568, or 18.555 when corrected for stem. 
 
 248.3 = 18.564, or 18.551 
 381.0 = 28.054, or 28.065 
 
 363.4 = 28.055, or 28.066 
 
 Specific heat 28 100 
 "Specific heal 18 ~^28" = 
 
 Other experiments gave 1.0015 and 1.0060, the mean of all of 
 which is 1.0033. Regnault's formula gives 1.005 ; but going directly 
 to his experiments, we get about 1.004, the other quantity being for 
 
 The agreement is very satisfactory, though one would expect my 
 
 small apparatus to lose more of the heat of the boiling water than 
 
 Regnault's. Indeed, for high temperatures my apparatus is much 
 
 r to Regnault's, and so I have not attempted any further 
 
 experiments at high temperatures.
 
 OP ARTS AND SCIENCES. 131 
 
 My only object was to confirm by this method the results deduced 
 from the experiments on the mechanical equivalent ; and this I have 
 done, for the experiments nearly all show that the specific heat of 
 water decreases to about 30, after which it increases. But the 
 mechanical equivalent experiments give by far the most accurate 
 solution of the problem ; and, indeed, give it with an accuracy hitherto 
 unattempted in experiments of this nature. 
 
 But whether water increases or decreases in specific heat from to 
 30 depends upon the determination of the reduction to the air ther- 
 mometer. According to the mercurial thermometers Nos. 6163, 6165, 
 and 6166, treating them only as mercurial thermometers, the specific heat 
 of water up to 30 is nearly constant, but by the air thermometer, or by 
 the Kew standard or Fastre, it decreases. 
 
 Full and complete tables of comparison are published, and from 
 them any one can satisfy himself of the facts in the case. 
 
 I am myself satisfied that I have obtained a very near approxi- 
 mation to absolute temperatures, and accept them as the standard. 
 And by this standard the specific heat of water undoubtedly decreases 
 from to about 30. 
 
 To show that I have not arrived at this result rashly, I may men- 
 tion that I fought against a conclusion so much at variance with my 
 preconceived notions, but was forced at last to accept it, after studying 
 it for more than a year, and making frequent comparisons of ther- 
 mometers, and examinations of all other sources of error. 
 
 However remarkable this fact may be, being the first instance of 
 the decrease of the specific heat with rise of temperature, it is no 
 more remarkable than the contraction of water to 4. Indeed, in 
 both cases the water hardly seems to have recovered from freezing. 
 The specific heat of melting ice is infinite. Why is it necessary that 
 the specific heat should instantly fall, and then recover as the tempera- 
 ture rises ? Is it not more natural to suppose that it continues to fall 
 even after the ice is melted, and then to rise again as the specific heat 
 approaches infinity at the boiling point? And of all the bodies which 
 we should select as probably exhibiting this property, water is cer- 
 tainly the first. 
 
 (J.) Heat Capacity of Calorimeter. 
 
 During the construction of the calorimeter, pieces of all the material 
 were saved in order to obtain the specific heat. The calorimeter which 
 Joule used was put together with screws, and with little or no solder. 
 But in my calorimeter it was necessary to use solder, as it was of a
 
 132 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 much more complicated pattern. The total capacity of the solder 
 used was only about ^ of the total capacity including the water; 
 and if we should neglect the whole, and call it copper, the error would 
 be only about T sW Hence H was considered 8ufficient to wei S h the 
 solder before and after use, being careful to weigh the scraps. The 
 error in the weight of solder could not possibly have been as great as 
 ten per cent, which would affect the capacity only 1 part in 12,000. 
 
 To determine the nickel used in plating, the calorimeter was weighed 
 before and after plating; but it weighed less after than before, owing 
 to the polishing of the copper. But I estimated the amount from the 
 thickness of a loose portion of the plating. I thus found the approxi- 
 mate weight of nickel, but as it was so small, I counted it as copper. 
 The following are the constituents of the calorimeter : 
 
 Thick sheet copper . 
 
 25.1 per cent. 
 
 Thin . 
 
 45.7 " 
 
 Cast brass 
 
 17.9 " 
 
 Rolled or drawn brass 
 
 5.7 " 
 
 Solder . . ' . 
 
 4.0 " 
 
 Steel . .. ' . 
 
 1.6 " 
 
 100.0 
 Nickel . . . . .3 
 
 To determine the mean specific heat, the basket of a Regnault's 
 apparatus was filled with the scraps in the above proportion, allowing 
 the basket of brass gauze, which was very light, to count toward the 
 drawn brass. The specific heat was then determined between 20 
 and 100, and between about 10 and 40. Between 20 and 100 
 the ordinary steam apparatus was used, but between 10 and 40 a 
 special apparatus filled with water was used, the water being around 
 the tube containing the basket, in the same manner as the steam is in 
 the original apparatus. In the calorimeter a stirrer was used, so that 
 the basket and water should rapidly attain the same temperature. 
 The water was weighed before and after the experiment, to allow for 
 evaporation. A correction of about 1 part in 1,000 was made, on 
 account of the heat lost by the basket in passing from the apparatus 
 to the calorimeter, in the 100 series, but no correction was made in 
 the other series. The thermometers in the calorimeter were Nos. 6163 
 and 6166 in the different experiments. 
 
 The principal difficulty in the determination is in the correction for 
 radiation, and for the heat which still remains in the basket after some
 
 OF ARTS AND SCIENCES. 133 
 
 time. After the basket has descended into the water, it commences 
 to give out heat to the water ; this, in turn, radiates heat ; and the 
 temperature we measure is dependent upon both these quantities. 
 
 Let T = temperature of the basket at the time t 
 
 " T' = " " " " 
 
 " T" = " " " " oo 
 
 " 6 " water " t 
 
 u Ql , 
 
 0" = " C 
 
 0" = T" 
 We may then put approximately 
 
 T T" = (T' T") e~7, 
 where c is a constant. But 
 
 T' T" TIT 
 
 hence t 
 
 0' = (d 6') (1 fi-^T). 
 
 To find c we have 
 
 1 0" tf 
 
 where 0" can be estimated sufficiently accurately to find C' approxi- 
 mately. 
 
 These formulae apply when there is no radiation. "When radiation 
 takes place, we may write, therefore, when t is not too small, 
 
 where C is a coefficient of radiation, and t is a quantity which must be 
 subtracted from t, as the temperature of the calorimeter does not 
 rise instantaneously. To estimate < , T a being the temperature of the 
 air, we have, according to Newton's law of cooling, 
 
 C(t < ) = jr=jr f( d ~ Ta) dt nearly ' 
 
 6" 6' 
 
 - D y> 
 
 where it is to be noted that _ , is nearly a constant for all values 
 of 0" T a according to Newton's law of cooling.
 
 134 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 The temperature reaches a maximum nearly at the time 
 
 and if O m is the maximum temperature, we have the value of 6" as 
 
 follows : 
 
 6" = T" = 6 m + (t m + c * ); 
 
 and this is the final temperature provided there was no loss of heat. 
 
 When the final temperature of the water is nearly equal to that 
 of the air, G will be small, but the time t m of reaching the maximum 
 will be great. If a is a constant, we can put G = a (0" T a ), and 
 O(t m _j- c t ) will be a minimum, when 
 
 That is, the temperature of the air must be lower than the tempera- 
 ture of the water, so that T a = 0" as nearly as possible ; but the for- 
 mula shows that this method makes the corrections greater than if we 
 make T a = 6', the reason being that the maximum temperature is 
 not reached until after an infinite time. It will in practice, however, 
 be found best to make the temperature of the water at the beginning 
 about that of the air. It is by far the best and easiest method to 
 make all the corrections graphically, and I have constructed the fol- 
 lowing graphical method from the formulae. 
 
 First make a series of measurements of the temperature of the 
 water of the calorimeter, before and after the basket is dipped, together 
 with the times. Then plot them on a piece of paper as in Fig. 5, 
 making the scale sufficiently large to insure accuracy. Five or ten 
 centimeters to a degree are sufficient. 
 
 n abed is the plot of the. temperature of the water of the calo- 
 rimeter, the time being indicated by the horizontal line. Continue 
 the line dc until it meets the line la. Draw a horizontal line 
 through the point /. At any point, b, of the curve, draw a tangent 
 and also a vertical line b g ; the distance e g will be nearly the value of 
 the constant c in the formulas. Lay off //equal to c, and draw the 
 line fh k through the point h, which indicates the temperature of the 
 atmosphere or of the vessel surrounding the calorimeter. Draw a 
 vertical line, j k, through the point k. From the point of maximum,
 
 OF ARTS AND SCIENCES. 
 
 135 
 
 c, draw a line, j c, parallel to dm, and where it meets kj will be the 
 required point, and will give the value of d". Hence, the rise of tem- 
 perature, corrected for all errors, will be kj. 
 
 This method, of course, only applies to cases where the final tem- 
 perature of the calorimeter is^ greater than that of the air ; otherwise 
 there will be no maximum. 
 
 In practice, the line dm is not straight, but becomes more and more 
 nearly parallel to the base line. This is partly due to the constant 
 decrease of the difference of temperature between the calorimeter and 
 the air, but is too great for that to account for it. I have traced it to 
 the thin metal jacket surrounding the calorimeter, and I must con- 
 demn, in the strongest possible manner, all such arrangements of calo- 
 rimeters as have such a thin metal jacket around them. The jacket is 
 
 of an uncertain temperature, between that of the calorimeter and the 
 air. When the calorimeter changes in temperature, the jacket follows 
 it, but only after some time ; hence, the heat lost in radiation is uncer- 
 tain. The true method is to have a water jacket of constant tempera- 
 ture, and then the rate of decrease of temperature will be nearly 
 constant for a long time. 
 
 The following results have been obtained by Mr. Jacques, Fellow 
 of the University, though the first was obtained by myself. Correc- 
 tions' were, of course, made for the amount of thermometer stem in 
 the air. 
 
 Temperature. Mean Specific Heat. 
 
 24 to 100 .0915 
 
 26 " 100 .0915 
 
 25 " 100 .0896 ,-' 
 
 13 " 39 .0895 
 
 14 " 38 .0885 
 
 9 " 41 .0910
 
 136 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 To reduce these to the mean temperature of to 40, I have used 
 the rate of increase foundry B^de for copper. They then become., . 
 for the mean from to 40, 
 
 .0897 
 
 .0897 . 
 
 .0878 
 
 .0893 
 
 .0883 
 
 .0906 
 Mean .0892 dc .00027 
 
 As the capacity of the calorimeter is about four per cent of that of 
 the total capacity, including the water, this probable error is about 
 J fo T of the total capacity, and may thus be considered as satisfactory. 
 
 I have also computed the mean specific heat as follows, from other 
 observers : 
 
 Copper between 20 and 100 nearly. 
 
 .0949 Dulong. 
 
 .0935 ) ^ 
 >0952 }Regnault. 
 
 .0933 Be-de. 
 .0930 Kopp. 
 ' .0940 
 
 This reduced to between and 40 by Bede's formula gives .0922. 
 Hence we have the following for the calorimeter: * 
 
 Per cent. Specific Heat between and 40 0. 
 
 Copper 91.4 .0922 
 
 Zinc .7 .0896 
 
 Tin 3.6 .0550 
 
 Lead 2.7 .0310 
 
 Steel 1.6 .1110 
 
 Mean .0895 
 
 The close agreement of this number with the experimental result 
 can only be accidental, as the reduction to the air thermometer would 
 decrease it somewhat, and so make it even lower than mine. How- 
 
 * The cast brass was composed of 28 parts of copper, 2 of tin, 1 of zinc, and 
 1 of lead. The rolled brass was assumed to have the same composition. The 
 solder was assumed to be made of equal parts of tin and lead.
 
 OF ARTS AND SCIENCES. 137 
 
 ever, the difference could not at most amount to more than 0.5 per 
 cent, which is very satisfactory. 
 
 The total capacity of the calorimeter is reckoned as follows : 
 
 Weight of calorimeter 3.8712 kilogrammes. 
 
 " screws ' .0016 " 
 
 " part of suspending wires .0052 " 
 
 Total weight 3.8780 " 
 
 Capacity = 3.878 X -0892 = .3459 kilogrammes. 
 
 To this must be added the capacity of the thermometer bulb and sev- 
 eral inches of the stem, and of a tube used as a safety valve, and we 
 must subtract the capacity of a part of the shaft which was joined to 
 the shaft turning the paddles. Hence, 
 
 .3459 
 
 + .0011 " 
 -f .0010 
 
 - .0010 Jtyr 
 
 Capacity = .3470 
 
 As this is only about four per cent of the total capacity, it is not 
 necessary to consider the variation of this quantity with the tempera- 
 ture through the range from to 40 which I have used. 
 
 IV. DETERMINATION OF EQUIVALENT. 
 (a.) Historical Remarks. 
 
 The history of the determination of the mechanical equivalent of 
 heat is that of thermodynamics, and as such it is impossible to give 
 it at length here. 
 
 I shall simply refer to the few experiments which a priori seem to 
 possess the greatest value, and which have been made rather for the 
 determination of the quantity than for the illustration of a method, 
 and shall criticise them to the best of my ability, to find, if possible, 
 the cause of the great discrepancies. 
 
 1. GEXERAL REVIEW OF METHODS. 
 
 Whenever heat and mechanical energy are converted the one into 
 the other, we are able by measuring the amounts of each to obtain 
 the ratio. Every equation of thermodynamics proper is an equation
 
 138 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 between mechanical energy and heat, and so should be able to give us 
 the mechanical equivalent. Besides this, we are able to measure a 
 certain amount of electrical energy in both mechanical and heat units, 
 and thus to also get the ratio. Chemical energy can be measured in 
 heat units, and can also be made to produce an electric current of 
 known mechanical energy. Indeed, we may sum up as follows the 
 different kinds of energy whose conversion into one another may fur- 
 nish us with the mechanical equivalent of heat. And the problem in 
 general would be the ratio by which each kind of energy may be con- 
 verted into each of the others, or into mechanical or absolute units. 
 
 a. Mechanical energy. 
 
 b. Heat. 
 
 c. Electrical energy. 
 
 d. Magnetic " 
 
 e. Gravitation u 
 /. Radiant " 
 g. Chemical " 
 h. Capillary " 
 
 Of these different kinds of energy, only the first five can be meas- 
 ured other than by their conversion into other forms of energy, 
 although Sir William Thomson, by the introduction of such terms as 
 " cubic mile of sunlight," has made some progress in the case of radia- 
 tion. Hence for these five only can the ratio be known. 
 
 Mechanical energy is measured by the force multiplied by the dis- 
 tance through which the force acts, and also by the mass of a body 
 multiplied by half the square of its velocity. Heat is usually referred 
 to the quantity required to raise a certain amount of water so many 
 degrees, though hitherto the temperature of the water and the reduc- 
 tion to the air thermometer have been almost neglected. 
 
 The energy of electricity at rest is the quantity multiplied by half 
 the potential ; or of a current, it is the strength of current multiplied 
 by the electro-motive force, and by the time; or for all attractive 
 forces varying inversely as the square of the distance, Sir William 
 Thomson has given the expression 
 
 where It is the resultant force at any point in space, and the integral 
 is taken throughout space. 
 
 These last three kinds of energy are already measured in absolute
 
 OP ARTS AND SCIENCES. 139 
 
 measure, and hence their ratios are accurately known. The only 
 ratio, then, that remains is that of heat to one of the others, and this 
 must be determined by experiment alone. 
 
 But although we cannot measure /, g, h in general, yet we can 
 often measure off equal amounts of energy of these kinds. Thus, 
 although we cannot predict what quantities of heat are produced when 
 two atoms of different substances unite, yet, when the same quantities 
 of the same substances unite to produce the same compound, we are 
 safe in assuming that the same quantity of chemical energy comes 
 into play. 
 
 According to these principles, I have divided the methods into direct 
 and indirect. 
 
 Direct methods are those where b is converted directly or indirectly 
 into a, c, d, or e, or vice versa. 
 
 Indirect methods are those where some kind of energy, as g, is con- 
 verted into 6, and also into a, c, d, or e. 
 
 In this classification I have made the arrangement with respect to 
 the kinds of energy which are measured, and not to the intermediate 
 steps. Thus Joule's method with the magneto-electric machine would 
 be classed as mechanical energy into heat, although it is first con- 
 verted into electrical energy. The table does not pretend to be com- 
 plete, but gives, as it were, a bird's-eye view of the subject. It could 
 be extended by including more complicated transformations ; and, in- 
 deed, the symmetrical form in which it is placed suggests many other 
 transformations. As it stands, however, it includes all methods so far 
 used, besides many more. 
 
 In the table of indirect methods, the kind of energy mentioned first 
 is to be eliminated from the result by measuring it both in terms of 
 heat and one of the other kinds of energy, whose value is known in 
 absolute or mechanical units. 
 
 It is to be noted that, although it is theoretically possible to measure 
 magnetic energy in absolute units, yet it cannot be done practically 
 with any great accuracy, and is thus useless in the determination of 
 the equivalent. It could be thus left out from the direct methods 
 without harm, as also out of the next to last term in the indirect 
 methods.
 
 140 
 
 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 TABLE XXV. SYNOPSIS OP METHODS FOB OBTAINING THE MECHANICAL 
 ' EQUIVALENT OF HEAT. 
 
 
 re. Expansion or compression ac- 
 cording to adiabatic curve. 
 
 
 b. Expansion or compression ac 
 
 
 cording to isothermal curve. 
 
 
 1. Reversible Process 
 
 c. Expansion or compression ac- 
 cording to any curve with re- 
 
 
 
 generator. 
 
 f-Heat| M -So a n E " rgy - 
 
 
 d. Electro-magnetic engine driven 
 by thermo-electric pile in a cir- 
 
 
 
 
 cuit of no resistance. 
 
 
 
 ( n. Friction, percussion, etc. 
 
 
 * 
 
 2. Irreversible Process \ b. Heat from magneto-electric cur- 
 
 
 ( rents, or electric machine. 
 
 f 
 
 
 a Thermo-electric currents 
 
 
 
 {1. Reversible Process 
 
 b, Pyro-electric phenomena (prob- 
 
 2 
 
 
 ably). 
 a. Heating of wire by current, or 
 
 <4 
 
 2. Irreversible Process 
 
 heat produced by discharge of 
 electric battery. 
 
 I y. Heat, Magnetic Energy 
 
 1. Reversible Process 
 
 o. Thermo-electric current magnet- 
 izing a magnet in a circuit of 
 
 {a. Mechanical Energy. 
 b. Electrical " 
 M * 
 c. Magnetic 
 d. Gravitation 
 
 Crooke's radiometer. 
 Thermo-electric pile. 
 Thermo-electric pile with electro- 
 magnet in circuit. 
 
 
 
 a Mechanical Energy 
 
 ( 1. Cannon. 
 < 2. Electro-magnetic machine run by 
 
 
 
 
 ( galv. battery.. 
 
 . Chemical Energy, Heat . . 
 (Combustion, etc.) 
 
 b. Electrical " 
 c. Magnetic 
 
 Current from battery. 
 ( Electro-magnet, magnetized by a 
 I battery current. 
 
 
 
 d. Gravitation " 
 
 
 
 
 a. Mechanical Energy. 
 
 Movement of liquid by capillarity. 
 
 y. Capillary energy, Heat . . . 
 
 6. Electrical 
 
 {Electrical currents from capillary 
 action at surface of mercury. 
 
 
 (Heat produced when a liq- 
 
 c Magnetic " 
 
 
 
 uid is absorbed by a po- 
 
 it. Gravitation " 
 
 Raising of liquid by capillarity. 
 
 
 rous solid.) 
 
 
 
 
 
 (a. Mechanical Energy 
 
 {Magneto-electric or electro-magnetic 
 machine. Electric attraction. 
 
 S. Electrical energy, Heat . . ( 
 (Heat generated in a wire by 
 
 6. Magnetic " 
 c. Gravitation " 
 
 Electro-magnet. 
 
 
 an electrical current.) 
 
 
 
 
 
 ia. Mechanical Energy 
 
 {Armature attracted by a permanent 
 magnet. 
 
 t 
 
 Magnetic Energy, Heat . . 
 (Heat generated on demag- 
 
 6. Electrical " 
 
 {Induced current on demagnetizing 
 a magnet. 
 
 
 netizing a magnet.) 
 
 c. Gravitation " 
 
 
 (. Gravitation Energy, Heat . 
 (Heat generated by a fall- 
 
 a. Mechanical Energy. 
 6. Electrical 
 c. Magnetic " 
 
 Velocity imparted to a falling body. 
 
 
 ing.body.) 
 
 
 
 2. RESULTS OF BEST DETERMINATIONS. 
 
 On the basis of this table of methods I have arranged the following 
 table, showing the principal results so far obtained. 
 
 In giving the indirect results, many persons have only measured 
 one of the transformations required ; and as it would lengthen out the
 
 OP ARTS AND SCIENCES. 
 
 141 
 
 TABLE XXVL-H,.*,., 
 
 OF Exr ERIMENTAL RE8ULm 
 
 Me 
 0< 
 
 thod in 
 neral. Method in Particular. 
 
 . 
 
 Observe 
 
 Date 
 
 Result. 
 
 
 Compression of air 
 Expansion " . ! . 
 
 Joule 2 
 
 184 
 
 443.8 
 
 
 Theory of gases (see below) . 
 
 Joule 2 
 
 184 
 
 437.8 
 
 
 ' vapors (see below) . 
 
 
 
 
 
 Experiments on steam-engine 
 
 Him 7 
 
 185 
 
 413.0 
 
 
 
 Him 7 
 
 186 
 
 420-432 
 
 
 Expansion and contraction of meta 
 
 Edlund 8 
 
 1865 
 
 443.6 
 4301 
 
 
 
 
 
 428.3 
 
 
 Boring of cannon .... 
 Friction of water in tubea . 
 in calorimeter 
 in calorimeter 
 . . in calorimeter 
 Junction of mercury in calonmete 
 plates of iron . . . 
 
 Rumforc 
 Joule 3 
 Joule* 
 Joule 5 
 Joule 6 
 Joule 6 
 Joule 6 
 
 1798 
 1843 
 1845 
 1847 
 1850 
 1850 
 1850 
 
 940ft.lbs. 
 424.6 
 488.3 
 428.9 
 423.9 
 424.7 
 4252 
 
 
 metals 
 metals in mercury calo 
 
 Him 7 
 Favre 9 
 
 1857 
 1858 
 
 371.6 
 413.2 
 
 
 metals .... 
 Joring of metals 
 Vater in balance a frottement . '. 
 
 Flow of liquids under strong pressur 
 Crushing of lead . 
 
 Him 7 
 Him 7 
 Him? 
 Him 7 
 
 TT- _T 
 
 1858 
 1858 
 1860 
 1860- 
 
 400-450 
 425.0 
 4320 
 432.0 
 
 
 Friction of metals . 
 
 -TUttl' 
 
 1860- 
 
 425.0 
 
 
 Water in calorimeter .... 
 
