UW1 ?\3H

HARVARD UNIVERSITY

LIBRARY

OF THE

Museum of Comparative Zoology

The Library "'
Museum of Comparative Zoology
Harvard University

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

MUS. COMP. ZOOL
LIBRARY

NOV 6 1969

HARVARD
UNIVERSITY

THE FLORA OF THE KANSAS FLINT HILLS

By
William T. Barker

Vol. XLVIII Pages 525-584 October 17, 1969 No. 14

S- NJ{ - Ld*,

e^

ANNOUNCEMENT

The University of Kansas Science Bulletin (continuation of the Kansas Uni-
versity Quarterly) is an outlet for scholarly scientific investigations carried out
at the University of Kansas or by University faculty and students. Since its in-
ception, volumes of the Bulletin have been variously issued as single bound vol-
umes, as two or three multi-paper parts or as series of individual papers. In each
case, issuance is at irregular intervals, with each volume approximately 1000
pages in length.

The supply of all volumes of the Kansas University Quarterly is now ex
hausted. However, most volumes of the University of Kansas Science Bulletin
are still available and are offered, in exchange for similar publications, to learned
societies, colleges and universities and other institutions, or may be purchased
at $15.00 per volume. Where some of these volumes were issued in parts, in-
dividual parts are priced at the rate of IV2 cents per page. Current policy, ini-
tiated with volume 46, is to issue individual papers as published. Such separata
may likewise be purchased individually at the rate of 1% cents per page. Sub-
scriptions for forthcoming volumes may be entered at the rate of $15.00 per
volume. All communications regarding exchanges, sales and subscriptions should
be addressed to the Exchange Librarian, University of Kansas Libraries, Law-
rence, Kansas 66044.

Reprints of individual papers for personal use by investigators are available
gratis for most recent and many older issues of the Bulletin. Such requests should
be directed to the author.

Editor
Charles R. Wyttenbach

Editorial Board
Kenneth B. Armitage
Richard F. Johnston

Paul A. Kitos

Charles D. Michener

Delbert M. Shankel

George W. Byers, Chairman

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLVIII Pages 525-584 October 17, 1969 No. 14

The Flora of the Kansas Flint Hills

William T. Barker1

ABSTRACT

This paper reports 987 species of vascular plants for the Kansas Flint Hills.
Descriptions of the topography, drainage, climate, original vegetation, disturbance
of the original vegetation and the present vegetation of the region are included.

The Kansas Flint Hills area is a region of transition where eastern distributed
and western distributed species meet. Approximately 25°^ of the native species
that occur in the Kansas Flint Hills reach either their western or eastern limit of
distribution in Kansas within the Hills. In addition to those species that reach
their limit of distribution within the Flint Hills region, there are 100 species that
are limited in their westward or eastward distribution in Kansas by the Flint Hills
but do not occur in the Hills.

INTRODUCTION

The Flint Hills, as they are commonly called due to the cherty nature of
the soil, are a range of hills which cross Kansas in a north-south direction
and extend into Oklahoma. On the map these hills appear as a somewhat
elongated oval-shaped area about 200 miles from tip to tip, being some 50
miles or somewhat more than two counties in width (Map I).

No consistency in designating the area is found, either in professional or
in popular literature. Various names such as Kansas Mountains, Permian
Mountains, Flint Hill Pastures, Flint Hill Uplands, Bluestem Pastures and
Flint Hills-Bluestem Pastures have been used in referring to the area. As
pointed out in Kollmorgen and Simonett's (1955) comprehensive survey of
the bluestem grazing operations in Chase County, the precise boundaries of
the Flint Hills have not been defined to everyone's satisfaction. The name
Flint Hills apparently implies different things to different people. To the
geologist, the Flint Hills constitute an area where certain bedrock, notably

1 Present Address: Department of Botany, North Dakota State University, Fargo, North
Dakota 58102.

526

The University Science Bulletin

• <
i °

I

i!
I?

i.L.

» w

— I

15

• a
I?

j

i

is

• a
I u

1 u»

in..

