454 A. B. MAOALLUM. 



interlobular canals the connective tissue also is not demonstrated 

 by the gold, so that one finds no difficulty in tracing a nerve- 

 fibre for a long distance, providing it lies in the plane of the 

 section. 



Nuclei were rarely observed on the largest fibres, and whether 

 these belonged to the sheath of the fibre or to nerve-corpuscles 

 it was impossible to determine. The division of the fibres in 

 the interlobular canals is not common, but branching occurs 

 more frequently in the capillaries of the lobules. Here 

 they give rise to fibrils of two sorts ; those which form the 

 perivascular plexus surrounding the capillary walls, and those 

 which, few in number, apparently course between the hepatic 

 cells. I am unable to say whether there is any morphological 

 or physiological diS'erence between these kinds of fibres. I am 

 inclined, however, to think they are one and the same, for they 

 terminate in the majority of observed cases in a like way. The 

 intercellular fibrils arise from fibres which serve as origins to 

 the perivascular plexus; this also supports the conclusion 

 already stated as to their physiological value. In a thin section 

 the intercellular fibrils are the most commonly visible. I have 

 applied the term intercellular to them because, although they 

 do not always run between the cells, they unquestionably lie 

 outside the capillary channels for the greater part of their 

 course. I have followed them in some cases for a distance equal 

 to the combined diameters of over twenty of the cells and have 

 found them to accommodate themselves but very little to the 

 windings and tortuous course of the capillaries. In thin sec- 

 tions, where the fragment of one cell is seen to lie over that of 

 another, one of the fibrils in question has been again and again 

 observed to pass between their contiguous walls. Part of their 

 course is in the capillary channel, i.e. between the wall of the 

 blood-capillary and the adjacent liver-cells. 



The meshes of the perivascular plexus vary much in size and 

 form, being usually less than the area covered by a hepatic 

 cell and of an irregular quadrangular or triangular shape. The 

 fibrils which form them are wavy in their course and apparently 

 anastomose completely with one another. Now and then only 