 Joule 
 
 1876 
 1878 
 
 426.6 
 423.9 
 
 a 
 
 Heating by magneto electric cur- ) 
 
 
 
 
 
 rents . .' t 
 
 Joule 8 
 
 1843 
 
 460.0 
 
 
 Heat generated in a disc between ) 
 the poles of a magnet . . . j 
 
 vww. 
 
 1870 
 
 435~.2 
 434.9 
 435.8 
 
 
 
 
 
 437.4 
 
 A ft 
 
 Heat developed in wire of known ( 
 absolute resistance . . . . " 
 
 Quintus 
 
 1857 
 
 399.7 
 
 
 ( 
 
 so Weber 
 
 
 
 
 D . do. do. J 
 
 .<enz, also 
 
 1859 . 
 
 396.4 
 
 
 _ | 
 
 Weber 
 
 
 478.2 
 
 ft 
 
 Do. do. do. 
 Do. do. do. 
 Diminishing of the heat produced ) 
 
 Joule 13 
 F.Weber 1 * 
 
 1867 
 1878 
 
 29.5 
 28.15 
 
 
 in a battery circuit when the > 
 current produces work . . ) 
 
 Joule 3 
 
 1843 
 
 99.0 
 
 
 Do. do. do. 
 
 Favre 15 
 
 1858 
 
 43.0 
 
 
 Heat due to electrical current, ] 
 
 
 
 
 
 electro-chemical equivalent of 
 
 Weber, 
 
 
 
 
 water = .009379, absolute re- 1 
 sistance electro-motive force of f 
 Daniell cell, heat developed by 
 action of zinc on sul. of copper I 
 
 Boscha, 
 ivre, and 
 jermann 
 
 857 
 
 32.1 
 
 
 eat developed in Daniell cell . . 
 Electro-motive force of Daniell cell 
 
 Joule 
 oscha 12 
 
 859 
 
 19.5
 
 142 
 
 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 table very much to give the complete calculation of the equivalent 
 from these selected two by two, I have sometimes given tables of these 
 parts. As the labor of looking up and reducing these is very great, 
 it is very possible that there have been some omissions. 
 
 I have taken the table published by the Physical Society of Berlin, 1 
 as the basis "down to 1857, though many changes have been made 
 even within this limit. 
 
 I shall now lake up some of the principal methods, and discuss them 
 somewhat in detail. 
 
 Method from Theory of Gases. 
 
 As the different constants used in this method have been obtained 
 by many observers, I shall first give their results. 
 
 TABLE XXVII. SPECIFIC HEAT OF GASES. 
 
 
 Limit to 
 Temperature. 
 
 Approximate 
 Temperature 
 of Water 
 
 Temperature 
 reduced to 
 
 Specific Heat. 
 
 
 Air 
 
 
 | 
 
 Mercurial 
 
 1 2669 I 
 
 Delaroche and 
 
 
 20 to 210 
 
 " I 
 
 *14.2 \ 
 ( 
 
 Thermometer 
 
 Air 
 Thermometer 
 
 r i 
 
 | .23751" 
 
 Be"rard. 
 Regnault. 
 
 
 20 to 100 
 
 20 { 
 
 Mercurial 
 Thermometer 
 
 I .2389" 
 
 E.Wiedemann. 
 
 Hydrogen 
 
 
 | 
 
 Mercurial 
 
 I 3 2936 \ 
 
 Delaroche and 
 
 
 15 to 200 
 
 ( 
 12.2 | 
 
 Thermometer 
 
 Air 
 Thermometer 
 
 j 3.4090' 6 
 
 Be'rard. 
 Regnault. 
 
 
 21 to 100 
 
 21 j 
 
 Mercurial 
 Thermometer 
 
 | 3.410" 
 
 E.Wiedemann. 
 
 TABLE XXVIII. COEFFICIENT OF EXPANSION OF Am UNDER CONSTANT 
 VOLUME. 
 
 
 Taking Expansion of Mercury 
 according to Regnault. 
 
 Taking Expansion of Mercury 
 according to Wiillner's Re- 
 calculation of Ragnault's Ex- 
 periments. 
 
 Regnault ..... 
 Magnus .... 
 Jolly 
 Rowland 
 
 .0036655 
 .0036678 
 .0036695 
 .0036676 
 
 .0036687 
 .0036710 
 .0036727 
 .0036707 
 
 Mean 
 
 .0036676 
 
 .0036708 
 
 * Taking mean of results on page 101 of Rel. des Exp., torn. ii.
 
 OP ARTS AND SCIENCES. 143 
 
 TABLE XXIX. RATIO OF SPECIFIC HEATS OF AIB. 
 
 Method. 
 
 Observer. 
 
 Date. 
 
 Ratio 
 of Specific 
 Heats. 
 
 Method of Clement & De'- ) 
 sormes, globe 20 litres . . ) 
 
 Never fully published . . . 
 Method of Clement & De'- I 
 sormes J 
 Using Breguet thermometer . 
 Cle'ment & Desormes, globe ) 
 
 Cle'ment & D^sormes' 8 ) 
 
 Gay-Lussac& Welter 19 . 
 Delaroche & Be'rard*' . 
 Favre & Silbermann 23 . 
 Masson 20 
 
 1812 
 Publishe'd in 
 1819 
 
 1853 
 
 1858 
 
 ( 1.354 
 
 1.3748 
 1.249 
 1.421 
 1.4196 
 
 Clement & De'sormes .... 
 Clement & Desormes, globe ) 
 
 Weisbach.2i 
 Him 22 
 
 1859 
 1861 
 
 1.4025 
 1.3845 
 
 Passage of gas from one ves- ) 
 
 Cazin 2 * 
 
 1862 
 
 141 
 
 litres ) 
 
 
 
 
 Pressure in globe changed by ) 
 aspirator, globe 25 litres . J 
 Heating of gas by electric \ 
 
 Dupre' 25 
 Jamin & Richard 28 . . 
 
 1863 
 1864 
 
 1.41 
 
 Clement & De'sormes . . . 
 Barometer under air-pump / 
 receiver of 6 litres . . . ) 
 Compression and expansion ) 
 
 Tresca et Laboulaye 29 . 
 Kohlrausch 26 .... 
 
 1864 
 1869 
 
 1871 < 
 
 1.302 
 Results lost 
 
 of gas by piston . . . . J 
 Clement & Desormes with ) 
 
 Rontgen 27 . . , 
 
 1878 
 
 of Paris. 
 14053 
 
 70 litres ) 
 
 
 
 
 Compression of gas by piston . 
 
 Amagat 30 
 
 1874 
 
 1.397 
 
 References. (Tables XXVI. to XXX.) 
 
 1 Physical Society of Berlin, Fort, der Phys., 1858. 
 
 2 Joule, Phil. Mag., ser. 3, vol. xxvi. See also Mec. Warmeaquivalent, 
 Gesammelte Abhandlungen von J. P. Joule, Braunschweig, 1872. 
 
 3 Joule, Phil. Mag., ser. 3, vol. xxiii. See also 2 above. 
 * " " " " " xxvi. 
 
 6 xxvji. 
 
 6 " " " xxxi. 
 
 7 Him, The"orie Me'c. de la Chaleur, ser. 1, 3 me ed. 
 
 8 Edlund, Pogg. Ann., cxiv. 1, 1865. 
 
 9 Favre, Comptes Rend., Feb. 15, 1858; also Phil. Mag., xv. 406. 
 i Violle, Ann. de Chim., ser. 4, xxii. 64. 
 
 11 Quintus Icilius, Pogg. Ann., ci. 69. 
 
 12 Boscha, Pogg. Ann., cviii. 162. 
 
 13 Joule, Report of the Committee on Electrical Standards of the B. A., Lon- 
 don, 1873, p. 175. 
 
 i* H. F. Weber, Phil. Mag., ser. 5, v. 30. 
 
 15 Favre, Comptes Rend., xlvii. 599. 
 
 16 Regnault, Rel. des Expe'riences, torn. ii. 
 E Wiedemann, Pogg. Ann., clvii. 1.
 
 144 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 
 *!.,- 
 
 CO t- 
 
 CO 
 
 NOD o ; .' o 
 
 : 
 
 
 
 !JB! 
 
 
 S S 
 CO CO 
 
 CO CO [ _' CO 
 
 Si : 
 
 
 
 
 
 
 
 
 
 
 ff!| 
 
 | j 
 
 
 CO CO 
 
 ; g 
 
 S 
 
 
 ^2 
 
 
 
 ea> -< 
 
 
 
 1 
 
 Velocity 
 reduced to 
 0and 
 Ordinary Air. 
 
 a s 
 
 OJ t- 
 
 s<i co 
 
 cS co 
 
 t333.0cm. 
 t329.36 m. 
 
 s s 
 
 CO S S S 
 
 CO CO CO CO CO 
 CO CO CO CO C0_ 
 
 c<i co 
 
 ^ Si 
 
 
 
 vi 
 
 H 
 
 ~ 
 
 S ; 
 
 
 ; 
 
 
 
 | 
 
 
 
 <N 
 
 , ! ? "? I 
 
 iq q 
 
 
 h 
 H 
 
 H 
 
 
 | 
 
 I'll' 
 
 I? Si 
 
 
 > 
 
 2 . 
 
 2-2 
 
 d : 
 
 lO 
 
 fc |* 
 
 o S g + 
 
 
 & 
 
 c3 
 
 
 
 t. 
 O 
 
 Tempera 
 Observ 
 
 6 to 7 
 
 o^ 2 
 
 00 * 
 
 o? ^ 2 P o 
 
 -^-1 
 
 2 5 
 
 2 " 
 
 2 
 
 D 
 
 *t 
 
 
 
 5 
 
 < 
 
 
 
 
 
 
 
 
 
 > 
 
 | J3.2 
 
 
 IH 
 
 
 
 "- 
 
 J 
 
 s 
 
 
 
 5 ^ 
 
 
 
 cw 
 
 
 M 
 
 
 s 
 
 
 
 y 
 
 
 O 
 
 
 s 
 
 
 
 iS 
 
 1 
 
 1 
 
 11 
 
 .2 .2 
 
 'O 13 
 
 c c 
 
 8 -S 1 S eq 
 
 I III s 
 
 PC, < K a P* 
 
 ' 
 < 
 
 
 
 % 
 
 1 
 
 1 S 
 
 i g 
 
 <N C, CO =0 3 
 
 1 S 
 
 s 
 
 H 
 
 
 
 
 
 
 
 hJ 
 
 
 
 
 
 
 
 - 
 
 
 
 
 s 
 
 
 
 j 
 
 
 
 
 c 
 
 
 
 H 
 
 
 
 
 s S 
 
 
 
 
 
 
 
 
 1 1 : 
 
 . s 
 
 
 
 
 
 I 4 . 
 
 T3 w 
 fcP 
 
 < 2 
 
 1 
 
 1 s i 
 
 o > B 
 
 >-> og a - S 
 'o u > PH 
 
 . S 
 S 
 =a 
 
 ; 
 
 
 
 
 bo 
 
 s *" og 
 
 1 -a 
 
 3 
 
 
 
 ~- c 
 (S 
 
 1 
 
 3 2 "0 S 
 
 B co S " (2 
 
 ( B 
 
 1 
 
 M 
 
 
 1 
 
 
 ^-v-w 
 
 N-*^-^X 
 
 
 
 
 a 
 
 rH (M 
 
 CO 
 
 ^ CD t- 
 
 QO C> 
 
 
 
 
 S 
 
 
 
 

 
 OP ARTS AND SCIENCES. 145 
 
 Estimating th< 
 follows: 
 
 3 weight 
 
 No. 
 
 1 
 
 rather arbitrarily, 
 
 Velocity at 0. C. 
 Dry Air. 
 
 332.6 
 
 I have combined them as 
 
 Estimated Weight 
 of Observation. 
 
 2 
 
 
 2 
 
 332.7 
 
 2 
 
 
 3 
 
 330.9 
 
 2 
 
 
 4 
 
 330.8 
 
 4 
 
 
 5 
 
 332.5 
 
 3 
 
 
 6 
 
 332.8 
 
 7 
 
 
 7 
 
 332.0 
 
 1 
 
 
 8 
 
 331.8 
 
 1 
 
 
 9 
 
 332.4 
 
 4 
 
 
 10 
 
 330.7 
 
 10 
 
 Mean 331.75 
 
 Or, corrected for the normal carbonic acid in the atmosphere, it be- 
 comes 331.78 meters per second in dry pure air at C. 
 
 18 Clement et Desdrmes, Journal de Physique, Ixxxix. 333, 1819. 
 is Laplace, Mec. Celeste, v. 125. 
 
 20 Masson, Ann. de China, et de Phys., ser. 3, torn. liii. 
 
 21 Weisbach, Der Civilingenieur, Neue Folge, Bd. v., 1859. 
 
 22 Him, The'orie Mec. de la Chaleur, i. 111. 
 
 23 Favre et Silbermann, Ann. de Chim., ser. 3, xxxvii. 1851. 
 z * Cazin, Ann. de Chim., ser. 3, torn. Ixvi. 
 
 25 Dupre, Ann. de Chim., 3 me ser., Ixvii. 359, 1863. 
 
 26 Kohlrausch, Pogg. Ann., cxxxvi. 618. 
 Rontgen, Pogg. Ann., cxlviii. 603. 
 
 28 Jamin and Richard, Comptes Rend., Ixxi. 336. 
 
 29 Tresca and Laboulaye, Comptes Rend., Iviii. 358. Ann. du Conserv. des 
 Arts et Me'tiers, vi. 365. 
 
 80 Amagat, Comptes Rend., Ixxvii. 1326. 
 Me'm. de 1'Acad. des Sei., 1738, p. 128. 
 82 Benzenberg, Gilbert's Annalen, xlii. 1. 
 38 Goldingham, Phil. Trans., 1823, p. 96. 
 
 8 * Ann. de Chim., 1822, xx. 210 ; also, CEuvres de Arago, Me'm. Sci., ii. 1. 
 85 Stampfer and Von Myrbach, Pogg. Ann., v. 496. 
 
 S6 Moll and Van Beek, Phil. Trans., 1824, p. 424. See also Shroder van der 
 Kolk, Phil. Mag., 1865. 
 
 37 Parry and Foster, Journal of the Third Voyage, 1824-5, Appendix, p. 86. 
 Phil. Trans., 1828, p. 97. 
 
 38 Savart, Ann. de Chim., ser. 2, Ixxi. 20. Recalculated. 
 89 Bravais and Martins, Ann. de Chim., ser. 3, xiii. 6. 
 
 Regnault, Rel. des Exp., iii. 533. 
 
 41 Delaroche and Berard, Ann. de Chim., Ixxxv. 72 and 113. 
 
 42 Puluj, Pogg. Ann., clvii. 656. 
 
 VOL. xv. (N. 8. vn.) 10
 
 146 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 From Regnault's experiments on the velocity in pipes I find by 
 graphical means 331. 4 m - in free air, which is very similar to the above. 
 
 Calculation from Properties of Gases. 
 
 K= specific heat of gas at constant pressure. 
 
 k = " " " volume. 
 
 p = pressure in absolute units of a unit of mass. 
 
 v = volume " " " " 
 
 p = absolute temperature. 
 
 J = Joule's equivalent in absolute measure. 
 
 y ~- 
 7 ~~k' 
 
 General formula for all bodies : 
 1 
 
 7 = 
 
 JK \dft) v \dn) p 
 
 . J= L _ JL y 
 K * 
 
 /dvy 
 
 Also, J= ^- \ dfl ' p 
 
 K d 
 
 - 
 \d P V 2 
 
 Application to gases; Rankine's formula is, _ 
 
 If a, is the coefficient of expansion between and 100, then 
 
 Mo = (l-|- .00635m); 
 
 r PH , . / 7 > 
 
 ^ F a " a "(^i,
 
 OF ARTS AND SCIENCES. 147 
 
 where a' p and a' p are the true coefficients of expansion at the given 
 temperature ; 
 
 /= 
 
 According to Thomson and Joule's experiments m = 0.33 C. for 
 air and about 2.0 for CO 2 . Hence /i - 272.99. 
 
 The equations should be applied to the observations directly at the 
 given temperature, but it will generally be sufficient to use them after 
 reduction to C. Using K= .2375 according to Regnault for air, 
 we have for the latitude of Baltimore, 
 
 From Rb'ntgen's value y = 1.4053 
 
 " Amagat's " 1.397 
 
 9 
 
 " velocity of sound 33 US- per sec. = 429.6. 
 Using Wiedemann's value for K, .2389, these become 
 
 = 427.8 ; = 434.0 ; L = 427.1. 
 
 9 y 9 
 
 As Wiedemann, however, used the mercurial thermometer, and as 
 the reduction to the air thermometer would increase these figures 
 from .2 to .8 per cent., it is evident that Regnault's value for K is 
 the more nearly correct. I take the weights rather arbitrarily as 
 follows : 
 
 Weight. J. 
 
 Rontgen 3 430.3 
 
 Amagat 1 436.6 
 
 Velocity of sound 4 429.6 
 
 Mean 430.7 
 
 And this is of course the value referred to water at 1 4 C. and in the 
 latitude of Baltimore. My value at this point is 427.7. 
 
 * Rontgen gives the value 428.1 for the latitude of Paris as calculated by a 
 formula of Shroder v. d. Kolk, and 427.3 from the formula for a perfect gas, 
 and these both agree more nearly with my result than that calculated from my 
 own formula.
 
 148 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 This determination of the mechanical equivalent from the proper- 
 ties of air is at most very imperfect, as a very slight change in either 
 y or the velocity of sound- will produce a great change in the mechan- 
 ical equivalent. 
 
 From Theory of Vapors. 
 
 Another important method of calculating the mechanical equivalent 
 of heat is from the equation for a body at its change of state, as for 
 instance in vaporization. Let v be the volume of the vapor, and 
 v 1 the volume of the liquid, and H the heat required to vaporize a 
 unit of mass of the water ; also let p be the pressure in absolute 
 units, and ^ the absolute temperature. Then 
 JH 
 
 The quantity Hand, the relation of p to p. have been determined with 
 considerable accuracy by Regnault. To determine J it is only 
 required to measure the volume of saturated steam from a given 
 weight of water ; and the principal difficulty of the process lies in 
 this determination, though the other quantities are also difficult of 
 determination. 
 
 This volume can be calculated from the density of the vapor, but 
 this is generally taken in the superheated -state. 
 
 The experiments of Fairbairn and Tate * are probably the best 
 direct experiments on the density of saturated vapor, but even those 
 do not pretend to a greater accuracy than about 1 in 100. With 
 Regnault's values of the other quantities, they give about Joule's value 
 for the equivalent, namely 425. Him, Herwig, and others have 
 also made the determination, but the results do not agree very well. 
 Herwig even used a Giessler standard thermometer, which I have 
 shown to depart very much from the air thermometer. 
 
 Indeed, the experiments on this subject are so uncertain, that 
 physicists have about concluded to use this method rather for the de- 
 termination of the volume of saturated vapors than for the mechanical 
 equivalent of heat. 
 
 From the Steam-Engine and Expansion of Metals. 
 The experiments of Him on the steam-engine and of Edlund on 
 the expansion and contraction of metals, are very excellent as illustrat- 
 
 * Phil. Mag., ser. 4, xxi. 230.
 
 OP ARTS AND SCIENCES. 149 
 
 ing the theory of the subject, but cannot have ajiy weight as accurate 
 determinations of the equivalent. 
 
 From Friction Experiments. 
 
 Experiments of this nature, that is, irreversible processes for con- 
 verting mechanical energy into heat, give by far the best methods for 
 the determination of the equivalent. 
 
 Rumford's experiment of 1798 is only valuable from an historical 
 point of view. Joule's results since .1843 undoubtedly give the best 
 data we yet have for the determination of the equivalent. The mean 
 of all his friction experiments of 1847 and 1850 which are given in 
 the table is 425.8, though he prefers the smallest number, 423.9, of 
 1850. This last number is at present accepted throughout the civil- 
 ized world, though there is at present a tendency to consider the 
 number too small. But this value and his recent result of 1878 have 
 undoubtedly as much weight as all other results put together. 
 
 As sources of error in these determinations I would suggest, first, 
 the use of the mercurial instead of the air thermometer. Joule com- 
 pared his thermometers with one made by Fastre. In the Appendix 
 to Thermometry I give the comparison of two thermometers made by 
 Fastre in 1850, with the air thermometer, as well as of a large number 
 of others. From this it seems that all thermometers as far as measured 
 stand above the air thermometer between and 1 00, and that the aver- 
 age for the Fastre* at 40 is about 0.l C. Using the formula given 
 in Thermometry this would produce an error of about 3 parts in 1,000 
 at 15 C., the temperature Joule used. 
 
 The specific heat of copper which Joule uses, namely, .09515, is 
 undoubtedly too large. Using the value deduced from more recent 
 experiments in calculating the capacity of my calorimeter, .0922, 
 Joule's number would again be increased 13 parts in 10,000, so that 
 we have, 
 
 Joule's value 423.9, water at 15.7 C. 
 
 Reduction to air thermometer . . -{-1.3 
 
 Correction for specific heat of copper -f- .5 
 
 " to latitude of Baltimore -\- .5 
 
 426.2 
 
 It does not seem improbable that this should be still further in- 
 creased, seeing that the reduction to the air thermometer is the small- 
 est admissible, as most other thermometers which I have measured 
 give greater correction, and some even more than three times as great
 
 150 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 as the one here used, .and would thus bring the value even as high 
 as 429. 
 
 One very serious defect in Joule's experiments is the small range 
 of temperature used, this being only about half a degree Fahrenheit, 
 or about six divisions on, his thermometer. It would seem almost im- 
 possible to calibrate a thermometer so accurately that six divisions 
 should be accurate to one per cent, and it would certainly need a very 
 skilful observer to read to that degree of accuracy. Further, the 
 same thermometer " A " was used throughout the whole experiment 
 with water, and so the error of calibration was hardly eliminated, the 
 temperature of the water being nearly the same. In the experiment 
 on quicksilver another thermometer was used, and he then finds a 
 higher result, 424.7, which, reduced as above, gives 427.0 at Baltimore. 
 
 The experiments on the friction of iron should be probably re- 
 jected on account of the large and uncertain correction for the energy 
 given out in sound. 
 
 The recent experiments of 1878 give a value of 772.55, which re- 
 duced gives at Baltimore 426.2, the same as the other experiment. 
 
 The agreement of these reduced values with my value at the 
 same temperature, namely 427.3, is certainly very remarkable, and 
 shows what an accurate experimenter Joule must be to get with his 
 simple apparatus results so near those from my elaborate apparatus, 
 which almost grinds out accurate results without labor except in re- 
 duction. Indeed, the quantity is the same as I find at about 20 C. 
 
 The experiments of Him of 1860-61 seem to point to a value of 
 the equivalent higher than that found by Joule, but the details of the 
 experiment do not seem to have been published, and they certainly 
 were not reduced to the air thermometer. 
 
 The method used by Violle in 1870 does not seem capable of 
 accuracy, seeing that the heat lost by a disc in rapid rotation, and while 
 carried to the calorimeter, must have been uncertain. 
 
