276 Experiments made by the late Dr. Henry Draper. 
3°583 inches.* As Mendenhall has shown that no change takes 
ing convection currents in the liquid. So long as the tempera 
tures of the prism and of its contents rise and fall together; 
* In the earlier series of experiments the change in position of the Na lines 
was very closely 0°1 inch for one degree Fahrenheit. In the later series, it was 
almost exactly U°3 inch for one degree Centigrade. The results of the two series 
cannot be compared together owing to a change in the adjustment of the a : 
tus during the interval. 
