2()2 VARIATIONS OF SPECIFIC ORAVITY. 



on 140 separate determinations with the spheronieter, and obtained 

 the vahie of the specific gravity as 8.;^7T4, which is a little less than 

 that found for any of the smaller samples. It might be supposed 

 ihat the dift'erence of 0.32 between this value and that obtained for 

 Sample II would be nearly the maximum obtainable. But Marchand 

 and Scheerer« by very accuilite determinations found for their high- 

 est value of the density of rolled and hanmiered copper plate 8.952, or 

 0.259 higher than my highest observed value, and, compared with 

 my lowest value, showing a difference of 6.2 per cent. 



It seems simple enough in the case of metals to do aAvay with these 

 differences of si^ecific gravity by employing high pressures sufficient 

 to close the faults produced in casting. But I have found in experi- 

 ments with lead, cadmium, copper, tin, antimony, gold, and silver 

 that up to a pressure of 10,000 atmospheres all these metals except 

 cadmium shoAv increased density with increased pressure, as was to 

 be expected, but on increasing the pressure to 20,000 atmosj^heres 

 these metals without exception diminished in density with the in- 

 creased pressure. It is not implied that the density rises in all cases 

 except that of cadmium up to 10,000 atmospheres pressure and then 

 diminishes, but that if any decrease occurred below 10,000 atmos- 

 pheres the density did not fall beloAV the original value.^' It would 

 doubtless be of great theoretical interest to determine the exact pres- 

 sure of maxiiuum specific gravity, but working as hitherto with 

 samples Avould consume much time, and it seemed to me that a shorter 

 method based on electrical conductivity measurements would prove 

 feasible. For if the electrical conductivity is measured continuously 

 while the pressure is steadily augmented there nuist be a change in 

 the march of the conductivity measurements when the pressure of 

 maxinuun specific gravity is reached. Without going into a descrip- 

 tion of the experiments, it will suffice to say that I was led in this 

 way to the examination of the specific gravity of wires. An altera- 

 tion of the density implies a change in the molecular structure of the 

 material. Such a change in the structure of wires is indicated by 

 the fact that cold-drawn wires are brittle and liable to faults, and 

 that in order to continue wire drawing a reheating is re(|uired.'' 



In wire drawing the metal unist be di-awn oi-, as in tlie case of 

 sodium joresses, squeezed thr-ough a hole with sharply defined edges. 

 In this process thei-e is formed at the back of the drawplate a ring of 

 material, which is held l)ack while the inner molecules by virtue of 

 their cohesion are pressed or draAvn through the opening. Metals 



" O. i:. Knlinaiin, Jourii. f. prakt. Clieni. 27. p. 20(), 1S42. 



''(;. W. A. Kalilbauni. VerliaiuU. d. Natnrf. (Jesellsch. in Basel, IT., p. 17, 1!»0.'k 

 f According to S. Kalischer ( P.er. Dcutsch. chem. Gesollsch. 14, p. 2750, 1881) 

 yAnc wire at temperatures above 300° becomes again brittle. 



