208 Benton — Experimental Method in the Flow of Solids. 



since certain points on each wire can always be recognized by 

 the intersections with other wires. 



I have made an attempt to apply this method to a study of 

 the flow in the interior of a cube compressed between two 

 parallel plates, the faces of the cube not in contact with the 

 plates being free. 



The frames were made of fuse- wire, which was melted 

 together at the joints. The material of the cubes experimented 

 on was Wood's fusible alloy, it being chosen simply on account 

 of the.greater convenience of working with a substance of low 

 melting point. Fuse-wire was used in preference to other 

 wire because it is soft enough to follow the flow, and not cut 

 through the Wood's metal, as harder wire might do. Only 

 one frame was inserted in each cube ; the frame was clamped 

 in the desired position inside a cubical mould, and molten 

 Wood's metal was poured in. After the metal had solidified, 

 the cube was compressed gradually between the plates of a 

 testing machine, the pressure being kept about 3000 lbs. per 

 sq. in. Then the metal was raised to a temperature sufficient 

 to melt it without melting the fuse-wire, and the frame was 

 taken out. Before melting the Wood's metal after the flow, 

 plaster of Paris was cast around it, which on hardening formed 

 a vessel containing the metal, and prevented it from flowing 

 away as it melted. Without this precaution there would have 

 been the danger that the metal might have melted irregularly, 

 and left solid masses sticking to the wire frame, which would 

 have deformed it by their weight. 



A slight difficulty occurred in the tendency of the molten 

 metal to adhere to the fuse-wire. Coating the wires with wax 

 did not obviate this ; but the Wood's metal could be dissolved 

 off in hot concentrated H 2 S0 4 . 



The edge of the cube was If in. = 4 - M cm in every case. 

 The friction between the plates of the testing machine and the 

 faces of the cube in contact with them was not sufficient to 

 prevent some flow along the plates. 



The framework of wire was always placed in a plane per- 

 pendicular to the plates of the testing machine, and midway 

 between and parallel to two faces of the cube. The original 

 intention was to construct frameworks consisting each of two 

 sets of parallel wires at right angles to each other. Preliminary 

 experiments showed that the wires parallel to the plates of the 

 testing machine remained nearly uniformly spaced and deviated 

 but slightly from straight lines. They were accordingly dis- 

 pensed with, and the frameworks used consisted each of a 

 single set of wires, perpendicular to the plates of the machine. 



Blue-prints of the frameworks were taken before and after 

 flow, by pinning them on pieces of sensitive paper and expos- 



