arborescent forms as are represented in Fig. 6. As these branches push into the yellow liquid, it becomes colorless even in advance of their points, and it frequently happens that yellow crystals of some salt shoot out in front of the crystallizing metal, which follows them and builds up its advancing fronds at their expense. This is shown in the figure. The gold will generally shoot its yellow branches rapidly round the margin of the drop. Such a running branch has been seen to stop on coming in contact with a loose piece of gold, which immediately in its turn becomes active, and commences to sprout on its farther side. Copper salts give round nodules, which have no crystal-line appearance when deposited from moderately weak solutions, but a very strong solution of the chloride—about forty per cent.—yields with zinc, first a thick, black growth, then arborescent fringes of red metal, terminating in crystals of very appreciable size.
The fringes referred to in the case of these three metals are still more characteristically developed by bismuth. When a solution of the 
Fig. 7. terchloride of bismuth acts on zinc there is an immediate outgrowth of black fringes, as in Fig. 7, where they are seen on an illuminated field. As they advance, these become more arborescent, and as the crystalline character becomes more developed they change from black to gray. Sometimes bismuth presents itself in botryoidal masses, but the tendency to form the fringes is very strong.
Chloride of antimony with zinc also gives black fringes. Lead salts yield crystals resembling those of silver, but leaves of irregular, hexagonal plates prevail, and frequently become of large size.
A solution of acetate of thallium, of twenty per cent, of salt, quickly gives a beautiful forest of thorny crystals. Sulphate of cadmium gives rise to a small, leaf-like growth on zinc; but a strong solution of the chloride produces an appearance of sticks covered with small spines or knobs. The new metal indium is thrown down upon zinc in the form of thick, white crystals. The deposition is promoted by touching the zinc with a piece of iron.
Tin gives beautiful results. If zinc be placed in a solution of stannous chloride it is quickly surrounded with a growth of prolonged octohedra, and as these advance into the liquid it is easy to observe that the additions of new metal commence at the apex, and that the wave of chemical change proceeds down the lateral edge, occupying some seconds of time in depositing the new layer of material. Frequently,
