266 



PROCEEDINGS OF THE AMERICAN ACADEMY 



In order that the air currents in the neighborhood of the bar dur- 

 ing the cooling experiments might be, as nearly as practicable, like 

 those prevailing during the steady flow observations, the Bunsen 

 lamp within the tin protecting case already described was kept 

 burning near one end of the cooling bar, the flame being of about 

 the same height in all cases. A screen was placed between the 

 box and the end of the bar to prevent direct radiation. 



No attempt was made to represent by equations the temperature 

 of the different parts of the bars during the condition of steady 

 flow. The observations for every case were merely plotted care- 

 fully, with distances along the bars for abscissae, and with tem- 

 peratures, minus temperature of the room, for ordinates. Through 

 the tops of the ordinates a smooth curve was drawn. The gradient 

 of temperature at any point of the bar was found by measurement 

 of the inclination of the curve at that point. The mean tempera- 

 ture of each division of the bar whose surface emission had to be 

 considered was easily estimated from the mean height of the curve 

 for that division. The amount of heat emitted from each such 

 division was then estimated by reference to the curves, already 

 mentioned, showing rate of cooling as a function of temperature of 

 metal minus temperature of room. Wherever the specific heat of 



