112 THE AMERICAN NATURALIST. [Vor. XXXIV. 
head of the species he examined. We have found that the 
arrangement of the twin and single cones is remarkably con- 
stant for any given species,! 
and that the arrangement dif- 
fers considerably in different 
groups of fishes, but that the 
patterns in all of the species 
examined can be derived seri- 
atim from one or the other of 
the patterns. 
Pattern A, the pattern from 
which all others can most 
readily be derived, has not 
been found in any species as 
yet, but slight modifications of it have. In this simple pattern 
there are twin cones only, and the axes of alternate rows of 
twins are parallel, while those 
of neighboring rows are at 
right angles to each other. A 
series of squares are formed 
in this way whose sides are 
the extended axes of four twin B 
cones. 
Pattern B is produced by a 
slight twisting of the axes of 
the twins so that if extended 
they would form a rhomb. 
This pattern was found in 
Scorpena porcus. The angles of the rhomb measure 53° and 
es e CX -) 127?. Theactual conditions 
| found in the eye are shown 
in Fig. i. 
Pattern C is produced 
from B by a still further 
ae twisting of the axes of the 
twins till they are parallel 
1 The variation of the pattern in different parts of the same eye will be dealt 
with in a later paper. 
