1886.] on Properties common to Fluids and Solid Metals. 409 



atmospheres, compelled to pass into the condition of octahedral 

 sulphur, an allotropic state which possesses a greater density. And 

 he points out that a solid metal (not powders of metals) may have 

 cavities obliterated by pressure, but that matter cannot be permanently 

 comj)ressed by pressure, unless it can assume an allotropic state of 

 greater density than the one it possesses at the moment of com- 

 pression.* Now let me point to the evidence these experiments afford 

 as to the relation between solid metals and fluids. Members of the 

 Royal Institution will know that Faraday discovered, in 1850, that 

 two fragments of ice pressed against each other will unite, tendency 

 to their union being considerable when the fragments are near their 

 melting point. We also know what splendid service the regelation 

 of ice has afforded in the hands of Dr. Tyndall, in explaining the 

 formation of glaciers. Ice owes its movement, not to viscosity, but 

 to regelation, and the union of fragments of ice under compression is 

 also due to regelation. The facts which have been aj)j)ealed to, and 

 the theories which have been formed, respecting the regelation of ice, 

 are too well known to you to demand lengthy notice from me. I will 

 only observe that bismuth, like ice, expands on solidifying, and 

 although Faraday failed to establish the existence of a property 

 similar to regelation in bismuth, an eminent engineer, Mr. Thomas 

 Wrightson, to whom we owe a series of experiments on the fluid 

 density of metals, has satisfied himself by experimental evidence, 

 that regelation exists in bismuth. Now, in explaining Spring's 

 results we are met by this difficulty ; the union of the particles of the 

 metals cannot, in all cases, be due to viscosity, because viscous bodies 

 are always capable of being stretched, and we find the welding taking 

 place between the compressed powders of bodies such as zinc and 

 bismuth, which, when submitted to traction, will not stretch. Spring 

 therefore asks, " Is it possible that regelation may have something to 

 do with the union of the powders ? " and he urges, " Is it safe to con- 

 clude that regelation is peculiar to water alone ? " " It is difficult to 

 believe," he adds, "that in the large number of substances which 

 nature presents to us, but one exists possessing a property of which 

 we can find only minute traces in other bodies. The sum of our 

 chemical and physical knowledge is against such a belief, and there- 

 fore the phenomenon of regelation may be pronounced in ice without 

 being absolutely wanting in other bodies. To ascertain whether this 

 is so, it is necessary to submit various bodies to the conditions under 

 which the phenomena can be produced." "What," he asks, "are 

 these conditions ? " and he answers, " The pressure supported by the 

 body, a certain degree of temperature, and time." 



Helmholtz and Tyndall have shown that when the pressure is 

 weak, the regelation of ice is effected slowly. Spring points out that 

 nitrate of sodium and phosphate of sodium, in powder, left to them- 



* " Sur I'elasticite parfaite des corps solides chemiquement definis." ' Bull. 

 Acad. Roy. Belgique,' [3] t. vi. 1883. 



