The Structure and Function of the Cutaneous Glands in Helix aspersa

Glands are present either in or below the epithelium of most external surfaces of molluscs. Beneath the shell and in the mantle groove lie cells which are concerned with shell formation, described by various workers in the past; this paper is concerned with those glands discharging on to outer surfaces which are not covered by the shell, and no mention will be made of either the pedal or the hypobranchial glands since they too have been previously considered.

The slimy covering of a mollusc is frequently referred to as ‘mucus’, a term with no precise meaning. Observation of a living specimen of Helix, however, suggests that the slime is not always of identical composition, and the histology of the structures concerned reveals that a variety of secretions is in fact produced.

Some detailed investigations have been made in the past: de Villepoix (1892), Prenant (1924), Heaysman (1951), and Russell (1954) all described H. pomatia, and other authors have worked on other genera. However, apart from that by Russell, most of these descriptions are of the histology of the glands and the appearance of the secretions, with little detail of the physical and chemical nature of the products. This investigation, although involving cell structure, has concentrated on the histochemistry of the glandular exudates. The attempt was made to determine the function of each kind, an aspect which has previously received little attention.

For histochemical work the choice of fixative is important. It is desirable to avoid mordants such as mercury, aluminium, or chromium, to avoid an acid medium in order to retain calcium salts, and to use formalin which is reputed to fix proteins without rendering them inert. The fixative most frequently used was that of Glick (1949), a 1:1 mixture of 15% formalin and 8% basic lead acetate, fixation being carried out for about 16 h. A white precipitate forms around the material with this mixture but usually this can be adequately removed by shaking or by wiping with a paint brush. Although this method gave reasonably good histochemical results it was unsatisfactory for critical histological work, being far too drastic in its shrinkage effects. For this part of the investigation either Susa or Flemming-without-acetic was used. By far the best results from the point of view of fixation were obtained by using an Edwards’s freeze-drier. By this method contact with chemicals is avoided until the sections are dewaxed, passed through the alcohols where fixation occurs, and hydrated before staining. Shrinkage and distortion are minimized (with the exception of one kind of mucus), and the protoplasm retains a much finer texture. There are, however, certain drawbacks. Serial sections cannot be guaranteed, blocks of tissue must not be larger than cubes of side 3 mm, and orientation of the blocks is therefore difficult. The histochemical tests revealed other intricacies which will be discussed later.

Large numbers of tests were carried out initially, including those to identify mucopolysaccharides, protein, fat, calcium, iron (both ferrous and ferric), strontium, copper, zinc, and phosphorus. For mucopolysaccharides aqueous toluidine blue (about o-i%) was found to give the best results. Staining time increased from 5 to 10 min with the age of the solution, which, however, gave good coloration for several months; this resulted in over-staining, differentiation occurring during dehydration. The material was inspected in distilled water immediately after staining because the metachromatic effect decreases in contact with alcohol, and a substance staining an intense pink may appear blue after dehydration. However, this reversal of metachromasia proved useful in comparative work since of two secretions both appearing pink in distilled water one might return to blue and the other retain a considerable amount of its pinkness after dehydration. Therefore, immersion in alcohol proved a useful way of differentiating between different kinds of mucus. All the results described here are those obtained after dehydration and the making of permanent mounts. It was found that viewing the sections in artificial light often enhanced the metachromatic effect. Alcian blue as a second stain for mucopolysaccharides was utilized as described by Steedman (1950), except that for more precise results a longer time of immersion in a much weaker solution was preferable since it completely avoided background staining. Occasionally the PAS technique was employed, according to the directions of Pearse (1954).

Heidenhain’s iron haematoxylin gave the best demonstration of the product from the protein glands; Millon’s reagent and ninhydrin stained positively, although rather evanescently, and only when large amounts of secretion were present.

Calcium could be displayed most readily by silver deposition by the von Kossa method (Gomori, 1952). Although the precise strength of the solution is not important, 5% for 5 min gives good results. An alkaline fixative should be used, although Russell (1954) considered one of pH 4·7 suitable. Alizarin red S or nuclear fast red also proved useful. Russell, working on H. pomatia, employed very long staining times with both von Kossa and nuclear fast red; these should be avoided, since not only is it unnecessary but it leads to darkening or colouring of the background. Control sections were treated with i N HC1 for 5 min before immersion in the staining solution.

Pearse’s molybdate test for phosphate, and Sudan black for lipids were also used..

For general histological investigation alcian blue was used in conjunction with Heidenhain’s haematoxylin and eosin; the more dilute the alcian blue the clearer and more distinct the result, the best results being obtained by immersion for 12 h in a scarcely coloured aqueous alcian blue after differentiation of the haematoxylin, followed by dehydration, with a little eosin in the first absolute alcohol. This gave a bluish-green colour to the mucus, and this contrasted well with the normal haematoxylin /eosin combination.

This paper is concerned with H. aspersa, a species not well described by earlier workers, but similar enough to H. pomatia for the latter to be used as a basis for comparison. Most, if not all, of the glands in the skin of gastropods are unicellular; each gland cell will therefore be referred to as a gland. In accordance with the terminology of Bowen (1929), ‘secretion’ will be restricted to the formation of the product from the cytoplasm within the cell, while ‘extrusion’ will be used for the removal of the product from the cell.