 The experiments of Him are of much interest from the methods 
 used, but can hardly have weight as accurate determinations. Some 
 of the methods will be again referred to when I come to the descrip- 
 tion of apparatus. 
 
 Method by Heat generated by Electric Current. 
 
 The old experiments of Quintus Icilius or Lenz do not have any 
 sxcept historical value, seeing that Weber's measure of absolute 
 resistance was certainly incorrect, and we now have no means of find- 
 ing its error.
 
 OP ARTS AND SCIENCES. 
 
 151 
 
 The theory of the process is as follows. The energy of electricity 
 being the product of the potential by the quantity, the energy ex- 
 pended by forcing the quantity of electricity, Q, along a wire of re- 
 sistance, R, in a second of time, must be Q*R, and as this must equal 
 the mechanical equivalent of the heat generated, we must have 
 JH= Q 2 Rt, where H is the heat generated and t is the time the 
 current Q flows. 
 
 The principal difficulty about the determination by this method 
 seems to be that of finding R in absolute measure. A table of the 
 values of the ohm as obtained by different observers, was published by 
 me in my paper on the " Absolute Unit of Electrical Resistance," 
 in the American Journal of Science, Vol. XV., and I here give it 
 with some changes. 
 
 TABLE XXXI. 
 
 Date. 
 
 Observer. 
 
 Value of 
 Ohm. 
 
 REMARKS. 
 
 1849 
 
 Kirehoff 
 
 .88 to .90 
 
 Approximately. 
 
 1851 
 
 Weber 
 
 .95 to .97 
 
 Approximately. 
 
 1862 
 
 Weber 
 
 j 1.088 
 | 1.075 
 
 From Thomson's unit. 
 From Weber's value of Siemens unit. 
 
 1863-4 
 
 B. A. Committee 
 
 ( 1.0000 
 \ .993 
 
 Mean of all results. 
 Corrected by Rowland to zero velocity 
 
 
 
 
 of coil. 
 
 1870 
 
 Kohlrausch 
 
 1.0193 
 
 
 1873 
 
 Lorenz 
 
 .975 
 
 Approximately. 
 
 1876 
 
 Rowland 
 
 .9911* 
 
 From a preliminary comparison with 
 
 
 
 
 the B. A. unit. 
 
 1878 
 
 H. F. Weber 
 
 1.0014 
 
 Using ratio of Siemens unit to ohm, 
 
 
 
 
 .9536. 
 
 The ratio of the Siemens unit to the ohm is now generally taken at 
 .9536, though previous to 1864 there seems to have been some doubt 
 as to the value of the Siemens unit. 
 
 Since 1863-4, when units of resistance first began to be made with 
 great accuracy, two determinations of fche heat generated have been 
 made. The first by Joule with the ohm, and the second by H. F. 
 Weber, of Zurich, with the Siemens unit. 
 
 Each determination of resistance with each of these experiments 
 gives one value of the mechanical equivalent. As Lorenz's result was 
 only in illustration of a method, I have not included it among the 
 exact determinations. 
 
 The result found by Joule was /= 25187 in absolute measure 
 
 -* Given .9912 by mistake in the other tables.
 
 152 
 
 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 using feet and degrees F., which becomes 429.9 in degrees C. on a 
 mercurial thermometer and in the latitude of Baltimore, compared 
 with water at 18. 6 C. 
 
 TABLE XXXII. EXPERIMENTS OF JOULE. 
 
 Observer. 
 
 Value of 
 B. A. Unit. 
 
 Mechanical Equivalent 
 from Joule's Exp. 
 
 Mechanical Equivalent 
 reduced to Air Ther- 
 mometer and cor- 
 rected for Sp. Ht. of 
 Copper. 
 
 B. A. Committee 
 Ditto corrected by Rowland 
 
 1.0000 
 .993 
 1.0193 
 
 429.9 
 426.9 
 438.2 
 
 431.4 
 
 428.4 
 439.7 
 
 
 .9911 
 
 426.1 
 
 427.6 
 
 H. F. Weber 
 
 1.0014 
 
 430.5 
 
 432.0 
 
 The experiments of H. F. Weber* gave 428.15 in the latitude of 
 Zurich and for 1 C. on the air thermometer and at a temperature of 
 18 C. This reduced to the latitude of Baltimore gives 428.45. 
 
 TABLE XXXIII. 
 
 EXPERIMENTS o 
 
 F H. F WEBEB 
 
 
 Mean of Joule and 
 Weber, giving Joule 
 twice the Weight of 
 Weber. 
 
 Observer. 
 
 Value of 
 B. A. Unit. 
 
 Mechanical Equivalent 
 of Heat from Weber's 
 Experiments. 
 
 Mean Equivalent re- 
 duced toAirThennom- 
 eter In the Latitude of 
 Baltimore. 
 
 B. A. Committee 
 Ditto corrected by Rowland 
 Kohlrausch 
 Rowland 
 
 1.000 
 .993 
 1.0193 
 9911 
 
 427.9 
 424.9 
 436.2 
 424 1 
 
 430.2 
 427.2 
 439.1 
 426 4 
 
 II . F. Weber. . . 
 
 1 0014 
 
 428 5 
 
 401 A 
 
 
 
 
 
 My own value at this temperatuj-e is 426.8, which agrees almost 
 exactly with the fourth value from my own determination of the ab- 
 solute unit.f 
 
 There can be no doubt that Joule's result is most exact, and hence 
 I have given his results twice the weight of Weber's. Weber used a 
 wire of about 14 ohms' resistance, and a small calorimeter holding only 
 250 grammes of water. This wire was apparently placed in the water 
 without any insulating coating, and yet current enough was sent through 
 
 * Phil. Mag, 1878, 5th ser, v. 135. 
 
 t The value of the ohm found by reversing the calculation would be .992, 
 almost exactly my value.
 
 OP ARTS AND SCIENCES. 
 
 153 
 
 it to heat the water 15 during the experiment. No precaution 
 seems to have been taken as to the current passing into the water, 
 which Joule accurately investigated. Again, the water does not seem 
 to have been continuously stirred, which Joule found necessary. And 
 further, Newton's law of cooling does not apply to so great a range 
 as 15, though the error from this source was probably small. Further- 
 more, I know of no platinum which Jias an increase of coefficient of 
 .001054 for 1 C., but it is usually given at about .003. 
 
 There can be no doubt that experiments depending on the heating 
 of a wire give too small value of the equivalent, seeing that the 
 temperature of the wire during the heating must always be higher 
 than that of the water surrounding it, and hence more heat will be 
 generated than there should be. Hence the numbers should be 
 slightly increased. Joule used wire of platinum-silver alloy, and 
 Weber platinum wire, which may account for Weber's finding a 
 smaller value than Joule, and Weber's value would be more in error 
 than Joule's. Undoubtedly this is a serious source of error, and I am 
 about to repeat an experiment of this kind in which it is entirely 
 avoided. Considering this source of error, these experiments confirm 
 both my value of the ohm and of the mechanical equivalent, and 
 unquestionably show a large error in Kohlrausch's absolute value of 
 the Siemens unit or ohm. 
 
 The experiments of Joule and Favre, where the heat generated by 
 a current, both when it does mechanical work and when it does not, 
 are very interesting, but can hardly have any weight in an estimation 
 of the true value of the equivalent. 
 
 The method of calculating the equivalent from the chemical action 
 in a battery, or the electro-motive force required to decompose any 
 substance, such as water, is as follows. 
 
 Let E be such electro-motive force and c be the quantity of chemi- 
 cal substance formed in battery or decomposed in voltameter per 
 second. Then total energy of current of energy per second is E Q, 
 where Q is the current, or cQ ffj, where ffis the heat generated by 
 unit of c, or required to decompose unit of c. Hence, if the process 
 is entirely reversible, we must have in either case 
 
 CHJ E. 
 
 But the process is not always reversible, seeing that it requires more 
 electro-motive force to decompose water than is given by a gas 
 battery. This is probably due to the formation at first of some un- 
 stable compound like ozone. The process with a battery seems to be
 
 154 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 best, and we can thus apply it to the Daniell cell. The following 
 quantities are mostly taken from Kohlrausch. 
 
 The quantity c has been found by various observers, and Kohl- 
 rausch* gives the mean value as .009421 for water according to his 
 units (mg., mm., second system). Therefore for hydrogen it is 
 .001047. 
 
 The quantity H can be observed directly by short-circuiting the 
 battery, or can be found from experiments like those of Favre and 
 Silbermann. 
 
 The electro-motive force E can be made- to depend either upon the 
 absolute measure of resistance, or can be determined, as Thomson has 
 done, in electro-static units. In electro-magnetic units it is 
 
 Absolute Measure 
 
 Siemens. Ohms. according to my 
 
 Determination. 
 
 After Waltenhofen 11.43 10.90 10.80 X 10 10 
 
 Kohlrausch f H-71 11.17 11.07 X 10 10 
 
 After Favre, 1 equivalent of. zinc develops in the Daniell cell 
 23993 heat units ; 
 
 On the mg., mm., second system, we have E = 10.935 X 10 10 , 
 c = .001047, H= 23993, g = 9800.5 at Baltimore. 
 
 .-. y = 444160 mm - = 444.2 meters. 
 
 Using Kohlrausch's value for absolute resistance, he finds 456.5, 
 which is much more in error than that from my determination. I do 
 not give the calculation from the Grove battery, because the Grove 
 battery is not reversible, and action takes place in it even when no 
 current flows. 
 
 Thomson finds the difference of potential between the poles of a 
 Daniell cell in electro-static measure to be .OQ374 on the cm., grm., 
 second system. $ Using the ratio 29,900 000 000 cm> per second, as I 
 have recently found, but not yet published, we have 111 800 000 011 
 the electro-magnetic system or 11.18 X 10 10 on the mm., mg., second 
 system. This gives 
 
 = 474.3 meters. 
 
 * Pogg. Ann., cxlix. 179. 
 
 t Given by Kohlrausch, Pogg. Ann., cxlix. 182. 
 
 t Thomson, Papers on Electrostatics and Magnetism, p. 246.
 
 OF ARTS AND SCIENCES. 155 
 
 General Criticism. 
 
 All the results so far obtained, except those of Joule, seem to be of 
 the crudest description ; and even when care was apparently taken in 
 the experiment, the method seems to be defective, or the determination 
 is made to rest upon the determination of some other constant whose 
 value is not accurately known. Again, only one or two observers have 
 compared their thermometers with the air thermometer, although I 
 have shown in " Thermometry " that an error of more than one per 
 cent may be made by this method. The range of temperatures is 
 also small as a general rule and the specific heat of water is assumed 
 constant. 
 
 Hence a new determination, avoiding these sources of error, seems 
 to be imperatively demanded. 
 
 (b.) Description of Apparatus. 
 1. PRELIMINARY REMARKS. 
 
 As we have seen in the historical portion, the only experiments of 
 a high degree of accuracy to the present time are those of Joule. 
 Looked at from a general point of view, the principal defects of 
 his method were the use of the mercurial instead of the air thermom- 
 eter, and the small rate at which the temperature of his calorimeter 
 rose. , 
 
 In devising a new method a great rise of temperature in a short time 
 was considered to be the great point, combined, of course, with an 
 accurate measurement of the work done. For a great rise of tem- 
 perature great work must be done, which necessitates the use of a 
 steam-engine or other motive power. For the measurement of the 
 work done, there is only one principle in use at present, which is, 
 that the work transmitted by any shaft in a given time is equal to 2 w 
 times the product of the moment of the force by the number of revo- 
 lutions of the shaft in that time. 
 
 In mechanics it is common to measure the amount of the force 
 twisting the shaft by breaking it at the given point, and attaching the 
 two ends together by some arrangement of springs whose stretching 
 gives the moment. Morin's dynamometer is an example. Hirn* 
 gives a method which he seems to consider new, but which is immedi- 
 ately recognized as Huyghens's arrangement for winding clocks with- 
 
 * Exposition de la Theorie Mecanique de la Chaleur, 3 me ed., p. 18.
 
 156 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 out stopping them. As cords and pulleys are used which may slip on 
 each other, it cannot possess much accuracy. I have devised a method 
 by cog-wheels which is more accurate, but which is better adapted for 
 use in the machine-shop than for scientific experimentation. 
 
 But the most accurate method known to engineers for measuring 
 the work of an engine is that of White's friction brake, and on this I 
 have based my apparatus. Him was the first to use this principle in 
 determining the mechanical equivalent of heat. In his experiment a 
 horizontal axis was turned by a steam-engine. On the axis was a 
 pulley with a flat surface, on which rested a piece of bronze which was 
 to be heated by the friction. The moment of the force with which 
 the friction tended to turn the piece of bronze was measured, together 
 with the velocity of revolution. This experiment, which Him calls a 
 balance de frottement, was first constructed by him to test the quality 
 of oils used' in the industrial arts. He experimented by passing a 
 current of water through the apparatus and observing the tempera- 
 ture of the water before and after passing through. He thus ob- 
 tained a rough approximation to Joule's equivalent. 
 
 He afterwards constructed an apparatus consisting of two cylinders 
 about 30- in diameter and lOO 01 "- long, turning one within the other, 
 the annular space between which could be filled with water, or through 
 which a stream of water could be made to flow whose temperature 
 could be measured before and after. The work was measured by the 
 same method as before. 
 
 But in neither of these methods does Hirn seem to have recognized 
 the principle of the work transmitted by a shaft being equal to the 
 moment of the force multiplied by the angle of rotation of the shaft. 
 In designing his apparatus, he evidently had in view the reproduction 
 in circular motion of the case of friction between two planes in linear 
 motion. 
 
 Since I designed my apparatus, Puluj * has designed an instrument 
 to be worked by hand, and based on the principle used by Hirn. He 
 places the revolving axis vertical, and the friction part consists of two 
 cones rubbing together. But no new principle is involved in his 
 apparatus further than in that used by Hirn.f 
 
 * Fogg. Ann., clvii. 437. 
 
 t Joule's latest results were published after this was written, and I was not 
 aware that he had made this improvement until lately. The result of his 
 experiment, however, reached me soon after, and I have referred to it in the 
 paper, but I did not see the complete paper until much later.
 
 OF AETS AND SCIENCES. 157 
 
 In my apparatus one of the new features has. been the introduction 
 of the Joule calorimeter in the place of the friction cylinders of Him 
 or the cones of Puluj. At first sight the currents and whirlpools in 
 such a calorimeter might be supposed to have some effect ; but when 
 the motion is steady, it is readily seen that the torsion of the calorim- 
 eter is equal to that of the shaft, and hence the principle must apply. 
 
 This change, together with the other new features in the experi- 
 ments and apparatus, has at once made the method one of extreme 
 accuracy, surpassing all others very many fold. 
 
 2. GENERAL DESCRIPTION. 
 
 The apparatus was situated in a small building, entirely separate 
 from the other University buildings, and where it was free from dis- 
 turbances. 
 
 Fig. 6 gives a general view of the apparatus. To a movable 
 axis, a b, a calorimeter similar to Joule's is attached, and the whole is 
 suspended by a torsion wire, c. The shaft of the calorimeter comes 
 out from the bottom, and is attached to a shaft, e f, which receives a 
 uniform motion from the engine by means of the bevel wheels g and h. 
 To the axis, a b, an accurately turned wheel, k I, was attached, and the 
 moment of the force tending to turn the calorimeter was measured by 
 the weights o and p, attached to silk tapes passing around the circum- 
 ference of this wheel in combination with the torsion of the suspend- 
 ing wire. To this axis was also attached a long arm, having two 
 sliding weights, q and r, by which the moment of inertia could be 
 varied or determined. 
 
 The number of revolutions was determined by a chronograph, which 
 received motion by a screw on the shaft ef, and which made one 
 revolution for 102 of the shaft. On this chronograph was recorded 
 the transit of the mercury over the divisions of the thermometer. 
 
 Around the calorimeter a water jacket, t M, made in halves, was 
 placed, so that the radiation could be estimated. A wooden box sur- 
 rounded the whole, to shield the observer from the calorimeter. 
 
 The action of the apparatus is in general as follows. As the inner 
 paddles revolve, the water strikes against the outer paddles, and so 
 tends to turn the calorimeter. When this force is balanced by the 
 weights o p, the whole will be in equilibrium, which is rendered stable 
 by the torsion of the wire c d. Should any slight change take place 
 in the velocity, the calorimeter will revolve in one direction or the 
 other until the torsion brings it into equilibrium again. The amount
 
 158 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 of torsion read off on a scale on the edge of k I gives the correction to 
 be added to or. subtracted from the weights op. 
 
 One observer constantly reads the circle k I, and the other con- 
 stantly records the transits of the mercury over the divisions of the 
 thermometer. 
 
 A series extending over from one half to a whole hour, and record- 
 ing a rise of 15 C. to perhaps 25 C., and in which a record was 
 made for perhaps each tenth of a degree, would thus contain several 
 hundred observations, from any two of which the equivalent of heat 
 could b,e determined, though they would not all be independent. 
 Such a series would evidently have immense weight ; and, in fact, I 
 estimate that, neglecting constant errors, a single series has more 
 weight than all of Joule's experiments of 1849, on water, put 
 together.* 
 
 The correction for radiation is inversely proportional to the ratio of 
 the rate of work generated to the rate at which the heat is lost ; and 
 this for equal ranges of temperature is only -fa as great in my 
 measures as in Joule's ; for Joule's rate of increase was about 0.62 C. 
 per hour, while mine is about 35 C. in the same time, and can be 
 increased to over 45 C. per hour. 
 
 3. DETAILS, 
 The Calorimeter. 
 
 Joule's calorimeter was made in a very simple manner, with few 
 paddles, and without reference to the production of currents to mix 
 up the water. Hence the paddles were made without solder, and 
 were screwed together. Indeed, there was no solder about the 
 apparatus. 
 
 But, for my purpose, the number of paddles must be multiplied, so 
 that there shall be no jerk in the motion, and that the resistance 
 may be great: they must be stronger, to resist the force from the 
 engine, and they must be light, so as not to add an uncertain quantity 
 to the calorific capacity. Besides this, the shape must be such as to 
 cause the whole of the water to run in a constant stream past the 
 thermometer, and to cause constant exchange between the water at 
 the top and at the bottom. 
 
 l^? 1 "? expertments > with an average rise of temperature of 0.56 F., equal 
 ' -31 C, gives a total rise of 12.4 C., which is only about two thirds the 
 erage of one of my experiments. As my work is measured with equal accu- 
 racy, and my radiation with greater, the statement seems to be correct.
 
 OP ARTS AND SCIENCES. 
 
 159
 
 160 
 
 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 Fig. 7 shows a section of the calorimeter, and Fig. 8 a per- 
 spective view of the revolving paddles removed from the appa- 
 ratus, and with the exterior paddles removed from around it; which 
 could not, however, be accomplished physically without destroying 
 
 them. 
 
 To the axis e b, Fig. 7, which was of steel, and 6 mm - in diame- 
 ter, a copper cylinder, a d, was attached, by means of four stout wires 
 at e, and four more at/. To this cylinder four rings, g, //, i,j, were 
 attached, which supported the paddles. Each one had eight paddles, 
 but each ring was displaced through a small angle with reference to 
 
 Fig. 7. 
 
 Fig. 8. 
 
 the one below it, so that no one paddle came over another. This was 
 to make the resistance continuous, and not periodical. The lower 
 row of paddles were turned backwards, so that they had a tendency 
 to throw the water outwards and make the circulation, as I shall 
 show afterwards. 
 
 Around these movable paddles were the stationary paddles, consist- 
 ing of five rows of ten each. These were attached to the movable 
 paddles by bearings, at the points c and k, of the shaft, and were 
 removed with the latter w^ien this was taken from the calorimeter. 
 "When the whole was placed in the calorimeter, these outer paddles 
 were attached to it by means of four screws, I and m, so as to be 
 immovable. 
 
 The cover of the calorimeter was attached to a brass ring, which 
 was nicely ground to another brass ring on the calorimeter, and which
 
 OF ARTS AND SCIENCES. 161 
 
 could be made perfectly tight by means of a little white-lead paint. 
 The shaft passed through a stuffing-box at the bottom, which was 
 entirely within the outer surface of, the calorimeter, so that the heat 
 generated should all go to the water. The upper end of the shaft 
 rested in a bearing in a piece of brass attached to the cover. In the 
 cover there were two openings, one for the thermometer, and the 
 other for filling the calorimeter with water. 
 
 From the opening for the thermometer, a tube of copper, perforated 
 with large holes, descended nearly to the centre of the calorimeter. 
 The thermometer was in this sieve-like tube at only a short distance 
 from the centre of the calorimeter, with the revolving paddles outside 
 of it, and in the stream of water, which circulated as shown by the 
 arrows. 
 
 This circulation of water took place as follows. The lower paddles 
 threw the water violently outwards, while the upper paddles were 
 
 Fig. 9. 
 
 prevented from doing so by a cylinder surrounding the fixed paddles. 
 The consequence was, that the water flowed up in the space between 
 the outer shell and the fixed paddles, and down through the central 
 tube of the revolving paddles. As there was always a little air at the 
 top to allow for expansion, it would also aid in the same direction. 
 These currents, which were very violent, could be observed through 
 the openings. 
 
 The calorimeter was attached to a wheel, fixed to the shaft a b, by 
 the method shown in Fig. 9. At the edge of the wheel, which was of 
 the exact diameter of the calorimeter, two screws were attached, from 
 which wires descended to a single screw in the edge of the calorimeter. 
 Through the wheel, a screw armed with^ a vulcanite point pressed 
 upon the calorimeter, and held it firmly. Three of these arrange- 
 ments, at distances of 120, were used. To centre the calorimeter, 
 a piece of vulcanite at the centre was used. By this method of 
 suspension very little heat could escape, and the amount could be 
 allowed for by the radiation experiments. 
 VOL. xv. (N. s. vii.) 11
 
 162 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 The Torsion System. 
 
 The torsion wire was of such strength that one millimeter on the 
 scale at the edge of the wheel signified 11.8 grammes, or about 7 ^ of 
 the weights o p generally used. There were stops on the wheel, so 
 that it could not move through more than a small angle. The weights 
 were suspended by very flexible silk tapes, 6 mm or 8 mm broad and 
 0.3 mra - thick. They varied from 4.5 k - to 8.5 k - taken together. The 
 shaft, a b, was of uniform size throughout, so that the wire c sus- 
 pended the whole system, and no "weight rested on the bearings. 
 
 The pulleys, m, n, Fig. 6, were very exactly turned and balanced, 
 and the whole suspended system was so free as to vibrate for a con- 
 siderable time. However, as will be shown hereafter, its freedom is 
 of little consequence. 
 
 The Water Jacket. 
 
 Around the calorimeter, a watef jacket, t u, was placed, so that 
 the radiation should be perfectly definite. During the preliminary 
 experiments a simple tin jacket was used, whose temperature was 
 determined by two thermometers, one above and the other below, 
 inserted in tubes attached to the jacket. 
 
 The Driving Gear. 
 
