The pelagic floating genera are also specially modified. Their shell is either thin or extended many times by long slender tapering spines, and the protoplasm outside has the same character as that of the Radiolaria (q.v.), being differentiated into jelly containing enormous vacuoles and traversed by reticulate strands of granular protoplasm. These coalesce into a peripheral zone from which protrude the pseudopods, here rather radiate than reticulate. Most genera and most species are cosmopolitan; but local differences are often marked. Foraminifera abound in the shore sands and the crevices of coral reefs. The membranous shelled forms decay without leaving traces. The sandy or calcareous shells of dead Foraminifera constitute a large proportion of littoral sand, both below and above tide marks; and, as shown in the boring on Funafuti, enter largely into the constituents of coral rock. They may accumulate in the mud of the bottom to constitute Foraminiferal ooze. The source of these shells in the latter case is double: (1) shells of bottom-dwellers accumulate on the spot; (2) shells of dead plankton forms sink down in a continuous shower, to form a layer at the bottom of the ocean, during which process the spines are dissolved by the sea-water. Thus is formed an ooze known as "Globigerina-ooze," being formed largely of that genus and its ally _Hastigerina_; below 3000 fathoms even the tests themselves are dissolved. Casts of their bodies in glauconite (a green ferrous silicate, whose composition has not yet been accurately determined) are, however, frequently left. Glauconitic casts of perforate shells, notably _Globigerina_, have been found in Lower Cambrian (e.g. Hollybush Sandstone), and the shells themselves in Siberian limestones of that age. It is only when we pass into the Silurian Wenlock limestone that sandy shells make their appearance. Above this horizon Foraminifera are more abundant as constituents, partial or principal of calcareous rocks, the genus _Endothyra_ being indeed almost confined to Carboniferous beds. The genus _Fusulina_ (fig. 20) and _Saccammina_ (fig. 19) give their names (from their respective abundance) to two limestones of the Carboniferous series. Porcellanous shells become abundant only from the Lias upwards. The glauconitic grains of the Greensand formations are chiefly foraminiferal casts. Chalk is well known to consist largely of foraminiferal shells, mostly vitreous, like the north Atlantic globigerina ooze. In the Maestricht chalk more littoral conditions prevailed, and we find such large-sized species as _Orbitoides_ (vitreous) and _Orbitolites_ (porcellanous; figs. 5, 6), &c. In the Eocene Tertiaries the Calcaire Grossier of the Paris basin is mainly composed of Miliolid forms. Nummulites occur in English beds and in the Paris basin; but the great beds of these, forming reef-like masses of limestone, occur farther south, extending from the Pyrenees through the