Casts of the kernels of nuts are formed in much the same way, and Professor E. H. Barbour has thus described the probable manner in which this was done. When the nuts were dropped into the water of the ancient lake the kernel rotted away, but the shell, being tough and hard, would probably last for years under favorable circumstances. Throughout the marls and clays of the Bad Lands (of South Dakota) there is a large amount of potash. This is dissolved by water, and then acts upon quartz, carrying it away in solution. This would find its way by infiltration into the interior of the nut. At the same time with this process, carrying lime carbonate in solution was going on, so that doubtless the stone kernels, consisting of pretty nearly equal parts of lime and silica, were deposited within the nuts. These kernels, of course, became hard and flinty in time, and capable of resisting almost any amount of weathering. Not so the organic shell; this eventually would decay away, and so leave the filling or kernel of chalcedony and lime.[1] [1] _Right here is the weak spot in Professor Barbour's explanation, and an illustration of our lack of knowledge. For it is difficult to see why the more enduring husk should not have become mineralized equally with the cavity within._ "Fossil leaves" are nothing but fine casts, made in natural moulds, and all have seen the first stages in their formation as they watched the leaves sailing to the ground to be covered by mud or sand at the next rain, or dropping into the water, where sooner or later they sink, as we may see them at the bottom of any quiet woodland spring. Impressions of leaves are among the early examples of color-printing, for they are frequently of a darker, or even different, tint from that of the surrounding rock, this being caused by the carbonization of vegetable matter or to its action on iron that may have been present in the soil or water. Besides complete mineralization, or petrifaction, there are numerous cases of incomplete or semi-fossilization, where modern objects, still retaining their phosphate of lime and some animal matter even, are found buried in rock. This takes place when water containing carbonate of lime, silica, or sometimes iron, flows over beds of sand, cementing the grains into solid but not dense rock, and at the same time penetrating and uniting with it such things as chance to be buried. In this way was formed the "fossil man" of Guadeloupe, West Indies, a skeleton of a modern Carib lying in recent concretionary limestone, together with shells of existing species and fragments of pottery. In a similar way, too, human remains in parts of Florida have, through the infiltration of water charged with iron, become partially converted into limonite iron ore; and yet we know that these bones have been buried within quite recent times.