 The cog-wheels, g, h, were made by Messrs-. Brown and Sharpe, of 
 Providence, and were so well cut that the motion transmitted to the 
 calorimeter must have been very uniform. 
 
 The Chronograph. 
 
 The cylinder of the chronograph was turned by a screw on the shaft 
 ef, and received one revolution for 102 of the paddles ; 155 revolutions 
 of the cylinder, or 15,810 of the paddles, could be recorded, though, 
 when necessary, the paper could be changed without stopping, and the 
 experiment thus continued without interruption. 
 
 The Frame and Foundation. 
 
 The frame was very massive and strong, so as to prevent oscillation; 
 and the whole instrument weighed about 500 pounds as nearly as 
 could be estimated. It was placed on a solid brick pier, with a firm 
 foundation in the ground. The trembling was barely perceptible to 
 the hand when running the fasteot.
 
 OP ARTS AND SCIENCES. 163 
 
 The Engine. 
 
 The driving power was a petroleum engine, which was very efficient 
 in driving the apparatus with uniformity. 
 
 The Balance. 
 
 For weighing the calorimeter, a balance capable of showing the 
 presence of less than ^ gramme with 15,000 grammes was used. 
 The weights, however, by Schickert, of Dresden, were accurate among 
 themselves to at least o"* for the larger weights, and in proportion for 
 the smaller. A more accurate balance would have been useless, as 
 will be seen further on. 
 
 Adjustments. 
 
 There are few adjustments, and they were principally made in the 
 construction. 
 
 In the first place, the shafts a b and ef must be on line. Secondly, 
 the wheels m n must be so adjusted that their planes are vertical, and 
 that the tapes shall pass over them symmetrically, and that their edges 
 shall be in the plane of the wheel k I. 
 
 Deviation from these adjustments only produced small error. 
 
 (c.) Theory of the Experiment. 
 1. ESTIMATION OF WORK DONE. 
 
 The calorimeter is constantly receiving heat from the friction, and 
 is giving out heat by radiation and conduction. Now, at any given 
 instant of time, the temperature of the whole of the calorimeter is not 
 the same. Owing to the violent stirring, the water is undoubtedly at 
 a very uniform temperature throughout. But the solid parts of the 
 calorimeter cannot be so. The greatest difference of temperature is 
 evidently soon after the commencement of the operation. But after 
 some time the apparatus reaches a stationary state, in which, but for 
 the radiation, the rise of temperature at all points would be the same. 
 This steady state will be theoretically reached only after an infinite 
 time ; but as most of the metal is copper, and quite thin, and as the 
 whole capacity of the metal work is only about four per cent of the 
 total capacity, I have thought that one or two minutes was enough to 
 allow, though, if others do not think this time sufficient, they can 
 readily reject the first few observations of each series. When there 
 is radiation, the stationary state will never be reached theoretically,
 
 164 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 though practically there is little difference from the case where there 
 is no radiation. 
 
 The measurement of the work done can be computed as follows. 
 Let M be the moment of the force tending to turn the calorimeter, 
 and d 9 the angle moved by tHe shaft. The work done in the time i, 
 will be f M d 6. If the moment of the force is constant, the integral 
 is simply M 9 ; but it is impossible to obtain an engine which runs 
 with perfect steadiness, and although we may be able to calculate the 
 integral, as far as long periods are concerned, by observation of the 
 torsion circle, yet we are not thus able to allow for the irregularity 
 during one revolution of the engine. Hence I have devised the follow- 
 ing theory. I have found, by experiments with the instrument, that 
 the moment of the force 1 is very nearly, for high velocities at least, 
 proportional to the square of the velocity. For rapid changes of the 
 velocity, this is not exactly true, but as the paddles are very numerous 
 in the calorimeter, it is probably very nearly true. We have then 
 
 where C is a constant. Hence the work done becomes 
 
 Yrfi, 
 
 As we allow for irregularities of long period by readings of the 
 torsion circle, we can assume in this investigation that the mean 
 velocity is constant, and equal to v . The form of the variation of 
 the velocity must be assumed, and I shall put, without further dis- 
 cussion, 
 
 ' d / . 2 IT t\ 
 
 -T = v ( 1 4- c cos ) . 
 
 V a / 
 
 We then find, on integrating from a to 0, 
 
 w= Cv s a (1 -|- f c 8 ), 
 
 which is the work on the calorimeter during one revolution of the 
 engine. 
 
 ^The equation of the motion of the calorimeter, supposing it to be 
 nearly stationary, and neglecting the change of torsion of the sus- 
 pending wire, is 
 
 where m is the moment of inertia of the calorimeter and its attach- 
 ments, f is the angular position of the calorimeter, W is the sum of
 
 OF AETS AND SCIENCES. 165 
 
 the torsion weights, and D is the diameter of the torsion wheel. 
 Hence, 
 
 When W D = 2 C v 2 (1 -f- c 2 ), the calorimeter will merely oscil- 
 late around a given position, and will reach its maximum at the times 
 t = 0, | a, a, &c. 
 
 The total amplitude of each oscillation will be very nearly 
 
 , _ ,/ __ C v^ga^c _ TF > gr a z c 
 V Y ~ * m 2ir* m ' 
 
 If a; is the amplitude of each oscillation, as measured in millimeters, 
 on the edge of the wheel of diameter D, we have \jr i// = -^. 
 
 where n is the number of revolutions of the engine per second. 
 Having found c in this way, the work will be, during any time, 
 
 W = TT WD N(l +c 2 ), 
 
 where N is the total number of revolutions of the paddles. 
 
 A variation of the velocity of ten per cent from the- mean, or 
 twenty per cent total, would thus only cause an error of one per cent 
 in the equivalent. 
 
 Hence, although the engine was only single acting, yet it ran easily, 
 had great excess of power, and was very, constant as far as long 
 periods were concerned. The engine ran very fast, making from 
 200 to 250 revolutions per minute. The fly-wheel weighed about 220 
 pounds, and had a radius of lj feet. At four turns per second, this 
 gives an energy of about 3400 foot pounds stored in the wheel. The 
 calorimeter required about one-half horse-power to drive it ; and, 
 assuming the same for the engine friction, we have about 140 foot 
 pounds of work required per revolution. Taking the most unfavorable 
 case, where all the power is given to the engine at one point, the 
 velocity changes during the revolution about four per cent, or c would 
 nearly equal .02, causing an error of 1 part in 2500 nearly. By 
 means of the shaking of the calorimeter, I have estimated c as follows, 
 the value of m being changed by changing the weight on the inertia 
 bar, or taking it off altogether. The estimate of the shaking was 
 made by two persons independently.
 
 166 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 m x observed. calculated. 
 
 2,200,000 grins, cm. 2 .6 mm. .016 
 
 3,100,000 " -36 " -013 
 
 11,800,000 " .13 " jO!7 
 
 Mean, c = .015 
 
 causing a correction of 1 part in 5000. 
 
 Another method of estimating the irregularity of running is to put 
 on or take off weights until the calorimeter rests so firmly against the 
 stops that the vibration ceases. Estimated in this way, I have found 
 a little larger value of c, namely, about .017. 
 
 But as one cannot be too careful about such sources of error, I 
 have experimented on the equivalent with different velocities and with 
 very different ways of running the engine, by which c was greatly 
 changed, and so have satisfied myself that the correction from this 
 source is inappreciable in the present state of the science of heat. 
 
 Hence I shall simply put for the work 
 w = -trNWD, 
 
 in gravitation measure at Baltimore. To reduce to absolute measure, 
 we must multiply by the force of gravity given by the formula 
 
 g = 9.78009 -f- .0508 sin 2 <, 
 
 which gives 9.8005 meters per second at Baltimore. If the calo- 
 rimeter moved without friction, no work would be required to cause it 
 to vibrate back and forth, as I have described; but when it moves 
 with friction, some work is required. When I designed the apparatus, 
 I thus had an idea that it would be best to make it as immovable as 
 possible by adding to its moment of inertia by means of the inertia 
 bar and weights. But on considering the subject further, I see that 
 only the excess of energy represented by c z wN W D can be used in 
 this way. For, when the calorimeter is rendered nearly immovable 
 by its great moment of inertia, the work done on it is, as we have 
 seen, TT N W D (1 -\- c 2 ) ; but if it had no inertia, it is evident that 
 the work would be only TrNWD. If, therefore, the calorimeter is 
 made partially stationary, either by its moment of inertia or by fric- 
 tion, the work will be somewhere between these two, and the work 
 spent in friction will be only so much taken from the error. Hence 
 in the latter experiments the inertia bar was taken off, and then the 
 calorimeter constantly vibrated through about half a millimeter on the 
 torsion scale. 
 
 Besides this quick vibration, the calorimeter is constantly moving to
 
 OF ARTS AND SCIENCES. 167 
 
 the extent of a few millimeters back and forth, according to the vary- 
 ing velocity of the engine. As frequent readings were taken, these 
 changes were eliminated. In very rare cases the weights had to be 
 changed during the experiment ; but this was very seldom. 
 
 The vibration and irregular motion of the calorimeter back and 
 forth served a very useful purpose, inasmuch as it caused the friction 
 of the torsion apparatus to act first in one direction and then in the 
 other, so that it was finally eliminated. The torsion apparatus moved 
 very freely when the calorimeter was not in position, and would keep 
 vibrating for some minutes by itself, but with the calorimeter there 
 was necessarily some binding. But the vibration made it so free that 
 it would return quickly to its exact 'position of equilibrium when drawn 
 aside, and would also quickly show any small addition to the weights. 
 This was tried in each experiment. 
 
 To measure the heat generated, we require to know the calorific 
 capacity of the whole calorimeter, and the rise of temperature which 
 would have taken place provided no heat had been lost by radiation. 
 The capacity of the calorimeter alone I have discussed elsewhere, 
 finding the total amount equal to .347 k ' of water at ordinary tempera- 
 tures. The total capacity of the calorimeter is then A -f- -347, where 
 A is the weight of water. Hence Joule's equivalent in absolute 
 measure is 
 
 ~~ (4 + .347) (t t>) y \ 
 
 where n is the number of revolutions of the chronograph, it making 
 one revolution to 102 of the paddles. 
 
 The corrections needed are as follows : 
 
 1st. Correction for weighing in air. This must be made to JF, the 
 cast-iron weights, and to A -j- .347, the water and copper of the calo- 
 rimeter. If A is the density of the air under the given conditions, the 
 correction is .835 A. 
 
 2d. For the weight of the tape by which the weights are hung. 
 
 This is '-rp; 
 
 3d. For the expansion of torsion wheel, D' being the diameter at 
 20 C. This is .000018 (t" 20). Hence, 
 
 nnofi 
 (1 -j_ .000018 (f '-20) 4- jp .835 A)/^.a 
 
 where t t 1 is the rise of the temperature corrected for radia- 
 tion. 
 
 . Oo 
 
 .
 
 168 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 2. RADIATION. 
 
 The correction for radiation varies, of course, with the difference of 
 temperature between the calorimeter and jacket; but, owing to the 
 rapid generation of heat, the correction is generally small in propor- 
 tion. The temperature , generated was generally about 0.6 per 
 minute. The loss of temperature per minute by radiation was approxi- 
 mately .0014 6 per minute, where 6 is the difference of the tempera- 
 ture. This is one per cent for 10.7, and four per cent for 14.2. 
 Generally, the calorimeter was cooler than the jacket to start with, 
 and so a rise of about 20 could be accomplished without a rate of 
 correction at any point of more than fpur per cent, and an average 
 correction of less than two per cenl. An error of ten per cent is thus 
 required in the estimation of the radiation to produce an average 
 error of 1 in 500, or 1 in 250 at a single point. The coefficients 
 never differ from the mean more than about two per cent. The 
 observations on the equivalent, being at a great variety of tempera- 
 tures, check each other as to any error in the radiation. 
 
 The losses of heat which I place under the head of radiation include 
 conduction and convection as well. I divide the losses of heat into the 
 following parts: 1st. Conduction down the shaft; 2d. Conduction by 
 means of the suspending wires or vulcanite points to the wheel above ; 
 3d. True radiation ; 4th, Convection by the air. To get some idea 
 of the relative amounts lost in this way, we can calculate the loss by 
 conduction from the known coefficients of conduction, and we can get 
 some idea of the relative loss from a polished surface from the experi- 
 ments of Mr. Nichol. In this way I suppose the total coefficient of 
 radiation to be made up approximately as follows : 
 
 Conduction along shaft . . .00011 
 ' " suspending wires .00006 
 
 True radiation 00017 
 
 Convection .00106 
 
 Total . . . .00140 
 
 The conduction through the vulcanite only amounts to .0000002. 
 From this it would seem that three fourths of the loss is due to 
 radiation and convection combined. 
 
 The last two 'losses depend upon the difference of temperature 
 between the calorimeter and the jacket, but the first two upon the 
 difference between the calorimeter and frame of the machine and the 
 wheel respectively. The frame was always of very nearly the same
 
 OF ARTS AND SCIENCES. 169 
 
 temperature as the water jacket, but the wheel was usually slightly 
 above it. At first its temperature was noted by a thermometer, and 
 the loss to it computed separately; but it was found to be unnecessary, 
 and finally the whole was assumed to be a function of the tempera- 
 ture of the calorimeter and of the jacket only. 
 
 At first sight it might seem that there was a source of error in 
 having a journal so near the bottom of the calorimeter, and joined to 
 it by a shaft. But if we consider it a moment, we shall see that the 
 error is inappreciable ; for even if there was friction enough in the 
 journal to heat it as fast as the calorimeter, it would decrease the 
 radiation only seven per cent, or make an average error in the experi- 
 ment of only 1 in 700. But, in fact, the journal was very perfectly 
 made, and there was no strain on it to produce friction ; besides which, 
 it was connected to a large mass of cast-iron which was attached to 
 the base. Hence, as a matter of fact, the journal was not appreciably 
 warmer after running than before, although tested by a thermometer. 
 The difference could not have been more than a degree or so at most. 
 
 The warming of the wheel by conduction and of the journal by 
 friction would tend to neutralize each other, as the wheel would be 
 warmer and the journal cooler during the radiation experiment than 
 the friction experiment. 
 
 The usual method of obtaining the coefficient of radiation would be 
 to stop the engine while the calorimeter was hot, and observe the 
 cooling, stirring the water occasionally when the temperature was 
 read. This method I used at first, reading the temperature at inter- 
 vals of about a half to a whole hour. But on thinking the matter 
 over, it became apparent that the coefficient found in this way would 
 be too small, especially at small differences of temperature ; for the 
 layer next to the outside would be cooled lower than the mean tem- 
 perature, and the heat could only get to the outside by conduction 
 through the water or by convection currents. 
 
 Hence I arranged the engine so as to run the paddles very slowly, 
 so as to stir the water constantly, taking account of the number of 
 the revolutions and the torsion, so as to compute the work. As I 
 had foreseen, the results in this case were higher than by the other 
 method. At low temperatures the error of the first method was 
 fifteen per cent ; but at high, it did not amount to more than about 
 three to five per cent, and probably at very high temperatures it would 
 almost vanish. 
 
 I do not consider it necessary to give all the details of the radiation 
 experiments, but will merely remark that, as the calorimeter was
 
 170 
 
 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 nickel-plated, and as seventy-five per cent of the so-called radiation is due 
 to convection by the air, the coefficients of radiation were found to be 
 very constant under similar conditions, even after long intervals of time. 
 
 The experiments were divided into two groups ; one when the 
 temperature of the jacket was about 5 C., and the other when it 
 averaged about 20 C. 
 
 The results were then plotted, and the mean curve drawn through 
 them, from which the following coefficients were obtained. These 
 coefficients are the loss of temperature per minute, and per degree 
 difference of temperature. 
 
 TABLE XXXV. COEFFICIENTS OF RADIATION. 
 
 Difference 
 
 
 
 between Jacket 
 
 Jacket 5. 
 
 Jacket 20. 
 
 and Calorimeter. 
 
 
 
 -5 
 
 .00138 
 
 .00134 
 
 
 
 .00135 
 
 .00130 
 
 +5 
 
 .00137 
 
 .00132 
 
 10 
 
 .00142 
 
 .00138 
 
 15 
 
 .00148 
 
 .00144 
 
 20 
 
 .00154 
 
 .00150 
 
 25 
 
 .00158 
 
 .00154 
 
 As the quantity of water in the calorimeter sometimes varied 
 slightly, the numbers should be modified to suit, they being true when 
 the total capacity of the calorimeter was 8.75 kil. The total surface 
 of the calorimeter was about 2350 sq. cm., and the unit of time 
 one minute. To compare my results with those of McFarlane and 
 of Nichol given in the Proc. R. S. and Proc. R. S. E., I will reduce 
 my results so that they can be compared with the tables given by 
 Professor Everett in his " Illustrations of the Centimeter-Gramme- 
 Second System of Units," pp. 50, 51. 
 
 The reducing factor is .0621, and hence the last results for the 
 jacket at 20 C. become : 
 
 TABLE XXXVI. 
 
 Difference of 
 Temperature. 
 
 Coefficient of Radiation 
 on the C. G. S. System. 
 
 McFarlane's Value. 
 
 K&tio. 
 
 3 
 
 .000081 
 
 .000168 
 
 2.07 
 
 6 
 
 .000082 
 
 . .000178 
 
 2.17 
 
 10 
 
 .OOOOSff 
 
 .000186 
 
 2.16 
 
 15 
 
 .000089 
 
 .000193 
 
 2.17 
 
 20 
 
 .000093 
 
 .000201 
 
 2.16 
 
 25 
 
 .000096 
 
 .000207 
 
 2.16
 
 OF ARTS AND SCIENCES. 171 
 
 The variation which I find is almost exactly that given by McFar- 
 lane, as is shown by the constancy of the column of ratios. But my 
 coefficients are less than half those of McFarlane. This may possibly 
 be due to the fact that the walls of McFarlane's enclosure were 
 blackened, and to his surface being of polished copper and mine of 
 polished nickel : his surface may also have been better adapted by its 
 form to the loss of heat by convection. The results of Nichol are 
 also much lower than those of McFarlane. 
 
 The fact that the coefficients of radiation are less with increased 
 temperature of jacket is just contrary to what Dulong and Petit found 
 for radiation. But as I have shown that convection is the principal 
 factor, I am at a loss to check my result with any other observer. 
 Dulong and Petit make the loss from convection dependent only upon 
 the difference of temperature, and approximately upon the square root 
 of the pressure of the gas. Theoretically it would seem that the loss 
 should be less as the mean temperature rises, seeing that the air be- 
 comes less dense and its viscosity increases./ Should we substitute 
 density for pressure in Dulong's law, we should have the loss by con- 
 vection inversely as the square root of the mean absolute tempera- 
 ture, or approximately the absolute temperature of the jacket. This 
 would give a decrease of one per cent in the radiation for about 6, 
 which is not far from what I have found. 
 
 To estimate the accuracy with which the radiation has been obtained 
 is a very difficult matter, for the circumstances in the experiment are 
 not the same as when the radiation was obtained. In the first place, 
 although the water is stirred during the radiation, yet it is not stirred 
 so violently as during the experiment. Further, the wheel above 
 the calorimeter is warmer during radiation than during the experi- 
 ment. Both these sources of error tend to give too small coefficients 
 of radiation, and this is confirmed by looking over the final tables. 
 But I have not felt at liberty to make any corrections based on the 
 final results, as that would destroy the independence of the observa- 
 tions. But we are able thus to get the limits of the error 'produced. 
 
 During the preliminary experiments a water jacket was not used, 
 but only a tin case, whose temperature was noted by a thermometer 
 above and below. The radiation under these circumstances was 
 larger, as the case was not entirely closed at the bottom, and so per- 
 mitted more circulation of air. 
 
 3. CORRECTIONS TO THERMOMETERS, ETC. 
 
 Among the other corrections to the temperature as read off from 
 the thermometers, the correction for the stem at the temperature of
 
 172 PEOCEEDINGS OP THE AMEEICAN ACADEMY 
 
 the air is the greatest. The ordinary formula for the correction is 
 .000156 n (t ") But, in applying this correction, it is difficult to 
 estimate n, the number of degrees of thermometer outside the calo- 
 rimeter and at the temperature of the air, seeing that part of the stem 
 is heated by conduction. The uncertainty vanishes as the thermometer 
 becomes longer and longer, or rather as it is more and more sensitive. 
 But even then some of the uncertainty remains. I have sought to 
 avoid this uncertainty by placing a short tube filled with water about 
 the lower part of the thermometer as it comes out of the calorimeter. 
 The temperature of this was indicated by a thermometer, by aid of 
 which also the heat lost to the water by conduction through the ther- 
 mometer stem could be computed ; this, however, was very minute 
 compared with the whole heat generated, say 1 in 10,000. 
 
 The water being very nearly at the temperature of the air, the stem 
 above it could be assumed to be at the temperature of the air indicated 
 by a thermometer hung within an inch or two of it. The correction 
 for stem would thus have to be divided into two parts, and calculated 
 separately. Calculated in this way, I suppose the correction is per- 
 fectly certain to much less than one hundredth of a degree : the total 
 amount was seldom over one tenth of a degree. 
 
 Among the uncertain errors to which the measurement of temper- 
 ature is subjected, I may mention the following : 
 
 1. Pressure on bulb. A pressure of 60 cm - of water produced a 
 change of about 0.01 in the thermometers. When the calorimeter 
 was entirely closed there was soon some pressure generated. Hence 
 the introduction of the safety-tube, a tube of thin glass about 
 lO 6 - long, extending through a cprk in the top of the calorimeter. 
 The top of the safety-tube was nearly closed by a cork to prevent . 
 evaporation. Had the tube been shorter, water would have been 
 forced out, as well as air. 
 
 2. Conduction along stem from outside to thermometer bulb. To 
 avoid this, not only was the bulb immersed, but also quite a length of 
 stem. As this portion of the stem, as also the bulb, was surrounded 
 by water in violent motion, there could have been no large error from 
 this source. The immersed stem to the top of the bulb was generally 
 about 5 cm - or more, and the stem only about .8- in diameter. 
 
 3. The thermometer is never at the temperature of the water, be- 
 cause the latter is constantly rising ; but we do not assume that it is 
 so in the experiment. We only assume that it lags behind the water 
 to the same amount at all parts of the experiment, -and this is doubt- 
 less true.
 
 OF ARTS AND SCIENCES. 173 
 
 To see if the amount was appreciable, I suddenly threw the appa- 
 ratus out of gear, thus stopping it. The temperature was observed to 
 continue rising about 0.02 C. Allowing 0.01 for the rise due to 
 motion after the word " Stop " was given, we have about 0.01 C. as 
 the amount the thermometer lagged behind the water. 
 
 4. Evaporation. A possible source of error exists in the cooling of 
 the calorimeter by evaporation of water leaking out from it. 
 
 The water was always weighed before and after the experiment in 
 a balance giving ^ gramme with accuracy. The normal amount of 
 loss from removal of thermometer, wet corks, &c. was about 1 
 gramme. The calorimeter was perfectly tight, and had no leakage at 
 any point in its normal state. Once or twice the screws of the stuffing- 
 box worked loose, but these experiments were rejected. 
 
