No. 3.] VISUAL. CELLS IN VERTEBRATES. 610 
takes a large part in the nutrition of these joined cells. This 
is suggested by the relations of its nucleus. The process 
which the latter sends distally down over the side of the 
paraboloid must be of significance. » The surface of the nuc- 
leus is increased thereby and noticeably that part of its sur- 
face which is in contact with the paraboloid and, through 
the paraboloid sheath, with the more distal cell organs. 
The nuclear sheath in this region seems to be very thin, or 
entirely absent, especially at the distal edge of the process. 
Here it is difficult to make out any distinction between nucleo- 
plasm and cytoplasm, as there is no sharp differentiation in 
the staining of the two. The dark staining nucleoplasm, 
however, shades off gradually, as if there were some limit, 
though an indefinite one The constant difference in stain- 
ing reactions between the two nuclei would seem a further 
indications of different functions. And again the loss of 
symmetry of the near-cone as to its nucleus and paraboloid 
must be associated with a loss of a light receiving office, as 
rays would not have the regular disposition that must result 
from the lens-shaped nuclei and paraboloid of the ordinary 
visual cells. 
If we seek between double and single cones a distinct 
difference that might be of physiological importance in relation 
to visual function, it would seem to be the greater distance 
of the outer segments of double cones from the source of 
light. There is evidently a rather small range of contraction 
in the double cones, so that their outer segments are more 
constantly remote from their nuclei. If the increased dis- 
tance were an advantage for a visual function, the distance 
from the nucleus might demand some new adaptation for 
more direct nuclear relations. 
Levi’s (:00) opinion that the double cones come from a 
single embryonic cell seems quite plausible. If that is the 
case, probably the nuclear division is complete while the 
cytoplasmic is less so. I have distinguished two nuclei 
in a great many cases and therefore do not agree with 
Schultze (’67) in the opinion that only one nucleus is present. 
