160 



PROCEEDINGS OF THE AMERICAN ACADEMY 



Fig. 7 shows a section of the calorimeter, and Fig. 8 a per- 

 spective view of the revolving paddles removed from the appa- 

 ratus, and with the exterior paddles removed from around it; which 

 could not, however, he accomplished physically without destroying 

 them. 



To the axis c b, Fig. 7, which was of steel, and G mn1, in diame- 

 ter, a copper cylinder, ad, was attached, by means of four stout wires 

 at e, and four more at/*. To this cylinder four rings, g, //, i,j, were 

 attached, which supported the paddles. Each one had eight paddles, 

 but each ring was displaced through a small angle with reference to 



Fig. 7. 



Fig. S. 



■cv>y, 



the one below it, so that no one paddle came over another. This was 

 to make the resistance continuous, and not periodical. The lower 

 row of paddles were turned backwards, so that they had a tendency 

 to throw the water outwards and make the circulation, as I shall 

 show afterwards. 



Around these movable paddles were the stationary paddles, consist- 

 ing of five rows of ten each. These were attached to the movable 

 paddles by bearing-;, at the points c and h, of the shaft, and were 

 removed with the latter when this was taken from the calorimeter. 

 When the whole was placed in the calorimeter, these outer paddles 

 were attached to it by means of four screws, / and m, so as to be 

 immovable. 



The cover of the calorimeter was attached to a brass ring, which 

 was nicely ground to another brass ring on the calorimeter, and which 