Although some glands are constant in appearance, considerable variation occurs among others. It is reasonable to assume that the physical and chemical nature of the secretion may change during its formation within the cell, especially as some of the substances involved may be excretory and vary with the conditions under which the animal had been living. The size and probably also the shape of the cell will vary according to its age and also perhaps with the stage of its secretion; after extrusion of the contents a decrease in volume would be expected. It is therefore perhaps not surprising that a certain amount of difficulty has been experienced in deciding how many distinct kinds of glands are found in Helix. However, taking all these factors into consideration, 8 types of gland cell, including 4 producing different sorts of mucus, have been distinguished; for reference purposes these have been calle:

mucus gland type A protein gland

mucus gland type B calcium gland

mucus gland type C pigment gland

mucus gland type D lipid gland

Three regions have been investigated, namely, the mantle-collar, the dorsal and lateral surfaces of the foot, and the sole of the foot. The glands in the mantle region of Helix are larger than those elsewhere and will be described first. Those on the dorsal and lateral surfaces of the foot resemble those of the mantle in many ways and will be considered later. Finally, the glands of the sole will be discussed.

This may be subdivided into two regions: the larger part adjoining the shell gland (the ventral surface), and the other lying approximately at right angles to it and adjacent to the sides of the body (the median surface).

The gross appearance of the collar varies from specimen to specimen, but it is always more or less glistening owing to the presence of its covering of slime. Under this it may be dark grey or almost cream, invariably mottled, and with the region around the pneumostome whiter than elsewhere. This surface of the mantle is covered by a single layer of columnar cells from 25 to 33 μ deep, bounded by a distinct cuticle. The presence or absence of cilia has received much comment in the past. Cilia are always to be found around the pneumostome, but elsewhere they occur only on small scattered groups of cells, and not continuously as described by Heaysman (1951). Beneath the epidermis lies the loose connective tissue in which the glands are found. All the gland cells are similar in consisting of a more or less swollen basal part opening by a narrow and well-differentiated duct or by a broad neck between the epidermal cells. Except in young stages very little cytoplasm can be seen.

The general shape of the mucus-glands of type A can be seen from fig. 1, their length being up to 800 μ. A nucleus is visible only in the smaller cells where it lies basally, becoming more and more compressed in larger glands until eventually it is invisible. The secretion is not granular, frequently appearing reticular or ‘bubbly’. After freezing-drying the mass of mucus often seems to have ‘exploded’, leaving trails of the material dispersed over the surrounding tissue; occasionally this happens after chemical fixation, and is probably due to the sudden effect of the fixing agent. The secretion stains with alcian blue, neutral red, Mayer’s mucicarmine, and (faintly) with PAS. It shows gamma metachromasia with toluidine blue, Terry’s polychrome blue, and thionin.

The distinguishing characteristic of the mucus-glands of type B is the clearly and evenly granular nature of their secretion. The cell-body is polygonal and comparatively small, opening by a fine duct which passes through the connective tissue and between the epidermal cells, though its position can be seen only when granules are present as the walls are too thin to be visible alone. The nucleus remains spherical and centrally situated. The whole cell including the duct may be 300 μ, long. The staining reactions of the secretion are similar to those for type A glands except that it is strongly positive with the PAS technique.

Previous workers have given various accounts of mucus-glands. Barfurth (1885) referred vaguely to mucus-glands, while Prenant (1924), using haemalum as his only stain for mucus, described two types; he did not find any protein glands and it seems more likely that he missed the second kind of mucus-gland and wrongly identified the protein. Roth (1929) distinguished, but not very clearly, two kinds of mucus-gland. Passing comment was made in 1951 by Heaysman about mucus-glands—presumably of only one kind; she alleged, however, that the secretion was found in the haemocoel and not in cells—although the associated nuclei can frequently be seen quite clearly. The most recent work (Russell, 1954) referred to one type of mucus-gland, though something which could well have been the type B cells was described but not identified either as mucus or as a gland.

The calcium-glands are the cause of the whitish mottling of the mantle and it is because they are particularly abundant around the pneumostome that that region appears whiter than the other parts. The cells themselves are similar in shape to the large mucus-glands, although rarely exceeding 550/z in length. The secretion consists of granules which are either spherical or very short rods, suspended in a faintly yellow matrix of basic protein. The granules frequently obscure the nucleus. When it is visible, however, it is often found along the side of the cell, although it may be terminal.

Positive results with von Kossa, alizarin, nuclear fast red, and Edward Gurr’s gallamine blue (all of which were prevented by previous immersion of the sections in 1 N HC1 for 5 min) indicate that the granules consist of a calcium salt; the fact that they effervesce readily with acids indicates a carbonate. The ammonium molybdate test for inorganic phosphate gives variable results with this material. The colour is evanescent and rather pale but in some sections a blue halo around the calcium glands has been obtained and in others a certain blueness actually within cells which were probably of this kind. Calcium phosphate is somewhat water-soluble and therefore very little would be left after hydration of the sections, regardless of the method of fixation; however, tests involving as little contact as possible with water contrasted with others left in running water for 20 min showed no detectable difference. It therefore seems likely that the calcium is transported to the glands as phosphate but that it is stored as carbonate. The secretion is certainly extruded from the cells as carbonate granules since they can be seen in sections lying in the mouth of the gland and in the slime clinging to the outer surface. Heaysman’s theory of transport through the epidermal cells and reconstitution on the surface of the mantle is therefore not supported.

Gray (1926), working on artificial membranes of mucin, found that dispersion was prevented by divalent cations. Further investigation on Mytilus gill tissue showed that the intercellular matrix, which is faintly positive with Millon’s reagent, was stabilized by calcium and magnesium ions. It is possible that a similar thickening of the protein occurs in the calcium glands of Helix. After chemical fixation the granules frequently appear to be arranged in bands as indicated in fig. x (for clarity fewer granules are shown than are in fact present). This may well be due to shrinkage causing the protein matrix to contract or pull away in strips from the cell boundary.

