646 The American Naturalist. [August, 
plest patterns, the corresponding bars on the two sides are 
exactly alike, and precisely meet each other in the mid-dorsal 
tine. This almost perfect symmetry is not so prevalent in the 
more complex patterns. The simplest cases of asymmetry 
are found in the simplest patterns when some of the bars do 
not exactly meet their fellows on the back. Fig. 8 shows an 
instance of this kind. Both the asymmetrical and the sym- 
metrical forms occur in the same locality, and the former 
seems purely accidental, but in all cases observed, it makes its 
first appearance in the bars along the spinous dorsal. From 
this point it spreads backward along the soft dorsal until we 
reach an extreme form of asymmetry, as represented in fig. 9. 
Here the first three and the last four bars, together with the 
bar between the dorsals, still preserve their symmetry, while 
those along the entire length of both dorsals are quite asym- 
metrica 
_ In regard to variations in parts other than in the color pat- 
tern, only those points of structure were examined that could 
be most accurately made out on alcoholic specimens. One 
very marked departure from the regular form exists in the 
specimens from San Marcos Spr., Texas. This departure con- 
sists, as shown in fig. 5, of an increase in the depth of the 
body in the region of the spinous dorsal, as a result of the 
unusual elevation of the back in this region. These belong 
to the variety carbonaria, described from Texa, and are more 
distinct in points of form than the varieties I examined from 
any other locality. 
No. 8 in Table I, taken by Dr. Meek at Cedar Rapids, Iowa, 
differs materially from any of the specimens from other local- 
ities. It approaches nearest the variety zebra in the color pat- 
tern, and in having no scales before the spinous dorsal. The 
scales, however, are larger, there being but 76 in the lateral 
line. The head measures 33 in body and the number of rays 
in anal is 12. 
The following table will give the number of specimens, their 
locality and the points of structure which have been examined. 
The spines in the dorsal and anal fins are indicated by Roman 
numbers and the rays by Arabic numbers. The length of the 
Te ea! ada Sire es ae irre 
