Two platinum tubes, closed at their lower ends, 6 centims. in length, 1.2 centim. wide, the walls of the tubes being 1.2 millim. thick, were placed side by side at a distance of 1.2 centim., and held in this position by means of a crosspiece of platinum, from which projected perpendicularly a platinum rod by means of which the two tubes could be moved about at will. Two pieces of platinum, of the total weight of 32-39 grms., were brought into one of the tubes; and in the other such a quantity of the substance under examination was placed as would have a heat-value nearly equal to that of the 32-39 grms. of platinum. The tubes were closed by asbestos stoppers, and placed in a small muffle made of sheet copper about 2 millims. thick, which was entirely closed and then placed at the back of the muffle of a newly constructed gas muffle-furnace, in which any temperature between 500° and 1000° could easily be obtained. From the arrangement and position of the tubes it might with safety be assumed that after the expiry of a sufficient time the contents of both would be at the same temperature. When the tubes had remained in the furnace for an hour, the perpendicular platinum rod was seized by a pair of very hot tongs and the tubes very quickly transported to a double water-calorimeter standing as close by as possible. At the proper moment, while the tubes were held in the operator's right hand, the asbestos stoppers were withdrawn simultaneously by the left hand, and the contents of the two tubes thrown, by a jerk of the right hand, the platinum into one, the carbon into the other calorimeter. After a little practice the time which elapsed from the withdrawal of the hot tubes from the furnace until the moment when the contents of the tubes found themselves in the calorimeter was reduced to 3 or at the most 4 seconds. The loss of heat during this brief time affected both tubes in an exactly similar manner, inasmuch as the tubes, in respect of size, shape, material, and temperature, were exact counterparts of one another. Considering the briefness of the time and also the thickness of the walls of the platinum tubes, the reduction in the temperature of the enclosed substances must have been extremely small. Inasmuch as the weights of the two substances were equalized so that the product of the weight into the specific heat was the same for each, the small reduction of temperature might be taken as the same in each case. Against the assumption that the temperature of the platinum was the same as that of the substance under examination at the moment when they both fell into the calorimeter, no argument of importance can be urged. The double calorimeter employed consisted of two identical vessels of the thinnest sheet copper placed parallel to one another at a distance of 2 millims. The cover of each instrument had two openings in it. Through one passed a thermometer, while the other, the larger of the two, served as the mouth of the calorimeter. The mouths of the two instruments were at such a distance from each other that the middle point between them coincided with the point midway between the two platinum tubes. Each calorimeter was provided with a copper stirrer; the two stirrers could, by a simple piece of mechanism, be set in motion simultaneously. The water-value of the first instrument, with thermometer and stirrer, was 3.15 grms.; the water-value of the second was 3.35 grms. Nearly the same quantity of water was used in each, in order that, by bringing into each equally heated substances whose heat-values were the same, the same rise of temperature might be obtained in each case. This was actually found to be the case, to within 0°.1, in all experiments. The thermometers used with these calorimeters were graduated to hundredths of a degree. The stirrers were set in motion 15 minutes before each trial, the alteration of temperature being noted every five minutes in each instrument. Both calorimeters showed always the same change of temperature, this change never amounting to more than +0°-006 per minute. The temperature was noted immediately before the heated substances were thrown in; as the substance touched the water, no development of vapour was noticed in any case, nor could even the slightest sound be heard indicative of the escape of gas-bubbles. The stirrers were kept in operation until fully ten minutes had elapsed from the time when the heated substances fell into the calorimeters, the temperature of each instrument being read off minute by minute, in order that the fall of temperature per minute might be obtained. The maximum of temperature always occurred before the expiry of the first minute, after which the temperature very steadily sunk. The temperature at the end of the second minute was taken as the definite final temperature; and to this was added the correction A (AO+3A6'), where AO and AO' represent the temperature-alterations per minute before and after the heated substance