1863.] 
233 
and consequently those eggs are not improbably carried down by the 
current from the upper waters of the river where they occur. Comes 
very near indeed to tricolor Burm., but differs in the % antennae 
being particolored not uniformly brown ; 27id^ in having a distinct late¬ 
ral green stripe on the dorsum of 9 thorax, whereas 9 tricolor is de¬ 
scribed as having none; 3?’c/, in the stripes of 9 pleura being bright- 
green, not “ brown4^A, in the abdomen % being much slenderer 
than in america^ia^ whereas in tricolor it is said to be “ a little more 
robust than in americanad' • 5th, in 9 abdominal joints 1—2 being 
bright-green above, not “ black 6th, in the inferior lamina of the % 
abdominal appendages terminating in a square truncation the apex of 
which is scarcely rounded off, whereas in tricolor it is described as ter¬ 
minating in “ an obtuse tooth.’’ 1th, in there being no tubercle at the 
superior base of the S appendage, as figured and described in tricolor ; 
Sth, in the 9 wings being within a minute fraction as broad in propor¬ 
tion as S wings; whereas in tricolor the 9 wings are described as ^ 
narrower; and in some other minor points. Limbata, which is given 
provisionally as a variety of tricolor, is said to be much slenderer than 
tricolor, but it has the cross-veins of the inferior surface of the basal 
spots of % wings “ not in the least whitish,” whereas they are conspicu¬ 
ously so on the hind wings of every one of my 36 % rupamnends. From 
morihunda Hag. (S. A.) it is separated at once by that species having- 
no distinct postoccipital tubercles, and from titia Drury (Mexico), by 
the wings of % titia being blackish-brown ” outside the basal carmine- 
red spots. 
The habits of the h h of this species are to traverse the water 
actively to and fro, and return from time to time to a favorite log or stick 
of wood lying in or overhanging the river. They are rather shy, but will 
come back repeatedly to their favorite station after being stricken at, 
unlike most of the larger Odonata. Their flight is strong and jerky, 
and resembles a good deal that of the lepidopterous genus Vanessa. 
The 9 9 are comparatively sluggish, sometimes hiding in the weeds, 
and the % % must certainly be at least four times as numerous as 9 9 . 
Americana,^ pseudamericana and rupamnensis all occurred within two 
hundred yards of each other. 
For americana read scelerata n. sp. See Appendix. 
