Discussion

The "Basement Membrane": It was suggested in an earlier paper by the present author (Wineera, 1969) that the "basement membrane" probably is collagenous. The results of the present study do not contradict this suggestion, but nor do they give it unequivocal support. The "basement membrane" exhibits a variable PAS reaction, and weak positive staining for protein, which could be expected of collagen (Pearse, 1960, p.162). It showed no metachromasia after staining with toluidine blue or azure A either before or after sulphation, suggesting that acidic carbohydrates are not present in this structure. There is the possibility that some are present but that the reactive groups which give rise to the metachromatic phenomenon are not present in sufficient quantity or are not placed at suitable intervals along the polysaccharide molecule. The weak positive staining with Mowry's colloidal iron reagent is best considered as non specific staining by the reagent and not indicative of the presence of acidic muco-polysaccharides. In the allochrome procedure the "basement membrane" coloured deep blue, which Lillie (1951) considers a characteristic of collagen when subjected to this procedure. Skaer (1961) describes in a fresh water triclad a thick basement membrane which he concludes contains collagen.

Parenchyma: The parenchyma of turbellarians continues to be an extremely difficult subject for study. The elements clearly distinguished in the present study were the neoblast cells (Pedersen, 1959) and the page 7 gland cells. As found by Pedersen for neoblast cells in a fresh water triclad, these cells in P. stephensoni contain large amounts of cytoplasmic RNA. The present study sheds no light on the problem of the origin of the "basement membrane", or the fibres continuous with it (Wineera, 1969). The extent to which lipids occur in the parenchyma, and their distribution, cannot be decided until further work involving electron microscopy is undertaken. The cells in the parenchyma which gave a strong positive reaction with the colloidal iron reagent are considered to be glands associated with the penis.

The Eosinophil Glands (Pl. 1, Fig. 3; Pl. 1, Fig. 7, S) : These are clearly protein in nature. The adhesive glands are shown to contain arginine, and the caudal glands tyrosine and tryptophan. They contain no polysaccharide material. The weak staining of the adhesive glands in the Mowry colloidal iron technique is considered non specific staining by the iron reagent. This view is supported by the fact that the glands do not colour with Mowry's reagent after sulphation, and do not exhibit metachromasia with azure A or toluidine blue. The Mowry colloidal iron method is based on the demonstration of bound ferric iron as prussian blue (Mowry, 1958). Wagner & Shapiro (1957) indicate that some proteins bind ferric iron and could give positive results in tests which demonstrate this ferric iron, as for example, the colloidal iron test for acidic polysaccharides. The staining of both types of gland by Sudan black is not considered "true sudanophilia" due to lipids, since staining was not diminished by hot pyridine extraction. The variable staining of the adhesive cells with pyronin indicates variable amounts of RNA in these cells. This can be correlated with the protein nature of the adhesive gland secretion. The variable nature of the reaction may indicate different stages of the secretion cycle in different cells. The cells showing "stranded" pyroninophilic cytoplasm probably represent those described by Wineera (1971) as possessing ergastoplasm. Pedersen (1959, 1963) records the presence of eosinophil adhesive glands in fresh water triclads. The distribution of the glands described by Pedersen is similar to the distribution of the eosinophil glands in P. stephensoni. Pedersen also describes the glands as being protein in nature, but states that they contain, in addition, a phospholipid component. The caudal glands referred to in this study are considered to be associated with the reproductive organs since they are found only in the region of the genital openings.

The Basiphil Glands: The reactions of these gland cells indicate that they contain neutral mucopolysaccharides: They exhibit diastase resistant PAS positivity, and are negative in tests for protein, lipid, and nucleic acid. Also they do not stain in the tests for acidic polysaccharides except after sulphation (Pl. 1, Fig. 5) which is a recognized procedure (Pearse, 1960; Moore and Schoenberg, 1957; Kramer and Windrum, 1954) for forming acidic polysaccharides from neutral ones present in tissue sections. The observed case of metachromatic staining of the basiphil glands with toluidine blue before sulphation seems anomalous in the light of the previous discussion, for it should indicate the presence of acidic mucopolysaccharides (Pearse, 1960; Moore and Schoenberg, 1957). Nucleic acids can display metachromasia under certain conditions (Pearse, 1960; Bergeron and Singer, 1958) but in the present case the gland cells gave page 8 negative results in tests for nucleic acids. It would appear best at this time to attach no significance to this case of metachromatic staining because (i) the staining was weak and (ii) the staining was an isolated case and could not be repeated in several subsequent attempts.

The nature of the basiphil secretion appears to change both in morphology and in staining reaction from the subepidermal parts of the glands to those in the epidermis: The subepidermal parts of these glands are granular and, as has already been mentioned, contain neutral mucopolysaccharides. The intraepidermal portions often appear homogenous, and stain in tests for protein (e.g. mercury/bromphenol blue test) (Pl. 1, Fig. 4, S) as well as in the PAS test. In the allochrome method the distinction between these two parts is seen clearly. The significance of these differences within different parts of the basiphil glands is not known. It is possibly that the protein part of the secretion is a product of the epidermal cells themselves, since these were found to contain varying amounts of RNA adjacent to their nuclear region.

The basiphil glands stain with Sudan black B, but this does not indicate the presence of lipids, as extraction of sections with hot pyridine does not lessen the staining reaction. Presumably the Sudan black acts, in this case, as a weak basic dye (as Casselman (1959) shows it is able to do), and not as a non-ionic fat soluble colourant.

In his studies of fresh water triclads, Pedersen (1959, 1963) describes three types of basiphil subepidermal gland. One type (called type 3) owes its basiphilia to RNA granules in its cytoplasm and is found immediately in front of the pharynx. This type appears to have no counterpart in P. stephensoni. The other two types, taken together, are similar in morphology, distribution, and staining reaction to the basiphil glands in P. stephensoni. In the present study slight differences in the basiphil gland cells occurred. For example the size and degree of packing of secretion granules varied slightly from some gland cells to others. But all basiphil glands exhibited the same staining reactions, and the variations observed were not thought sufficient to justify the division of the basiphil glands into two groups. However, it is possible as Skaer (1961) suggests, that the basiphil glands include more than one histochemically distinct type.

Pigment: The pigment in P. stephensoni may be a melanin. It is slowly bleached by hydrogen peroxide, and is bleached by 40% peracetic acid in a time which is very close to that given by Pearse (1960) for the bleaching of melanin in tissue sections by this substance. Skaer (1961) concludes that the pigment in the triclad Polycelis nigra is a melanin, but Needham (1965) states that P. nigra pigment can be extracted by dilute (4%) hydrochloric acid, which is not characteristic of melanin. In the present study the pigment granules were insoluble in 5% HCl.