405 



to the depth of half the diameter of the bore, so as to leave a shoul- 

 der near the end. A bar of any metal 6^ inches long is introduced 

 into the cavity, resting against its solid end, and a cylindrical piece 

 of porcelain, about 1| inch long, which he calls the index, is placed 

 upon the top of the bar, and projects beyond the open part of the 

 tube, being confined in its place by a ring or strap of platina passing 

 round it, and also round the end of the black-lead bar, and made 

 sufficiently tight by a small porcelain wedge inserted between them. 

 When the instrument thus prepared is subjected to heat, the porce- 

 lain index will be forced up, by the expansion of the metallic bar, to 

 a certain distance, where it will remain when the bar retires from it 

 on cooling. The distance it has been moved from its original po- 

 sition, will be the measure of the difference of expansion of the me- 

 tallic bar, and of an equal length of the black-lead in which it is con- 

 tained. This cannot be influenced by any permanent contraction 

 which the black-lead may undergo by intense heat, because any such 

 contraction will occur at the moment of the greatest expansion of 

 the metal, and the index will still mark its point of furthest exten- 

 sion upon this contracted basis. It remains then to measure accu- 

 rately the distance to which the index has been moved by the appli- 

 cation of the scale, which is a detached instrument constructed of 

 two rules of brass joined together at a right angle ; the one fitting 

 square upon two sides of the black-lead bar, the other resting on its 

 shoulder ; with these are connected two arms, which, acting on 

 the principle of proportional compasses, measure the distance of the 

 extremity of the index from the shoulder of the black-lead bar. The 

 spaces comprehended between the points of the shorter legs of the 

 compasses are magnified ten times by the longer legs, the angular 

 motion being measured by a graduated arc furnished with a vernier, 

 and capable of being easily read off to minutes. 



The author next enters into a comparison of the results afforded 

 by this instrument with those of former experimentalists ; and espe- 

 cially with the accurate determination of the expansions of metals by 

 Messrs. Dulong and Petit, with a view to show the degree of confi- 

 dence to which it is entitled. The close agreement in the results of 

 a great number of experiments upon metals which differ much in their 

 expansions is highly satisfactory in this respect. Differences having 

 been found in the expansibility of different specimens of black-lead, 

 it becomes necessary to ascertain the expansions of each register for 

 itself, by applying to it the heat of boiling mercury. The author 

 concludes with an account of some experiments which he made to 

 determine the fusing points of different metals, referred to the com- 

 mon thermometric scale. His calculations proceed upon the suppo- 

 sition that the equal amounts of expansion denote equal increments 

 of temperature ; and he is thus enabled to compare the series now 

 obtained with that which was obtained with his first pyrometer. A 

 remarkable accordance is found between the results with platina and 

 with iron, metals which differ widely in their expansions ; conform- 

 ably with the conclusions of MM. Dulong and Petit, the expansion 



