18 
G. H. Parker 
through cell 2 and reappears in most of the qnarter-plateS belonging 
to this cell, the other elements of the retinula remaining almost en- 
tirely unaffected; in other words, the silver stain here, as in other 
cases, effects a given cell and the structnre produced hy it, hut does 
not necessarily color neighhoring cells. It was from seeing pre- 
parations such as this (Fig. 24) that I was led to regard the dorsal, 
ventral, and anterior half-plates as composed each of two quarter- 
plates, though I could never find a plane of Separation hetween them 
as hetween the pairs of half-plates. 
The fihres helonging to the other cells, 1, 5, and 6, in the 
section just mentioned are parallel to those of cell 2, as may he 
demonstrated in corresponding sections in which the fihres of all 
four cells are colored, and, as all dorsoventral plates agree in this 
respect, the general Statement may he made that all the fihres in 
dorsoventral plates agree in having a common dorsoventral direction. 
In the anteroposterior plates the relation of the fihres to the 
cells and division-planes corresponds to that in the dorsoventral 
ones; i. e., the direction of the fihres in the former is anteroposte- 
rior, or at right angles to that in the latter. When a rhahdome in 
which the fihres are colored is cut so that the section includes 
parts of adjoining plates, the fihres appear to cross one another 
(Fig. 25) ; of course, in such cases the two sets lie at different levels. 
In longitudinal sections made parallel to one face of the rhah- 
dome, the fihres in one set of plates would he cut longitudinally, in 
the other transversely. Figure 23 represents such a section, in 
which, however, the fihres from only two' cells are colored; those 
helonging to the cell on the left of the rhahdome are seen in longi- 
tudinal section and appear as thick lines; those in the alternating 
plates of the right half of the rhahdome are cut transversely and 
appear as dots. 
The fihres, as may he seen in the figures given, are always 
unhranched; of their two ends, one is huried in the retinular cell 
and the other usually reaches the division-plane hetween the two 
half-plates. Not infrequently fihres are seen that are not so long 
as this; I am uncertain whether these are really short ones, or of 
normal length hut only partially stained (cf. Fig. 24). Occasionally 
some seem to pass through the division-plane and extend as rather 
delicate processes into the adjoining half-plate (Fig. 24); this I am 
inclined to regard as the result of the spreading of the silver from 
the deeply colored fihres of one half-plate to a few in the adjoining 