Previous workers have all recorded calcium-glands.

Protein-glands appear abundantly in the collar. Since their length (up to 850 /z) tends to be greater than that of the other glands, their bases frequently appear to form a band at the inner edge of the glandular mass. Their shape is cylindrical, appearing somewhat ribbon-like in section, tapering only very little and opening by a wide pore between the epidermal cells. Their nuclei are larger than those of the other glands; they tend to be ovoid and lie in a more plentiful mass of cytoplasm at the inner end of the cell. The secretion is usually homogeneous, although occasionally very finely granular; positive reactions with ninhydrin and with Millon’s reagent indicate that it is protein in nature. It stains with iron haematoxylin. This material in the fresh state has a distinct primrose colour; further comment on this pigment appears below.

In many specimens only these 4 types of gland are to be found in the mantlecollar; others, however, possess a fifth kind (labelled pigment-gland in fig. 1), and occasionally these occur in considerable numbers. In shape they are somewhat intermediate between the protein- and calcium-glands, though seldom as long as either. On rare occasions an ovoid nucleus is visible at the inner end of the gland, but their main characteristic is the deep yellow colour of the secretion.

Whereas in the other kinds of gland the secretion has been generally dispersed through the cell, either in single vacuoles or in granules distributed through the matrix, it exists in this type in various forms. Frequently a series of globules is present, of very variable size even within a single cell, each globule apparently with a distinct boundary since what appear to be cracks in it can sometimes be seen. Often one or more elongated structures are present which look very much as though they have been formed by the joining together end-to-end of several granules; the yellow pigment then seems to be concentrated on the surface and the inner part is paler. The rounding-off effect of these globules suggests an oily nature but no positive tests of any kind for lipid or any other substance have ever been obtained. The pigment probably masks any reaction to some extent. Although the pigment is evident in frozen-dried material before it is put into alcohol, none can be found later.

Pigment glands might perhaps be discharged protein-glands, in which case one would expect a positive reaction to be given at least by Heidenhain’s haematoxylin. Roth (1929) and Russell (1954) both recorded pigment-glands, without giving much comment on their nature; other workers have made no mention of them.

Finally, cells containing globules of lipid giving a positive reaction with Sudan black must be recorded. They are rarely found and have been seen only in frozen-dried material. Lipid would be dissolved out during the usual chemical fixation, but none was detected after formaldehyde-calcium, which normally preserves fats. Although the shape of the cells might be due to pressure of the surrounding tissue it suggests that of a gland, especially as in some instances what would appear to be a duct leading towards the surface can be seen. No lipid has, however, been found actually in the discharged slime, and lipid has not been recorded in the skin of Helix before.

Occasional cells very similar to the type B mucus-glands but with non-metachromatic granules have been found; they appear to be like the type D glands to be described later.

The main difference between this surface, which lies adjacent to the foot and the ventral surface, is in the relative abundance of the various kinds of glands; otherwise it is similar, though with a general tendency for the gland cells to be smaller.

Numerous mucus-glands of the A type occur here; indeed, there may be a cluster of a dozen or more lying pressed closely together, but there is considerable individual variation. Mucus-glands of the smaller type are scattered and not so frequent. Both protein and pigment glands are much rarer than on the ventral surface; although calcium-glands may be found on the lower edge of this part, where it joins the major surface of the mantle, they are never found higher up.

This change in distribution may well be correlated with epiphragm formation. Russell (1954) stated that the mantle-glands deposit the epiphragm, but gave no further details. In this species it is largely proteinaceous in composition with a small amount of calcium. When the body of Helix is retracted into the shell, the foot contracts to such an extent that it is withdrawn completely through the collar. The inner edges of the ventral surfaces of the mantle are drawn together below the contracted foot, while the outer edges retain more or less their original position around the edges of the shell mouth. This means that it is the ventral surfaces of the collar which occupy the position just within the opening of the shell, and it is presumably from this part of the mantle that the epiphragm material is extruded. The presence of protein and calcium glands on the ventral surfaces and their scarcity on the lateral ones is probably linked with their use in the laying down of the epiphragm.

The general appearance of these parts of a snail’s body is of a reticulum of grooves subdividing the surface into greyish polygonal areas. Microscopically, apart from the grooving, the structure is remarkably reminiscent of the mantlecollar, as may be seen from fig. 2, and the same kinds of glands are present, the main difference being in their smaller size, particularly on the dorsal aspect. There are isolated patches of ciliated cells in the epidermis, but the majority of the surface bears no trace of them.

Mucus-glands of type A occur commonly in this region, but there are fewer of type B; indeed, in some snails it is difficult to find examples. Occasionally intermediate forms are found, as well as specimens with metachromatic mucus. Protein-glands are less abundant than in the mantle, and this is also true, particularly on the dorsal surface, of calcium-glands; those containing either pigment or lipid occur sparsely as before. In each case the histochemistry of the secretions is identical with that in the mantle.

The sole of the foot is completely covered with fine cilia arising from elongated epidermal cells. This is the only region of the body, apart from the pneumostome, where cilia are invariably present. Beneath this layer lies a glandular zone extending to 450 μ, deep. The general shape of the individual cells can be seen from fig. 3.