 The evaporation of 1 gramme of water requires about 600 heat 
 units, which is sufficient to depress the temperature of the calorimeter 
 about 0.07 C. As the only point at which evaporation could take 
 place was through a hole less than I" 1 - diameter in the safety-tube, I 
 think it is reasonable to assume that the error from this source is in- 
 appreciable. But to be doubly certain, I observed the time which 
 drops of water of known weight and area, placed on the warm calo- 
 rimeter, took to dry. From these experiments it was evident that it 
 would require a considerable area of wet surface to produce an ap- 
 preciable effect. This wet surface never existed unless the calo- 
 rimeter was wet by dew deposited on the cool surface. To guard 
 against this error, the calorimeter was never cooled so low that dew 
 formed ; it was carefully rubbed with a towel, and placed in the appa- 
 ratus half an hour to an hour before the experiment, exposed freely to 
 the air. The surface being polished, the slightest deposit of dew was 
 readily visible. The greatest care was taken to guard against this 
 source of error, and I think the experiment is free from it. 
 
 (d.) Results. 
 1. CONSTANT DATA. 
 
 Joule's equivalent in gravitation measure is of the dimensions of 
 length only, being the height which water would have to fall to be 
 heated one degree. Or let water flow downward with uniform velocity 
 through a capillary tube impervious to heat ; assuming the viscosity 
 constant, the rate of variation of height with temperature will be 
 Joule's equivalent. 
 
 Hence, besides the force of gravity the only thing required in ab-
 
 174 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 solute measure is some length. The length that enters the equation 
 is the diameter of the torsion wheel. This was determined under a 
 microscope comparator by comparison with a standard meter belong- 
 ing to Professor Rogers of Harvard Observatory, which had been 
 compared at Washington with the Coast Survey standards, as well as 
 by comparison with one of our own meter scales which had also been 
 so compared. The result was .26908 meter at 20 C. 
 
 To this must be added the thickness of the silk tape suspending the 
 weights. This thickness was carefully determined by a micrometer 
 screw while the tape was stretched, the screw having a flat end. The 
 result was .0003 l m -. 
 
 So that, finally, J7 = .26939 meter at 20 C. Separating the 
 constant from the variable parts, the formula now becomes 
 
 f= 
 
 g = 9.8005 at Baltimore. 
 
 It is unnecessary to have the weights exact to standard, provided 
 they are relatively correct, or to make double weighings, provided the 
 same scale of the balance is always used. For both numerator and 
 denominator of the fraction contain a weight. 
 
 2. EXPERIMENTAL DATA AND TABLES OF RESULTS. 
 
 In exhibiting the results of the experiments, it is much more sat- 
 isfactory to compute at once from the observations the work neces- 
 sary to raise 1 UL of the water from the first temperature observed to 
 each succeeding temperature. By interpolation in such a table we 
 can then reduce to even degrees. To compare the different results I 
 have then added to each table such a quantity as to bring the result 
 at 20 about equal to 10,000 kilogramme meters. 
 
 The process for each experiment may be described as follows. 
 The calorimeter was first filled with distilled water a little cooler 
 than the atmosphere, but not so cool as to cause a deposit of dew. 
 It was then placed in the machine and adjusted to its position, 
 though the outer half of the jacket was left off for some time, so that 
 the calorimeter should become perfectly dry ; to aid which the calo- 
 rimeter was polished with a cloth. The thermometer and safety- 
 tube were also inserted at this time. 
 
 After half an hour or so, the chronograph was adjusted, the outer 
 half of the jacket put in place, the wooden screen fixed in position, 
 and all was ready to start. The engine, which had been running
 
 OP ARTS AND SCIENCES. 175 
 
 quietly for some time, was now attached, and the experiment com- 
 menced. First the weights had to be adjusted so as to produce equi- 
 librium as nearly as possible. 
 
 The observers then took their positions. One observer constantly 
 recorded the. transit of the mercury over the divisions of the ther- 
 mometer, making other suitable marks, so that* the divisions could be 
 afterwards recognized. He also read the thermometers giving the 
 temperatures of the air, the bottom of the calorimeter thermometer, 
 and of the wheel just above the calorimeter ; and sometimes another, 
 giving that of the cast-iron frame of the instrument. 
 
 The other observer read the torsion wheel once every revolution of 
 the chronograph cylinder, recording the time by his watch. He also 
 recorded on the chronograph every five minutes by his watch, and 
 likewise stirred the water in the jacket at intervals, and read its tem- 
 perature. 
 
 The recording of the time was for the purpose of giving the con- 
 necting link between the readings of the torsion circle and of the ther- 
 mometer. This, however, as the readings were quite constant, had 
 only to be done roughly, say to half a minute of time, though the 
 records of time on the chronograph were true to about a second. 
 
 The thermometers to read the temperature of the water in the 
 jacket were graduated to 0.2 C., but were generally read to 0.l C., 
 'and had been compared with the standards. There was no object in 
 using more delicate thermometers. 
 
 After the experiment had continued long enough, the engine was 
 stopped and a radiation experiment begun. The last operation was 
 to weigh the calorimeter again, after removing the thermometer and 
 safety-tube, and also the weights which had been used. 
 
 The chronograph sheet, having then been removed from the cylin- 
 der, had the time records identified and marked, as well as the ther- 
 mometer records. Each line of the chronograph record was then 
 numbered arbitrarily, and a table made indicating the stand of the 
 thermometer and the number of the revolutions and fractions of a 
 revolution as recorded on the chronograph sheet. The times at which 
 these temperatures were reached was also found by interpolation, and 
 recorded in another column. 
 
 From the column of times the readings of the torsion circle could 
 be identified, and so all the necessary data would be at hand for cal- 
 culating the work required to raise the temperature of one kilo- 
 gramme of the water from the first recorded temperature to any 
 succeeding temperature.
 
 ' 176 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 As these temperatures usually contained fractions, the amount of 
 work necessary to raise one kilogramme of the water to the even 
 degrees could then be found from this table by interpolation. Joule's 
 equivalent at any point would then be merely the difference of any 
 two succeeding numbers ; or, better, one tenth the difference of two 
 numbers situated 10 apart, or, in general, the difference of the num- 
 bers divided by the difference of the temperatures. 
 
 It would be a perfectly simple matter to make the record of the 
 torsion circle entirely automatic, and I think- I shall modify the 
 apparatus in that manner in the future. 
 
 It would take too much space to give the details of each experiment ; 
 but, to show the process of calculation, I will give the experiment of 
 Dec. 17, 1878 as a specimen. The chronograph sheet, of course, I 
 cannot give. The computation is at first in gravitation measure, but 
 afterwards reduced to absolute measure. 
 
 The calorimeter before the experiment weighed 12.2733 kil. 
 
 " " after " " " 12.2716 " 
 
 Mean 12.2720 
 
 Weight of calorimeter alone 3.8721 " 
 
 .-. Water alone 'weighed 8.3999 " 
 
 .3470 
 
 Total capacity 8.7469 " 
 
 The correction for weighing in air was .835 X = .00106. 
 The total term containing the correction is therefore .99878. 
 
 log 86.324 =1.9361316 
 
 log .99878 = 1.9994698 
 
 1.9*356014 
 
 log 8.7469" = .9418542 
 
 log const, factor = .9937472 = log 9.85706. 
 
 Hence the work per kilogramme is 9.85706 2JFn in gravitation 
 measure, the term 2 W n being used to denote the sum of products 
 similar to Wn as obtained by simultaneous readings of torsion circle 
 and records on chronograph sheet. 
 
 Zero of torsion wheel, 79.3 mm -. 
 
 Value of l mm - on torsion wheel .0118 kQ -. 
 
 The following were the records of time on the chronograph sheet :
 
 OP ARTS AND SCIENCES. 177 
 
 Time observed. 
 
 Revolutions of Chronograph. 
 
 Time calculated. 
 
 15 
 
 8.74 
 
 15.2 
 
 20 
 
 25.32 
 
 20.1 
 
 25 
 
 42.10 
 
 25.0 
 
 30 
 
 59.05 
 
 30.0 
 
 35 
 
 76.00 
 
 35.0 
 
 40 
 
 93.03 
 
 40.0 
 
 45 
 
 109.97 
 
 45.0 
 
 50 
 
 126.92 
 
 50.0 
 
 55 
 
 144.14 
 
 55.0 
 
 The times were calculated by the formula 
 
 Time = .294 X Revolutions -f 12.66, 
 
 which assumes that the engine moves with uniform velocity. As the 
 principal error in using an incorrect interpolation formula comes from 
 the calculation of the radiation, and as this formula is correct within 
 a few seconds for all the higher temperatures, we can use it in the 
 calculation of the times. 
 
 The records of the transits of the mercury over the divisions of 
 the thermometer were nearly always made for each division, but it is 
 useless to calculate for each. I usually select the even centimeters, 
 and take the mean of the records for several divisions on each side. 
 
 While the mercury was rising l cm on No. 6163, there would be 
 about seven revolutions of the chronograph, and consequently seven 
 readings of the torsion circle, each one of which was the average for 
 a little time as estimated by the eye. 
 
 I have obtained more than thirty series of results, but have thus 
 far reduced only fourteen, five of which are preliminary, or were made 
 with the simple jacket instead of the water jacket, the radiation to 
 which was much greater, as there was a hole at the bottom which 
 allowed more circulation of the air. The mean of the preliminary 
 results agrees so closely with the mean of the final results, that I have 
 in the end given them equal weight. 
 
 On March 24th, the same thermometer was used for a second ex- 
 periment directly after the first, seeing that the chronograph failed to 
 work in the first experiment until 8 was reached. The error from 
 this cause was small, as the first experiment only reached to 26 C., and 
 hence there could have been no change of zero, as this is very nearly 
 the temperature at which the thermometer was generally kept. 
 
 Having thus calculated the work in conjunction with the tempera- 
 ture, I have next interpolated so as to obtain the -work at the even 
 VOL. xv. (N. s. vn.) 12
 
 178 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 degrees. The tables so formed I have combined in two ways : first, 
 I have added to the column of* work in each table an arbitrary number, 
 such as to make the work at 20 about 10,000, and have then combined 
 them as seen in . Table LI. ; and, secondly, I have subtracted each 
 number from the one 1 farther down the table, and divided the num- 
 bers so found by 10, thus obtaining the mechanical equivalent of heat. 
 
 In these tables four thermometers have been used, and yet they 
 were so accurate that little difference can be observed in" the experi- 
 ments which can be traced to an error of the thermometer, although 
 the Kew standard has some local irregularities. The greatest difference 
 between any column of Table LI. and the general mean is only 10 
 kilogramme-meters, or 0.023 degree, and this includes all errors of 
 calibration of thermometers, radiation, &c. This seems to me to be a 
 very remarkable result, and demonstrates the surpassing accuracy of 
 the method. Indeed, the limit of accuracy in thermometry is the only 
 limit which we can at present give to this method of experiment. 
 Hence the large proportional time spent on that subject. 
 
 The accuracy of the radiation is demonstrated, to some extent, by 
 the agreement of the results obtained even with different temperatures 
 of the jacket. But on close observation it seems apparent that the 
 coefficients of radiation should be further increased as there is a ten- 
 dency of the end figures in each series to become too high. This is 
 exactly what we should suppose, as we have seen that nearly all 
 sources of error tend in the direction of making the radiation too 
 small. For instance, an error came from not stirring the water dur- 
 ing the radiation, and there must be a small residual error from not 
 stirring so fast during radiation as during the experiment. Besides 
 this, some parts around the calorimeter were warm during the radiation 
 which were 'cool during the experiment. And both of these make the 
 correction for radiation too small. However, the error from this 
 source is small, and cannot possibly affect the general conclusions. In 
 each column of Tables LI. and LII. a dash is placed at the tem- 
 perature of the jacket, and for fifteen degrees below this point the 
 error in the radiation must produce only an inappreciable error 
 in the equivalent: taking the observations within this limit as the 
 standards, and rejecting the others, we should still arrive at very 
 nearly the same conclusions as if we accepted the whole. 
 
 Most of the experiments are made with a weight of about 7.3 kiL as 
 everything seemed to work best with this weight. But for the sake 
 of a test I have run the weight up to 8.6 and down to 4.4 kil - by which 
 the mte of generation of the heat was changed nearly three times.
 
 OF ARTS AND SCIENCES. 179 
 
 By this the correction for the radiation and the error due to the 
 irregularity of the engine are changed, and yet scarcely an appreciable 
 difference in the results can be observed. 
 
 The tables explain themselves very well, but some remarks may be 
 in order. Tables XXXVII. to L. inclusive are the results of fourteen 
 experiments selected from the total of about thirty, the others not 
 having been worked up yet, though I propose to do so at my leisure. 
 
 Table LI. gives the collected results. At the top of each column 
 the date of the experiment and number of the thermometer are 
 given, together with Xhe .approximate torsion weight and the rate of 
 rise of temperature per hour. The dash in each column gives 
 approximately the temperature of the jacket, and hence of the air. 
 There are four columns of mean values, but the last, produced from 
 the combination of the table by parts, is the best. 
 
 Table LII. gives the mechanical equivalent of heat as deduced from 
 intervals of 10 on Table LI. The selection of intervals of 10 tends 
 to screen the variation of the specific, heat of water from view, but a 
 smaller interval gives too many local irregularities. In taking the 
 mean I have given all the observations equal weight, but as the Kew 
 standard was only graduated to ? F. it was impossible to calibrate.it 
 BO accurately as to avoid irregularities of 0.02 C. which would affect 
 the quantities 1 in 500. , Hence, in drawing a curve through the 
 results, as given in the last column, I have almost neglected the Kew, 
 and have otherwise sought to draw a regular curve without points of 
 inflection. The figures in the last column I consider the best. 
 
 Table LIII. takes the mean values as found in Tables LL and LII., 
 and exhibits them with respect to the temperatures on the different 
 thermometers, to the different parts of the earth, and also gives the 
 reduction to the absolute scale. I am inclined to favor the absolute 
 scale, using m = .00015, as given in the Appendix to Thermometry, 
 rather than .00018, as ufeed throughout the paper. 
 
 Table LIV. gives what I consider the final result of the experiment. 
 It is based on the result m = .00015 for the thermometers, and is 
 corrected for the irregularity of the engine by adding 1 in 4000. 
 
 The minor irregularities are also corrected so that the results signify 
 a smooth curve, without irregularity or points of contrary flexure. 
 But the curve for the work does not differ more than three kilo- 
 gramme-meters from the actual experiment at any point, and generally 
 coincides with it to about one kilogramme-meter. These differences 
 signify 0.007 C. and 0.002 C., respectively. The mechanical equiv- 
 alent is for single degrees rather than for ten degrees, as in the 
 other tables.
 
 180 
 
 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 TABLE XXXVII. FIRST SERIES. Preliminary. 
 January 16, 1878. Jacket and Air about 14 C. 
 
 ji 
 
 Ja 
 H 
 
 140 ' 
 160 
 180 
 203 
 220 
 240 
 259 
 289 
 
 | 
 
 S 
 
 Correction. 
 
 Corrected 
 Temperature. 
 
 Revolutions of 
 Chronograph 
 
 Mean Weight 
 W. 
 
 Work per 
 Kilogramme. 
 
 I 
 
 Work per 
 Kilogramme. 
 
 IWork per 
 Kilogramme 
 + 6380 
 
 1 
 
 1 
 
 52.0 
 '56.0 
 59.2 
 63.4 
 66.5 
 70.2 
 74.0 
 80.0 
 
 .005 
 .003 
 
 +.006 
 -j-.Oll 
 +.020 
 +.028 
 +.045 
 
 
 
 .017 
 .022 
 .015 
 .001 
 +.027 
 +.067 
 +.161 
 
 9.185 
 11.412 
 13.650 
 16.230 
 18.137 
 20.392 
 22.538 
 25.943 
 
 5.485 
 18.023 
 30.652 
 45.329 
 56.241 
 69.158 
 81.484 
 101.214 
 
 7.509 
 
 7.478 
 7.442 
 7.394 
 7.364 
 7.354 
 7.292 
 
 
 951 
 1906 
 3010 
 3825 
 4786 
 5702 
 7156 
 
 o 
 
 io 
 n 
 
 12 
 
 13 
 14 
 15 
 16 
 17 
 18 
 19 
 20 
 21 
 22 
 23 
 24 
 25 
 26 
 
 348 
 775 
 1202 
 1629 
 2056 
 2484 
 2012 
 3340 
 3767 
 4193 
 4619 
 5048 
 5472 
 5899 
 6326 
 6753 
 7180 
 
 5728 
 6155 
 6582 
 7009 
 7436 
 7864 
 8292 
 8720 
 9147 
 9573 
 9999 
 10428 
 10852 
 11279 
 11706 
 12133 
 12560 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 TABLE XXXVIII. SECOND SERIES. Preliminary. 
 March 7, 1878. Jacket 18.5 to 22.5. Air about 21 C. 
 
 j 
 
 
 Correction. 
 
 $1 
 
 If 
 
 i 
 
 MS 
 
 j 
 
 J 
 
 jj 
 
 
 
 
 SI 
 
 fijfj 
 
 
 
 
 P.S 
 
 
 o 
 
 
 - 
 
 
 ifc 
 
 IS" 
 
 0^ 
 
 &te 
 
 1 
 
 S& 
 
 Je 
 
 Ji 
 
 | 
 
 1 
 
 1 
 
 n 
 
 Is 
 
 E 
 
 * S 1 
 
 1 
 
 g 
 
 ^ + 
 
 
 
 
 
 
 
 
 
 
 
 
 170 
 
 19.9 
 
 .016 
 
 
 
 12.537 
 
 6.03 
 
 
 
 
 13 
 
 198 
 
 7010 
 
 180 
 190 
 
 
 
 
 13.646 
 14 755 
 
 11.12 
 17 22 
 
 7.710 
 
 474 
 
 947 
 
 14 
 15 
 
 625 
 1052 
 
 7437 
 
 
 
 
 200 
 210 
 
 26.8 
 
 
 
 15.863 
 16.972 
 
 23.36 
 29.55 
 
 7.666 
 7.642 
 
 1421 
 1897 
 
 16 
 17 
 
 1480 
 1909 
 
 8292 
 8721 
 
 .010 
 
 .036 
 
 220 
 230 
 
 
 
 .... 
 
 18.085 
 19.196 
 
 35.70 
 41.90 
 
 7.630 
 
 2369 
 2845 
 
 18 
 19 
 
 2333 
 2761 
 
 9145 
 9573 
 
 24.0 
 250 
 260 
 270 
 280 
 290 
 
 33.8 
 
 +.003 
 
 .036 
 
 20.305 
 21.419 
 22.533 
 23.642 
 24.754 
 25.867 
 
 48.09 
 64.30 
 
 66.69 
 
 72.92 
 79.16 
 
 7.600 
 7.596 
 7.582 
 7.552 
 7.547 
 
 3319 
 3794 
 
 4740 
 5213 
 5687 
 
 20 
 21 
 22 
 23 
 24 
 25 
 
 3189 
 3615 
 4041 
 4467 
 4892 
 6318 
 
 10001 
 10427 
 10853 
 11279 
 11704 
 12130 
 
 
 
 *"* 
 
 40.8 
 
 +.020 
 
 .001 
 
 300 
 
 
 
 
 
 26.990 
 
 85.42 
 
 
 6164 
 
 26 
 
 6744 
 
 12556 
 
 * In the calculation of this column, more exact data were used than given in the other 
 two columns, seeing that the original calculation was made every 6 mm. of the thermometer. 
 Hence the last figure may not always agree with the rest of the data.
 
 OP ARTS AND SCIENCES. 
 
 181 
 
 TABLE XXXVIII. Continued. 
 
 ji 
 
 1 
 
 Correction. 
 
 Corrected 
 Temperature. 
 
 1 
 
 Mean Weight 
 TV. 
 
 8 
 
 1 
 
 Work per 
 
 Kilogramme. 
 
 1 
 
 1 
 
 I 
 
 310 
 320 
 330 
 340 
 350 
 360 
 370 
 3.80 
 390 
 
 4V.8 
 61.4 
 
 +.044 
 
 +.073 
 
 28.119 
 29.253 
 30.393 
 31.540 
 32.689 
 33.842 
 34.998 
 36.158 
 37.321 
 
 91.67 
 97.98 
 104.28 
 110.67 
 117.12 
 123.54 
 130.04 
 136.56 
 143.08 
 
 7.611 
 7.604 
 7.611 
 7.617 
 7.602 
 7.592 
 7.576 
 7.550 
 7.550 
 
 6643 
 7125 
 7608. 
 8097 
 8590 
 9081 
 9576 
 10071 
 10567 
 
 28 
 29 
 30 
 31 
 
 33 
 34 
 35 
 36 
 37 
 
 6168 
 6593 
 7017 
 7441 
 7867 
 8294 
 8722 
 9149 
 9577 
 10004 
 10430 
 
 12980 
 13405 
 13829 
 14253 
 14679 
 15106 
 15534 
 15961 
 16389 
 16816 
 17242 
 
 65.0 
 
 58.7 
 
 +.072 
 
 +588 
 
 +.184 
 +.261 
 
 
 
 
 
 
 
 
 
 
 
 
 
 TABLE XXXIX. THIRD SERIES. Preliminary. 
 March 12, 1878. Jacket 13.2 to 16.6. Air about!5 C. 
 
 Is 
 
 p 
 
 
 
 g 
 
 p 
 
 Correction. 
 
 Corrected 
 Temperature. 
 
 Revolutions of 
 Chronograph 
 2n. 
 
 i- 
 
 
 
 & |l 
 S3 g>^? 
 
 p S w 
 II 
 
 Temperature. 
 
