Scientific Intelligence. Meteorology. 187 



Of these latter a tolerably accurate idea may be formed, by 

 conceiving two rectangular rods of the same metal, at the 

 four extremities of which four distinct arcs are attached, each 

 of which is formed of two metals unequally dilatable, rivet- 

 ed at their extremities. These bi-metallic arcs must necessarily 

 change their curve, and consequently their position, on a varia- 

 tion of temperature. It is equally evident that the movement 

 of the free extremity of each arc will be towards the metal which 

 is least dilatable when the temperature increases, and in the op- 

 posite direction when the temperature diminishes. Hence it re- 

 sults, that at the same time when in the way of dilatation con- 

 sequent on the augmentation of temperature, the radii of the 

 balance-wheel separate from their point of intersection or axis 

 of rotation the extremities of the arcs which are attached to 

 them, the other extremities of these same arcs, or their free ex- 

 tremities, will, on the contrary, be borne towards the centre, 

 if the most dilatable metal is the more external. When the 



Z/ 



more dilatable metal is more central or within, the play of these 

 compound bi-metallic arcs, far from diminishing or entirely com- 

 pensating the effects of the dilatation of the radii, would con- 

 siderably increase them. After these statements, it will be very 

 manifest, that when we wish to give a watch the power of indi- 

 cating the smallest variations of temperature, the relative posi- 

 tion which it was necessary for M. Jurgensen to assign to the 

 bi-metallic arc was the reverse of that which is adopted in our 

 common chronometers, and the most dilatable metal of the com- 

 pound arcs must be put the most internal. But the artist has 

 done more ; the free extremity of each of the four arcs has been 

 selected as the point of attachment of another entirely similar 

 arc, the variations of the curve of which increase the effects of 

 the analogous changes of the other. The watch-thermometer 

 which M. Jungensen has made, according to these princi- 

 ples, is of the same size as a common watch. The outer arc 

 of the bi-metal balance-wheel is made of platina, the inner one 

 of brass. The variation in its going is very nearly 3%' in the 

 twenty-four hours for every single degree of temperature. When 

 this instrument is put in an exposed situation, it will evidently 

 go quicker or slower, according as the atmospheric temperature 

 diminishes or increases. When, then, it has once been dete r- 