HJ

l-H

<5

-— K-J

PU

tn

u

3

3

1— 1

i
Is

I

K

!s

i_t_.._

j<

il

u.

i

is

1

- At

id

\'i
i

L

il

•I—
i

!«

5

12
J--.-

I

IS

I?

■H-*.

is

cr-

o

'J

o

'-M

0)

T3
O

a

05

a

U

00

O

ft

2

..„i:

The Flora of the Kansas Flint Hills 527

cherty limestones, outcrop. Kollmorgen and Simonett (1955) point out that
Jewett (1941) and Schoewe (1949) do not agree on the extent of the Flint
Hills. To the botanist and the range scientist, the Flint Hills constitute an
area of natural vegetation whose substrate only incidentally contains chert
rock (McGregor, 1955, and Bidwell, 1966). To the northwest of the Flint
Hills one encounters the Dakota Sandstone Formation which is covered with
a tall grass vegetation with the same dominant bluestems as found in the
Flint Hills. The same is true to the east, where much of the Osage Plains is
covered by tall grass vegetation with the same dominant bluestems. The area
designated as the Flint Hills is by no means coterminous with big pasture
country in Kansas. To the zoologist, the Flint Hills are a natural habitat for
certain animal species and a barrier to the westward distribution of eastern
species and to the eastward distribution of western species (Gier, 1967). To
the native Kansan, they constitute an unusual bluestem mantled landscape
whose steep rocky slopes and accordant summits create a panoramic beauty
that constantly changes with the seasons. For this study, a broad interpreta-
tion of the Flint Hills was followed. It includes not only the outcrops of the
Florence, Schroyer and Three-Mile cherty limestone strata, but also the ad-
jacent shallow soils on massive limestone cuestas, the associated steep slopes
and the areas of deep, nearly level, clayey soils of the broad ridges. With this
interpretation, the Flint Hills include parts of the following counties: Cowley,
Butler, Chase, Marion, Lyon, Morris, Geary, Wabaunsee, Riley, Pottawatomie
and Marshall. The total area studied includes approximately 4.5 million acres.

This study of the vascular flora of the Kansas Flint Hills was made with
two objectives in mind. The first was to compile a list of the vascular plants
for the area and to support this list with specimens filed in the Herbarium at
the University of Kansas. The second objective was to study the individual
species in the environment in which they grow and to record this information.

Other workers have compiled and in some cases published lists of vascular
plants from portions of the area. Carruth (1877) listed 52 species from the
area. Most of these were from the vicinity of Irving, a small town in Marshall
County. Hitchcock (1899) recorded some 600 species for the Kansas Flint
Hills. Stevens (1917) lists plants he collected from the Manhattan (Riley
County) and Blue Rapids (Marshall County) area which included 600 spe-
cies. Maus (1919) lists 495 flowering plants for Wabaunsee County. Gates
(1940) listed 602 taxa for the Kansas Flint Hills. Fish (unpublished thesis,
1953) listed 813 species for Riley County, 693 species for Pottawatomie County
and 528 species for Wabaunsee County. There are other papers which list
species for the area but none of these studies involved systematic inventories
(Aldous, 1934; Anderson, 1951, 1953, 1965; Anderson & Fly, 1955; Weaver,
1954). From this review of the literature, it is evident that a complete study
of the vascular flora of the Kansas Flint Hills had not been done. Since this

528 The University Science Bulletin

area is economically important as a grazing area and is a region of transition
where eastern distributed and western distributed species meet, it was felt that
a floristic study of the Kansas Flint Hills would be a worthwhile addition to
the knowledge of Kansas plants.

Prior to this study extensive fieldwork had been done in the area by Dr.
R. L. McGregor and Mr. Steve Stephens of the Department of Botany,
University of Kansas. As a result, many plant specimens deposited in the
University of Kansas Herbarium have supplemented the writer's collections.

ACKNOWLEDGMENTS

The writer is indebted to Dr. R. L. McGregor for first suggesting the
problem studied, for many suggestions and criticisms, and for arranging
financial assistance for the necessary fieldwork. Special thanks are due to
Drs. McGregor, Baxter and Thompson, of the University of Kansas Botany
Department, for their allotment of time in editing of this paper. An acknowl-
edgment is also due Mr. Steve Stephens, of the University of Kansas Botany
Department, whose aid in collecting has contributed greatly to this study.

