JOIJSSEAUMIA. 249 



safranin or hsematoxylin, and the byssus itself is similarly unaffected. The interest 

 of this observation consists in the demonstration that the byssus gland-cells, like 

 those of sebaceous follicles, are broken up to form the secretion, and that the secretum 

 travels in and among the epithelial cells for relatively long distances until it reaches 

 the lumen into which it is finally discharged. Thus the existence of a great mass of 

 gland cells, forming a central core to the foot, and apparently distant from the byssus 

 cavity and groove, is satisfactorily explained. The secretion is not confined to the 

 byssus cavity, but throughout my sections I find the same indications of granules 

 streaming between the ciliated cells of the cresceutic demi-canal, but not between the 

 non-ciliated epithelial cells of the furrow. In spite of some differences in detail, 

 which can be accounted for by the widely different genera examined by us, my 

 observations agree in all fundamental particulars with those of Horst (8). The 

 byssus, as he maintained, is undoubtedly a secretion product and not a cuticular 

 structure. Comparing Horst's figures of Dreissensia polymorpha (loc. cit., plate xi., 

 figs. 2, 3, and 4) with mine, it will be seen that in the latter species the byssogenous 

 glands are concentrated in the region of the demi-canal, and that there are numerous 

 mucus glands, of which I could find no trace in Jousseaumia. And whereas he shows 

 numerous branching and anastomosing canals passing from the cells of the byssus 

 gland between the epithelial cells of the demi-canal, these canals becoming narrower 

 in diameter as they approach the lumen of the byssus cavity, and gives no indication 

 of the breaking up of the cells themselves to form the secretum, I find that the cells 

 are broken up and the secretion travels (probably) between the epithelial cells in the 

 form of streams or strings whose ends nearest the lumen are swollen. It appears 

 that in both cases the secretum follows intercellular paths, and that in both cases 

 it has the form of granules which are converted in the lumen of the byssus cavity, 

 probably by the action of a ferment, into the material of the byssus. 



The Alimentary Canal. The labial palps are relatively large, and the upper 

 and lower palps pass respectively into the upper and lower lips. Posteriorly the 

 labial palps are continuous with the anterior ends of the gill plates. The palps are 

 richly ciliated, the cilia being borne by large cubical epithelial cells with a very 

 distinct limiting membrane, but the surfaces of the palps appear to be smooth, and 

 not thrown into ridges as is usually the case in Lamellibranchia. The mouth leads 

 into a buccal cavity lined by somewhat elongated columnar epithelial cells con- 

 tinuous with those covering the labial palps, and provided, like the latter, with a 

 very distinct limiting membrane or cuticle, through which the cilia project. The 

 pharyngeal cavity is wide and strongly compressed dorso-ventrally. As it passes 

 back into the oesophagus, the shape of the lumen, as seen in transverse section, alters. 

 There is a diamond-shaped central lumen (fig. 8), the lateral angles of which are 

 produced into lateral diverticula, suggestive of a comparison with the oesophageal 

 pouches of Gastropoda. Such pouches are only known in the Protobranchia among 

 the Lamellibranchia, and in them, e.g., in Leda pella, as figured by Pelseneer (11), 



2 K 



