ON THE GENERAL MORPHOLOGY OF THE MYXINOID FISHES. 331 



the bile duct — without any connection of the two cavities — whilst the duct portion of 

 the same gland was at the opposite end and turned away from the bile duct. In fact, I 

 find it to be the rule rather than the exception for the duct portions of the lobules to 

 be directed away from the cavity into which they are supposed to open. Indeed, in 

 practically all my preparations the pancreas is a ductless gland, although in saying this 

 I do not wish to claim that such a condition is any other than a secondary and 

 acquired one. The lobules with the largest cavities are often those which are absolutely 

 and indubitably shut otf from any connection with the bile duct. Owing to the strong 

 selective staining of the connective-tissue framework, there can be no question here of 

 any error of observation. Further, I have never found that injections thrown into the 

 lumen of the bile duct pass into the cavities of the pancreatic follicles. If the latter 

 were in connection with the lumen of the bile duct, such connections could easily be 

 observed where the connective-tissue sheath is penetrated. In the very few cases where 

 the connection was found, it was easily traced even with the low power of the microscope. 



Several pancreatic lobules may fuse and have a perfectly defined epithelial duct 

 with a diameter of 144 m, and formed by a single layer of epithelium 36 m high, enclos- 

 ing a cavity 72 /x across. Such a duct may lose and acquire its lumen in a very curious 

 manner, and finally terminate in a cul-de-sac quite apart from the lumen of the 

 bile duct. 



Usually the lobules are not isolated, but dichotomise so as to form connected groups 

 after the manner of an acinous gland. Some, however, may be quite independent. 

 There is, in fact, an absence of structural stability one does not usually find in an organ 

 of functional importance. The pancreas occurs only on one side of the bile duct, until 

 shortly before the latter opens into the gut, when it completely surrounds the bile duct 

 like a ring. The alveoli contain sometimes a colourless liquid, and sometimes a granular 

 substance including variously-sized particles — some of them quite large, and others 

 having a concentric appearance — staining an intense black with iron hsematoxylin. 

 Here and there one finds a single disintegrating cell in the alveolus. 



The glandular as distinct from the epithelial portion of the lobule has a somewhat 

 complex structure. The nuclei are oval or round and have precisely the same appear- 

 ance as those of the duct portion of the lobule and of the bile duct itself, except that 

 the latter are larger and the chromatic network is somewhat looser and therefore stains 

 more lightly. In some cases a well-defined nucleolus can be seen lying in a clear area. 

 Very few mitoses were, however, observed. The cell outlines are sometimes difficult, or 

 even impossible, to define. In other cases they appear quite clearly, and the gland cells 

 are then seen to be irregular in shape and to vary in size. The cytoplasm is a loose, 

 faintly staining reticulum, and not granular. It may contain large non-staining spaces. 

 Inter- or intra-cellular cavities as described by Maas I have not observed ; but these, 

 like bile capillaries, may depend on the state of activity of the gland. A tangential 

 section of a lobule shows what might be interpreted as inter-cellular cavities, but these 

 on examination turn out to be the inter- lobular blood capillaries. I am speaking now of 



