112 Trans. Acad. Sci. of St. Louis. 
has little significance. Indeed, it seems that a return to com- 
plete regularity would not seriously disturb the present insect 
relations of the flower. Insects can land indefinitely upon the 
head, approach the flowers with equal convenience from 
almost any side, and can receive the pollen on almost any part 
of their bodies. 
The tubes measure 18 or 19 mm., which‘indicates an adap- 
tation to long tongues. The form of the tube, the bilabiate 
corolla, and the positions of the stamens and style indicate 
that the flower is a modification of a nofotribe flower origi- 
nally adapted to bumble-bees. The level-topped heads, the 
erect corollas, the exposed organs and the rose color indicates 
a tendency to suit butterflies, which in fact are the principal 
guests. , 
I have found the flowers in bloom from July 9 to Sept. 19. 
On 19 days, July 16-Aug. 28, I observed the following 
visitors :— 
Lepidoptera — Rhopalocera: (1) Papilio philenor L., ab.; (2) P. asterias 
F. — (3) P. troilus L.; (4) P. re Cram.; (5) Colias philodice 
Goa -; (6) Danais archippus F.; (7) Argynnis cybele F.; (8) Limenitis 
ane Godt.; (9) Pamphila ashulon B.-L.; (10) P. peekttia Kby.; (11) 
P. otho Bok, v. egeremet Scud.; (12) Pholisora hsyhurstii Edw.; 
(18) Eudamus bathyllus §.-A.; (14) E. tityrus F.; Sphingide: (15) 
Hemaris thysbe F 
Birds— Trochilide: (16) Trochilus colubris L. 
Hymenoptera — Apidw: (17) Bombus vagans Sm. 8; (18) B. ameri- 
Canorum F. ¢' 2 8, ab.; (19) B. pennsylvanicus De G. 79 8,ab; (20) B- 
separatus Cr. <j’, ab.; (21) Melissodes comanche Cr. ¢'; (22) M. bimaculata 
p.?. 
Diptera — Bombylide: (23) Exoprosopa fasciata Mcq., freq. — allsucking 
In addition to the above, which visit the flowers only for 
nectar, which they obtain in a legitimate way, the flowers are 
sometimes sought by small bees in search of pollen, as in the 
ease of M. Bradburiana. In this case I have observed : (24) 
Ceratina dupla Say ¢; (25) Halictus confusus Sm. 9, ab. 
The corollas are even more abundantly perforated by the 
same wasp, (26) Odynerus foraminatus Sauss., and by (27) 
O. dorsalis F. To obtain the nectar these wasps always cut 
new holes at the base of the tube, apparently never using the 
old holes. After this comes a lot of insects which do not cut 
holes themselves, but take advantage of the holes cut by 
