between Metallic Masses having different Temperatures. 455 



cess which he had traced out. I consider it essential to point 

 out on what grounds I dissented from a theory supported by 

 two of the first names in British science, before I proceed to give 

 any opinion of my own, which may perhaps be liable to equally 

 strong objections, but the data of which are not the less valuable 

 as physical facts. 



55. Waving all minor objections, I conceive that the process 

 of the communication of heat, and consequently its effects, would 

 be very different from what has been stated in the passage just 

 quoted. Let Fig. 9. represent on an exaggerated scale the pre- 

 sumed state of the apparatus in the middle of an oscillation ; the 

 hot bar A, whilst performing its vibration upon one of the solid 

 angles a, has expanded a portion of the cold block BC into a hil- 

 lock at d ; when the semi-vibration is completed, the angle b of 

 the bar will touch the block, and raise a new hiUock at the cor- 

 responding point c, whilst the elevation at d subsides ; and so 

 on alternately. Let us conceive that de is the finite depth to 

 which heat is communicated in the minute portion of time oc- 

 cupied by a semi- vibration, a depth so small as to be inappre- 

 ciable by the senses, and insignificant compared to the distances 

 between the points of impact dc. The elevation or height of the 

 hillock da is the amount of expansion of the element de, by the 

 accession of temperature received during a semi-vibration ; the 

 question is, what relation will this expansion, or acquired van- 

 tage-ground for the commencement of a new vibration, bear to 

 the nature of the block BC, considering the nature and tempera- 

 ture of the bar A, and the initial temperature of the block, to 

 be constant ? It surely requires no elaborate demonstration to 

 prove that the amount of caloric which passes into the block 

 must increase with the conducting power of the material. Up- 

 on the very fundamental axioms of the theory of heat, the 

 amount of caloric which passes from a molecule A into a mole- 

 cule B in an infinitely short interval of time, is proportional to 



