In the case of fossil wood greater changes have taken place than in the fossilization of bone, for there is not merely an infiltration of the specimen but a complete replacement of the original vegetable by mineral matter, the interior of the cells being first filled with silica and their walls replaced later on. So completely and minutely may this change occur that under the microscope the very cellular structure of the wood is visible, and as this varies according to the species, it is possible, by microscopical examination, to determine the relationship of trees in cases where nothing but fragments of the trunk remain. The process of fossilization is at best a slow one, and soft substances such as flesh, or even horn, decay too rapidly for it to take place, so that all accounts of petrified bodies, human or otherwise, are either based on deliberate frauds or are the result of a very erroneous misinterpretation of facts. That the impression or cast of a body _might_ be formed in nature, somewhat as casts have been made of those who perished at Pompeii, is true; but, so far, no authentic case of the kind has come to light, and the reader is quite justified in disbelieving any report of "a petrified man." Natural casts of such hard bodies as shells are common, formed by the dissolving away of the original shell after it had become enclosed in mud, or even after this had changed to stone, and the filling up of this space by the filtering in of water charged with lime or silica, which is there deposited, often in crystalline form. In this way, too, are formed casts of eggs of reptiles and birds, so perfect that it is possible to form a pretty accurate opinion as to the group to which they belong. [Illustration: Fig. 2.--Bryozoa from the Shore of the Devonian Sea that Covered Eastern New York. _From a specimen in Yale University Museum, prepared by Dr. Beecher._] Sometimes it happens that shells or other small objects imbedded in limestone have been dissolved and replaced by silica, and in such cases it is possible to eat away the enveloping rock with acid and leave the silicified casts. By this method specimens of shells, corals, and bryozoans are obtained of almost lace-like delicacy, and as perfect as if only yesterday gathered at the sea-shore. Casts of the interior of shells, showing many details of structure, are common, and anyone who has seen clams dug will understand how they are formed by the entrance of mud into the empty shell.