Glands of type C are very numerous in the sole, and often occur in clusters, although there is never any indication of syncytium-formation, as clear cell boundaries remain. The secretion in the cells occurs as distinct granules; that in the ducts, particularly near the surface, often appearing somewhat reticular. After chemical fixation it stains with alcian blue, faintly with PAS; with toluidine blue it shows gamma metachromasia. It is similar therefore to the mucus of the type A glands, but it differs in taking up Heidenhain’s haematoxylin. The latter result suggests the presence of carboxyl groups, which, although probably in protein, might be in mucopolysaccharide. Tests with Millon’s reagent and ninhydrin were, however, negative; but since these methods never give a strong coloration and since the granules are minute, this is perhaps not surprising. Mucopolysaccharides are frequently, if not always, associated with a small and varying amount of protein; indeed, Sherwin (1935), Hempelmann (1940), Ewer and Hanson (1945-6), and Pigman and Goepp (1948) have all considered mucoitin-sulphuric-acid and chondroitin-sulphuric-acid to be mucoproteins, although they should correctly, according to Meyer (1945), be termed mucopolysaccharides. One is therefore tempted to suggest that the secretion from these glands contains more protein than that from the types A and B mucus-glands. This would seem to be supported by the work of Suzuki (1940-1) and of Masamune, Yasuoka, Takahasi, and Asagi (1947), who investigated the chemical composition of mucus (presumably meaning slime) from the sole and other parts of the body of H. laeda and found two distinct types. Suzuki’s figures show 13·9% nitrogen in mucus from the foot, presumably the sole, compared with 12-9% for the rest. Assuming that the gland population of the species is similar to that in H. aspersa, the 12.9% would include exudate from the protein glands as well as mucus; this means that the 13’9% for the sole glands is very much higher than the percentage of nitrogen in the actual mucus-glands elsewhere.

This work must, however, be treated with caution since the composition of the slime in H. aspersa varies with the strength of excitation—as shown later —and this could affect these results considerably.

According to Walton and Ricketts (1954) the binding of protein to mucopolysaccharides results in decreased metachromasia, in which case one would expect the sole glands to display less metachromasia than the A and B type glands, which is not the case. Certainly mucoproteins would not give gamma metachromasia, and one can only conclude that this secretion is a mucopolysaccharide, probably linked with some protein, and is distinct from that produced by the other mucus-glands.

Since these are the most numerous glands in the sole, they must be regarded as the ones which Roth (1929) and subsequent workers called soleglands. However, Roth was wrong in saying that they also occur in the mantle: the secretion of the type B glands found there is quite distinct in histochemical properties.

Scattered among the sole-glands, particularly within 265 μ. of the surface, lie occasional gland cells of type D. These appear very similar to those described above in general shape and in the granular nature of their secretion. However, on chemically fixed material the granules are positive to alcian blue and to PAS, but are not metachromatic with toluidine blue. In this they resemble the cells found infrequently in the other areas studied.

Some doubt exists as to whether the type D glands are a distinct kind. Cells can sometimes be found containing a very little secretion which has the type D characteristics; others are occasionally present which contain this material with some granules positive with Heidenhain’s haematoxylin, from which it might seem that type D glands were simply early stages of type C. However, non-metachromatic material has sometimes been seen in ducts passing between the epidermal cells and it therefore seems likely that type D glands do discharge their contents, which means that they are unlikely to be young stages. It may be that both C and D mucus-glands have a common origin.

It is perhaps pertinent here to mention the development of the other kinds of gland. Opinion in the past has been divided into two schools of thought: that they are epidermal or that they are of connective tissue origin. Roth (1929) maintained that it is rare to see developmental stages of glands in mature specimens of Helix but that they occur in the unhatched egg and young animal. From this one is led to assume that some at least of the glands presumably function throughout the life of the animal. Certainly not many convincing developmental stages have been seen during this work, which has been mainly on snails more than one year old. However, periodically enlarged connective tissue-cells with obvious early stages of protein secretion or the various kinds of mucus secretion around the nucleus have been seen. Steps in the formation of a duct or elongation of the cell towards the surface have also been observed. It seems likely that at least some of the skin glands of Helix are of connective tissue origin.

The results of histochemical tests on the calcium secretion and various kinds of mucous secretions with frozen-dried material have been the same as after chemical fixation. The apparent absence of pigment glands has already been commented upon. However, the reactions of the protein secretions are completely different. Whereas after chemical fixation a positive result was obtained with Millon’s reagent, ninhydrin, and iron haematoxylin, after freezing-drying no staining occurred.

This suggests that the protein is not bound during the process and is therefore washed away during hydration, or that it is bound so firmly that it cannot react with the stain (although this is unlikely), or that it is denatured in some other way. Formalin is reputed to bind protein, but immersion of the frozendried sections does not change the result. Little seems to be known of the effect of freezing-drying on protein. Proom and Hemmons (1949) showed that temperatures down to —78° C (that used here was —170° C) do not seem any worse than —17° C from the point of view of survival of bacteria. Haines (1938), working on freezing alone, showed that the critical temperature for bacterial survival, although varying with the species, appears to be about —2° C and that even after 8 days at this temperature only 50% of the coagulable protein had coagulated. An interesting point is that the secretion of the sole-glands, which was positive with Heidenhain’s haematoxylin after chemical fixation, remains positive with it after freezing-drying. If the haematoxylin is demonstrating protein, this suggests that that bound to mucopolysaccharide is not affected in the same way as unbound protein.

Therefore, although freezing-drying offers many advantages, histochemical results must be interpreted with caution.

Robertson (1941) postulated that the skin glands of Helix were organs for the excretion of excess calcium. The presence of a yellow pigment in the slime suggested that it, too, might be a waste product. An experiment was therefore carried out to see what properties of the secretions depend on the food of the animal.

