120 Scientific Proceedings, Royal Dublin Society. 
tion into the general statements that the brittleness of the bars 
at + 65° and + 40° was little, if at all, increased at — 53°. 
I hoped to extend the experiments by operating on the frag- 
ments of the bars, but the alteration of the apparatus delayed me 
till the sledging work began. 
The little stock of cast iron bars was supplemented by cast 
steel and brass bars of the same size. These were tested with 
similar precautions at temperatures of + 50° and — 55° Fah. 
The bars bent but did not break at either temperature, and there 
was no appreciable difference in the amount of flexion at + 50° 
and — 55° Fah. 
