158 



Mr. J. G. Wilson. 



[Jan. 21, 



for investigation. All three varieties of cells are found (see Plate 5) unipolar 

 (fig. 3), bipolar (fig. 4), and multipolar (fig. 5). Ganglionic groups are 

 scattered in the connective tissue not only around the muscle band but also 

 in the interstices between the muscle fibres (fig. 4). They are not especially 

 located in the subendocardial area ; when the bundle reaches the surface of 

 the heart, they are seen not only between its tissue and the endocardium, but 

 are also equally conspicuous within the bundle and in the fibrous tissue on 

 the side remote from the endocardium. 



They can be seen all the way from the coronary sinus to the distribution in 

 the right and left walls of the ventricular septum. In the particular part 

 examined they appear to be most abundant near the point of bifurcation and 

 in the course of the right and left ventricular divisions. Their abundance 

 may be roughly estimated by saying that in a series of sections of one of 

 these arms, especially the left, cut 50 microns thick, it is no unusual thing 

 to find in a section three or four ganglionic masses containing from five to 

 nine nerve cells. Were one roughly to indicate any one area more than 

 another where they are conspicuous, one might select the upper border of the 

 bundle just before its division into right and left limbs. Here there is a 

 large group, easily seen in the calf and sheep, lying in the fibrous tissue out- 

 side the bundle whose processes pass into the muscular pathway. 



The processes of these nerve cells can often be traced for a long distance 

 gradually dividing in their course (fig. 3). Many of them ultimately become 

 varicose fibres, and some at least go into the nerve plexus around the muscle 

 cells. But the mode of termination of the majority, from the distance they 

 traverse and from the failure to stain referred to above, I have so far been 

 unable to determine. Fig. 6 shows a large nerve cell in some respects akin 

 to Dogiel's type I. It has one long process and several smaller ones, which 

 latter project only a short distance from the cell-like sharp prickles. The 

 long process can be traced for a very considerable distance ; it frequently 

 divides and ultimately ends as very fine varicose fibrils which enter into the 

 plexus around the muscle bundle. Fig. 5 shows the interlacing of the 

 processes of several nerve cells in one pericellular plexus. Here a multipolar 

 cell, G.i, gives off a branch, one of whose rami enters into the pericellular 

 nest of cell G.2, into which also enter twigs from two other more distant 

 nerve cells by fibres C and D. 



II. In the atrio-ventricular bundle the nerves present themselves 



(a) as strands of fibres ; 



(b) as plexuses of fibrils. 



(a) The nerve strands have a general course along the length of the bundle. 

 In preparations of the entire bundle they are seen to break into the fibrous 