 Work per 
 
 Kilogramme. 
 
 ijs 
 *&* 
 
 *a+ 
 
 00 
 
 I 
 
 205 
 210 
 220 
 230 
 240 
 250 
 260 
 270 
 280 
 290 
 300 
 310 
 320 
 330 
 340 
 350 
 360 
 370 
 380 
 390 
 400 
 410 
 420 
 
 28.0 
 28.6 
 29.9 
 31.1 
 32.4 
 33.6 
 34.9 
 36.2 
 37.4 
 38.7 
 39.9 
 41.2 
 42.5 
 43.7 
 45.0 
 46.3 
 47.6 
 48.9 
 50.1 
 51.4 
 52.7 
 64.0 
 55.3 
 
 
 
 
 +.663 
 
 
 +.002 
 
 +.6io 
 
 14.368 
 14.754 
 15.529 
 16.307 
 17.090 
 17.875 
 18.662 
 19.452 
 20.242 
 21.029 
 21.825 
 22.619 
 23.418 
 24.220 
 25.023 
 28.825 
 26.628 
 27.438 
 28.253 
 29.069 
 29.884 
 30.703 
 31.519 
 
 3.156 
 6.334 
 9.770 
 14.184 
 18.642 
 23.080 
 27.550 
 32.014 
 36.474 
 40.924 
 45.424 
 49.838 
 54.302 
 58.844 
 63.366 
 67.874 
 72.403 
 76.987 
 81.550 
 86.100 
 90.720 
 95.316 
 99.920 
 
 7.5167 
 
 7.5462 
 7.5668 
 7.5875 
 7.5763 
 I 7.5872 
 7.5801 
 
 
 164 
 
 495 
 827 
 1160 
 1495 
 1831 
 2167 
 2504 
 2840 
 3179 
 3514 
 3853 
 4194 
 4536 
 4876 
 5219 
 5565 
 5910 
 6255 
 6604 
 6951 
 7299 
 
 
 
 15 
 16 
 
 17 
 18 
 19 
 20 
 21 
 22 
 23 
 24 
 25 
 26 
 27 
 28 
 29 
 30 
 31 
 
 
 
 269 
 696 
 1122 
 1548 
 1975 
 2401 
 2828 
 3253 
 3676 
 4101 
 4526 
 4951 
 5378 
 5803 
 6226 
 6653 
 7078 
 
 7868 
 8295 
 8721 
 9147 
 9574 
 10000 
 10427 
 10852 
 11275 
 11700 
 12125 
 12550 
 12977 
 13402 
 13825 
 14252 
 14677 
 
 +.009 
 
 +.021 
 
 +.014 
 
 +.038 
 
 +.019 
 
 +.055 
 
 +.024 
 +.030 
 +.038 
 +.047 
 +.056 
 +.066 
 
 +.089 
 
 +.i20 
 
 +.159 
 +.202 
 +.251 
 +.304 
 
 * As this table was originally calculated for every 5 mm. on the thermometer, I have given the 
 weights which were used to check the more exact calculation.
 
 182 
 
 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 TABLE XL. FOURTH SERIES. Preliminary* 
 March 24, 1878. Jacket 5.4 to 8.2. Air about 6 C. 
 
 ft 
 
 
 Correction. 
 
 
 1| 
 
 t - 
 
 $ 
 
 
 
 
 
 
 fg 
 
 
 
 f 
 
 JL 
 
 
 S.ss 
 
 I 
 
 c 
 
 If 
 
 Rag 
 
 I s 
 
 Jafc 
 
 1 
 
 03 
 
 t 
 
 
 
 Si* 
 
 u 
 
 9* 
 
 *ly 
 
 ! 
 
 II 
 
 lit 
 
 130 
 140 
 150 
 160 
 
 27.4 
 29.2 
 31.0 
 32.9 
 
 +.002 
 
 
 
 8?071 
 9.204 
 10.340 
 11.480 
 
 42.364 
 48.898 
 55.438 
 62.066 
 
 7.471 
 
 7.446 
 7.442 
 
 
 
 485 
 968 
 1458 
 
 8 
 9 
 ]0 
 11 
 
 30 
 398 
 823 
 1252 
 
 4872 
 5300 
 5725 
 6154 
 
 +.010 
 
 +.019 
 
 170 
 180 
 
 34.7 
 36.6 
 
 +.017 
 
 +.050 
 
 12.620 
 13.763 
 
 68.669 
 75.330 
 
 7.390 
 
 1944 
 2433 
 
 12 
 13 
 
 1680 
 2107 
 
 6582 
 7009 
 
 190 
 200 
 
 38.4 
 403 
 
 +.025 
 
 +.093 
 
 14.908 
 16.054 
 
 81.973 
 88.597 
 
 7.431 
 
 2921 
 3410 
 
 14 
 15 
 
 2534 
 3960 
 
 7436 
 8862 
 
 210 
 220 
 
 42.2 
 44.2 
 
 +.034 
 
 +.150 
 
 17.202 
 18.350 
 
 95.264 
 101.941 
 
 7.437 
 
 3902 
 4395 
 
 16 
 17 
 
 3387 
 3815 
 
 8289 
 8717 
 
 230 
 
 46.1 
 
 +.046 
 
 +.222 
 
 19.604 
 
 108.588 
 
 
 4886 
 
 18 
 
 4245 
 
 9147 
 
 240 
 
 
 
 
 
 
 !. 
 
 
 19 
 
 4672 
 
 9574 
 
 250 
 
 
 
 
 
 
 7.4617 
 
 
 ''0 
 
 5098 
 
 10000 
 
 260 
 
 
 
 
 
 
 
 
 21 
 
 5524 
 
 10426 
 
 270 
 280 
 
 53.6 
 
 55.7 
 
 +.073 
 
 +.399 
 
 24.124 
 
 25.288 
 
 135.158 
 141.803 
 
 7.509 
 
 6855 
 7350 
 
 22 
 23 
 
 5950 
 6376 
 
 10852 
 11278 
 
 290 
 
 57.7 
 
 +.084 
 
 +.524 
 
 26.456 
 
 148.427 
 
 
 7844 
 
 24 
 
 6802 
 
 11704 
 
 
 
 
 
 
 
 .... 
 
 
 25 
 
 7228 
 
 12130 
 
 
 
 
 
 
 
 
 
 26 
 
 7651 
 
 12553 
 
 TABLE XLI. FIFTH SERIES. Preliminary. 
 March 24, 1878. Jacket 5.4 to 8.4. Air about 6 C. 
 
 1 . 
 
 
 Corre 
 
 tion. 
 
 4 
 
 * M 
 
 1 
 
 4 
 
 1 
 
 dj 
 
 
 p 
 
 1 
 
 I 
 
 i 
 
 l! 
 
 Ei 
 
 lf H 
 Ir 
 
 I* 
 
 IS? 
 
 I 
 
 If 
 
 f" 
 
 75 
 80 
 
 0.9 
 1.7 
 
 .003 
 
 
 
 L891 
 2451 
 
 3.154 
 
 6 118 
 
 8.1544 
 
 
 
 239 
 
 2 
 g 
 
 46 
 477 
 
 2296 
 2727 
 
 90 
 100 
 110 
 120 
 130 
 140 
 
 3.4 
 
 5.1 
 6.8 
 8.5 
 10.2 
 12,0 
 
 .002 
 
 b 
 
 +.003 
 
 .012 
 
 .61 7 
 .612 
 
 3.569 
 4.690 
 5.810 
 6.936 
 8.060 
 9190 
 
 12.174 
 18.172 
 24.212 
 30.397 
 36.621 
 42 854 
 
 8.0900 
 8.0409 
 8.0074 
 7.9170 
 7.8973 
 7.8786 
 
 723 
 1200 
 1677 
 2161 
 2647 
 3132 
 
 4 
 6 
 6 
 
 7 
 8 
 g 
 
 906 
 1332 
 1759 
 2189 
 2621 
 3050 
 
 3156 
 3582 
 4009 
 4439 
 4871 
 5300 
 
 150 
 160 
 
 13.7 
 15.5 
 
 +.007 
 
 +.005 
 
 10.323 
 11 459 
 
 49.068 
 55 398 
 
 7.8512 
 7.8061 
 
 3614 
 
 10 
 
 3477 
 
 5727 
 
 170 
 180 
 
 17.2 
 190 
 
 +.015 
 
 +.032 
 
 12.600 
 13 742 
 
 61.707 
 68036 
 
 '7.7799 
 7.7622 
 
 4588 
 
 12 
 
 4333 
 
 6583 
 7009 
 
 190 
 200 
 210 
 
 20.8 
 22.6 
 243 
 
 +.024 
 +.028 
 
 +.068 
 +.092 
 
 14.882 
 16.025 
 17 170 
 
 74.358 
 80.716 
 87 064 
 
 7.7643 
 7.7807 
 7.8419 
 
 5558 
 6047 
 
 14 
 
 15 
 
 5183 
 5608 
 
 7433 
 7858 
 
 QOQO 
 
 220 
 230 
 
 26.1 
 27.9 
 
 +.039 
 
 +.150 
 
 18.316 
 19.467 
 
 93.402 
 99.677 
 
 7.8468 
 7.8579 
 
 7030 
 7518 
 
 17 
 18 
 
 6466 
 6895 
 
 8716 
 9146 
 
 * The first part of the experiments were lost, as the pen of the chronograph did not work.
 
 OF ARTS AND SCIENCES. 
 TABLE XLI. Continued. 
 
 183 
 
 Thermometer 
 No. 6163. 
 
 , 
 
 Correction. 
 
 Corrected 
 Temperature. 
 
 Revolutions of 
 Chronograph 
 2n. 
 
 ^' 
 
 Work per 
 Kilogramme 
 = 2 9.8816 Wn. 
 
 Temperature. 
 
 Work per 
 Kilogramme. 
 
 1 
 
 I 
 
 I 
 
 240 
 250 
 260 
 270 
 280 
 290 
 300 
 310 
 
 29.6 
 
 +.050 
 
 +.270 
 
 20615 
 
 105.950 
 
 7.8802 
 f 7.8980 
 
 7.9038 
 7.9091 
 7.8979 
 7.8974 
 
 8006 
 
 9482 
 9976 
 10474 
 10974 
 11481 
 
 19 
 20 
 21 
 22 
 
 24 
 
 25 
 26 
 27 
 
 7320 
 7745 
 8170 
 8597 
 9024 
 9451 
 9878 
 10305 
 10733 
 11160 
 
 9570 
 9995 
 10420 
 10847 
 11274 
 11701 
 11128 
 12555 
 12983 
 13410 
 
 34'.9 
 36.7 
 38.5 
 40.2 
 42.1 
 
 
 
 
 
 +.069 
 +.087 
 +.109 
 
 +.351 
 +.450 
 +.583 
 
 24.072 
 25.231 
 26.395 
 27.565 
 28.748 
 
 124.863 
 131.181 
 137.560 
 143.972 
 150.467 
 
 
 
 
 
 
 
 TABLE XLIL SIXTH SERIES. 
 May 14, 1878. Jacket 12.l to 12.4. Air about 13 C. 
 
 lo 
 
 
 Correction. 
 
 if 
 
 ^ 
 
 sg- 
 
 %* 
 
 
 
 
 
 e 
 
 & 
 
 i 
 
 5 
 
 1.1 
 
 !ii 
 
 B* 
 
 
 
 
 I 
 
 
 5 is 
 
 Ii* 
 
 2 
 
 8 
 
 
 Jl 
 
 1 
 
 a 
 
 
 
 1 
 
 3J 
 
 II 
 
 i 
 
 *g 
 
 I 
 
 *I 
 
 if + 
 
 140 
 
 46.4 
 
 .002 
 
 
 
 9319 
 
 1.93 
 
 
 
 
 9 
 
 137 
 
 5296 
 
 150 
 
 47.9 
 
 
 
 10.178 
 
 7.07 
 
 7.2291 
 
 370 
 
 10 
 
 293 
 
 6726 
 
 160 
 
 49.4 
 
 ' '.000 
 
 .007 
 
 11.032 
 
 12.19 
 
 . 
 
 735 
 
 11 
 
 721 
 
 6154 
 
 170 
 
 50.9 
 
 
 
 11.886 
 
 17.37 
 
 } 7.1608 
 
 1102 
 
 12 
 
 1151 
 
 6584 
 
 180 
 
 62.5 
 
 +.002 
 
 .008 
 
 12.740 
 
 22.52 
 
 
 1467 
 
 13 
 
 1579 
 
 7012 
 
 190 
 
 54.0 
 
 
 
 13.596 
 
 27.70 
 
 7.1600 
 
 1835 
 
 14 
 
 2007 
 
 7440 
 
 200 
 
 55.5 
 
 +.006 
 
 .002 
 
 14.454 
 
 32.88 
 
 
 2201 
 
 15 
 
 2434 
 
 7867 
 
 210 
 
 57:0 
 
 
 
 15.314 
 
 38.07 
 
 7.1512 
 
 2568 
 
 16 
 
 2863 
 
 8296 
 
 220 
 
 68.5 
 
 +.6io 
 
 +.6ii 
 
 16.174 
 
 43.29 
 
 
 2938 
 
 17 
 
 3290 
 
 8723 
 
 230" 
 
 60.0 
 
 
 
 17.037 
 
 48.50 
 
 7.1446 
 
 8306 
 
 18 
 
 3716 
 
 9149 
 
 240 
 
 61.6 
 
 +.015 
 
 +.031 
 
 17^093 
 
 63.70 
 
 
 3675 
 
 19 
 
 4142 
 
 9575 
 
 250 
 
 
 
 
 
 
 7.1536 
 
 
 20 
 
 4567 
 
 10000 
 
 260 
 
 
 
 
 
 
 
 
 21 
 
 4993 
 
 10426 
 
 270 
 
 280 
 
 66.2 
 67.7 
 
 +.024 
 
 +.075 
 
 20.500 
 21.362 
 
 69.27 
 74.50 
 
 7.1230 
 
 4778 
 5148 
 
 22 
 23 
 
 6420 
 
 5846 
 
 10853 
 11279 
 
 290 
 300 
 
 69.2 
 70.7 
 
 +.031 
 
 +.113 
 
 22220 
 23.076 
 
 79.69 
 84.84 
 
 7.1344 
 
 5514 
 
 5878 
 
 24 
 25 
 
 6271 
 6696 
 
 11704 
 12129 
 
 310 
 320 
 
 72.2 
 73.7 
 
 +.039 
 
 +158 
 
 23.928 
 24.774 
 
 89.97 
 95.05 
 
 7.1302 
 
 6240 
 6600 
 
 26 
 27 
 
 7121 
 7547 
 
 12554 
 12980 
 
 330 
 340 
 
 75.2 
 76.2 
 
 +.047 
 
 +.2i2 
 
 25.624 
 
 26.467 
 
 100.19 
 105.27 
 
 7.1117 
 
 6962 
 7319 
 
 28 
 29 
 
 7973 
 8400 
 
 13406 
 13833 
 
 350 
 360 
 
 78.2 
 79.7 
 
 +.056 
 
 +.272 
 
 27.309 
 28.147 
 
 110.39 
 115.44 
 
 7.0958 
 
 7680 
 8035 
 
 30 
 31 
 
 8829 
 9259 
 
 14262 
 14692 
 
 370 
 380 
 
 81.2 
 
 82.7 
 
 +.065 
 
 +.341 
 
 28.990 
 29.825 
 
 120.57 
 125.66 
 
 7.1076 
 
 8396 
 8754 
 
 32 
 33 
 
 9678 
 10096 
 
 15111 
 15529 
 
 390 
 
 84.2 
 
 +.076 
 
 +.417 
 
 30.663 
 
 130.78 
 
 7.1088 
 
 9115 
 
 
 .... 
 
 
 400 
 
 85.7 
 
 
 
 31.505 
 
 135.90 
 
 
 9475 
 
 
 .... 
 
 
 410 
 420 
 
 87.2 
 88.7 
 
 +.087 
 
 +.504 
 
 32.377 
 33.226 
 
 140.98 
 146.08 
 
 7.1064 
 
 9833 
 10192 
 
 
 '.'.'.'. 

 
 184 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 TABLE XLIII. SEVENTH SERIES. 
 May 16, 1878. Jacket 1P.8 to 12. Air about 12 C. 
 
 Thermometer 1 
 No. 6163. 
 
 1 
 
 Correction. 
 
 if 
 1! 
 
 Revolutions of 
 Chronograph 
 
 Mean Weight 
 W. 
 
 1 
 
 Temperature. 
 
 Work per 
 Kilogramme. 
 
 1 
 
 1 
 
 i 
 
 130 
 140 
 
 30.9 
 32.2 
 
 .004 
 
 
 
 8538 
 9.315 
 
 5.07 
 9.73 
 
 7.2350 
 
 
 335 
 
 9 
 
 199 
 
 6296 
 
 150 
 160 
 170 
 
 33'.6 
 350 
 36.3 
 
 .002 
 
 .006 
 
 10.094 
 10.875 
 11.654 
 
 14.36 
 18.98 
 23.56 
 
 7.3011 
 
 668 
 1003 
 1335 
 
 10 
 11 
 12 
 
 628 
 1056 
 1484 
 
 5725 
 6153 
 
 6581 
 
 
 
 .010 
 
 180 
 
 37.6 
 
 
 
 12.433 
 
 28.16 
 
 7.3165 
 
 1670 
 
 13 
 
 1913 
 
 7010 
 
 190 
 200 
 
 38.9 
 40.2 
 
 +.003 
 
 .008 
 
 13.209 
 13.984 
 
 32.74 
 37.31 
 
 7.3460 
 
 2003 
 2337 
 
 14 
 15 
 
 2344 
 2770 
 
 7441 
 
 7867 
 
 210 
 220 
 
 41.5 
 
 42.8 
 
 +.006 
 
 .000 
 
 14.758 
 15.536 
 
 41.84 
 46.38 
 
 7.3094 
 
 2667 
 2998 
 
 16 
 17 
 
 3196 
 3623 
 
 8293 
 8720 
 
 230 
 240 
 
 44^2 
 45.5 
 
 +.010 
 
 +.013 
 
 16.317 
 17.103 
 
 50.99 
 55.62 
 
 i 7.2846 
 
 3332 
 3667 
 
 18 
 19 
 
 4052 
 4478 
 
 9149 
 9575 
 
 250 
 260 
 
 46.9 
 48.3 
 
 +.014 
 
 +.032 
 
 17.891 
 18.682 
 
 60.29 
 
 } 
 7.2822 
 
 4005 
 
 20 
 21 
 
 4906 
 5324 
 
 10003 
 10421 
 
 270 
 280 
 
 49.6 
 50.9 
 
 +.619 
 
 +.056 
 
 19.475 
 20.269 
 
 69.63 
 74.34 
 
 7.2610 
 
 468i 
 . 5021 
 
 22 
 23 
 
 6754 
 6179 
 
 10851 
 11276 
 
 290 
 300 
 
 52.3 
 53.6 
 
 +.025 
 
 +.090 
 
 21.079 
 21.866 
 
 79.01 
 83.71 
 
 7.2504 
 
 5358 
 5697 
 
 24 
 25 
 
 6603 
 7028 
 
 11700 
 12125 
 
 310 
 320 
 
 65.0 
 56.4 
 
 +.032 
 
 +.127 
 
 22.665 
 23.471 
 
 88.42 
 93.14 
 
 7.2893 
 
 6037 
 6379 
 
 26 
 27 
 
 7454 
 
 7883 
 
 12551 
 12980 
 
 330 
 
 67.8 
 
 +.039 
 
 +.172 
 
 24.281 
 
 97.88 
 
 
 6722 
 
 28 
 
 8307 
 
 13404 
 
 340 
 
 59.2 
 
 
 
 25.088 
 
 102.61 
 
 7.3047 
 
 7065 
 
 29 
 
 872'.) 
 
 13826 
 
 350 
 
 60.5 
 
 +.046 
 
 +.222 
 
 25.896 
 
 107.36 
 
 
 7410 
 
 30 
 
 9157 
 
 14254 
 
 360 
 
 61.9 
 
 
 .... 
 
 26.706 
 
 112.14 
 
 7.3389 
 
 7759 
 
 31 
 
 9582 14679 
 
 370 
 
 63.2 
 
 +.055 
 
 +.279 
 
 27.523 
 
 116.88 
 
 1 ' 
 
 8104 
 
 32 
 
 10009 
 
 15106 
 
 380 
 
 64.6 
 
 
 .... 
 
 28.346 
 
 121.62 
 
 \ 7.4109 
 
 8454 
 
 
 .... 
 
 .... 
 
 390 
 
 66.0 
 
 +.065 
 
 +.345 
 
 29.172 
 
 126.34 
 
 1 . , 
 
 8801 
 
 
 
 
 400 
 
 67.4 
 
 
 
 29^996 
 
 13L12 
 
 7.4356 
 
 9155 
 
 
 
 
 
 410 
 
 68.8 
 
 +.075 
 
 +.419 
 
 30.827 
 
 135.90 
 
 
 9508 
 
 
 
 
 420 
 
 70.1 
 
 +.080 
 
 +.456 
 
 31.653 
 
 140.66 
 
 
 9861 
 
 
 .... 
 
 
 TABLE XLIV. EIGHTH SERIES. 
 May 23, 1878. Jacket 16.2 to 16.5. Air about 20 C. 
 
 ii 
 
 Ji 
 
 1 
 
 Correction. 
 
 Corrected 
 Temperature". 
 
 Revolutions of 
 Chronograph 
 2n. 
 
 Mean Weight 
 
 ,|* 
 
 IgjW 
 
 i 
 1 
 
 Work per 
 Kilogramme. 
 
 *9* 
 
 & 
 
 
 S 
 
 1 
 
 230 
 240 
 250 
 260 
 270 
 280 
 290 
 
 23.9 
 25.4 
 26.8 
 28.3 
 29.7 
 31.2 
 32.7 
 
 .007 
 
 
 
 16287 
 17.063 
 
 39.120 
 43.982 
 
 6.9137 
 > 6.9358 
 
 6.9007 
 6.9125 
 
 
 333 
 
 i338 
 1673 
 2010 
 
 
 
 17 
 18 
 19 
 20 
 21 
 22 
 
 306 
 
 735 
 1163 
 1592 
 2019 
 2446 
 
 8715 
 9144 
 9572 
 10001 
 10428 
 10866 
 
 
 
 
 
 .000 
 
 +.005 
 
 19.405 
 20.190 
 20.978 
 
 58.602 
 63503 
 68.428 
 
 .... 

 
 OP ARTS AND SCIENCES. 
 TABLE XLIV. Continued. 
 
 185 
 
 I 
 
 
 Correction. 
 
 . 
 
 ** 
 
 
 
 | 
 
 . 
 
 
 
 5 
 
 
 
 ll 
 
 1!.- 
 
 f 
 
 !i| 
 
 I 
 
 la 
 
 ig 
 
 jM 
 
 i 
 
 1 
 
 i 
 
 l| 
 
 r 
 
 1* 
 
 ||| 
 
 | 
 
 : 
 
 Iff 
 
 
 H 
 
 3 
 
 
 
 EH 
 
 PSO 
 
 s 
 
 n 
 
 fcH 
 
 P M 
 
 N 
 
 300 
 
 34.2 
 
 
 
 21/765 
 
 73.351 
 
 6.8878 
 
 2346 
 
 23 
 
 2871 
 
 11280 
 
 310 
 320 
 
 35.6 
 37.1 
 
 +.008 
 
 +.040 
 
 22.554 
 23.350 
 
 78.283 
 83.245 
 
 6.8866 
 6.8594 
 
 2682 
 3020 
 
 24 
 25 
 
 3298 
 3722 
 
 11707 
 12131 
 
 330 
 
 38.6 
 
 
 
 24.151 
 
 88.314 
 
 6.8358 
 
 3363 
 
 26 
 
 4150 
 
 12559 
 
 340 
 350 
 
 40.1 
 41.6 
 
 +.6i7 
 
 +.085 
 
 24.952 
 25.751 
 
 93.294 
 98.275" 
 
 6.8748 
 6.9184 
 
 3702 
 4044 
 
 27 
 28 
 
 4574 
 4999 
 
 12983 
 14408 
 
 360 
 
 43.1 
 
 
 
 26.552 
 
 103.232 
 
 6.9444 
 
 4385 
 
 29 
 
 5423 
 
 13832 
 
 370 
 380 
 
 44.6 
 46.0 
 
 +.028 
 " " 
 
 +.144 
 '*" 
 
 27.361 
 28.175 
 
 108.216 
 113.269 
 
 6.9291 
 6.9338 
 
 4727 
 5074 
 
 30 
 31 
 
 5851 
 6275 
 
 14260 
 14684 
 
 390 
 
 47.5 
 
 
 
 28.989 
 
 118.231 
 
 6.9385 
 
 5418 
 
 
 
 .... 
 
