imperfectly crystalline structure, and had been deposited in irregular laminæ, beginning at the sides of the chambers, and filling them toward the middle, and had afterward been cracked by shrinkage, and the cracks filled with a second deposit of serpentine.[65] Now, serpentine is a hydrous silicate of magnesia, and all that we need to suppose is that in the waters of the Laurentian sea magnesia was present instead of iron, alumina or potash, and we can understand that the Laurentian fossil has been petrified by infiltration with serpentine, as more modern Foraminifera have been with glauconite, which, though it does not contain magnesia, often has a considerable percentage of alumina. Further, in specimens of Eozoon from Burgess, the filling mineral is loganite, a compound of silica, alumina, magnesia and iron with water, while in other specimens the filling mineral is pyroxene. In like manner, in certain Silurian limestones from New Brunswick and Wales, in which the delicate microscopic pores of the skeletons of stalked starfishes or crinoids have been filled with mineral deposits, so that when decalcified these are most beautifully represented by their casts, Dr. Hunt has proved the filling mineral to be[66] intermediate between serpentine and glauconite. We have, therefore, ample warrant for adhering to his conclusion that the Laurentian serpentine was deposited under conditions similar to those of the modern greensand. Indeed, independently of Eozoon, it is impossible that any geologist who has studied the manner in which this mineral is associated with the Laurentian limestones could believe it to have been formed in any other way. Nor need we be astonished at the fineness of the infiltration by which these minute tubes, perhaps 1/10000 of an inch in diameter, are filled with mineral matter. The micro-geologist well knows how, in more modern deposits, the finest pores of fossils are filled, and that mineral matter in solution can penetrate the smallest openings that the microscope can detect. Wherever the fluids of the living body can penetrate, there also mineral substances can be carried, and this natural injection, effected under great pressure and with the advantage of ample time, can surpass any of the feats of the anatomical manipulator. Fig. 16 represents a microscopic joint of a Crinoid from the Upper Silurian of New Brunswick, injected with the hydrous silicate already referred to, and Fig. 17 shows a microscopic chambered or spiral shell, from a Welsh Silurian limestone, with its cavities filled with a similar substance. [64] Very fine specimens can be produced by polishing thin slices, and then etching them slightly with a very weak acid. (Plate prefixed.)