Eighty-four snails which had been well fed with cabbage, rolled oats, and chalk were distributed between 7 glass tanks. Each series was kept under different conditions, thus:

Group A received a full diet of cabbage, rolled oats, and chalk.

Group B received carrot, rolled oats, and chalk.

Group C received cabbage and rolled oats.

Group D received rolled oats and chalk.

Group E received cabbage, chalk, and fibrin.

Groups F and G were starved.

Groups A to F were maintained at 18° to 21° C. They were moistened regularly, and given fresh food every 2 or 3 days; group G was kept in a refrigerator at 4° C.

The central part of the mantle-collar of one snail from each group was fixed, after 1, 2, 3, 4, 6, 8, 11, 15, 20, 42, 75, and 93 days, in formalin /lead acetate and embedded. Because quantitative results were desired, all the material was sectioned together and as far as possible stained together or for identical lengths of time.

The pigment could be observed in unstained sections; tests were needed for mucus, protein, and calcium. Those used were toluidine blue, Heidenhain’s haematoxylin, and von Kossa respectively. It was obviously impossible to obtain precisely quantitative results in this way, but all the slides were examined in random order and the amounts of each kind of secretion estimated as normal, more, much more, less, and much less.

There were no apparent changes in the amounts of mucus present in these snails except a possible increase in those starved in the refrigerator. The only variation in protein was a slight drop in the starved, warm specimens and in those fed but without fresh plant food. The presence or absence of calcium in the food appeared to have no effect on the glands, although the snails readily ate the chalk; but starvation, especially in the refrigerator, produced a decrease in quantity. Perhaps the most interesting result was a decrease, particularly after a fortnight, of the amount of pigment in the glands from the group D snails—i.e. those which had had no fresh plant material in their diet. As well as this, however, there seemed to be a rather large amount in those kept in the refrigerator.

These results do not support Robertson’s hypothesis that the glands excrete excess calcium; they coincide however with those of Wagge (1951), who regarded the shell-gland and digestive gland as stores. On this hypothesis the amount in the skin glands would be unrelated to diet.

The results from the snails kept in the refrigerator must be treated with caution since under such condition’s metabolism would be slow and there would be a long period during which the materials already in the body at the beginning of the experiment could be utilized.

In summary, it seems that only the amount of pigment in the skin-glands bears any relationship to the diet of the animal.

This provokes the question of the nature of the pigment.

Yellow pigments in animal cells are likely to be either carotenoids, flavines, or flavones. Tests were carried out on fresh slime collected from irritated snails, since fixatives have a very marked effect on the colour, which is almost or quite missing from tissues treated with Susa, Zenker, or Carnoy, but is retained after formalin /lead acetate right through the hydration process, and after freezing-drying until the tissue reaches the alcohols.

The standard tests for carotenoids all gave negative results, and similarly those for flavines. However, the pigment shows the following characteristics of a flavone: it is water- and alcohol-soluble (hence its absence from most chemically-fixed material), is bleached by dilute acids, becomes brown with ferric chloride, deepens in colour when subjected to ammonia fumes, and evidently forms an insoluble yellow salt with lead as shown by the fixation in formalin /lead acetate.

The pigment has therefore been identified as a flavone. The histochemistry of the tissues showed at least most of the pigment to occur in glands whose contents were also positive with tests for protein. According to Fox (1953), carotenoids and flavines may be associated with protein, but he makes no mention of a combination with flavone. He states, however, that these substances are assumed to be taken into the animal body from plants. Flavones are very widely distributed in all organs of plants and this would be correlated with the observation that the quantity of pigment in Helix appears to vary with the amount of fresh plant food eaten. Fox recorded flavones only in insects and hydroids.

Reference was made in the introduction to the apparent change in the kind of slime released under different circumstances. To obtain slime, stimulation with electric shocks of constant duration and controlled frequency was used. The specimens were stimulated by holding the electrodes in contact with the body, as nearly as possible at the anterior and posterior extremities of the animal. Shocks were administered every minute, long enough for complete retraction into the shell, followed by re-emergence of the body. The voltages tested were 2, 5, 8, 10, 14, and 20, each snail receiving 4 similar shocks.

The type of slime released showed a direct correlation with the strength of the stimulus. At 2 to 5 volts it was usually colourless and markedly viscous— the normal slimy covering of a snail. At about 8 volts a whitish secretion sometimes, although not always, appeared—whitish because of the calcium carbonate granules present. Subsequently a yellow, much clearer slime was released, which was less viscous and rather watery in consistency; it was this type of secretion which accounted for the largest proportion of the total volume produced.

Dexheimer (1951) recorded two types of slime for H. pomatia: a frothy, clear secretion produced under normal conditions and on slight stimulation, and a sticky, creamy kind which appeared under long-continued or violent stimulation. As he did not specify the nature of the stimulus, it seems likely that he did not irritate sufficiently to release the third type of slime, if this species reacts in the same way as aspersa.

From the previous experiment it appeared that the slime released after irritation was more watery than the normal covering. In an attempt to prove this, snails were carefully cleaned of extraneous material clinging to them and wiped very gently with a clean, weighed glass slide. Each such slide was reweighed with its adherent slime and then left so that evaporation could occur until its weight was constant. This gave the water-content of the ‘normal’ slime as 86·1%.

After stimulation with 20 volts the thinner exudate was allowed to drip on to other slides and its water-content calculated as before. It was 96-3%. The results are statistically significant at well over the 1 in 1,000 level.

This proved therefore that the slime produced by an irritated snail is more watery than that normally released.

