TRANSACTIONS OF SECTION A. 517 



fiieeta from 0-22 millimetre to 1"40 millimetre thick. The rubber was alternated 

 ■with sheets of tinfoil and heated for two hours in a steam-jacket at 100° C. A 

 jnodification of Regnault's method of mixtures was employed to find the specific 

 fceat of the hot rubber. Owin<^ to the non-conducting nature of the substance tlie 

 time of the calorimeter attaining its maximum temperature may be as long as ten 

 minutes ; hence it has been necessary to apply special formulae for the correction due 

 to cooling. Those described by Pfaundler, Pape, and Schuster have been used, and 

 the results calculated in accordance with them. The mean of the best-conducted 

 experiments gives for the Para rubber the number •481. The investigation is 

 being extended to allied bodies, especially the different forms of vulcanised rubber 

 and gutta-percha. 



4. On the Temporary Thermo-current in Iron. By F. T. Trouton. 



If a portion of an iron wire connected up with a galvanometer be heated red- 

 Ikot and the heated portion be caused to travel along the wire by moving the 

 'fiame a current is produced. This current is generally greatest in a newly takea 

 wire, but ultimately settles down to a constant thing on repeated heatings. 



It is this residual effect which I refer to as a temporary thermo-current in 

 •order to emphasise the fact that unlike, in the case of other thermo-currents, the 

 eonsideration of time here enters. At first sight, however, I did not see that this 

 •was essential, and I attributed the occurrence of the current to dis-symmetry intro- 

 •duced by the temperature gradient along the wire in front of the moving flame 

 being steeper than that behind, as it evidently must be, owing to the cooling 

 •taking place more slowly than the heating. But this can easily be shown not to 

 be the case, for artificially cooling the wire as the flame passes along, so as to 

 make the temperature gradient behind steepest, does not alter the direction of 

 the current. Also if the occurrence of the current depended merely on the tem- 

 perature gradients it should be possible to obtain a current without moving the 

 flame by arranging that the gradient on each side should be different, but this is 

 not the case. Thus it is seen that the movement along the wire is essential, and 

 that the electromotive force cannot be represented as a function of the temperature 

 gradient alone, but that the rate of change of the gradient xvith time is involved. 



However, what seems to be the chief interest attaching to this phenomenon is 

 its connection with the phenomenon known as recalescence. Representing the 

 temperature by the ordinates of the curve shown in the figure drawn along the 

 wire the state of things is exhibited which exists when the flame is stationary. If 

 the height of the dotted line above the wire represents the temperature of reca- 



lescence we may consider the portion of the wire lying between the points of inter- 

 section of the dotted line with the curve as being thermo-electrically a different 

 metal to the rest of the wire, for from the phenomenon of recalescence we know 

 that it is in a different molecular state. Let us call the unaltered wire to right 

 und left ' metal A ' and that in the centre ' metal B.' As long as the flame is 

 Stationary the two thermo-electric junctions A B, and B A, are at the same 

 temperature, and no current will occur, but this will not be the case when the 

 flame is moved if there is any delay in changing from the ' A ' state to the ' B ' 

 state, or back, that is to say if the temperature at which the change of state occurs 

 ■on the rise is higher than that at which it occurs on the fall, as is the case with 

 the melting-point of many organic bodies, especially when they are rapidly heated 

 or cooled, the point of liquefying being higher than the point of solidifying. Thus 

 the junction behind the flame will be at a lower temperature than the junction in 

 front, as is intended to be shown in the figure by the dotted vertical lines; the 

 arrow shows the direction the flame moves ; a current will thus ensue, provided 

 * tnetal A ' is as above stated thermo-electrically different from ' metal B.' The 



