Summary.1The oviduct is lined by a single‐layered columnar epithelium which varies somewhat in height and in its detailed characters in the different parts of the tube. In the funnel it is of the ciliated columnar variety; then it increases in height to high columnar in the upper region of the Fallopian tube, where it is composed of two types of cells, one of which (Type I.) is non‐ciliated and secretory, the other (Type II.) ciliated and non‐secretory. Towards the glandular region of the tube the epithelium decreases in height. In this latter region, Type I. cell is no longer secretory, although both types of cells are still recognizable, whilst a third variety of cell (Type III.) makes its appearance. These latter cells elaborate a globular secretion at a much later period in gestation. In the uterus proper the epithelium is of the low columnar form; most of the cells are ciliated, except in the case of a few secretory cells which resemble the Type III. cells of the tube.2The secretory cells (Type I.) which occur throughout the upper two‐thirds of the Fallopian tube except in the region of the funnel form a fine granular secretion. This is in process of formation during the growth of the Graafian follicle and its contained Ovum in the ovary and the cells are in active secretion when the ovum is shed. This secretion is regarded as forming the albumen coat of the egg.3In the glandular region of the tube (forming its lower third) the formation of albumen granules ceases, the great majority of epithelial cells being ciliated and non‐secretory. Intermixed with these non‐secretory cells, however, are a few cells of the third variety (Type III. cells). These cells are not actively secretory until late in gestation, when they pour out a globular secretion which is regarded as being concerned with the calcification of the shell. These Type III. secretory cells are also found in the uterine epithelium, but they are not very numerous.4In the glandular region of the tube the mucosa contains numerous convoluted glands (tubal glands) lined by tall columnar cells with reticular cytoplasm, the meshes of which are occupied by fluid secretion. It is suggested that this secretion is concerned with the formation of the homogeneous basal layer of the shell (layer 1). The secretion is discharged through the cell‐membrane in the form of small droplets and may either pass directly into the lumen of the oviduct or it may become stored in dilatations formed by the enlarged basal ends of the glands. These dilatations are specially distinct on the glands situated in the lower part of the glandular region, though they also occur sporadically on the glands of the upper region.5In the junctional region between the tubal and uterine segments of the oviduct tubal glands are still present in the mucosa, but in addition glands similar to those of the uterus make their appearance, whilst the dilatations of the basal ends of the tubal glands are more numerous. Certain of the cells lining these dilatations form a finely granular secretion, whilst those lining the uterine type of gland form a coarsely granular secretion comparable to that formed in the basal ends of the uterine glands proper. They become actively secretory sotnewhat later than the tubal glands and their dilatations. This secretion, together with that of the dilatations of the basal ends of the tubal glands, is, we suggest, concerned with the formation of the rodlet layer of the shell (layer 2,videp. 436).6The uterine glands differ from the tubal in that they are lined by a ciliated epithelium composed of two types of granule‐secreting cells: of these the more numerous type possesses a pale‐staining oval nucleus which lies in the lower third of the cell, its chromatin being clumped round the nuclear membrane; the other type, of much less frequent occurrence, possesses a darkly staining nucleus, shrunken and pycnotic in appearance, and situated close to the base of the cell. Both types of cell form secretory granules which appear to be similar, but the cells can be distinguished from each other throughout the entire secretory phase, The secretory granules vary in their detailed characters in different regions of the gland. In the superficial and middle parts of the glands they are finely granular and begin to be shed at a relatively early stage—e. g.,PlatypusVIII., with two intra‐uterine eggs in a late cleavage (blastodisc) stage. This secretion is extremely plentiful and furnishes, we suggest, the nutritive fluid which is absorbed by the egg during its growth in the uterus, and which is utilized along with the ovular yolk for embryonic growth during the intrauterine and incubatory phases of the life‐cycle. It is formed at a much earlier stage than that of the deeper parts of the glands, and its production ceases before these latter parts become active.In the deeper parts of the glands, the secretory granules are larger, more numerous, and more variable in size than those produced in their upper segments. They first appear in minute vacuoles in the cytoplasm in close proximity to the nucleus, and spread from there throughout the surrounding cytoplasm to become aggregated at the apical end of the cell. During their formation the chromatin granules of the nucleus are closely adherent to the nuclear membane, but no actual nuclear extrusions were seen. The accumulation of the secretory granules causes the apical end of the cell to project into the lumen of the gland, and results in the disappearance of the cilia and their basal granules. Sooner or later the projection with its granules becomes constricted off from the main body of the cell, and so becomes free in the gland‐lumen. Eventually the granules liquefy, and the fluid, diffusing out, appears in the sections as a coagulum, whilst the cytoplasmic matrix is left as a minute granule‐free, light‐staining mass, which presumably disintegrates and becomes added to the secretion. These more deeply situated gland‐cells do not begin to shed their secretion until the egg has attained a diameter of about 9 mm. In the stage of egg DQ, which measures 16×15 mm. in diameter and so has reached its full size, the cells lining the basal ends of the glands are still laden with secretory granules, and the latter are also abundantly present in the gland‐lumina. It is suggested that this secretion is utilized in the formation of the dense massive protective layer (layer 3) of the shell, which first begins to appear in eggs of about 12 