 400 
 410 
 
 49.0 
 50.6 
 
 +.039 
 
 +217 
 
 29.800 
 30.624 
 
 123.329 
 128.399 
 
 6.9444 
 6.9467 
 
 5766 
 6115 
 
 
 
 .... 
 
 .... 
 
 420 
 
 52.1 
 
 +.047 
 
 +.281 
 
 31.445 
 
 133.480 
 
 6.9314 
 
 6464 
 
 
 
 .... 
 
 TABLE XL V. NINTH SERIES. 
 May 27, 1878. Jacket 19.6 to 20. Air about 23 C. 
 
 I 
 
 
 Correction. 
 
 g* 
 
 J 
 
 ^ 
 
 life 
 
 4 
 
 
 
 C 6 
 
 
 
 
 II 
 
 lii 
 
 t 
 
 ll 
 
 3 
 
 M 
 
 *l 
 
 tii 
 
 
 
 
 1 
 
 
 1 
 
 fl 
 
 (Sw 
 
 i 
 
 " S ll 
 
 1 
 
 
 Mr 
 
 200 
 
 380 
 
 .015 
 
 
 
 15890 
 
 6.33 
 
 
 
 
 16 
 
 47 
 
 8293 
 
 210 
 
 394 
 
 
 
 17:000 
 
 11.74 
 
 8.8108 
 
 473 
 
 17 
 
 473 
 
 8719 
 
 220 
 
 409 
 
 .011 
 
 .010 
 
 18.106 
 
 17.17 
 
 
 946 
 
 18 
 
 901 
 
 9147 
 
 230 
 
 49, 3 
 
 
 
 19.219 
 
 22.62 
 
 8.7341 
 
 1419 
 
 19 
 
 1326 
 
 9572 
 
 240 
 
 438 
 
 .005 
 
 .011 
 
 20.329 
 
 28.13 
 
 
 1895 
 
 20 
 
 1754 
 
 10000 
 
 250 
 
 45.3 
 
 
 
 21.442 
 
 33.68 
 
 
 2368 
 
 21 
 
 2180 
 
 10426 
 
 260 
 270 
 
 
 +.002 
 
 .004 
 
 22.552 
 23659 
 
 
 [ 8.4800 
 
 .... 
 
 22 
 23 
 
 2606 
 3031 
 
 10852 
 11277 
 
 280 
 290 
 
 49.8 
 51.3 
 
 +.009 
 
 +.012 
 
 24.771 
 
 25.885 
 
 60.55 
 56.25 
 
 ) 
 
 8.4399 
 
 3785 
 4263 
 
 24 
 
 25 
 
 3457 
 3883 
 
 11703 
 12129 
 
 300 
 310 
 
 52.9 
 544 
 
 +.019 
 
 +.037 
 
 27.006 
 28.133 
 
 61.93 
 67.63 
 
 8.4765 
 
 4737 
 5215 
 
 26 
 27 
 
 4312 
 4734 
 
 12558 
 12980 
 
 320 
 330 
 
 56.0 
 
 575 
 
 +.029 
 
 +.072 
 
 29.264 
 30.404 
 
 73.36 
 79.15 
 
 8.4552 
 
 5697 
 6182 
 
 28 
 29 
 
 5159 
 5584 
 
 13405 
 13830 
 
 340 
 350 
 
 59.1 
 60 fi 
 
 +.042 
 
 +.118 
 
 31.552 
 32.702 
 
 84.97 
 90.85 
 
 } 8.4015 
 
 6669 
 7159 
 
 30 
 31 
 
 6010 
 6435 
 
 14256 
 14681 
 
 360 
 370 
 
 62.2 
 638 
 
 +.056 
 
 +.173 
 
 33.853 
 35.011 
 
 96.78 
 102.66 
 
 [ 8.4222 
 
 7652 
 8143 
 
 32 
 33 
 
 6860 
 7286 
 
 15106 
 15532 
 
 380 
 390 
 
 65.4 
 670 
 
 +.071 
 
 +.242 
 
 36.170 
 37.331 
 
 108.59 
 114.45 
 
 ) 
 
 8.4706 
 
 8638 
 9128 
 
 34 
 35 
 
 7714 
 8138 
 
 15960 
 16384 
 
 400 
 410 
 
 68.6 
 
 70.2 
 
 +.088 
 
 +.322 
 
 38.497 
 39.664 
 
 120.36 
 126.33 
 
 8.4316 
 
 9626 
 10126 
 
 36 
 37 
 
 8565 
 8988 
 
 16811 
 17234 
 
 420 
 
 71.8 
 
 +.105 
 
 +.419 
 
 40.833 
 
 132.26 
 
 
 10620 
 
 38 
 39 
 
 9414 
 9842 
 
 17660 
 18Q88 
 
 
 
 
 
 
 
 
 
 40 
 
 10268 
 
 18514 
 
 
 
 
 
 
 
 
 
 41 
 
 10691 
 
 18937
 
 186 
 
 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 TABLE XL VI. TEHTH SERIES. 
 June 3, 1878. Jacket 18.l to 18.4. Air about 20 C. 
 
 
 
 
 
 
 
 g 
 
 
 
 
 b 
 
 
 Correction. 
 
 
 
 "Sji 
 
 ^ 
 
 |P 
 
 S 
 
 i 
 
 
 
 if 
 
 
 
 fl 
 
 If" 
 
 1 
 
 ||1 
 
 5 
 g 
 
 11 
 
 Iff 
 
 
 
 H 
 
 I 
 
 1 
 
 1 
 
 k 
 
 II 
 
 |v 
 
 *$% 
 
 H 
 
 *B 
 
 
 250 
 
 4.1 
 
 .007 
 
 
 
 17838 
 
 7.82 
 
 
 
 
 18 
 
 69 
 
 9145 
 
 260 
 
 7.0 
 
 
 
 18.617 
 
 
 4.3899 
 
 
 19 
 
 496 
 
 9572 
 
 270 
 
 9.9 
 
 .003 
 
 +.004 
 
 19.401 
 
 23.'l9 
 
 
 '667 
 
 20 
 
 925 
 
 10001 
 
 280 
 
 12.8 
 
 
 
 20.188 
 
 3095 
 
 4.3919 
 
 1005 
 
 21 
 
 1350 
 
 10426 
 
 290 
 
 15.7 
 
 +.003 
 
 +.020 
 
 20.978 
 
 38.70 
 
 
 1341 
 
 22 
 
 1778 
 
 10854 
 
 300 
 
 18.7 
 
 
 
 21.763 
 
 46.41 
 
 i 4.3912 
 
 1676 
 
 23 
 
 2204 
 
 11280 
 
 310 
 320 
 
 21.6 
 24.5 
 
 +.008 
 
 +.037 
 
 22.551 
 23.354 
 
 54.21 
 62.04 
 
 ) 
 { 4.3907 
 
 2014 
 2354 
 
 24 
 25 
 
 2627 
 3054 
 
 11703 
 12130 
 
 330 
 340 
 
 27.5 
 30.5 
 
 +.014 
 
 +.078 
 
 24.162 
 24 970 
 
 69.92 
 
 77.92 
 
 J 
 4.3624 
 
 2696 
 3041 
 
 26 
 27 
 
 3479 
 3904 
 
 12555 
 12980 
 
 350 
 360 
 
 336 
 36.6 
 
 +.020 
 
 +.132 
 
 25.780 
 26.593 
 
 85.89 
 93.94 
 
 4.3542 
 
 3385 
 3731 
 
 28 
 29 
 
 4332 
 4852 
 
 13408 
 13828 
 
 370 
 
 396 
 
 +.028 
 
 +.198 
 
 27.415 
 
 102.05 
 
 
 4081* 
 
 30 
 
 5179 
 
 14255 
 
 380 
 
 42.7 
 
 
 
 28.246 
 
 110.34 
 
 4.3362 
 
 4437 
 
 31 
 
 5604 
 
 14680 
 
 390 
 
 45.8 
 
 +.036 
 
 +.281 
 
 29.079 
 
 118.49 
 
 
 4786 
 
 
 .... 
 
 .... 
 
 400 
 
 48.9 
 
 
 
 29.911 
 
 126.66 
 
 4.3978 
 
 5141 
 
 
 
 
 410 
 
 52.0 
 
 +.044 
 
 +.377 
 
 30.754 
 
 134.89 
 
 
 5499 
 
 
 .... 
 
 .... 
 
 TABLE XLVII. ELEVENTH SERIES. 
 June 19, 1878. Jacket 19.6 to 20. Air about 23 C. 
 
 Thermometer 
 No. 6163. 
 
 
 
 5 
 H 
 
 Correction. 
 
 g 
 SI 
 
 
 
 ReTolutions of 
 Chronograph 
 2n. 
 
 Mean Weight 
 
 III 
 
 O C OJ 
 
 s w 
 
 Temperature. 
 
 Work per 
 Kilogramme. 
 
 *3 + 
 
 1 
 
 I 
 
 250 
 260 
 270 
 280 
 290 
 300 
 310 
 320 
 330 
 340 
 350 
 360 
 370 
 380 
 390 
 400 
 410 
 420 
 
 ... 
 
 .002 
 
 +.002 
 
 
 
 +.006 
 
 +.029 
 +.063 
 
 -f.iis 
 
 2L450 
 22.562 
 
 24.789 
 25.007 
 27.032 
 28.168 
 29.307 
 30.456 
 31.612 
 32 774 
 
 8.933 
 16.087 
 
 30.281 
 37.439 
 44.655 
 61.848 
 69.098 
 66.390 
 73.724 
 81.163 
 88.462 
 95.734 
 103.093 
 110.560 
 118.121 
 126.693 
 133.250 
 
 '6.7572 
 6.7678 
 
 6.7749 
 6.7896 
 6.7973 
 | 6.8188 
 6.9165 
 6.7876 
 6.7808 
 
 
 
 476 
 
 i42J 
 
 1899 
 2379 
 286p 
 3344 
 3832 
 4323 
 4817 
 5311 
 5807 
 6307 
 6808 
 7311 
 7815 
 8321 
 
 21 
 22 
 23 
 
 24 
 
 26 
 27 
 28 
 29 
 30 
 31 
 32 
 33 
 34 
 35 
 36 
 37 
 38 
 39 
 40 
 41 
 
 192 
 235 
 602 
 
 1087 
 1511 
 1939 
 23*55 
 2789 
 3214 
 3638 
 4003 
 4488 
 4913 
 5337 
 6760 
 6187 
 6614 
 7040 
 7465 
 7891 
 8317 
 
 10428 
 10855 
 11282 
 11707 
 12131 
 12569 
 12985 
 13409 
 13834 
 14258 
 14683 
 15108 
 15533 
 15957 
 16380 
 10807 
 17234 
 17660 
 18085 
 18511 
 18937 
 
 
 +.010 
 +.019 
 +.031 
 
 
 +.043 
 
 +.177 
 
 
 +.058 
 +.072 
 
 +.257 
 +.351 
 
 33.939 
 35.110 
 36.280 
 37.456 
 38.637 
 39.821 
 41.010 
 
 
 +.087 
 +.106 
 
 +.463 
 +.596 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

 
 OF ARTS AND SCIENCES. 
 
 187 
 
 0961 + 
 tauiBjSonji 
 
 J3d Jt.lO.ii. 
 
 ~ ~ X ^1 1^ I-H i 
 
 work pe 
 
 ilogram 
 
 CC W CO C3i * CC CO O * rti <N t- 
 
 !M (M O5 CC O 
 
 ^io *^io^ 
 
 coco 'T v i'M~>j 
 
 co co ' co ao co 
 
 a>Q5oa>Qajao0Q S 
 
 
 qdBoSouoaqo 
 jo 
 
 SiggggSSSSScSiSSSg g S5 : : : 
 
 (6 co cc t-^ t^ CD co to o so d o co t-^ t-^ co t^ 05 *i t^ r I I 
 
 jo suonn[OA8H 
 
 
 
 1 1 1 ++++++++++++++++++ 
 
 .I._I_._I_._L._L.O --^ -co - 
 
 \^ . T^ . l^ . |7 . |^ .(?q CO '-^t* 'i 
 
 ' "M 7* * 
 
 .. 1 .. .. .. .. . 
 
 '8919 J 
 
 ffl 
 
 JB 
 
 CO O lO O O O 'CO -CD 
 CO 'CO CO 'CO '"^ *O "lO '*O 
 
 Ico Ico Ico led Ico" Ico Ico !cd I I I 
 
 8919 -ON *q 
 
 J3J3UIUOIBOJO 
 
 oanjK.iadiu.JX * 
 
 r-! cs co * vo co r^ oJ o r cc 
 
 'S9T9 'ON g ; 
 J3;3uiouijaqx 
 
 oo 
 to .
 
 188 
 
 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 TABLE XLIX. THIRTEENTH SERIES. 
 Dee. 19, 1878. Jacket 3.2 to 3.5. Air 4.2 to 5.2 C. 
 
 55 
 
 Corrections. 
 
 
 
 ^ 
 
 ^ 
 
 ,-i 
 
 
 
 
 S 
 
 |i 
 
 
 if 
 
 IL 
 
 I 
 
 f|* 
 
 I 
 
 1 
 
 El 
 
 1 
 
 
 ; 
 
 
 H 
 
 I 
 
 I 
 
 || 
 
 If 
 
 G**" 
 
 * 6 a! 
 
 1 
 
 1 
 
 H 
 
 I 
 
 70 
 
 
 
 
 
 1248 
 
 1.72 
 
 
 
 
 
 Oj 
 
 106 
 
 1858 
 
 
 
 
 
 
 8.6610 
 
 485.0 
 
 
 
 
 
 80 
 
 .... 
 
 ' 
 
 2.378 
 
 7.38 
 
 
 
 485.0 
 
 2 
 
 +323 
 
 2287 
 
 
 
 
 
 
 8.5571 
 
 485.1 
 
 
 
 
 
 90 
 
 
 
 -.003 
 
 3.500 
 
 13.11 
 
 
 
 970.1 
 
 3 
 
 754 
 
 2718 
 
 
 
 
 
 
 8.4325 
 
 482.2 
 
 
 
 
 
 100 
 
 
 
 4.626 
 
 18.89 
 
 
 
 1452.3 
 
 4 
 
 1184 
 
 3148 
 
 
 
 
 
 
 8.3688 
 
 481.1 
 
 
 
 
 
 110 
 
 +.001 
 
 +.003 
 
 5.751 
 
 24.70 
 
 
 
 1933.4 
 
 5 
 
 1612 
 
 3576 
 
 
 
 
 
 
 8.4155 
 
 487.1 
 
 
 
 
 
 120 
 
 
 
 6.881 
 
 30.55 
 
 
 
 2420.5 
 
 6 
 
 2041 
 
 4005 
 
 
 
 
 
 
 8.4189 
 
 485.6 
 
 
 
 
 
 130 
 
 +.005 
 
 +.019 
 
 8.013 
 
 36.38 
 
 
 
 2906.1 
 
 7 
 
 2472 
 
 4436 
 
 
 
 
 
 
 8.3953 
 
 489.2 
 
 
 
 
 
 140 
 
 
 
 9.148 
 
 42.27 
 
 
 
 3395.3 
 
 g 
 
 9001 
 
 .Q_ 
 
 
 
 
 
 
 8.4366 
 
 486.6 
 
 
 
 
 
 150 
 
 +.009 
 
 +.044 
 
 10.284 
 
 48.10 
 
 
 
 3881.9 
 
 9 
 
 3331 
 
 5295 
 
 
 
 
 
 
 8.4484 
 
 486.5 
 
 
 
 
 
 160 
 
 
 
 11.424 
 
 53.92 
 
 
 
 AnpQ A 
 
 - -. 
 
 Q7Pf\ 
 
 _^ . 
 
 
 
 
 
 
 8.4189 
 
 490.6 
 
 
 
 
 &12A 
 
 170 
 
 +.016 
 
 +.080 
 
 12.569 
 
 59.81 
 
 
 
 4859.0 
 
 11 
 
 4187 
 
 6151 
 
 
 
 
 
 
 8.3988 
 
 491.1 
 
 
 
 
 
 180 
 
 
 
 13.713 
 
 65.72 
 
 
 
 5350.1 
 
 12 
 
 4615 
 
 6579 
 
 
 
 
 
 
 8.4153 
 
 487.1 
 
 
 
 
 
 190 
 
 +.023 
 
 +.126 
 
 14.859 
 
 71.57 
 
 
 
 5837.2 
 
 13 
 
 5045 
 
 7009 
 
 
 
 
 
 
 8.3811 
 
 491.7 
 
 
 
 
 
 200 
 
 
 
 
 
 16.005 
 
 77.50 
 
 
 
 6328.9 
 
 14 
 
 5472 
 
 7436 
 
 210 
 
 +.033 
 
 +.183 
 
 17.154 
 
 83.40 
 
 8.3835 
 
 489.4 
 
 6818.3 
 
 15 
 
 5898 
 
 7862 
 
 
 
 
 
 
 8.3976 
 
 490.2 
 
 
 
 
 
 220 
 
 
 
 
 18.300 
 
 89.30 
 
 
 
 7308.5 
 
 16 
 
 6327 
 
 8291 
 
 
 
 
 
 
 8.4035 
 
 493.0 
 
 
 
 
 
 230 
 
 +.044 
 
 +.251 
 
 19.452 
 
 95.23 
 
 
 
 7801.5 
 
 17 
 
 6753 
 
 8717 
 
 
 
 
 
 
 8.4460 
 
 496.4 
 
 
 
 
 
 240 
 
 
 
 
 20.604 
 
 101.17 
 
 
 
 8297.9 
 
 18 
 
 7180 
 
 9144 
 
 250 
 
 +.056 
 
 +.332 
 
 21.760 
 
 
 
 f- 8.4555 
 
 981.3 
 
 
 
 19 
 
 7608 
 
 9572 
 
 260 
 
 
 
 
 22.912 
 
 112.90 
 
 
 
 9279.2 
 
 20 
 
 8038 
 
 10002 
 
 270 
 
 +.069 
 
 +.424 
 
 24.065 
 
 118.81 
 
 8.4602 
 
 494.7 
 
 9773.9 
 
 21 
 
 8465 
 
 10429 
 
 280 
 
 
 
 25.221 
 
 124.70 
 
 8.4779 
 
 494.0 
 
 10267.9 
 
 22 
 
 8891 
 
 10855 
 
 
 
 
 
 
 
 
 
 
 
 23 
 
 9317 
 
 11281 
 
 
 
 
 
 
 
 
 
 25 
 
 9746 
 10173 
 
 11710 
 12137 
 
 
 
 
 
 
 

 
 OF ARTS AND SCIENCES. 
 
 189 
 
 TABLE L. FOURTEENTH SERIES. 
 December 20, 1878. Jacket 1.5 to 1.9. Air about 3.4 C. 
 
 Temperature 
 by Kew 
 Standard. 
 
 1 
 
 Corrections. 
 
 Corrected Tem- 
 perature, Abso- 
 lute Scale. 
 
 Revolution of 
 Chronograph 
 
 Mean Weight 
 TV. 
 
 Work per 
 Kilogramme 
 = 2 9.7832 Wn. 
 
 
 
 I 
 
 Work per 
 Kilogramme. 
 
 Work per Kilo- 
 gramme -f 2210. 
 
 Reduction 
 to Absolute 
 Scale. 
 
 I 
 
 1 
 
 36.0 
 
 66.0 
 
 .00 
 
 
 
 
 
 1.82 
 
 8.03 
 
 7.3682 
 
 
 
 
 
 2 
 
 77 
 
 2287 
 
 38 5 
 
 584 
 
 
 
 
 3.23 
 
 1637 
 
 7.3458 
 
 '601 
 
 3 
 
 * 503 
 
 2713 
 
 41.0 
 
 .9 
 
 .01 
 
 .00 
 
 +.01 
 
 4.62 
 
 24.78 
 
 7.3705 
 
 1206 
 
 4 
 
 936 
 
 3146 
 
 43.5 
 
 3.3 
 
 
 .... 
 
 .... 
 
 6.02 
 
 33.19 
 
 7.4012 
 
 1812 
 
 6 
 
 1370 
 
 3580 
 
 46.0 
 
 5.8 
 
 .02 
 
 +.01 
 
 +.04 
 
 7.43 
 
 41.48 
 
 7.4142 
 
 2412 
 
 6 
 
 1803 
 
 4013 
 
 485 
 
 8 2 
 
 
 
 
 884 
 
 4981 
 
 7 4177 
 
 3016 
 
 7 
 
 2226 
 
 4436 
 
 51.0 
 
 10.7 
 
 .03 
 
 +.02 
 
 +.09 
 
 10.26 
 
 68.18 
 
 7.4390 
 
 3624 
 
 8 
 
 2656 
 
 4866 
 
 535 
 
 13 2 
 
 
 
 
 11 68 
 
 66 56 
 
 7 4107 
 
 4234 
 
 9 
 
 3084 
 
 5294 
 
 56.0 
 
 15.6 
 
 .04 
 
 +.03 
 
 +.16 
 
 13.12 
 
 74.95 
 
 7.3493 
 
 4842 
 
 10 
 
 3513 
 
 5723 
 
 58.5 
 
 18? 
 
 
 
 
 14.56 
 
 83.56 
 
 7.3269 
 
 5461 
 
 11 
 
 3942 
 
 6152 
 
 61.0 
 
 20.7 
 
 .04 
 
 +.05 
 
 +.25 
 
 16.01 
 
 92.27 
 
 7.2335 
 
 6085 
 
 12 
 
 4369 
 
 6579 
 
 63.5 
 
 23.3 
 
 
 .... 
 
 
 
 17.46 
 
 100.99 
 
 7.1603 
 
 6703 
 
 13 
 
 4790 
 
 7000 
 
 66.0 
 
 25.9 
 
 .05 
 
 +.06 
 
 +.38 
 
 18.92 
 
 109.95 
 
 7.2076 
 
 7330 
 
 14 
 
 5220 
 
 7430 
 
 685 
 
 28 5 
 
 
 
 
 2039 
 
 118 84 
 
 7 1839 
 
 7957 
 
 15 
 
 5650 
 
 7860 
 
 71.0 
 
 31.2 
 
 .05 
 
 +.08 
 
 +.62 
 
 21.86 
 
 127.83 
 
 7.2122 
 
 8689 
 
 16 
 
 6081 
 
 8291 
 
 73.5 
 
 33.8 
 
 
 
 ..:. 
 