It was then desirable to determine, if possible, the relative amounts of the different kinds. A series of snails was stimulated 10 times with 20 volts while supported individually in funnels running into small specimen tubes of known weight; the thinner exudate dripped freely into the tubes, the thicker was removed from the creatures with forceps and added to the rest. Because of the movements of the snail in response to the shocks the slime was rather bubbly. It was weighed and left corked until most of the froth had disappeared and then the height in each tube was marked. After drying, the weight of the slime was redetermined. The dried residue was removed and water run into the tubes to the points marked from a graduated pipette, giving the volume of the slime.

If one knew the volume, total water-loss, and water-loss expected from each kind of slime, the relative amounts could be calculated. These figures gave a value of 98-6% for the proportion of thin yellow exudate in the total slime of the irritated animal. It must be emphasized, however, that these results are not very accurate, because of the difficulties in removing the thicker slime from the body.

Further properties were determined with slime collected from electrically stimulated snails.

The pH of this material, as shown by a series of indicator solutions, was between 8·5 and 9·0. Similar tests carried out on fresh-frozen sections gave the same figures, with some indication that the mucus was slightly more acid than the protein. This is in agreement with Prenant (1924), who recorded that the mucus (presumably meaning slime) of H. pomatia is slightly alkaline.

Although the protein-glands appear to discharge as part of a defence mechanism, their secretion is neither highly alkaline nor acidic. Tests with 35 people of different ages and both sexes suggest that there is no very strong flavour associated with the exudate.

Essex (1945) recorded a protein-splitting enzyme in the venom of some reptiles. On the chance that Helix might be similarly supplied, drops of slime were put on to sterile 10% gelatine films in Petri dishes. Controls with slime kept at 100° C for 10 min to activate any enzymes present were also set up.

After 24 h the gelatine below the untreated slime had been partly digested away, whereas it remained unaffected beneath the control drops.

The slime of H. aspersa has thus been shown to possess an active proteinase. Its importance is difficult to determine and it is unlikely to be strong enough to harm any predators. However, Russell (1954) reported that the epiphragm of this species appears to be dissolved away by mucus (presumably he meant the whole exudate), and the epiphragm is formed largely of protein. There would be little need for the breakdown process to be rapid and it may be that this is the function of the enzyme which has been detected.

The body of a snail normally appears clear of either bacterial or fungal growth, although constantly in contact with such organisms. Fischer (1948) suggested that the slime is actively antiseptic, and experiments were therefore devised to test this hypothesis.

Sterile Petri dishes of potato dextrose agar and peptone glucose agar were exposed overnight at ground level; half of the surface of each was then coated with the exudate from H. aspersa which had been collected with as little contamination as possible. Other plates with soil mixed with the agar as it cooled and therefore containing a wide variety of micro-organisms were treated similarly. After 3 days at 24° C in all cases fungi and bacteria were generally distributed over the plates and in 3 cases appeared more numerous on the side with the slime. The growth was so prolific, however, that similar tests with streaks of pure cultures of fungi and bacteria (species common in soil) counterstreaked with exudate were set up, but again the results showed no inhibition. Clearly, then, the slime does not act as an antiseptic.

To further the suggestion of the first experiment, that growth was actually enhanced, slime was smeared over half the surface of a series of plates of nutrient agar, which were then exposed at ground level overnight. Large numbers of colonies appeared and counts were made of those on and away from the smear. Analysed statistically at the 1 in 100 level, the population on the slime was bigger than that away from it.

There are two possible explanations—that the exudate retained spores settling on it more easily than the surface of the agar, and that it exerted a definite stimulatory effect. Although the first hypothesis is possible, since in the first trial the slime was administered after exposure to the air and the results appeared similar, this seems unlikely. The second may well be true. Perhaps the protein or carbohydrate offers extra nutrients to the microorganisms. The possibility that the slime contains growth-promoting substances cannot be discounted.

This is an aspect that nas received hardly any attention from previous workers.

Helix has an extensive system of haemocoel spaces and these might be assumed to take some part in causing discharge. Branches of these bloodspaces ramify throughout the body. The connective tissue of the mantlecollar is well supplied but the haemocoel rarely penetrates between the gland cells, although many cavities occur around their bases, and the same is true of the dorsal and lateral surfaces of the foot. In the sole of the foot, however, branches penetrate the whole glandular area (figs. 1-3).

Such a system could work by changing the pressure in the blood-spaces, which would in turn cause pressure on the glands and thus initiate their discharge. These changes in pressure would be brought about by movement of the animal, and on this would depend the discharge of the glands.

In the sole of the foot, changes in blood-pressure would result in the glands being compressed from all directions; the cell-bodies would respond to this pressure by ejecting their contents through their ducts to the surface. During locomotion a constant series of rhythmical contractions passes along the sole, producing rhythmical changes in pressure in the haemocoel, and on the glands, and a rhythmical discharge of their secretion. When the animal ceased to move, the secretion would cease to be released. From the functional point of view this would meet the snail’s requirements, for only during locomotion would replenishment of the slime be necessary. In the sole of H. aspersa, therefore, the discharge of the glands could be accounted for quite adequately by changes in the blood-pressure.

Since the haemocoel does not penetrate so far between the glands in the mantle-collar and dorsal and lateral surfaces of the foot, pressure changes would constrict the glands less than in the sole. Although changes in bodyform occur here too as the animal moves into and out of its shell, they are by no means so definite or continuous as on the sole. Thus, if dependent on blood-pressure, exudation would be spasmodic and perhaps not plentiful. However, since these surfaces are exposed to the air they need a good supply of exudate and experiment has shown that its volume and kind may vary with the situation in which the animal finds itself. It is therefore difficult to envisage blood-pressure alone as being responsible for discharge of the glands in these regions.

