The imperforate Keramosphaera is spherical, and consists of chambers arranged in concentric layers; there is no corresponding form known in the perforate series. An altogether irregular arrangement of chambers obtains in the perforate Tinoporinae, and irregular layers of chambers are found on each side of the disc in Orbitoides.

1As a rule no signs of structure are to be discerned in the test save the lamination of many perforate forms. In a few instances, the perforate Amphistegina, Operculina, Heterostegina, Cydoclypeus, it is composed of vertical hexagonal prisms traversed each by a pore; and its outer or secondarily deposited layers in Orbulina, Globigerina, etc, are said to be made up of vertical wedge-shaped masses. The siliceous skeletons of diatoms, sand-grains, etc, have been detected in the tests of the perforate Amphistegina and Orbulina during decalcification (Folin, C. R. 102, 1886, pp. 1575-6). The specific gravity of the imperforate test is higher than that of the perforate; the former varies from 27 to 2722, the latter from 2626 to 2674 (Sollas, Sci. Proc. Royal Dublin Soc. iv. 1885, P. 39°). Little is known as to the way in which new chambers are added. The protoplasm grows in amount and protrudes from the last chamber, and the walls of the new chamber are at first very thin; the pores are evident at an early period; see Max Schultze, Organismus, etc, pp. 29-30. Growth in size must take place by resorption within and addition without.

According to Moebius (Beitrage, etc, p. 83), the pores in the arborescent stem of Carpenteria raphidodendron are formed soon after the calcareous matter is first laid down.

1The term 'spiral axis' applied to a coiled test denotes the longitudinal axis of the convolutions; 'plane of convolution' a vertical plane coiled like the test; 'axis of convolution' a line drawn through the point at which the test commences, and round which it is imagined to be coiled. See on the subject of the spiral coiling, von Moller, Mem. Imp. Acad. St. Petersburg (7), xxv. No. 9, pp. 27-40. The perforate family Cheilostotnellidae is remarkable for the degree to which involution is carried. In the extinct Ellipsoidina every new chamber completely incloses its predecessors, all being connected at their bases; i. e. the test resembles a number of flasks standing one within the other. So, too, in Cheilostomella, but every new chamber has its aperture turned in the opposite direction to its immediate predecessor; whilst in Allomorphina the chambers are disposed in cycles of three, the last chamber leaving a portion of its two predecessors exposed.

2 The cyclical Orbitolites has a very perfect series of species showing its evolution from Orbi-culina, which is itself derived from Peneroplis. The Orbiculine portion of the test is extremely large in O. tenuissima, and is very decidedly excentric, features very much less marked in 0. margi-nalis, whilst in 0. duplex and 0. complanata the excentricity is lost, and the cyclical mode of growth established almost from the first, the last named being the most advanced. There is also in the two last named species a complication in the structure of the chamberlets, O. duplex leading from the simple type of 0. marginalis to the complicated type of 0. complanata. See W. B. Carpenter, 1Challenger Reports,' vii. 'On the genus Orbitolites.' The spiral origin of the test is indicated plainly but slightly in the perforate Cycloclypeus, and apparently not at all in Orbitoides. The most remarkable feature in Cycloclypetis is the addition of a fresh calcareous lamina to each surface of the test when a new cycle of chamberlets is formed; and in Orbitoides the presence of irregularly arranged chambers on either aspect of the disc.

Dimorphism exists in some cases, but the term has been applied in two different senses. In one it denotes that a given test follows in its individual growth two different types of structure, in the other that it is found under two different forms. The first mode is exemplified in the Miliolid Hanerininae, some Textularidae, and Lagenidae, e.g. Hauerina, with its first chambers arranged as in Miliolina, their successors in a planospiral with more than two in each turn of the coil. Trimorphism in the same sense is very rare. The second mode has been observed in Nummulites and its congener Assilina. The tests, e.g. of Nummulites, occur in pairs, one constantly smaller than the other and possessed of a large primordial chamber, the other of larger size but with a small primordial chamber or none at all recognisable. The Miliolininae combine both modes. One set of forms is small, the other large; the former have a large, the latter a small primordial chamber, with, moreover, the first chambers differently grouped to their successors which are typical. The initial chambers, moreover, in one and the same species may be disposed in different ways, leading to a polymorphism.

The significance of these facts is not known.

As to structure, the following points may be noticed. The terminal aperture of the test is rarely absent as in many specimens of Orbulina; it may be simple, its shape sometimes depending on that of the test, sometimes independent of it; or it may be radiate, owing to the projection across it of calcareous bars; finally, it may be replaced by pores. In Lagena great variety prevails; it may be situated at the extremity of a tube (ectosolenian), or the tube is prolonged inwards into the test (ento-solenian); whilst in other examples the two modes may be combined (ecto-ento-solenian), and more rarely a second aperture is presentl. The spherical or pyriform chambers of Ramulina have several tubular orifices, and in the arborescent Carpenteria raphidodendron and Polytrema the branches terminate in simple apertures. In a few chambered tests each chamber has its own separate aperture, e. g. in some species of Globigerina; and in Cymbalopora not only is this the case, but accessory apertures may be present as well. The chambers in polythalamous tests may be separated from one another by slight constrictions, by imperfect septa, or by perfect but porous septa.