Several accounts of regeneration, of viscera in holothurians have shown that it occurs along the whole length of the free edge of the mesenteric remnants remaining after autoevisceration in Thyone (Scott, 1914; Kille, 1935) and in Stichopus (Bertolini, 1930; Dawbin, 1949). Only part of the total length of the mesentery is involved during regeneration in Holothuria tubulosa Bertolini (1932). The two modes of regeneration are quite distinct and apparently constant for each species. There is no evidence showing that different methods of regeneration occur in the same species.
Regeneration of viscera in Stichopus mollis normally occurs along the complete length of the mesenteric remnants (Dawbin, 1949), but three specimens showed regeneration on a different pattern. In these specimens, the alimentary canal passes from one portion of the mesentery across an adhesion to a more posterior portion, thus omitting a length of the mesentery.
The methods used for inducing autoevisceration and for the keeping of specimens during regeneration are those previously described. For the examination of stages reached in regeneration Bouin's fluid was injected to harden structures, and an incision was first made along the right dorsal interambulacrum. The state of the organs and the distribution of mesenteries was first determined by gently reflecting the flaps on either side, and removing portions of the body wall which were clearly not in contact with mesenteric attachments. In each of the three cases considered in this study, there
was an adhesion between the dorsal and ventral mesenteries (Fig. 2). This would have been torn if the flaps had been completely reflected in the usual manner. A new incision was therefore made along the left dorsal interambulacrum and the flaps were then reflected. This left the mesenteric edge in a looped condition (Fig. 2) instead of the S-shaped condition found after opening along the right dorsal interambulacrum. (Fig. 1.)
Fig. 1.—Dissection of an eviscerate specimen opened along the right dorsal interambulacrum, showing the S-shaped course of the mesentery.
Fig. 2.—Dissection opened along the left dorsal interambulacrum, showing the relationship of the regenerating alimentary canal to the mesenteries. A., anus; AL.C., regenerating alimentary canal; CL., cloaca; D.M., dorsal mesentery; L.D.I., left dorsal interambulacrum; L.M., lateral mesentery; L.MU., longitudinal muscle band; OES., oesophagus; P.V., polian vesicle; R.D.I., right dorsal interambulacrum; R.T., regenerating respiratory tree; V.M., ventral mesentery.
The organs remaining and the distribution of the mesenteric remnants after autoevisceration in Stichopus mollis have been described previously (Dawbin, 1949). Usually there is no cross connection between any portions of the mesentery from the oesophageal remnant to the cloaca, as the mesentery follows its S-shaped course along the dorsal, lateral and ventral loops (Fig. 1.).
Before reflexion of any portion of the body wall, the dorsal remnant of the mesentery hangs freely into the body cavity from the mid-dorsal interambulacrum, and its free edge can closely approach the free edge of the ventral mesentery. The length of the mesenteric remnants from the point of attachment at the body wall to its free edge was found to vary. In three cases out of 299 examined, it was found that the dorsal mesentery had contacted and fused with the ventral mesentery at a point almost level with the anterior free edge of the lateral mesentery (Fig. 2). This established a direct connection of the mesentery from the oesophageal remnant to the cloaca, without the necessity of following the whole length of the dorsal, lateral and ventral borders. The posterior portion of the dorsal mesentery and the whole of the lateral mesentery is thus eliminated from the direct course. Unfortunately the three specimens were all at a relatively similar stage of regeneration on examination, having undergone 44, 47 and 50 days' regeneration respectively. It was therefore not possible to determine whether the adhesion had taken place immediately after autoevisceration, or whether there had been a later increase in depth of the mesenteries permitting contact between the dorsal and ventral loops. From the state of the regenerating tissue along its edges (see below) it is apparent that contact was made at some early stage in regeneration.
Regeneration occurred along the dorsal mesentery between the oesophageal remnant and the point of contact with the ventral mesentery, and along the edge of the latter to the cloaca. Along this length the lining epithelium on either side of the mesentery had fused over the torn edge, and the mesenchyme cells between had increased to form a rod-like thickening. This is the primordium of the alimentary canal, and its structure showed that its formation followed the same course as that described previously for Stichopus mollis (Dawbin, 1949). The alimentary canal of the specimens examined after 47 and 50 days' regeneration tapered from a diameter 1·5 mm. at the oesophageal remnant to 0·8 mm. at the cloaca. Histological study showed that an irregular lumen was present between the mesenchyme cells along the whole length. A distinct inner epithelium had not been formed, but there was a concentration of nuclei around the lumen indicating its initial development. Longitudinal and circular muscle fibres had not been developed at this stage.
The specimen examined after 44 days' regeneration showed the greatest diameter of the alimentary canal anteriorly, being 0·5 mm. wide at the oesophageal remnant and 0·4 mm. wide at the point of contact with the ventral mesentery. From this level it tapered posteriorly towards the cloaca. Just anterior to the junction with the cloaca, it narrowed to a width of 0·1 mm. This specimen was smaller than the preceding, being 47 gms. as compared with 150 gms. and 170 gms., and the alimentary canal was correspondingly of smaller diameter. A lumen was present from the oesophageal remnant to the junction of the dorsal and the ventral mesenteries. There were irregular spaces among the mesenchyme cells along the anterior third of the ventral mesentery, followed by a solid cord of cells from this region to the cloaca.
In all three specimens, the mesenteric edge not included in the direct connection from the oesophageal remnant to the cloaca remained unthickened, and showed no evidence of cell proliferation at any point. In addition there was no forward growth of the mesentery to form the pocket-like extension normally present in the angle between the dorsal and lateral mesenteries (see Fig. 1) at this stage of regeneration.
The separation of the dorsal and ventral intestinal haemal vessels as extensions from the regenerating alimentary canal was similar to that described previously (Dawbin, 1949). The transverse connecting vessel had not been split off in any of the specimens.
In these three specimens the respiratory trees were 2·0 × 1·O mm., 3·0 × 1·0 mm., and 10·0 × 1·5 mm. in size respectively. Similar variability in dimensions during the early stages of regeneration has already been recorded (Dawbin, 1949) and there was no difference in the mode of origin and the stages of differentiation of the layers.