The majority of the glands found there are of a different type from those found on the sole: they are larger and have no distinct duct. The presence of muscle-fibres in the connective tissue around these cells has already been mentioned. When a surface view of part of the bounding membrane of one of these glands has been obtained it can be seen to be covered by an irregular network of fibres (fig. 4), which run obliquely around the inner part of the cell, and more longitudinally in the outer part. No sphincter or other structure has been seen on the duct.

Surface views of the network have been observed only round the protein- and the calcium-glands, although the appearance of sections is similar also round the large mucus-glands. The best material for investigating these fibres is that which has been frozen-dried. This process has a drastic effect on the mucus, causing it to be dispersed over the surrounding tissues and obliterating any vestiges of the original bounding membrane which might remain. Contraction of such a network round a cell of this shape would cause the gland to shrink not only in diameter but also in length, thus bringing about exudation. Although nothing is known about innervation of these structures, this method of discharge would seem to fit the situation better than bloodpressure, because the amount extruded could be controlled by nervous stimulation without any appreciable movement being necessary; and this could easily explain the increase in the volume of calcium and protein secreted when the animal is irritated.

No evidence of investing muscle-fibres has been seen in the glandular region of the sole: there the gland-cells are often found in clusters lying adjacent to one another, with apparently no space between in which fibres might be present.

Previous comment on discharge was made by Jones (1935), who, working on the mantle of Anguispira alternons, stated that ‘Sometimes the large mucus glands may have a musculature enveloping the gland in a loose network of fibres, probably only a differentiation of closely adhering interglandular muscle fibres’. According to him similar structures are present in H. pomatia and Polygyra thyroides. Fretter and Graham (1949) recorded a basket-work of muscle-fibres around gland cells in the foot of pyramidellids, and they too thought they were involved in causing release of the secretion.

Thus one is led to the conclusion that the skin glands of H. aspersa are not all induced to discharge in the same way. It seems likely that in the sole blood-pressure is adequate to cause the release of the slime from the glands, but that in other regions, where most of the gland cells are club-shaped, this alone could not suffice, although, nevertheless, it may well play an important part. Here some glands, and probably all, are enveloped in a network of fine muscle-fibres, which on contraction cause exudation of the enclosed secretion.

A certain amount of confusion exists in the terminology applied to mucoid substances, which makes precise interpretation of staining results difficult. This arises because some mucopolysaccharides may be firmly or loosely bound to varying amounts of protein, and although, according to Meyer (1945), chondroitin-sulphuric-acid and mucoitin-sulphuric-acid should properly be termed mucopolysaccharides even though they are firmly bound to protein, Hempelmann (1940) and Grishman (1952) described them as mucoproteins.

The amount of this associated protein and the way in which it is linked to the polysaccharide might be expected to affect the staining reactions of the carbohydrate; it may or may not itself be demonstrable. Thus it may be possible to be certain of the presence of protein by its staining reaction but it is not possible to be certain of its absence.

The standard test for a ‘mucus’ is the development of gamma metachromasia with a dye such as toluidine blue, but this result may be given by a variety of substances: thus Hempelmann (1940) regarded it as characteristic of most mucoproteins, including MSA and CSA, while Grishman (1948) listed only MSA and CSA. Lison and Mutsaars (1950) reported that nucleic acid may sometimes react metachromatically, but Sibatini (1952) stated that nucleic acids produce beta but not gamma metachromasia, and that the latter is characteristic only of mucopolysaccharides. Further detail was added by Hale in 1953, who regarded gamma metachromasia as indicative of polysaccharides with uronic acid as well as sulphate groups, or with polymeric phosphate groups.

Further comments on metachromasia include those of Landsmeer (19512), who discussed the importance of electrostatics, especially with respect to salt concentration, mono- and divalent ions suppressing staining; Sibatini (1952) agreed that electrostatic charges are significant. Grossfield (1954) noted that metachromasia changes on the death of the cell. According to Walton and Ricketts (1954) it is the attachment and dissociation of acidic radicles which is important and not polymerization of either the substrate molecules or the dye on the substrate, as was thought by Michaelis (1947) according to Pearse (1954). Clearly, then, it is impossible to be certain of the chemical significance of gamma metachromasia. It seems likely that both MSA and CSA always react in this way, although other substances may do likewise.

With regard to other methods, it should be noted that Steedman (1950) reported that all kinds of ‘mucus’ pick up alcian blue, and although Lison (1954) treats the test with caution because the mechanism is not understood, he too agreed that only acid mucopolysaccharides are stained and that mucoproteins and neutral mucopolysaccharides react faintly if at all. During this investigation, however, alcian blue has been taken up by a wider range of substances than has given gamma metachromasia.

Comments on the significance of positive PAS reaction also vary: McManus (1948) states that mucin and sometimes glycogen stain, Grishman (1952) regards it as demonstrating glycogen, some mucoproteins such as MSA but not others including CSA (using his terminology), and various other substances; Glegg, Clermont, and Leblond (1952) agreed that CSA is negative but some other mucins positive as well as a very few proteins. Against this must be put Hale (1953), who obtained a strongly positive result with free CSA.

It is difficult to distinguish between CSA and MSA, and attempts on Helix have been unsuccessful. CSA is usually a connective-tissue component, while MSA occurs in epithelia; in view of the apparent development of these glands from the connective tissue it would be interesting to know which occurs here.

