744 BILE-CAPILLARIES. [BOOK n. 



blue line between the hepatic cells will be seen to be continuous 

 with a larger quantity of the same blue material occupying the 

 lumen of one of the minute bile-ducts as it abuts on the margin 

 of the lobule. These minute canals are therefore continuous with 

 the bile-ducts ; they are the terminations of the bile-ducts within 

 the lobules, and indeed not only may they be injected during life 

 Avith sodium sulphindigofcate, but injection material may, though 

 with difficulty, be driven into them backwards along the bile- 

 ducts. They are spoken of as bile-capillaries, or more fitly 

 bile-canaliculi. 



We said just now that each hepatic cell touched a blood vessel 

 by at least one of its surfaces, we may now add that each hepatic 

 cell has at least one side, and generally more than one side, 

 grooved to form a bile-canaliculus. Since each side thus grooved 

 is in contact with the corresponding side of a neighbouring cell, 

 it cannot run alongside a blood vessel. Hence between a bile- 

 canaliculus and a blood vessel some amount of cell-substance is 

 always interposed. The relative position of the bile-canaliculi and 

 blood vessels may be illustrated by taking a cube and converting 

 it into a polygon by bevelling down the angles of the sides, leaving 

 in the first instance those of the upper and lower faces untouched. 

 The blood vessels may then be considered as running down the 

 bevelled edges, while bile-canaliculi run along the middle lines of 

 the sides. Two such cubes placed end to end might represent a 

 thin small islet of cells in one of the smaller shorter radial meshes 

 of the vascular network; and then the angles of the upper and 

 lower face of the conjoined cubes would have also to be bevelled 

 for the cross bars of the network. Frequently, as we have said, 

 the cells lie two abreast in a mesh of the vascular network ; then 

 of course in the model the angles of the surfaces in contact would 

 not have to be bevelled since no blood vessels run between them. 

 If several such bevelled cubes were built up into a model, it would 

 be seen that the network of bile-capillaries runs along the middle 

 of the surfaces between the blood vessels, forming nodal points 

 where cells are in contact with each other by their surfaces, and 

 leaving some amount of cell-substance between the bile-canaliculus 

 and the blood vessels. This at least may be taken as the typical 

 arrangement, when the network of bile-canaliculi is most complex. 

 But many cells have the lumen of a bile-canaliculus on one side 

 only; and occasionally a bile-canaliculus is seen in section at the 

 point of convergence of three cells after the fashion of an ordinary 

 alveolus. 



When a bile-duct abuts on the margin of a lobule the lumen, 

 as we have previously said, seems suddenly to come to an end. 

 The flattened cells lining the ductule or terminal portion of the 

 duct suddenly change into large hepatic cells, marginal cells of 

 the lobule, which appear to be completely in contact with each 

 other and to block up the ductule. But along the sides of these 



