The manner in which these forests were fossilized may next be considered. Though the whole history of the process is not fully understood, it was undoubtedly dependent on or at least greatly facilitated by the presence of volcanic and hydrothermal activity, which was doubtless then, as it is to some extent now, a marked feature of the park region. At least a hint of the probable process is afforded by the action now going on in the hot spring areas. Many of those areas are closely surrounded by forests, and unless the action of the springs is very violent the trees may be growing only a short distance away. Occasionally a hot spring may break out near the edge of a forest, the first effect being, of course, to kill the trees. In a few years, by the action of the ordinary processes of decay, a tree so killed may have lost its bark and most of its smaller branches. The hot water which constantly or intermittently surrounds the tree contains a considerable amount of silica in solution, and as this hot silica-charged water is drawn up into the wood by capillarity the silica may be deposited in the cells of the wood after the water cools or evaporates. The first result will be a more or less complete cast of the interior of the cells and vessels of the wood. This much of the process has actually been observed, but as decay is more rapid than silicification, the wood crumbles to dust before petrifaction is complete. If the trunk could be surrounded by ashes or mud and thus protected from atmospheric action, it might in time be completely turned to stone.

The fossil forests are surrounded by a matrix that is known as an acidic lava—that is, a siliceous lava—which contains abundant silica in solution. The first part of the process of silicification may well have been that above described as taking place in the hot spring areas at the present day—that is, the silica would be deposited in all the cells and vessels of the wood, making an accurate cast of all open spaces. Then, while the slow process of decay went on, as each particle of organic matter was removed its place was taken by the silica, until, finally, all the wood substance had disappeared and its place atom by atom had been taken by silica.

By this or a similar process the wood has been preserved or fossilized with remarkable fidelity: in fact, thin sections or slices of the fossil wood may be studied under the higher powers of the microscope with almost or quite as much completeness and satisfaction as if they were sections cut from a piece of living wood. Each cell and vessel, with its characteristic pits and markings, is preserved exactly as it grew. Some of the wood, however, was evidently more or less decayed before it was fossilized, or else decay worked faster than replacement, so that in some fragments the structure is not so clearly preserved. Many of the trunks were subjected to pressure before replacement was complete, and this has crushed or distorted the cells. On the whole, however, the wood is exceptionally well preserved, as may be seen in figures 12, 13, 14, and 15. These are all magnified 100 diameters and were photographed directly from the thin sections—that is, they are photomicrographs—and have not been retouched in any manner. Figure 12 shows a transverse section of the wood of the large redwood trunk that has been so often mentioned (see title page). The section is cut through one of the growth rings, which consists of 12 or 15 rows of very thick-walled cells. The large, regular thin-walled cells, which begin abruptly above the growth ring, belong to the spring wood—that is, the wood first formed after growth starts in spring, when the supply of nourishment is abundant. If there is sufficient moisture and all conditions are favorable this vigorous growth of wood cells may continue without interruption until the approach of cold or dry weather, but not infrequently there may be a brief shortage of moisture, and this is reflected in the formation of a few rows of thicker-walled cells. Such a condition may be observed in the present specimen, in which a slight, partial ring may be seen at some distance above the main ring.