Near to the Lake of Tartarus is the _Solfatara_ lake which contains similar calcareous water, but strongly impregnated with sulphureted hydrogen; it consequently deposits a mixture of carbonate and sulphide of calcium, a sort of porous tufa, some of it so porous that it floats like a stony scum, forming what the cicerone call “floating islands.” Lyell, in his “Principles of Geology,” confounds these lakes, and describes Tartarus under the name of Solfatara. The travertine used as a building stone is chiefly derived from the quarries of Ponte Lucano, and is the deposit that was formed on the bed of a lake like that of Tartarus. The celebrated cascade of the Anio at Tivoli forms calcareous stalactites, and all the country round has rivulets, caverns, and deposits, where this formation may be seen in progress or completed. It varies considerably in structure, some specimens are compact and smooth, others have the appearance of a petrified moss, and great varieties may be found among the materials of a single building. It is, however, usually rough and more or less spongy-looking, as above stated, but this structure does not seem to affect its stability, at least, not in the climate of Italy. Whether it would stand long frosts is an open question. The night frosts at and about Rome are rather severe, but usually followed by a warm sunny day; thus there is no great penetration of ice. Every specimen I have examined shows a remarkable compactness of _molecular_ structure in spite of visible porosity. All give out a clear metallic ring when struck, and the intimate surface, if I may so describe the surface of the warm-like structure it sometimes displays, is always clear and smooth as though varnished. To this I attribute its durability. Lest the above description should appear self-contradictory, I will explain a little further. If melted glass were run into threads, and those threads while soft were allowed to agglomerate loosely into a convoluted mass, it would, as regarded in mass, have a porous or spongy-looking structure, but nevertheless its _molecular_ structure would be compact and vitreous; there would be mechanical but not molecular, porosity. Travertine is similar.