new Species of Earthworms. 61 



oval one. It is while fixing the animals tLat the body 

 becomes four-cornered, and in the struggles snaps occur at 

 dillerent parts, leading to eonn)lcte separation. The quadri- 

 lateral nature of the body is produced by the sudden con- 

 tractions of the transverse bands of coelomic muscles, which 

 extend from seta-space be to below the intestine. These 

 muscles (fig. 3, A & B), which start from the body-svall in 

 the median line, spread outwards in the form of a cone on 

 either side, and a pull on the body-wall on the sides accounts 

 for the lateral canals and dorsal and ventral corners. The 

 contraction of the vertical muscles of the septa, which 

 extend beyond the grooves on either side of the body-wall, 

 would produce the dorsal and ventral canals. The secondary 

 sets of coelomic muscle-bundles do not occur in front of the 

 segment 1-1, and hence this region remains round. 



But the most interesting fact connected with these 

 muscles is that they almost form, being connected here and 

 there in their course by patches of peritoneal membrane, 

 secondary ventral chambers, in which the nephridia are 

 lodged. Tliey have to be disengaged from these chambers 

 for a more detailed examination. Numerous bubbles of air 

 escape from these secondary chambers, as in other freshwater 

 species, while the worms are opened and spread out, and, 

 where the pressure has not ruptured the investing membrane 

 of the ventral coelomic chamber, very large air-bubbles are 

 noticed, being entangled w'ithin the muscular mesh. Such 

 air-bubbles are found throughout the worm. The chamber 

 on one side communicates with its fellow on the opjjosite 

 side below the intestine and the nerve-cord, the associated 

 vessels and the nephridia lie within the secondary cavity. 

 In cross-section the chambers lock like two cones, their 

 apices meeting in the middle. The upper wall is composed 

 of several bands of muscles, held together imperfectly in 

 most j)laces by the reflected portion of the peritoneum from 

 the intestine. On the body-wall {he) the bundles of muscles 

 at their point of insertion spread out in the form of a 

 fan. Occurring in the narrower middle portion of the 

 chamber, and also in the outward part, are clumps of large 

 polygonal cells with one or more deeply staining nuclei 

 associated with the muscle-bundles. ]\Iixed up with these 

 clumps are ccelomocytes and a fine plexus of blood-capil- 

 laries. In any teased and stained preparation of the muscle- 

 fibres, the cell-clumps are a striking feature, and occur 

 uniformly. I am unable to trace the source or determine at 

 present the nature of the origin of these cells, which may be 

 due perhaps to the proliferation of peritoneal cells, which 