TOPOGRAPHY

In terms of major physical divisions of the United States, Kansas lies
almost wholly within the Interior Plains (Fenneman, 1939). The Interior
Plains Division is comprised of two provinces in Kansas and both of these
extend far beyond the borders of the State. These are the Central Lowland,
including roughly the eastern third of Kansas, and the Great Plains, to which
the remaining two-thirds of Kansas belongs. The Kansas Flint Hills are a
part of the Central Lowland. The Central Lowland has the following units
in Kansas: Dissected Till Plains, Osage Plains, Chautauqua Hills and the
Flint Hills (Map I).

In general, the Flint Hills, as well as the Osage Plains, consist of a series
of escarpments between which are flat to gently rolling plains. The escarp-
ments are east-facing and tend irregularly from southwest to northeast. The
underlying strata are made up of unequally resistant alternating hard and
soft Pennsylvania!! formations of limestones and shales which are gently
inclined to the west and northwest. Due to differential erosion in these alter-
nating hard limestones and soft shales, the series of cuestas have developed in
the area. In the Flint Hills these Pennsylvanian deposits are capped by Perm-
ian strata and erosion of these strata accounts for the cherty nature of the soil.

The surface of Kansas gradually slopes downward from the western
border eastward, at a rate of 10 to 15 feet per mile. Along the western edge
of the Flint Hills the topographic relief is almost imperceptible. By contrast,
it is the highly dissected east-facing escarpment with its terraced slopes that
makes possible the separation of the Flint Hills from the lower or east division

The Flora of the Kansas Flint Hills 529

of the Osage Plains (Schoewe, 1949). This eastern border of the Flint Hills
is probably the most rugged surface feature in Kansas. The eastward-facing
escarpment is not everywhere one great step but is more commonly made up
of two or three closely spaced rock benches, with the intervening slopes rising
with steep gradients to the highest bench which everywhere forms the broad
upland of the Flint Hills. The relief of this east-facing escarpment ranges
from 300 to 350 feet. The surface of the uplands slopes gently toward the
Arkansas River Valley. There are places where the uplands are pitted by sink
holes. The uplands generally range between 1,500 and 1,600 feet above sea
level. The streams in the Flint Hills have deep precipitous channels lined
with outcropping rock ledges. In places, where these have cut into limestones,
narrow box-like channels have been produced, but where they have cut into
the less resistant shales, the valleys immediately open out and the valley slopes
are much gentler. A relief of 300 to 350 feet is often encountered from the
stream valley floor to the uplands. In Cowley County, Grouse Creek (un-
labeled stream, Map II), a south-flowing left-handed tributary of the Arkan-
sas River, divides the Flint Hills into two ridges, the easternmost one of
which is known as the Big Flint Hills, and the western one as the Little
Flint Hills (Adams, 1903). North of the Kansas River, the Flint Hills
diminish in height and become a less conspicuous topographic feature.
This northern part of the Flint Hills has been glaciated, as evidenced by
scattered erratics and till deposits, and is included in the Attenuated Drift
Border unit of the Dissected Tills Plains section (Schoewe, 1949).

DRAINAGE

The only river to cut through the Flint Hills upland from west to east is
the Kansas River (Map II). The Kansas River flows in a well-defined, rock-
walled, terraced valley from two to three miles wide and from 150 to 200 feet
deep. The northern part of the Flint Hills is drained by the Kansas River and
its tributaries. Further south, numerous spring fed creeks give rise to
several permanent rivers that drain the southern part of the Flint Hills. The
Marais des Cygnes courses generally eastward roughly paralleling the Kansas
River to join the Missouri River in Missouri. The Neosho and its principal
affluent, the Cottonwood, head in the Flint Hills and flow southeast to join
the Arkansas River in eastern Oklahoma. Paralleling the Neosho to the west
is the Verdigris, with the Elk and Fall Rivers as its main tributaries. It unites
with the Arkansas a few hundred feet above the Neosho in Oklahoma. The
Walnut River rises on the west, or back slope of the Flint Hills and empties
into the Arkansas a few miles above its point of exit from the state. Each of
the rivers mentioned above have numerous tributaries which arise in the
uplands. These are too numerous to mention here. All of these rivers and
creeks run in wide, flat-floored valleys with high bordering bluffs and ter-

530 The University Science Bulletin

races. The valleys are subject to major floods that often spread widely. The
bottom land soils are typically heavy though fairly productive. Numerous
impoundments are found along the drainage areas throughout the area.

