232 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
In Figure 14 (Plate 3) several structures are shown in which the 
neurofibrillae could be distinguished. As a rule, these appear to be 
independent of one another, and to take more or less parallel courses. 
In some preparations, however, the neurofibrillae appear to form a 
terminal basket work. Examples of this are shown in Figures 21, 
22, 26 (Plate 4). Whether this structure is in reality a basket work, 
formed by the interlacing of neurofibrillae, or a structure formed by 
the anastomosis of neurofibrillae, is a question which I cannot now 
answer. I know of no way to settle such a question except by careful 
focusing, and in this case the diameter of the neurofibrillae is too small 
to permit of answering the question by that method. Such structures 
are rarely seen in my preparations, there being only two sections in 
which I have observed them. When I first noticed them I was in- 
clined to regard them as the equivalent of the structures described by 
Kolmer under the name of pericellular networks. By comparing the 
sections in which these structures occurred, however, with the adjacent 
sections in the series, I came to the conclusion that they were incom- 
plete parts of larger terminal structures, the main bodies of which lay 
in the next section, and that they had been obtained by cutting thin 
slices from the surfaces of these structures. 
‘lhe terminal expansions which have been designated as ‘ Kelch- 
bildungen” may be regarded as a type of free termination, if I am 
correct in my belief that they are related to the sense cells only by 
contact, and that no nervous structures occur within the sense cells. 
They are so characteristic in their form, however, that it seems wise 
to class them in a category by themselves. Their expansions are 
chiefly horizontal, as in Figures 12, 14 (Plate 3); but may tend toward 
the vertical, as in Figures 11, 16 (Plate 3). 
‘There is not a pronounced condition of secondary branching, such 
as exists in the terminal structures of the maculae. Such branches as 
are given off are much shorter than those of the maculae, and do not 
penetrate far between the sense cells. The sense cells, rather, rest 
upon the nervous material of the calyx, or are slightly imbedded in it. 
There is another type of free termination in the cristae, which differs 
widely from the form just described, and corresponds more closely to 
what is generally understood by free terminations, in that the terminal 
twigs penetrate farther into the stratum of the sensory epithelium. 
In some cases they may be traced to fine fibres in the eighth nerve, 
which, after penetrating the basement membrane, make their way to 
their terminations between the sense cells. Examples of such fine 
fibres, which end free without branching, are represented in Figures 
