Discussion

The system of channels in the outer region of the epidermis could be part of a mechanism that functions to protect the animal from damage to its epidermis, as there is no cuticle. These animals are probably exposed to many kinds of environmental conditions such as the abrasive power of silt laden water, and the washing of the alga on which the worms are situated against rocks, which are potentially damaging. It seems reasonable to suppose that damage to the surface of an epidermis which has an internal system of channels forming "compartments" (such as that in P. stephensoni) could be repaired relatively easily by "partitioning off" the damaged "compartments". This "protective mechanism" could be especially effective in the case of unicellular epidermal gland cells, for if part of the surface of such a cell was damaged, the ease of isolation of the damaged region from the rest of the cell could allow the cell to continue functioning as a secretory structure. If such a mechanism did not exist, it is probable that any damage done to the surface of epidermal cells, and in particular to the surface of epidermal gland cells would result in a great loss of cell contents. This, in turn, would probably lead to the death of the cells concerned.

It must be noted that the precise morphology of these channels in the outer regions of epidermal cells could not be determined in the present study. The above theory, therefore, is largely conjecture, and must wait until the structures upon which it is based are examined by electron microscopy before it can be fully determined.

There is evidence which suggests that the structure in marine triclads which is termed a basement membrane (Hyman, 1951, Gamble, 1896, Grassé, 1961) should not be called by this name until its ultrastructure can be elaborated. Fawcett (1966, p. 353) states that the basement membrane as seen with the electron microscope (which he terms "basal lamina" to avoid the connotations of the word "membrane") is a moderately dense band 500-700A thick. It is separated from the basal cell membrane by a light zone about 400A wide. This relatively thin layer would not be visible page 6 with the light microscope (Bloom & Fawcett, 1966, p. 5). The thickness of the basal lamina varies with different types of epithelia. Fawcett suggests that the layer between an epithelium and the underlying connective tissue to which light microscopists originally assigned the term "basement membrane" included some of the underlying collagenous fibres of the connective tissue. In the present study this "basement membrane" layer is very thick and has the staining properties of collagen, that is to say, it stains red with Van Gieson's picrofuchsin stain, and is deeply coloured by the acid dyes aniline blue and fast green FCF which Lillie (1945) considers to be excellent collagen stains. Moreover, it is continuous with fine similarly staining fibres which course throughout the body parenchyma and between muscle fibres, and which enclose specialised organs such as the gut and the reproductive organs. This suggests that it has the same functions as the connective tissue of higher animals. However, with the techniques used in this study no cellular material was demonstrated in this layer, and because of the difficulty of defining cell boundaries within the parenchyma, the relationships between the cells of the parenchyma and (a) the fibres of the basement membrane and (b) the connective tissue fibres deep in the parenchyma, could not be evaluated. Therefore, the term "connective tissue", which implies cellular material as well as extracellular fibres and other substances, is best avoided for the present. The terms "connective layer" and "connective fibres" seem appropriate.

Fibrous proteins belonging to the collagen class have been identified from the skeletal tissues of most invertebrate groups of animals (Rudall, 1955; Gross, Sokal, and Rougvie, 1956). Ruddall shows that among the various groups there generally is an inverse relation between α chitin and collagen production. That is, where collagen is reduced to a minimum as for example in hydrozoan polyps and in insects, chitin usually forms a high proportion of the total bulk of the animal; and conversely, where collagen forms the bulk of the skeletal system α chitin is absent or present in only small amounts. There are no records of the presence of collagen in turbellarians. However, the work of Rudall and of Gross et al suggests that it probably would be present, because it occurs widely in other lower metazoan groups, and because chitinous material is generally lacking in turbellarians except in the reproductive organs of some species.

According to Hyman (1951), the basiphil gland cells in triclads are for the secretion of "slime" to aid in locomotion, and the eosinophil ones are for the secretion of adhesive substances. The position of these gland cells in P. stephensoni appears to confirm this conclusion. The former type are most abundant in the ventral parenchyma of the anterior end of the animal, while the latter occur as a zone extending around the lateral body margin. Hyman adds that both secretions probably also serve to protect the surface of the animal from harmful environmental substances and to entangle prey. The present study supports the views of Hyman (1951, p. 74) that the fine necks of the subepidermal eosinophil glands pierce the epidermal cells as they open to the surface (Pl. 3), and also that these fine gland cell necks terminate in "papillae" which extend beyond the epidermis (Pl. 3, Figs. 1, 2). Once again, it must be noted that the detailed structure of these cells could not be determined. It is clear, however, that for the gland cell necks to open through the epidermal cells, a system of internal membranes must be present in the epidermal cells to enclose the secretory substance of the gland cells and to "pipe" it to the surface. It was mentioned earlier that the epidermis over the general body surface possesses a system of internal "channels" (Pl. 1, Fig. 1), but it is not clear whether this existing system is utilised by the subepidermal glands or whether the fine ducts of the glands remain separate from these internal epidermal "channels". It is possible that the epidermal cells through which these glands open are specialised for this task, but in the present study no morphological differences were observed betwen these and other epidermal cells.