which proceed from tfie Sun* 315 



in proportion as the image itself and corresponding circle 

 of calorific rays are enlarged by increasing the focal di- 

 stance. Even though the velocity of the rays last men- 

 tioned fell but little short of the vast rapidity of light, still 

 the crescents might be measurable on the screen, if we 

 had it in our power to extend the focal distance at pleasure, 

 with a suitable enlargement of aperture, so as to give an 

 image of the sun a little hotter than common sunshine. 



Part IL 



To enter now, a little, upon a practical view of the sub- 

 ject, it comes to be considered how much slower the velo- 

 city of the sun's calorific rays ought to be, compared to that 

 •of his light, before we might expect to discover, by the most 

 delicate thermometric trials, the presence of the two cres- 

 cents susceptible of being measured. In order to this, it is 

 apprehended that a focal distance of at least fifty feet could 

 be so managed, by a proper apparatus, as to give sufficient 

 steadiness to the image, whose diameter on the screen would 

 be about five inches and a half, as in fig. 2.* Next sup- 

 pose that, by thermometrical trials, we found out the two 

 crescents, the bright one from B to A relatively cold, and 

 the opposite dark one hot to the distance of A, and that 

 AB, the greatest thickness of either, measured 2-10ths of 

 an inch. From this last datum, the velocity of the calorific 

 rays, corresponding to it, would be determined by the fol- 

 lowing analogy : As six thousand, the focal distance ex^ 

 pressed in tenths of an inch, is to two-tenths the greatest 

 breadth of either crescent, so is radius to a fourth propor- 

 tional, the tangent of an angle of sixty-eight seconds, or 

 l' 8". This would be the difference between the angles of 

 aberration of light and the rays in question, which when 

 increased by twenty seconds, the known aberration of the 

 former, would give an angle of eighty-eight seconds for the 

 real aberration of the calorific rays. Lastly, as it is well 

 known that the velocities of any two sets of rays must be 

 inversely as the tangents of the angles of aberration corre- 

 sponding to them, or to the angles themselves when very 

 small, it follows that, in the present example, the velocity 

 of the calorific rays would be to that of light in the ratio of 

 £0 to 88 ; that is, only between fo'ur and five times slower. 



It is of importance here particularly to remark, that 



* In the original diagrams sent with this article the diameter of tr.e 

 circles was 5-^ inches, but. in the engraving this has necessarily been re- 

 duced to 3 inches.— Edit. 



Y 4 though, 