 23.34 
 
 136.75 
 
 7.2252 
 
 9218 
 
 17 
 
 6507 
 
 8717 
 
 76.0 
 
 36.5 
 
 .06 
 
 +.10 
 
 +.69 
 
 24.84 
 
 145.78 
 
 7.2134 
 
 9857 
 
 18 
 
 6935 
 
 9145 
 
 785 
 
 392 
 
 
 
 
 26 33 
 
 15480 
 
 
 10493 
 
 19 
 
 7364 
 
 9574 
 
 
 
 
 
 
 
 
 
 
 ?0 
 
 7791 
 
 10001 
 
 
 
 
 
 
 
 
 
 
 ftl 
 
 8219 
 
 10429 
 
 
 
 
 
 
 
 
 
 
 9,9: 
 
 8648 
 
 10858 
 
 
 
 
 
 
 
 
 
 
 TO 
 
 9074 
 
 11284 
 
 
 
 
 
 
 
 
 
 
 ?4 
 
 9499 
 
 11709 
 
 
 
 
 
 
 
 
 
 
 fln 
 
 9925 
 
 12135 
 
 
 
 
 
 
 
 
 
 
 26 
 
 10352 
 
 12562 
 
 
 
 
 
 

 
 190 
 
 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 ifc 
 
 h 
 
 O 
 
 M 
 
 u 
 
 II 
 
 ERS AT BALTIMORE TO HEA 
 GIVEN TEMPERATURE ON TH 
 
 anoq aad 
 1!^ Si *q3 
 IS *<X 'OS 
 
 nn f 8 ?u 3 PA 
 
 8919 '61 ^ 
 
 -jnoq aa 
 
 un z'L 
 
 jnoq Jad C J8 
 'lin 8'9 I^PM 
 8919 6t 3un f 
 
 jnoq jad 
 
 til -^ >qa 
 
 jnoq aad 
 'III 9'8 
 
 jnoq jad ( 
 R1 6'9 ?U a 
 
 anoq aad 
 1PI 8'9 ? 
 9919 
 
 anoq aad 
 IW I'i 
 
 9919 
 
 anoq jd 
 '111 8'i 'R 3 
 '8919 '88U 
 
 <M O S 
 
 s s 
 
 oocoSiSoo^oo 
 
 
 
 & S ^ o !? 
 
 I- rH O O * 
 
 . cc CO ^ "^ I'M O t^- 
 " * oa K i i o IS T* GO 
 
 . . o 5 S c 1 S S 
 
 I^-O 
 
 O " 
 
 '8919 -sawasi 
 
 jnoq jad 
 HI 
 
 'ilDllI 
 
 Jnoq jad o8 8 
 Til 9'i m3\A\. 
 
 jnoq jad c 
 I!3J Vi iqSp 
 '8919 -9I-n 
 
 S S SIS 
 S g &fe 
 
 1-1 (M CO *! U3 CO
 
 OF ARTS AND SCIENCES. 
 
 191 
 
 O O i (^H 
 
 oaoo 
 
 
 ills 
 
 T< CO CO CO 
 CO O CO O 
 
 l^- I-H 
 O CO 
 
 l> lt* i 'COTfOCOO^OCOCOOO 
 TJI r~ Q <N tO CO O CO O CO O <M "-O CO 
 
 ri |>o S S 51 r- -. 3 S 5 cp .c* S 
 
 
 t- <M |Q tp SN 
 
 ^i-OO'*<CO!Mt^i-l>OO5^fc6c^COr-IOC5COCO<M'Ol.O 
 "'^ISSSS^SS^SS^^IfHSlSSSrt^SS 
 
 HO^CMi-iCOOOt^i-iOSWCOlMOTtl 
 
 -.^t^OCMOCOOCOOCOOCCOCO 
 
 
 CN !M ^H CJ 
 
 
 M iO ISO i-^ 
 
 OSCOGicO 
 
 t~ co a 
 
 s i 
 
 S 8 

 
 192 
 
 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 -u3iu mcud 
 
 wrapf 
 IBjaaoo 
 
 oj oq oq p 
 
 : |j ^ S 
 
 . iO *< iO (N O 
 ! C<1 <M ' (M <N (M 
 
 rH CO
 
 OF ARTS AND SCIENCES. 
 
 193 
 
 ' 8 
 
 CO O CO CO CO >O >6 
 
 3 5J | Si 3 $ 3 
 
 O r-4 
 
 Si Si Si 
 
 q <=>. -i 
 
 8 
 
 cs co cq p >o i 
 
 s 8 s s 
 
 1-4 co 10 
 
 t- Tt< <M CO 
 
 5 8 
 
 o r-j IN q 
 o o u5 u6 
 
 3 $ $ 5! 
 
 ir-O-^cspeqNCNieoia 
 
 ^CNGMCMCMeScMCMCNffj 
 
 8 
 
 . 
 
 q en cq 
 
 CM CM CM 
 
 cot 
 
 ScN 
 
 VOL. xv. (N. s. vn.) 
 
 13
 
 194 
 
 PROCEEDINGS OF THE AMERICAN ACADEMY 
 
 tric 
 to th 
 
 Th 
 lass 
 
 CO uOiCO<M<N ^00000 
 
 - ^^^Oit^-*O < ^'- H 
 OOO3OCiCVC5 
 
 1000 
 Sai 
 
 sisggggggs 
 
 SIOOO' = "* 
 
 gfOOO' = 
 3aiBfl 
 
 i^covO^ 
 tDGOO C 1 ! 
 
 g- 'aiotn 
 
 t^or--oco 
 
 06 o o i-t 
 
 CO Tf< 
 
 0; P 
 
 'Z9I9 nipmjg 
 
 8 8 8 2 S 
 
 22 
 
 i~i" 
 
 "SIOOO" = 
 8ajsn 'ajnjB 
 
 o o 
 o o
 
 OF ARTS AND SCIENCES. 
 
 195 
 
 js p p p i-j c<; G* 
 
 $ 8 s3 S 8 
 
 CSppOOOrHCNI 
 
 c^coco'cococococo 
 
 CSCNIC3'M-M<M<NCNI 
 
 to t-^ oi p c^i "^ 
 
 C<l <M (M (?) -M (N 
 
 2f^t2^"?it^S 
 
 8 S 2 2' 
 
 CO <N 00 CN O 
 
 o t^ oo o :N 
 
 (M O) 
 
 t--co^ccoocog 
 
 t^rJ(O*-DCMaOiOCOiOC5-^<OCOCOO>a'- l ?DlMCO^OiiOC5 
 ^t*t^-O^iOt^-OC^iOt OCNtOr^-O^or-OC^Ot OT*i 
 
 OJOSOOO' ; -'^SN<NIMCOCOT}H-<lliaibiOc6ot~r^COCOCO 
 
 5S8S228S 
 
 5S8S2JSS 
 
 co co r~ t^ co co 
 
 S o S S S S 
 SS 2 g =5 S 
 
 C50>CiOOOOOO 
 .pppr-jr-;^-;^-.-;^ 
 
 e4eo4ig<DroooaQ>H 
 
 COCOCOCOCOCOCOCO-tl-* 
 
 gggggggo 
 
 sssssssssss 
 
 ^5cie<S'fli?strotfc6QTH 
 
 S ? 
 
 CO **<< i.O 
 
 CO CO CO 5! 
 
 c<i CN o^ c<) 
 
 co as o .-j 
 
 t- t-- CO CO CO 
 
 . *: ^: -: -j : *. ; ; : 
 
 i <C x 1CO^ l ^COt^GOOO* * 
 COCOCOCOCOCOCOCOCO-*T}1 
 
 oooooooo^ 
 pppppppp 
 
 SSg?5^?5^S 
 
 t-H (N CO ^ 10 ?6 t^ GO C: O O 
 
 COCOCOCOCOCOCOCOCO'^^
 
 196 
 
 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 TABLE LIV. FINAL MOST PROBABLE RESULTS. 
 
 J. 
 
 Work. 
 
 Mechanical 
 Equivalent. 
 
 
 
 Work. 
 
 Mechanical 
 Equivalent. 
 
 
 | 
 
 sf 
 
 ft,* 
 
 6 a 
 
 || . 
 
 h| 
 
 B 
 ftj? 
 
 be 
 
 ^-. 
 
 g? II 
 
 ill 
 
 j2oa 
 
 UJ 
 
 t*l 
 
 
 I e 1 
 
 loi 
 
 P a 
 
 Ji^K 
 
 ll 5 
 
 III 
 
 d 
 
 gii 
 
 II* 
 
 H 
 
 Pi 
 
 M 
 
 
 PI 
 
 Jr 
 
 
 
 00000. 
 
 
 0000. 
 
 
 
 00000. 
 
 
 0000. 
 
 s 
 
 2289 
 
 2443 
 
 
 
 22 
 
 10852 
 
 10835 
 
 426.1 
 
 4176 
 
 3 
 
 2720 
 
 2865 
 
 
 
 23 
 
 11278 
 
 11253 
 
 426.0 
 
 4175 
 
 4 
 
 3150 
 
 3286 
 
 
 
 24 
 
 11704 
 
 11670 
 
 425.9 
 
 4174 
 
 5 
 
 3580 
 
 3708 
 
 429.8 
 
 4212 
 
 25 
 
 12130 
 
 12088 
 
 425.8 
 
 4173 
 
 6 
 
 4009 
 
 4129 
 
 429.5 
 
 4209 
 
 26 
 
 12556 
 
 12505 
 
 425.7 
 
 4172 
 
 7 
 
 4439 
 
 4550 
 
 4293 
 
 4207 
 
 27 
 
 12982 
 
 12922 
 
 425.6 
 
 4171 
 
 8 
 
 4868 
 
 4970 
 
 429.0 
 
 4204 
 
 28 
 
 13407 
 
 13339 
 
 425.6 
 
 4171 
 
 9 
 
 6297 
 
 5390 
 
 428.8 
 
 4202 
 
 29 
 
 13833 
 
 13756 
 
 425.5 
 
 4170 
 
 10 
 
 5726 
 
 6811 
 
 428.5 
 
 4200 
 
 30 
 
 14258 
 
 14173 
 
 425.6 
 
 4171 
 
 11 
 
 6154 
 
 6230 
 
 428.3 
 
 4198 
 
 31 
 
 14684 
 
 14950 
 
 425.6 
 
 4171 
 
 12 
 
 6582 
 
 6650 
 
 428.1 
 
 4196 
 
 32 
 
 16110 
 
 15008 
 
 425.6 
 
 4171 
 
 13 
 
 7010 
 
 7070 
 
 427.9 
 
 4194 
 
 33 
 
 15535 
 
 15425 
 
 425.7 
 
 4172 
 
 14 
 
 7438 
 
 7489 
 
 427.7 
 
 4192 
 
 34 
 
 15961 
 
 15842 
 
 425.7 
 
 4172 
 
 15 
 
 7865 
 
 7908 
 
 427.4 
 
 4189) 
 
 35 
 
 16387 
 
 16259 
 
 425.8 
 
 4173 
 
 16 
 
 8293 
 
 8327 
 
 427.2 
 
 4187 
 
 36 
 
 16812 
 
 16676 
 
 425.8 
 
 4173 
 
 17 
 
 8720 
 
 8745 
 
 427.0 
 
 4185 
 
 37 
 
 17238 
 
 17094 
 
 
 
 18 
 
 9147 
 
 9164 
 
 426.8 
 
 4183 
 
 38 
 
 17664 
 
 17511 
 
 
 
 19 
 
 9574 
 
 9582 
 
 426.6 
 
 4181 
 
 89 
 
 18091 
 
 17930 
 
 
 
 20 
 
 10000 
 
 10000 
 
 426.4 
 
 4179 
 
 40 
 
 18517 
 
 18347 
 
 
 
 21 
 
 10426 
 
 10418 426.2 
 
 4177 
 
 41 
 
 18943 
 
 18765 
 
 
 
 
 
 TABLE LV. QUANTITY TO ADD TO THE EQUIVALENT AT BALTIMOBE TO 
 EEDUCE TO ANT LATITUDE. 
 
 Latitude. 
 
 Addition in 
 Kilogramme-Meters. 
 
 
 
 + 0.89 
 
 10 
 
 + 0.82 
 
 20 
 
 + 0.63 
 
 30 
 
 + 0.34 
 
 40 
 
 + 0.08 
 
 50 
 
 0.41 
 
 60 
 
 -0.77 
 
 70 
 
 1.06 
 
 80 
 
 1.26 
 
 90 
 
 . 1.33 
 
 Manchester 0.5; Paris 0.4; Berlin 0.6.
 
 OF ARTS AND SCIENCES. 197 
 
 V. CONCLUDING REMARKS. AND CRITICISM OF RESULTS 
 AND METHODS. 
 
 On looking over the last four columns of Table LIU., which gives 
 the results of the experiments as expressed in terms of the different 
 mercurial thermometers, we cannot but be impressed with the unsatis- 
 factory state of the science of thermometry at the present day, when 
 nearly all physicists accept the mercurial thermometer as the standard 
 between and 100. The wide discrepancy in the results of calori- 
 metric experiments requires no further explanation, especially when 
 physicists have taken no precaution with respect to the change of zero 
 after the heating of the thermometer. They show that thermometry 
 is an immensely difficult subject, and that the results of all physicists 
 who have not made a special study of their thermometers, and a com- 
 parison with the air thermometer, must be greatly in error, and should 
 be rejected in many cases. And this is specially the case where 
 Geissler thermometers have been used. 
 
 The comparison of my own thermometers with the air thermometer 
 is undoubtedly by far the best so far made, and I have no improve- 
 ments to offer beyond those P have already mentioned in the " Ap- 
 pendix to Thermometry." And I now believe that, with the improve- 
 ment to the air thermometer of an artificial atmosphere of constant 
 pressure, we could be reasonably certain of obtaining the tempera- 
 ture at any point up to 50 C. within O.01 C. from the mean of two 
 or three observations. I believe that my own thermometers scarcely 
 differ much more than that from the absolute scale at any point up to 
 40 C., but they represent the mean of eight observations. However, 
 there is an uncertainty of O.01 C. at the 20 point, owing to the un- 
 certainty of the value of m. But taking m = .00015, I hardly think 
 that the point is uncertain to more than that amount for the ther- 
 mometers Nos. 6163, 6165, and 6166. 
 
 As to the comparison of the other thermometers, it is evidently 
 unsatisfactory, as they do not read accurately enough. However, the 
 figures given in Table LIU. are probably very nearly correct. 
 
 The study of the thermometers from the different makers intro- 
 duces the question whether there are any thermometers which stand 
 below the air thermometer between and 100. As far as I can find, 
 nobody has ever published a table showing such a result, although 
 Boscha infers that thermometers of " Cristal de Choisy-le-Roi " should 
 stand below, and his inference has been accepted by Regnault. But 
 it does not seem to have been proved by direct experiment. My
 
 198 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 Baudin thermometers seem to contain lead as far as one can tell from 
 the blackening in a gas flame, but they stand very much above the air 
 thermometer at 40. I have since tried some of the Baudin ther- 
 mometers up to 300, and find that they stand below the air thermom- 
 eter between 100 and 240 ; they coincide at about 240, and stand 
 above between 240 and 300. This is very nearly what Regnault 
 found for " Verre Ordinaire." It is to be noted that the formula 
 obtained from experiments below 100 makes them coincide at 233, 
 which is remarkably close to the result of actual experiment, especially 
 as it would require a long series of experiments to determine the 
 point within 10. 
 
 The comparison of thermometers also shows that all thermometers 
 in accurate investigations should be used as thermometers with arbitrary 
 scales, neither the position of the zero point nor the interval between 
 the and 100 points being assumed correct. The text books only 
 give the correction for the zero point, but my observations show that 
 the interval between the and 100 points is also subject to a secular 
 change as well as to the temporary change due to heating. Of all 
 the thermometers used, the Geissler is the worst in this as in other 
 respects, except accuracy of calibration, in which it is equal to most of 
 the others. 
 
 The experiments on the specific heat of water show an undoubted 
 decrease as the temperature rises, a fact which will undoubtedly sur- 
 prise most physicists as much as it surprised me. Indeed, the dis- 
 covery of this fact put back the completion of this paper many months, 
 as I wished to make certain of it. There is now no doubt in my mind, 
 and I put the fact forth as proved. The only way in which an error 
 accounting for this decrease could have been made appears to me to 
 be in the determination of m in " Thermometry." The determination 
 of m rests upon the determination of a difference of only O.05 C. be- 
 tween the air thermometer and the mercurial, the and 40 points 
 coinciding, and also upon the comparison of the thermometers with 
 others whose value of m was known, as in the Appendix. Although 
 the quantity to be measured is small, yet there can be no doubt at least 
 that m is larger than zero ; and if so, the specific heat of water 
 certainly has a minimum at about 30. 
 
 One point that might be made against the fact is that the Kew 
 standard, Table L., gives less change than the others. But the cali- 
 bration of the Kew standard, although excellent, could hardly be 
 trusted to 0.02 or 0.03 C., as the graduation was only to J F. In 
 drawing the curve for the difference between the Kew standard and
 
 OP ARTS AN.D SCIENCES. 199 
 
 the air thermometers, I ignored small irregularities and drew a regu- 
 lar curve. On looking over the observations again, I see that, had I 
 taken account of the small irregularities, it would have made the ob- 
 servations agree more nearly with the other thermometers. Hence 
 the objection vanishes. However, I intend working up some obser- 
 vations which I have with the Kew standard at a higher temperature, 
 and shall publish them at a future time. 
 
 There is one other error that might produce an apparent decrease 
 in the specific heat, and that is the slight decrease in the torsion weight 
 from the beginning to the end of most of the experiments, probably due 
 to the slowing of the engine. By this means the torsion circle might 
 lag behind. I made quite an investigation to see if this source of error 
 existed, and came to the conclusion that it produced no perceptible 
 effect. An examination of the different experiments shows this also ; 
 for in some of them the weight increases instead of decreases. See 
 Tables XXXVII. to L. 
 
 The error from the formation of dew might also cause an apparent 
 decrease ; but I have convinced myself by experiment, and others can 
 convince themselves from the tables, that this error is also inappre- 
 ciable. 
 
 The observations seem to settle the point with regard to the specific 
 heat at the 4 point within reasonable limits. There does not seem 
 to be a change to any great extent at that point, but the specific 
 heat decreases continuously through that point. It would hardly be 
 possible to arrive at this so accurately as I have done by any method 
 of mixture, for Pfaundler and Platter, who examined this point, could 
 not obtain results within one per cent, while mine show the fact 
 within a fraction of one per cent. 
 
 The point of minimum cannot be said to be known, though I have 
 placed it provisionally between 30 and 35 P C., but it may vary much 
 from that. 
 
 The method of obtaining the specific heat of the calorimeter seems 
 to be good. The use of solder introduces an uncertainty, but it is too 
 small to affect the result appreciably. The different determinations 
 of the specific heat of the calorimeter do not agree so well as they 
 might; but the error in the equivalent resulting from this error is very 
 small, and, besides, the mean result agrees well with the calculated 
 result. It may be regarded as satisfactory. 
 
 The apparatus for determining the equivalent could scarcely be 
 improved much, although perhaps the record of the torsion might be 
 made automatic and continuous. The experiment, however, might be
 
 200 PROCEEDINGS OP THE AMERICAN ACADEMY 
 
 improved in two ways ; first, by the use of a motive power more 
 regular in its action ; and, second, by a more exact determination of 
 the loss due to radiation. The effect of the irregularity of the engine 
 has been calculated as about 1 in 4,000, and I suppose that the error 
 due to it cannot be as much as that after applying the correction. 
 The error due to radiation is nearly neutralized, at least between 
 and 30, by using the jacket at different temperatures. There may 
 be an error of a small amount at that point (30) in the direction of 
 making the mechanical equivalent too great, and the specific heat may 
 keep on decreasing to even 40. 
 
 Between the limits of 15 and 25 I feel almost certain that no 
 subsequent experiments will change my values of the equivalent so 
 much as two parts in one thousand, and even outside those limits, say 
 between 10 and 30, I doubt whether the figures will ever be changed 
 much more than that amount. 
 
 It is my intention to continue the experiments, as well as work up 
 the remainder of the old ones. I shall also use some liquids in the 
 calorimeter other than water, and so have the equivalent in terms of 
 more than one fluid. 
 
 BALTIMORE, 1878-79. Finished May 27, 1879.
 
 NOTE FROM PROFESSOR ROWLAND. 
 
 [Comparison with Dr. Joule's Thermometer.] 
 
 Dr. Joule has kindly sent me the comparison of my thermometer, No. 
 6166, with his, and the result will be published in full in the " Proceed- 
 ings of the American Academy of Sciences." In this manner I have 
 been able to make the exact reduction of his results to the air thermo- 
 meter. The following are the results: 
 
 DATE. 
 
 METHOD. 
 
 Temperature 
 of water. 
 
 o a 
 
 t 
 
 Joule's value re- 
 duced to the air 
 thermometer,and 
 to the latitude of 
 Baltimore. 
 
 Kowland's 
 value. 
 
 . 
 Diflerence. 
 
 English 
 System. 
 
 Metric 
 System. 
 
 1847 
 
 Friction of water. . . 
 
 15. 
 
 781.5 
 
 787.0 
 
 442.8 
 
 427.4 
 
 +15.4 
 
 1850 
 
 " .. 
 
 14. 
 
 772.7 
 
 778.0 
 
 426.8 
 
 427.7 
 
 .9 
 
 " 
 
 " " mercury 
 
 9. 
 
 772.8 
 
 779.2 
 
 427.5 
 
 428.8 
 
 1.3 
 
 
 
 . 
 
 9. 
 
 775.4 
 
 781.4 
 
 428.7 
 
 428.8 
 
 1 
 
 
 
 " " iron 
 
 9. 
 
 776.0 
 
 782.2 
 
 429.1 
 
 428.8 
 
 + .3 
 
 " 
 
 " " .... 
 
 9. 
 
 773.9 
 
 780.2 
 
 428.0 
 
 428.8 
 
 - .8 
 
 1867 
 
 Electric heating... 
 
 18.6 
 
 
 
 428.0 
 
 426.7 
 
 + 1.8 
 
 1878 
 
 Friction of water. . . 
 
 14.7 
 
 772.7 
 
 776.1 
 
 425.8 
 
 427.6 
 
 1.8 
 
 u _ 
 
 12.7 
 
 774.6 
 
 778.5 
 
 427.1 
 
 428.0 
 
 .9 
 
 ,. 
 
 15.5 
 
 773.1 
 
 776.4, 
 
 426.0 
 
 427.3 
 
 1.3 
 
 ' . j .< <. .. m 
 
 14.5 
 
 767.0 
 
 770.5 
 
 422.7 
 
 427.5 
 
 4.8 
 
 ' .. 
 
 17.3 
 
 774.0 
 
 777.0 
 
 426.3 
 
 426.9 
 
 .6 
 
 The mean difference of the two amounts to only 1 in 430, almost exactly 
 the same as I estimated in the body of my paper, an extremely satisfac- 
 tory result. 
 
 H. A. ROWLAND. 
 
 BALTIMORE, February 16th, 1880.
 
 lilfiill 
 
 AA 001204057 2