was thrown into the calorimeter. This correction never reached 2 per cent. of the temperature-increase At of the calorimeter. k If G, and G, represent the weights of platinum and of the other substance, K, used, Q, and Q, the water-values, At aud At the rise of temperature, and t, and t, the corrected final temperatures in the respective calorimeters, then the common initial temperature of the platinum and of the substance K, and also the average specific heat, C-T, of the substance K may be obtained by solving the two equations At QA0-03232 (T-1) +0.0000041 (T-t,2), (1) G It were an impossible request to demand for experiments such as these, made at temperatures so high, an accuracy equal to that which attaches to measurements conducted at low temperatures; sources of error will unavoidably creep in. If by more exact researches Pouillet's expression T W=0-03237(T-T) +0.0000041(T2-To2) be found not wholly correct, this will but slightly alter the results of the experiments described in the present paper. The numerical value attaching to the specific heat of carbon at high temperatures may be somewhat modified; yet the general result, that the two modifications of carbon have for about 600° an almost identical specific heat, will remain unchanged. II. THE SPECIFIC HEAT OF CARBON. A. Specific Heat of Diamond. In my former estimation of the specific heat of diamond* I made use of two clear crystals belonging to the Berlin Mineralogical Museum, of the total weight of 1.061 grm. When this small weight of substance was heated 20° or 30° and brought into the ice-calorimeter, the mercury thread receded not more than 20 to 40 divisions of the scale-tube. From further experiments I found that the mercury thread is never moved to the exact point corresponding to the amount of heat developed in the calorimeter, but that it always stops a little before reaching this point. If, then, the total displacement of the mercurythread be small, this error in the proper position of the thread will cause the results calculated therefrom to be too low. On this account my former determinations of the specific heat of diamond from 0° to 30° were too small. In order to reduce this error to the smallest possible limits, the only plan was to use a larger amount of diamond. Through the kindness of the late Prof. G. Rose, the entire diamond collection belonging to the Berlin University was placed at my disposal. I selected the following stones : * Ber. der deut. chem. Ges. vol. iv.; and Pogg. Ann. 1872, part ii. The total weight was therefore 3.562 grms. The thermic relations of these six specimens were not altog ther identical. From a number of determinations, the avera specific heats for the temperatures 0° to 100° were deduced The evidently not perfectly homogeneous masses D, E, and would therefore have been rejected had it not been for the cor sideration that the determination of the specific heat for th temperature-interval -30° to +30° could not be trusted beyon at the furthest, the third place of decimals when working wit the quantities of diamond A, B, and C only. This degree of certainty I did not, however, deem sufficien for determining the existence of an inflection-point in the cur of the specific heat indicated by preliminary experiments. As th determination of this point appeared to me of the utmost impor tance, I sacrificed the advantage of working with absolutely pur material for the gain of being able to determine with the greates possible accuracy the specific heat of the diamond at and neart the temperature 0°. In all the series of experiments the tota quantity of diamond mentioned above was used. The following Tables give the data and results of 27 experi ments, having for their object the determination of the specifi heat of diamond for nine different temperatures between -80 and +280°. G=the weight of diamond used. T = the temperature, measured by the air-thermometer, o the diamond at the moment when it touched the water in the calorimeter. N=the number of divisions on the scale-tube through which the mercury thread was displaced after bringing the substance heated to T into the instrument. (The direct readings were reduced to the average bore of the tube, and then corrected for the independent oscillation of the mercury thread before the introduction of the hot substance.) No=the number of divisions indicating the displacement of the mercury thread when one heat-unit was developed in the calorimeter; W=the number of heat-units, calculated from N and No, developed by the bringing of G grms. of the substance, heated to the temperature T, into the calorimeter ; Co-r=the average specific heat between the temperature 0° and T, deduced from W, T, and G, and W=the amount of heat required to raise a unit-weight of the To substance from To to T. a. Experiments within the Temperatures -80° to 280°, carried out by means of the Ice-calorimeter. W214-2-414. Phil. Mag. S. 4. Vol. 49. No. 324. March 1875. N |