BULLETIN OF THE ALLYN MUSEUM
THE ALLYN MUSEUM OF ENTOMOLOGY
Number 37 28 May 1976
gen. n., sp. nov.
F. Martin Brown
Research Associate, Allyn Museum of Entomology, and
6715 South Marksheffel Rd., Colorado Springs, CO 80911
Several years ago a very fine fossil butterfly was recovered from the trench on
the Singer Ranch at Florissant, Teller County, Colorado. This raises the total number
of such specimens to about twenty and of named fossils of Lepidoptera from the
trenches to about a dozen. When the National Park Service established the Floris-
sant Fossil Beds National Monument, they acquired the Singer property and with
it a few fossils. These included the butterfly in question. Recently I have become
a volunteer consultant to the Monument staff. Upon examining the fossil butterfly
I became convinced that sufficient evidence was present to warrant naming the
At first glance the fossil looks as though it may be a second specimen of tBarbaro-
thea florissanti Scudder. As can be seen from Figure 1. it lies in just about the same
position as Scudder's insect but as a mirror image. The length of the costa of the
forewing is 2.0 millimeters less than that of tBarbarothea. The palpus, although
distorted, is proportionally smaller and not at all like the palpus of a libytheid.
The venation of the apical area of the forewing is markedly different from tBar-
barothea as can be seen in Figure 2. It is this character alone that made me
reject the idea that I had a second specimen of tBarbarothea. Unless Scudder's
interpretation of the venation of the forewing is quite erroneous, which I doubt,
the two butterflies belong to different families.
The shape of the wing outline suggested the Indo-Australian genus Dercas
Boisduval. Comparison of the branching and position of the radius on the forewing
immediately ruled that out. I next turned to Leodonta Butler. The crenulation of the
wing outline for this genus is more extreme than in the fossil but the venation
of it is at odds since the fossil shows clearly four branches of the radius. I can
match the venation of the fossil with no modern pierid genus.
Several things militate against a clear-cut positive decision about relationship
to modern genera. First, much of the cell area of both wings is undecipherable. This
because the wings are folded and in the cell area there are four thickness and four
sets of veins. Second, the genitalia cannot be studied. Third the specimens represents
a situation that is 38 million years old. It is my belief that the portion of Pieridae
in which lies Leodonta is undergoing dynamic development at present. When R2
was lost was long enough ago so the loss is shared with several other genera, in-
cluding Delias Huebner of the Old World. Figure 1 shows the fossil with the best
illumination to reveal the venation. Along the margin of the hind wingin the cubital
area the preservation is so good that the rows of individual scales can be seen.
The fringe is long and suggests a fresh specimen.
tOligodonta, new genus
The squashed condition of the head, thorax and abdomen makes study somewhat
difficult. Approximately 9.5 mm. of the basal portion of the shaft of one antenna
is visible. It is a typical butterfly antennal shaft. What appears to be the club and
a short piece of shaft of the second antenna lies a little before and above the palpus.
Careful study proved this to be the wrong interpretation. What has happened is
that the second antenna is fragmentary and not at all well-preserved except for a
short stretch of a dozen or so segments which in flattened condition suggest a club.
The palpus that can be seen seems to have a distal joint about half as long as the
second. Both are heavily scaled and haired as is the basal curved segment. The
position of the terminal joint lies between erect and porrect. I can make out nothing
of the eye except its position. The proboscis is coiled just below the palpi.
The legs appear to be quite shaggy. I think that I can make out the first and third
FLCORIS3, A T hN'3.I3j n.sp. ,
-. ... -. .- .L /'
"* s'^ ^' f "''
'f-i '^'. .
Figure 1. The type of Oligodonta florissantensis, n.g., n. sp. The bar under the
label is 1 cm. long. Brown negative no. 2
rather well but cannot see any tarsal joints, claws, paronychia, etc. The abdomen
seems to be enveloped in the anal margin of the hind wings.
The venation of the fore wing, so far as I can decipher it after more than a week of
study, is shown in figure 2. The costal margin is about 26.5 mm long. R4+5 branches
5.1 mm. from the apex, M, does so 8.4 mm. from the apex. R2 and R, originate close
together, 12.8 and 15.7 mm respectively from the apex with R2 rising just basad
of the origin of udc. This is something like the branching in Aporia Huebner or
Neophasia Behr but with a much straighter radial stalk. The drawing of M2
and M3 is a little erroneous. The outer half and inner quarter of each can be seen.
The connecting of these parts is believed correct. The origins of the subcostal vein
and of the radial stalk are lost at the base of the wing in squashed thorax.
The arrangement of the veins of the hind wings is rather well preserved. Figure 3
shows those parts that I was able to find. The missing sections are present but
lost in a tangle of superimposed veins and edges of the fore wings. I could detect
Figure 2. The venation of the forewing of 0. florissantensis as far as can be
deciphered. The total length of the scale bar is 1.5 cm. Drawn from the fossil.
Figure 3. The venation of the hind wing of 0. florissantensis as far as it can be
deciphered. Drawn from the fossil. The scale is the same as for Figure 1.
no evidence of Idc and thus thought for a long time that I was dealing with a
nymphalid. I could find no trace of the humeral vein, lost in the same confusion as
the base of the fore wing.
Since I could find no evidence of basally swollen veins and there is a clear-cut
3A I have eliminated Satyridae and Papilionidae from consideration. The size of
the insect I believe eliminates Lycaenidae, Riodinidae and Libytheidae. Its habitus
is quite unlike Hesperioidea. This leaves me only Pieridae and Nymphalidae, both
in a broad sense, to consider. The presence of only four branches of the radius on the
fore wing points strongly to Pieridae. The possibly open cell on the hind wing suggests
Nymphalidae but there are pierids with a weak Idc that would be totally lost in the
confusion of four superimposed wings.
The name Oligodonta alludes to the age of the specimen, late Oligocene, and
to the blunt tooth on the margin of the forewing.
tOligodonta florissantensis, new species
This is the genus-type of Oligodonta and has all of the structural features
described for the genus. Unlike the two well-known Florissant butterflies,
Hypanartia (tProdryas) persephone (Scudder) and Chlorippe twillmetae
Cockerell, this specimen shows no pattern features that can be recognized. It is
quite possible that if it had been preserved expanded as the two noted are, it would
have. There is a dark mottling over much of the fossil that could be the result of
superimposing the patterns of four wings. Figure 1 does not show this will
since the lighting used reflected from the rather rough surface of the shale in which
the fossil is imbedded.
Holotype: a well preserved folded specimen with the radius of the fore wing about
26.5 mm., and of the hind wing about 22 mm. Found in the late Oligocene volcanic
lake shales of Florissant, Teller Co., Colorado. Excavated from the Singer trench,
now a part of the Florissant Fossil Beds National Monument.
Repository of the type: The type of Oligodonta florissantensis Brown is held in
the collection of the National Park Service at the Monument. The photographic
negatives are on deposit at the Allyn Museum of Entomology, Sarasota, Florida.
Numbers 2, 7 and 9 are the better ones.
Scudder, S. H., 1892, U. S. Geological Survey, Bulletin 93, pp. 21-24, P1. III; original
description and figures of tBarbarothea florissanti.