THIRD SERIES— REFRANGIBILITY OF HEAT. 193 



the pile was at d. The other parts maintaining the same positions, it is evident 

 that the pile must be moved into the position cV, if the source of heat be now one 



yielding rays of greater refrangibility. Although the radius of the circular arc 

 was (if I understand the account rightly) eleven inches, but little deviation of 

 position was required for heat from different sources ; and M. Melloni admits 

 that, whilst his experiment indicates the difference of refrangibility, it is inade- 

 quate to measure it. 



49. There are many reasons why such a form of apparatus must be rejected 

 for accurate observations. I mil mention only the impossibility of obtaining a 

 beam of heat which shall preserve the same breadth at different distances from 

 its source (of course supposing the rays rendered as parallel as possible by re- 

 traction through a rock-salt lens), arising, 1. from the angular magnitude of the 

 source ; 2. from the scattered reflection and refraction at the surfaces of the lens 

 and prism ; 3. from the want of homogeneity of the ray. On all these accounts, 

 the beam must have acquired a very sensible breadth at the distance of the pile, 

 and consequently the effect of heat must be perceptible, and even nearly uni- 

 form, through a certain space. I may also add from experience, that the diffi- 

 culty of varying the arrangement of an experiment, so as to get a maximum heat- 

 ing effect at the pile, is so considerable, that no delicate result can be deduced 

 from the merely tentative procedure. Finally, the smallness of the variation of 

 refrangibility, seems to require some more critical method of ascertaining its mea- 

 sure. On all these grounds, it seemed to me desirable to discover a method in 

 some degree less open to objection. 



50. The phenomenon of total reflection, successfully employed by Dr Wol- 

 LASTON in the measurement of refractive indices in the case of light,* presents the 

 advantage of being (theoretically at least) abrupt in its action, the transition from 

 partial to total reflection being (with the necessary exception arising from the 

 want of homogeneity) an instantaneous change, amounting in the case of light to 

 many times the intensity of the smaller effect. It seemed reasonable to expect, 

 that an apparatus constructed on the principle of determining the critical angle 



* Phil Trans. 1802. 

 VOL. XIV. PART I. B b 



