179 
[Garman. 
than that of the copperhead. The cap and button are horizontal, 
figs. 13, 14, that is, they lie in the axis of the vertebral column. 
They resemble those of the common water snakes,. Nerodia , the 
habits of which are similar. A pair of scales usually overlie the cap 
and are fused with it ; another pair is often present on the lower 
side, but at times the traces are almost obliterated. Occasionally,, 
on the cap just below the edges of the upper pair of scales and par- 
allel with them, longitudinal on the button, there is a shallow fold,, 
fig. 14, in character like the transverse folds on the cap of the 
following species. It is evident, from shape folds and axis, that 
when the moccasin is excited the button strikes on the lower side 
and not on the apex. 
The copperhead, Ancistrodon contortrix , is closely allied to the 
preceding. Its habits, however, are very different, and their effects- 
have impressed themselves on the structure. This snake inhabits 
rocky or dry places, more or less free from vegetation. The tail is 
rather more round, slender and flexible than that of the moccasin, 
and the button, instead of being directed backward in the axis of 
the column, is inclined obliquely back and downward ; so that in 
striking to produce a warning rattle the extreme end is struck upon 
the ground. The stroke itself is peculiar, in that the terminal inch 
or more of the tail is swung over in the arc of a circle and brought 
down from side to side. Such a method is suited to localities des- 
titute of leaves or grasses. The cap on the young, fig. 15, is 
bluntly rounded, somewhat compressed, and with or without a shal- 
low fold near the edge of the scales. Afterward the button becomes 
longer, more pointed, and the cap acquires one or more constric- 
tions or shoulders similar to but shallower than those of the rattle- 
snake, figs. 16-18. This modification with age in the caps of these 
snakes is in direct comparison. On the other hand to compare 
with the moccasin is sufficient to establish the fact that the differ- 
ences in the cap are due to the different methods of striking with the 
tail ; in the one, the concussion affects the lower side of the button, 
in the other the apex. Each manner is best suited to its own local- 
ity. Fig. 15 is taken from a small one, fig. 16 from a larger one, fig. 
17 from an average specimen and fig. 18 from the largest of the cop- 
perheads at hand. On very young individuals the declination of the 
axis of the button is comparatively little, not more than may have been 
handed down from progenitors that were expert climbers and used 
the tail for grasping the branches. If the young snake were born 