mm. diameter and which is already well developed in the full‐grown intra‐uterine egg DQ.Summary.The observations recorded above may be summarized as follows:—1The Monotreme shell when first laid down appears as a very thin, transparent, structureless membrane. It soon begins to thicken, and on its outer side and in continuity with it a second layer is deposited, which, from its first appearance, is distinguishable from the primary membrane by its staining reactions, and later on also by its structural characters. These two layers form layers 1 and 2 of the definitive shell. In the blastodisc stage of the egg (PtatypusB) the shell has a thickness of about 0.0036 mm., its two layers being of about equal thickness, but layer 2 now exhibits differentiation into a matrix‐substance in which close‐set, radially disposed, rod‐like bodies are visible. By the time the egg has attained a diameter of 6.25 mm. (PlatypusS) the shell‐thickness has increased to 0.011 mm., layer 1, to 0.0026 mm., whilst layer 2 shows a much more marked increase to 0.009 mm. In its outer half or thereabouts the rod‐like bodies have now become definite radially disposed rodlets, separated by narrow cleft‐like intervals and ending below in an undifferentiated zone resting on the basal layer. The shell of the 10 mm. egg (PlatypusDD) is just about twice as thick as that ofPlatypusS. The basal layer is slightly thinner than in the latter, and from now on it shows no further increase, but rather a very slight decrease in thickness. In the full‐grown intra‐uterine egg UQ (16.5×15 mm. in diameter) it still has a thickness of about *diameter of the egg, the basal layer does not simply undergo mere mechanical stretching (as Nathusius (13) contended), but must grow in surface‐extent as the result of the intussusception of new material.Layer 2 in egg DD, on the other hand, shows a very definite increase in thickness and in differentiation. The rodlets have practically reached their definitive form and size, and now reach from the surface down to the basal layer, with which they appear to be in continuity. They have attained a length of about 0.019 mm. Moreover, they are no longer uniformly disposed, but show a definite tendency to an arrangement in groups separated by spaces much wider than the normal intervals between adjacent rodlets. This grouping is, no doubt, related in some way to the extension of the layer which results from the increase in size of the egg; but just how the change comes about is not easy to understand, since the rodlets, at all events in the earlier stages, seem to be in continuity with the basal layer. But there can be no doubt the rodlets do undergo displacement and dislocation as the egg grows in size (cf:Pl. XXXIV. figs. 44&45).In the matrix‐substance between and around the rodlets are present numerous fine granules, the origin and significance of which are not clear. They are stilb recognizable round the rodlets in the laid eggshell.2In the interval between the 10 mm. and 12 mm. eggs a new layer (layer 3) begins to be deposited on the outside of the rodlet‐layer. In eggs of about 12 mm. diameter, it takes the form of a loose irregular layer composed of granules mostly aggregated to form coarse rounded or ovalish masses. By the time the intra‐uterine egg has attained its full size (egg DQ, 16.5 ×15 mm. in diameter) layer 3 has attained massive proportions, having an average thickness of about 0.14 mm. Layers 1 and 2 possess a combined thickness of only 0.016 mm., so that layer 3 constitutes by far the greater part of the thickness of the shell. It is distinguishable into inner and outer zones, both composed of the same amorphous refractive material, and only differing in: their degree of compactness. The inner zone adjoining the rodlet‐layer is loose and open, being composed of coarse granules, loosely arranged, the spaces between them being in communication with the inter‐rodlet spaces. The outer zone is composed of irregularly oblong masses, compactly arranged, and evidently formed by the fusion of coarse granules such as form the inner zone, these latter granules themselves representing the transformed granular aggregates seen in the earlier eggs (PP and E). In the outer zone, in addition to fine clefts and fissures between its constituent masses there are present irregular channels opening on the surface. These represent the pore‐canals of the laid egg‐shell, and serve to place the spaces in the inner zone and rodlet‐layer in communication with the exterior.The shell of the laid egg ofPlatypusdiffers from that of the full‐grown intrauterine egg mainly in the dense, extremely compact, and resistant character of the outer zone of layer 3, in the possession of fully established pore‐canals, and in the presence of a thin superficial layer outside layer 3. The shell of the pouch‐egg ofEchidnadiffers from that ofPlatypusmainly in the more open, less compacted character of layer 3.One interesting fact emerges from the foregoing observations, and that is that the Monotreme shell, like that of the typical Sauropsidan, is laid down in two parts. We have (1) the formation of the two‐layered shell, which topographically at least represents the shell‐membrane of the latter, and which is well adapted by its structure (a) to increase in surface‐areapari passuwith the increase in size of the egg, and (b) to transmit to the interior of the egg the nutritive uterine secretion which is necessary to supplement the yolk‐content of the same. Then (2) we have the laying down of the protective layer of the shell (layer 3), a process which does not commence until after the stage of the 10 mm. egg, and which is only finally completed after the intrauterine egg has attained its full size and has accordingly taken up its full complement of the above‐mentioned nutritive secretion. During the earlier stages in its deposition layer 3, owing to the loose arrangement of its constituent granules, is evidently capable of transmitting fluid‐material to the egg interior, but in its fully consolidated state, as seen in thePlatypusshell, it mould seem to be impermeable, thus necessitating the presence of definite pore‐canals through which the respiratory exchanges can be effected. Layer 3, which is stated to contain calcareous salts inPlatypus, would appear to represent the calcareous shell of the Sauropsidan, and layer 4 the cuticular layer of the same.It is worthy of note that the shell of the Marsupial egg does not advance beyond the stage of