CLIMATE

The climate of the Kansas Flint Hills is one of the area's most variable
features. Sudden changes in temperature, wind and precipitation are com-
mon. Temperature extremes vary from ~32°F. to +118°F. with the averages
being rather moderate (Flora, 1948). From April to December the prevailing
winds come from the south, whereas during January through March north-
erly winds prevail. The periods of heaviest rainfall come in the fall and the
spring with the average annual precipitation ranging from 30 inches in the
northern part of the area to 35 inches in the southern part. The average
growing season is 189 days in the south as compared to 176 days in the north
part of the area. The average relative humidity during January ranges from
60 to 70 percent, whereas during July it ranges from 45 to 50 percent. These
data were obtained by averaging relative humidity readings taken daily at
noon (Flora, 1948). Snowfall is moderate and snow seldom remains for long
on the ground.

Ktichler (1967) points out that the mean values of annual and monthly
precipitation have been recorded for many decades but they conceal the most
critical of all climatic features, the rainfall reliability. In any given season the
rains may come early or late; they may be evenly distributed throughout the
growing season or be bunched together in a few hard showers; it may rain
amply or very little. The seasonal aspect of the flora is greatly influenced by
the seasonal distribution of rainfall. It is evident from the recorded clima-
tological data that periodic drouths are characteristic of the region's climate.
During the recorded history of the state there have been seven periods of dry
weather of exceptional note: 1859-1868, 1872-1875, 1892-1894, 1901, 1915-1917,
1930-1940, and 1952-1956. Albertson and Tomanek (1965) have presented
carefully documented observations of vegetational changes near Hays, Kan-
sas, located in Ellis County, during 30 years which illustrate complex features
of the ecosystems of the prairie. Unfortunately, there has not been a com-
parable study for the Kansas Flint Hills, but the effects of drouths in the area
have not gone completely unnoticed. Hoyle (1938) stated that during the five
year period, 1931-1936, one third of the maples and elms died due to dry
weather. Gates (1945) stated that due to the dying out of prairie grasses
during the drouth (1930's), Gutierrezia dracunculoides spread into the
prairie lands at an amazing rate but by 1945, after the drouth had broken,
the prairie grasses had again replaced the Gutierrezia. Agrelius (1945) also
commented on the disappearance of Gutierrezia dracunculoides after the
drouth broke, and mentioned that Anemone caroliniana and Viola pedatifida
had also become scarce in the more mesic times.

The Flora of the Kansas Flint Hills

531

IS ' *

CO
CO

C

V-l

O

03
<D
W
(h

o
U

cu
M
+j
W

o

••-I

2

2

532 The University Science Bulletin

ORIGINAL VEGETATION

From geologic and paleontologic studies (Frye & Leonard, 1952, 1957) it
seems clear that the grassy plains of Kansas are not a recent development.
It is thought that the climatic control of the Kansas vegetation has been the
same as it is today since Pleistocene times. Wedel (1959) feels that the grass-
land vegetation antedates man's demonstrable presence on the scene.

A knowledge of the vegetation of the Kansas Flint Hills before the time
of settlement must be gained largely from notes of explorers, travelers and
surveyors. These written records occasionally include mention of native
plants. Malin (1967) points out that the explorers and travelers usually went
west after the grass started in the spring, and, if they returned, usually did so
in the late summer and fall, and that the reports on the vegetation of these
seasons tended to develop into a stereotype for the grass country. It is evident
as one reads the notes of these early people that they were tree conscious and
the grasslands were unfamiliar and quite monotonous to them. Tree species
are mentioned often in their notes but only occasionally are the herbaceous
species mentioned.

