LONG-CONTINUED ELECTRIC CURRENTS OF LOW TENSION. 
43 
smaller quantity ; differences arising in all probability from the smaller quantity of 
silicon present in the solution. 
14. I have frequently had occasion to observe that when an aqueous solution of 
hydrofluosilicic acid has been submitted to the action of currents of low tension (from 
the battery already described (4.)) continued during two or three weeks, a consider- 
able deposition of gelatinous silicic acid takes place around the reduced silicon ; 
mixed with which, or precipitated in a zone on the sides of the tube, especially if of 
small diameter, or even upon the platinum electrode itself, frequently appear minute 
crystalline grains of sufficient hardness to scratch glass, and appearing translucent 
under the microscope. These minute crystals I have no hesitation in stating to be 
crystallized silicic acid, closely resembling its natural form of quartzose sand. 
15. I next attempted to form potassium with the same apparatus, but failed, as I 
had anticipated, from the presence of water, which indeed would react on the potas- 
sium as soon as reduced. I therefore endeavoured to form its amalgam, well knowing 
that when dissolved in mercury a very weak electric current is sufficient to prevent 
the oxidating influence of water upon it ; and by using a modification of the decom- 
posing apparatus before described (/.) I succeeded perfectly. In place of the smaller 
tube containing the metallic solution, I used a small glass funnel*, the beak of which 
was carefully filled up with plaster of Paris : on this plaster floor I placed a piece of 
glass tube closed at one end, about 0‘5 inch in length and 0*2 inch in diameter, and 
half filled with pure mercury ; this tube was not placed vertically, but inclined so as 
to form an angle of about 40° with the plaster floor of the funnel, which with its 
contents was partly immersed in the weak brine contained in the larger cylinder of 
the decomposing apparatus. The external cylinder communicated as before with the 
copper plate of the battery, by means of a slip of amalgamated zinc dipping into the 
brine it contained. The funnel was then nearly filled with a solution of chloride of 
potassium, and a piece of platinum wire connected with the zinc plate of the battery 
being twisted into a flat spiral at one end so as to present a larger surface, was im- 
mersed in the mercury contained in the little tube submerged in the saline contents 
of the funnel. The circuit being thus completed, galvanic action soon became ap- 
parent, bubbles of hydrogen being evolved from the surface of the mercury (which 
now formed the negative electrode) in a very curious manner, not in confused and 
rapid streams, but in large and distinct bubbles, which very slowly appeared, and 
performed several gyratory movements on the surface of the fluid metal before they 
were evolved. Not unfrequently a single bubble only was seen, which continued 
* This variety of the apparatus is shown in fig. 3. of the outline sketch. 
A. The external vessel containing the salt and water. 
B. The funnel containing the alkaline salt required, with its plaster floor F. 
C. The positive electrode of amalgamated zinc communicating with the copper plate of the battery (fig. 1.). 
D. The little tube containing mercury immersed in the alkaline solution with a platinum wire E, connected 
with the zinc plate of the battery (fig. 1.) dipping into it. 
G 2 
