like cilia reaching as much as two-thirds the height of the stout columnar cells. Along the ventral wall of this sac, the shorter ciliated longitudinal tract is continued from behind, interrupting the denser cilia, and passing imperceptibly into the intestinal epithelium in front. There is a single digestive diverticulum, opening just behind the ciliated sac, on the ventral wall of the stomach, close alongside the longitudinal ciliated tract. Its mouth is surrounded by ciliated cells whose currents appear to aid the entry of fine particles, and perhaps also the outward passage of egesta. At the angle of the stomach and along the margin of the ciliated groove, the cuticle becomes heavily thickened, forming a stouter ridge representing the gastric shield of style-bearing molluses.

In the living Murdochia, the stomach is filled with a viscid mucous rope, continuous behind with the less coherent food string entering from the oesophagus, and passing in front directly into the narrower string occupying the lumen of the intestine. The mucous contents of the stomach are thus divisible into two portions. The ciliated anterior chamber is filled with a short, opaque rod, whitish and quite free from food particles. This is inserted behind into a longer cord of mucus, permeated with dark-coloured ingested material, and extending through the rest of the stomach. It is constantly augmented by the inward passage of particles from the oesophagus. In addition to a good deal of unrecognisable detritus, material of plant origin such as leaf fragments, broken fern sporangia, and—in particular—numerous short segments of fungal mycelia, are almost always present. Murdochia must be to some extent non-selective, using the radula for scraping up the rich nutritive film overlying the damp surfaces within the leaf sheaths and similar locations. The general movement of stomach contents is apparent from figure 5.

The chief impetus to particles is conveyed by the colourless anterior portion of the mucous rod, which is rotated by the transverse beat of the “style sac” cilia. This has the effect of rotating the whole attached mass of food contents within the cuticulate chamber, while in addition, the shorter cilia of the ventral tract evidently impart sufficient forward motion to carry the contents slowly towards the intestine. This movement must be very gradual, permitting of sufficient time for such preliminary digestion as may occur while the mucous rod is rotated within the stomach. As suggested by Graham (1939) in the case of the land operculate Pomatias elegans, the long posterior chamber of the stomach evidently serves as a storage region, and—though feeding seems to be almost continuous in Murdochia—may help to even out irregularities in the flow of food through the stomach.

It is evident that the functioning of the stomach in Murdochia typifies what must have been the simplest and earliest condition of the style sac form of molluscan stomach. The mass of whitish mucus within the style sac—if we may call it such—is an example of what the writer has called elsewhere (1951b) a “protostyle.” Its chief role appears to be to transmit rotatory movement, imparted by the style sac cilia, backwards to the rest of the stomach contents—a function which is of obvious importance in the activity of the stomach. By this means, ingested food is drawn through the stomach at a slow, uniform speed by the action of the very robust cilia, concentrated in the style sac. Stomach contents are thoroughly admixed with such digestive enzymes as may be present. The smallest, finely divided particles are repeatedly rotated alongside the opening of the digestive diverticulum, into which they are received, in part by ciliated action, in part probably by compression movements of the sparsely muscular

Fig. 5—The stomach in longitudinal Section, somewhat diagrammatic showing the disposition of contents and the arrangement of cilliary currents.
Fig. 6—Transverse section of stomach, through the region of the gastric shield, just behind the digestive divertrculum.

[ unclear: ]

stomach wall. In addition, the larger, coarse particles near the periphery of the food string are swept across the longitudinal groove (IGR) and caught by its cilia to be rapidly carried forward into the intestine. Though the clear mucus of the protostyle must also be slowly carried into the intestine, where it forms the matrix of the faecal string, yet the strongest forward flow is along the ventral ciliated groove. This forms a channel whereby particles too large to enter the digestive diverticulum are able to by-pass the protostyle, and find their way directly into the intestine. The longitudinal groove in Murdochia is the sole representative of the ciliary sorting area in this snail. Pomatias, similarly, shows an almost total reduction of the sorting plicae typical of the stomachs of marine prosobranchs. The mode of feeding of marine microherbivorous gastropods, generally on a continuous stream of poorly sorted bottom deposits, is a habit calling forth a high development of sorting and grading mechanisms within the stomach. In terrestrial prosobranchs, such as Murdochia or Pomatias, the problem is not so much one of sorting, as of breaking up larger fragments of plant material and rendering them finally suitable for absorption by the digestive gland epithelium.

According to Graham (1939), the opaque mucous contents of the style sac in Pomatias constitute a true crystalline style: this is gradually thrust into the stomach towards the gastric shield, by a backwardly directed ciliary current within the style sac. No such current could be detected within the much smaller stomach of Murdochia; from the whole appearance of the mucous rod, including its direct merging into the faecal string, we may conclude it is not properly speaking a crystalline style. Amylase tests are not mentioned in Graham's account of Pomatias. In Murdochia the spotting tests for amylase were difficult to carry out satisfactorily because of minute size, but gave negative results. No doubt the chief function of the protostyle in Murdochia is its mechanical role of rotating the stomach contents.