It would seem likely that the secretion of the type A mucus-glands is an acid mucopolysaccharide containing no or only a small amount of protein. It may well be MSA or CSA. That from the B type of mucus-glands must be similar, since its reaction with toluidine blue and alcian blue is the same, but it picks up PAS far more readily, indicating some difference in molecular structure.

The sole-glands offer a problem. Their gamma metachromasia would seem to indicate an acid mucopolysaccharide character, although the result with Heidenhain’s haematoxylin might mean that acidic protein is present, which is not found in the other kinds of mucus-gland. On Walton and Rickett’s theory, protein should decrease the degree of metachromasia; yet this did not occur here. The safest description for this secretion is that it is a mucopolysaccharide probably associated with a considerable amount of protein, but not a mucoprotein as shown by the toluidine blue.

Similarly, the type D glands cannot be readily identified by their secretion. It is positive with PAS and alcian blue, but is not metachromatic with toluidine blue. This suggests, since it gives no reaction for protein, that it is a neutral mucopolysaccharide, although, according to Pearse (1954), alcian blue should give only a faint result in this case. The only other possibility is that the granules are composed of glycogen.

The functions of the various secretions must now be discussed. Of these lubrication is perhaps the most obvious.

Throughout the animal kingdom the main lubricating agents are mucopolysaccharides. They are found in the lining of the alimentary canal and genital tracts from worms to mammals, and in the skin from Cnidaria to Amphibia. Histochemically these secretions are similar to that found in the large type A mucus-glands and it is logical to assume, therefore, that the main action of this exudate is lubrication. The distribution of these glands in Helix coincides with the areas that would need an agent to reduce friction. They are most numerous on the median surface of the mantle skirt, an area where this would be of great importance.

Lubrication of a particularly well-controlled kind is essential on the sole of a snail because of the method of locomotion. Cilia cannot beat without fluid over them and one might expect that the muscular contractions which move the creature would result in a far less smooth kind of progression without the aid of some kind of mucopolysaccharide on the surface. Some of the mucus used in locomotion is released from the pedal-gland, which has not been considered here; it is logical to assume, however, that the great array of soleglands also contributes to the supply. Ideally, the exudate should be fairly thin so that it would flow readily and offer no resistance to the waves of muscular activity. The protein associated with these glands may perhaps lower the surface tension of the mucus (which in other respects stains similarly to that of the type A glands) and enable it to flow more readily.

If the slime-trail from a snail which has moved over a slide is stained, the secretion from the sole-glands can be seen as a dense mass of extremely fine threads lying more or less parallel to one another. These are probably formed from the secretion droplets as the animal drags itself over the substratum. The very regular appearance of the trail suggests that the glands release their exudate gradually and more or less continuously. This would be expected if, as has been suggested, discharge is caused by changes in the blood-pressure as the animal moves. In contradistinction, threads are not evident in slime smears from the other surfaces of the body, which appear far more patchy, as though the secretion was released more spasmodically. It may be mentioned here that patches of irregular shape and size which stain pale blue with toluidine blue and often have an oval, darker area within, are usually apparent as well as the exudate; these are probably epidermal cells or parts of spent gland cells. The latter is the less likely, since one fails to find the developmental stages of glands to replace them.

Arey and Crozier (1921) stated that adhesion in pulmonates is due to slime and one must therefore assume that the secretion of the sole-gland must also be used for this. Gray (1926) showed that calcium makes intercellular matrices more viscous. Calcium is present in small amounts in the sole and this might be adequate to change the viscosity of the sole-gland secretions. Far more is present in the sides of the foot, which would mean that a rim of thicker mucus might be found round the edges of the sole, which may help in adhesion.

Robertson (1941) considered the calcium in the glands of Helix to be an excretory product, but this has been shown to be erroneous. It appears that the granules are released more copiously when the animal is irritated, and this would suggest that it has some protective action. The very fact of mixing granules with a fluid would make that fluid thicker, and Gray showed that divalent cations increase the viscosity of some protein or mucopolysaccharide substances. If the same effect occurred here, it would produce a deeper layer of slime over the body. The absence of calcium-glands from the median surface of the mantle, where thick slime is not desirable and where predators are unlikely to attack, supports this hypothesis.

A clue to the function of the protein may be its release in response to stimulation, from which one might conclude that it has some defensive action. Tests of pH, taste, and the effect on micro-organisms do not support the idea of active defence or attack. However, it is a very watery secretion, of much lower surface tension than the mucus, and this is probably the keynote of its action. A snail that has been irritated by powdered metaldehyde, the well-known slug poison, responds by extruding vast quantities of this protein secretion; the result is that the slime, carrying the power, runs off the body, and it seems likely that this is the normal mode of action.

It is difficult to decide the function of the two other types of mucus-gland, B and D; neither occurs abundantly, and their action may well be supplementary to that of the others producing mucopolysaccharides.

Hogben and Kirk (1944-5) suggested that the slime is important in temperature regulation by evaporation. Certainly it is far from hygroscopic and its production must cause considerable dehydration of the animal, but Helix is rarely found in dry or hot situations where evaporation would have its maximum effect on temperature regulation.

One final function of the slime must be mentioned—that of the formation of the epiphragm, which is thought to be produced from the combined secretions from the ventral surface of the mantle-collar. The presence of a proteinase in the slime and its possible link with the dissolution of the epiphragm has already been discussed.

I am very grateful to Professor A. Graham for providing me with facilities for research, including the use of the Edwards’s freeze-drier, in his Department of the University of Reading, and especially for numerous discussions during the progress of the investigation. I am also deeply indebted to the Mrs. Smith Trust Fund for a grant which enabled me to continue the work.