MULLENIX: EIGHTH CRANIAL NERVE. 235 
stead of expanding calyx-like, however, they often divide, after enter- 
ing the subepithelial stratum, giving fibres which pass horizontally, 
and finally terminate in close relation to the sense cells. I have seen 
no cases in which the bases of the sense cells present the appearance 
of resting in concavities of these terminal structures, “as an egg rests 
in an egg cup.” ‘The sense cells appear rather to rest upon the sur- 
faces of the fibres. ‘This class of terminal structure is more suggestive 
of the branching of an extremely low tree than of the terminal expan- 
sion seen in the flower calyx, and may be regarded as a representative 
dendritic structure, or ‘‘end brush.” Figures 29, 30, 32, 34 (Plate 5) 
show such structures from the sacculus, and Figures 35, 36, 38 (Plate 5) 
from the lagena. In some cases the fibres turn toward the sensory 
epithelium soon after leaving the parent fibre (Figs. 29, 30, Plate 5). 
In other instances they are greatly extended in the plane of the epithelial 
layer (Figs. 32, 34, 35, 36, Plate 5). As a rule the tips of these 
branches are turned somewhat toward the periphery, though there are 
cases in which this is not apparent. Often there are severed parts of 
such dendritic structures in close proximity to more complete ones, 
as in Figures 29, 30, 35 (Plate 5). 
Again, it is not uncommon to find examples of these structures in 
which one or more branches have been severed, leaving blunt cut ends, 
as in Figure 34 (Plate 5). While cases exist where there is no evidence 
that the tips of the fibres penetrate between the cells, there are also 
many instances in which they extend for a greater or less distance into 
the epithelium, between the cells. There are apparent cases of the 
secondary branching of the fibres (Figs. 34, 36, Plate 5). 
I have found a very few instances in which there is at the termina- 
tion of fibres a knob-like structure, in close proximity to the nucleus of 
a sense-cell. In some instances this presents the appearance of a 
solid mass of black, while in other cases it is possible to make out what 
seems to be a network of neurofibrillae. Both of these conditions are 
shown in Figure 36 (Plate 5). When such a network has been ob- 
served its compactness has been such as to render its demonstration 
difficult. These structures have so rarely appeared in my prepara- 
tions that I have not undertaken to form any opinion as to their sig- 
nificance. 
There are other fibres of fine calibre, which penetrate the basement 
membrane and make their way between the supporting cells and 
through the subepithelial plexus to the sensory epithelium, where 
they end free between the sense-cells. ‘They appear to terminate at 
various levels in the sensory epithelium. I have found some which 
