COLLARETTE FLOWERS 
T. D. A. CocKERELL, Boulder, Colo. 
HE first collarette dahlia was ex- 
hibited in 1900, and placed on the 
market in 1901. This apparently 
unique variety, called President 
Viger, possessed the peculiarity of having 
supplementary lobes, of variable shape, 
at the base of the ray-corollas. These 
lobes or processes being white or nearly 
so, against the dark crimson back- 
ground of the rays, were very effective, 
and the variety was at once recognized 
as an important addition to horticul- 
ture. During the last fifteen years the 
original collarette dahlia has been 
crossed with numerous other varieties, 
giving a long list of collarettes, of 
various colors. 
It might readily be supposed that the 
collarette form was something abso- 
lutely new when it appeared in the 
dahlia less than twenty years ago. 
Experience with various compositae 
throws doubt on such an opinion, be- 
cause we find that certain variations 
crop out here and there in diverse but 
more or less related genera, at different 
times and places. The variations of 
Dahlia, Helianthus, Ratibida, etc., run so 
closely parallel that whatever occurs in 
one we begin to expect in the others, 
while it appears increasingly probable 
that the whole group has been producing 
the known series of variations at 
intervals during past ages. The “‘origin”’ 
of a variety, as recorded, thus has to do 
only with the first occasion when it 
chanced to be detected on coming to 
the surface, as it were, of the stream of 
heredity. 
The first collarette sunflower, of the 
chestnut and vinous forms of Helianthus 
annuus, was obtained in our cultures at 
Boulder, Colo., in 1915. Quite un- 
expectedly, a considerable number of 
plants showed this character, but the 
supplementary lobes were narrow and 
1Arnica pedunculata is common at Boulder. 
very variable, not nearly as showy as in 
the dahlia. The essential structure was, 
however, quite the same. This year we 
have a piece of ground devoted to 
collarettes, and by continued breeding 
and selection it is probable that the 
type will eventually be much improved. 
The evidence seemed to indicate that 
the collarette sunflower had arisen for 
the first time in cultivation; but on 
September 24, 1915, I found a wild plant 
(H. annuus lenticularis) in Boulder, 
having genuine collarette characters. 
There was only one head on the plant, 
the last of the season, and in this two 
of the rays were tubular. There is 
apparently a connection between the 
variation with tubular rays and the 
collarette, as is especially indicated by 
Arnica pedunculata Rydberg! var. n. 
tubularis, found by my wife at Boulder, 
June, 1915. In this variety the ligules 
are tridendate: at the end, as in the 
typical form, but are variably tubular 
below, with a liguliform process on each 
side of the mouth of the tube. From 
this condition to a collarette Arnica 
would be no great step. 
The latest and in some ways most 
surprising collarette to be discovered 
is in the long-headed cone-flower, Rati- 
bida columnifera.2 This form (var. nov. 
appendiculata) was found by the writer 
at the base of Flagstaff Hill, Boulder, 
July 8, 1916. The rays possess long 
appendages, usually a pair, arising from 
the throat. 
Although Ratibida is so different in 
many respects from Helianthus, having 
a long conical receptacle, the pappus a 
rudimentary crown, and the very pecu- 
liar dise-bracts with a long dark crim- 
son spot on each side, the variations 
run parallel in the most astonishing way. 
The form with wholly or partly chest- 
nut red rays, parallel with the red sun- 
Apparently Nelson is correct in considering 
A, monocephala Rydb. to be a form of pedunculata, but wrong in referring them to A. fulgens. 
“Wooton and Standley write columnifera in place of columnaris, the former name having 
a year’s priority. 
428 
