184 ANNUAL OF SCIENTIFIC DISCOVERY. 



shorter legs dipped in a solution of gutta percha and chloroform a 

 sort of collodion, which incrusted them with an impermeable envelope. 

 After drying, the gutta percha cuticle was scraped from the end of 

 one bar, and from the end and a nearly equal portion of the side of the 

 other. The shorter legs of both were placed in the same cup of mer- 

 cury, and the large legs in other weighed cups. Two drops fell 

 from the bar having the larger surface before any fell from the other. 

 After nine days the quantities were weighed. Through the bar hav- 

 ing the greater absorbing surface there had flowed . . 3.8902 gr. 

 Through that having the less ..... 2.1285 gr. 



By increasing the length of the shorter leg beyond a certain meas- 

 ure the syphon action ceased. Some further results may be of inter- 

 est. Two syphon bars were placed in mercury that had once run 

 through lead ; in three days drops fell from both. Mercury in which 

 lead had been standing for months, and which was viscid from 

 the crystallized amalgam, was taken, and two bar syphons, one satur- 

 ated with mercury and the other pure, were placed in it. In due time 

 the amalgam fell from both. A drawn bar saturated with mercury 

 became brittle, as Daniel has observed. It was so brittle as to be read- 

 ily broken by an effort suddenly to bend it. Such a bar, scraped 

 brightly and laid aside, in a few weeks lost its brittleness and peculiar 

 texture, and recovered the properties of the original lead. It had lost 

 its mercury by evaporation. A cast bar, the surface of which was not 

 scraped, after a little time lost no more of its mercury. I have made 

 successful experiments as to the permeability to mercury of gold, 

 silver, cadmium and zinc, and have obtained, at ordinary temperature, 

 negative results, with platinum, palladium, iron, copper and brass. 

 The permeability of several of the latter metals to molten tin, silver and 

 gold, and of iron to copper, is well known. The experiments with tin 

 were of an unanticipated interest, although my research is not con- 

 cluded. Mercury penetrates tin more rapidly than lead, and exhibits 

 the syphon action. As the bar of tin becomes saturated, the whole 

 mass begins to crystallize, and splits into irregular longitudinal fissures. 

 If, at an early stage in this crystallization, the bar is bent, the outside 

 cracks off, revealing a pith as distinct as if it had been at first cast and 

 then a sheath cast around it. If the crystallization be permitted to go 

 on, the fissures penetrate to the centre of the bar. Daniel observed, 

 that a square bar split into triangular prisms, the separating fissures 

 following the diagonal planes. If the top and bottom of the bar were 

 right-angled terminal planes, the crystallization freed a pyramid at 

 either extreme. 



The bar being irregularly cylindrical, the fissures were formed, as in 

 the case of the prism, along the lines of least resistance. In looking 

 at these fissures, and the pith just referred to, and at the septaria 

 which abound in the shales of the Genesee valley, in Livingston 

 county, of which numerous specimens occur in various other widely 

 separated localities, it is impossible to resist the conviction first, 

 that the concentric arrangement in the latter case may have been pro- 

 duced by a process illustrated in the experiment with the tin bar 

 showing the ulterior pith, and not necessarily by aggregation ; and, 