It is unlikely that any digestion takes place in Murdochia in the part of the gut anterior to the stomach. Particles of food, including fragments of fungal hyphae, are too large to enter the digestive diverticulum intact, and recgnisable hyphal or other food particles were never detected either within the lumina of the digestive tubules, or within the cytoplasm of the digestive cells. The structure of the digestive gland is in no way dissimilar from that of other phytophagous prosobranchs, or lamellibranchs already investigated; ingestion of fine particles evidently occurs, followed by intracellular digestion. The question arises, how—and to what extent—are large particles, such as fragments of hyphae, first broken up within the stomach? There is little evidence that mechanical trituration occurs. The muscle layer of the stomach wall is not conspicuous, and the cuticle is for the most part not robust. It is possible that particles are pressed, by the rotation of the prostyle, against the stout ridge of cuticle (GSH) and thus become slowly comminuted, perhaps assisted by a squeezing action of the stomach wall as suggested by Yonge in the protobranchiate bivalves (1939).

Further, in the cyclophorids, such as Murdochia, which feed to a large extent on fungal mycelia (see also de la Torre et al., 1942), a high proportion of cellulose-like substances must be included in the diet. If this carbohydrate is to be successfully utilised, a digestive problem is presented that has in only a few cases been successfully surmounted by molluses. A possible clue to the means of cellulose break-down is obtained by microscopic examination of the opaque mucoid mass of the style sac. The colloidal matrix is here invariably packed with long, highly

motile spirochaetes, progressing through the medium in constant vibratile movement, and especially densely concentrated at the boundary between clear mucus and food material. It may be suspected that these organisms, occurring in pure state with such regularity, are not without some role in the digestive process of the molluse.

Large spirochaetes assigned to the genus Cristispira have been briefly recorded as living “commensally” in the stomachs and crystalline styles of a fairly long list of lamellibranchs. Yonge (1926) refers to the presence of spirochaetes in the style substance of Ostraea edulis, while Breed, Murray and Hitchens (1948) list occurrences in no less than 13 families of bivalves. There would appear to be no previous record of spirochaetes from a gastropod style sac; nor has any explanation been hazarded as to the role they may play. Results of great interest may be expected from further inquiry into the spirochaetes of style sacs, especially from a survey of their occurrence among molluscs, and their powers of breaking down complex carbohydrates. An analogy might be suggested with Spirochaeta cytophaga, a free-living cellulose-digesting organism, whose activity has been investigated by Walker and Warren (1938). As regards the snail, Murdochia, some amount of simpler carbohydrate is probably available witin the lumen of the gut for direct absorption, after external bacterial or autolytic breakdown of cellulose and other plant substances; but the problem of the fungal hyphae—so conspicuous in stomach contents but apparently quite absent in the faecal pellets—remains to be solved.

The digestive diverticula in Murdochia contain both digestive and excretory cells. The former are wedge-shaped or clavate, filled with granular cytoplasm, consisting of colourless refractile particles easily liberated by cell maceration. Ingested material appears to be taken up in an extremely fine state of division, and the digestive cells contain little or no coloured egesta for return to the stomach. Enzyme droplets are not detectable.

The excretory cells are broad-based and pyramidal, brown-staining with Van Giesen's. They apparently extract chlorophyllous or other absorbed pigments from the blood, passing their contents to the lumen in the form of large droplets, sometimes equipped with little tails, or drumstick-shaped, as they emerge from the cell. The intestine is of simple structure throughout, describing a short S-shaped course after leaving the stomach, and crossing to the right side around the edge of the renal sac. The rectum is slightly wider (70μ in diameter).

Each faecal pellet occupies the whole width of the intestine, and is stoutly ovoid in shape, being nipped off from the mucous string by repeated peristaltic constrictions of the intestinal wall, which is provided with a narrow coat of circular muscle, several fibres thick. Muscular action plays the most important part in faeces formation, both in Murdochia and evidently in mesogastropoda in general. While it is broadly true—as stated by Yonge (1936 and elsewhere)—that the gut in microphagous prosobranchs performs little muscular action, yet the small but definite peristaltic movements of the intestine have generally been overlooked. As well as in Murdochia, peristalsis has been observed by the present writer in Struthiolaria (1951), Suterilla (unpublished) and in Serpulorbis (1951a). Murdochia is reminiscent of marine mesogastropods in the care taken for the consolidation of faecal pellets, as a precaution against the fouling of the pallial cavity in gill-breathing animals. Mucus is liberally secreted by fusiform gland cells, interspersed between the narrow ciliated cells of the intestinal epithelium.