The Flint Hills vegetation prior to the time of white settlement must have
been similar to what one sees in the area today. While in Pottawatomie
County, Fremont (1845) described the many tributaries of the Kansas River
as looking like "trenches in the prairie usually well timbered."

Mollhausen (Taft, 1948) traveled from west to east across Kansas in 1858
and he relates how just west of Marion County the grass became more fresh
and luxuriant than the short, insignificant, but nonetheless nourishing buffalo
grass (Buchloe dactyloides). At about the same time he relates how the
depressions became deeper, the elevations higher, and the springs and brooks
more numerous.

Mollhausen wrote on July 15, 1858, "We camped several miles east of
Diamond Springs (undoubtedly within the Flint Hills region) on an un-
named brook. Tall grass surrounded us, with an annoyance which we could
not avoid." It is interesting to note that in reference to the country between
Council Grove and Kansas City, Mollhausen wrote, "I shall avoid describing
the rest of the journey in diary form for on the entire route from the Neosho
to the Missouri (Rivers) we were constantly in surroundings whose characters
remained unchanged." If originally unbroken prairie extended all the way
from Kansas City through the Flint Hills region, as Mollhausen's statement
indicates, it is not surprising that the early people did not describe the Flint
Hills in more detail. Undoubtedly, the region is a more prominent region in
the landscape now than it was in 1858.

Early observations of the vegetation in the southern part of the Flint Hills
were recorded by members of a surveying party that surveyed the southern
boundary line of Kansas in 1857 (Caldwell, 1937; Miller, 1931). These people

The Flora of the Kansas Flint Hills 533

also indicate that the stream valleys were well-timbered and that the hillsides
and uplands were covered with tall grasses of a high quality.

The Kansas Flint Hills counties were surveyed during the year 1856-1871.
The original surveyor's notes, taken in the field during the surveys of town-
ship and section lines, contain a general description of the land and often
observations of the vegetation. These original notes are on file at the State
Auditor's Office in the State House in Topeka, Kansas. It is evident from
these notes that the surveyors were tree conscious and were mainly interested
in land suited for farming and therefore were usually very unimpressed with
the vast areas of rolling prairie. As one reads these notes he is impressed by
the uniformity of observations of the landscape. The ravines, creeks, streams
and rivers are described as being "well timbered," while the slopes and up-
lands are described as being covered with "prairie or grassland." Frequent
mention was made of the following trees: elm, oak, burr oak, hickory, cotton-
wood, walnut, hackberry, sycamore, redbud, willow, hazel, box elder, ash,
cedar, linden, dogwood, plum and prickly ash. This list of trees compiled
from the historical records corresponds rather well with the list of the com-
mon trees of the present woodlands.

These first observations, prior to or at the beginning of white settlement
in the region, indicate that most of the uplands and hillsides were covered by
tall grass prairie vegetation and that the creeks, streams and rivers were lined
with trees.

SETTLEMENT AND DISTURBANCE OF THE
ORIGINAL VEGETATION

Disturbances of the original vegetation by the plow, ax and grazing
accompanied white settlement of the area and have increased and continued
to the present day. The first major steps in which occupation were initiated
in the early 1840's when missions were established at Council Grove (Morris
County) and St. Mary's (Pottawatomie County) to teach general farming
and cattle grazing to the Pottawatomie Indians (Malin, 1942). During the
next several years, more small farmers settled in the Big Blue, Kansas, Neosho
and Cottonwood river valleys. These first settlers, being farmers, were mainly
interested in the bottom lands and were of the opinion that the uplands were
unproductive. Timber was cut from the bottom lands to build houses and
other improvements and to clear fields in preparation for general farming.
After the Civil War, many more small farmers came to settle, first in the
bottom lands and finally in some of the uplands. Much of the native grassland
was plowed and more timber was cut from the river valleys.

During the decade of the 1870's, the agricultural interests of the region
were relatively diversified. There were at least four types of activities repre-
sented in the area: general farming, which emphasized grain crops; farming,

534 The University Science Bulletin

which emphasized the raising of corn; the breeding of fine stock; and the
maturing and grazing of transient cattle (Malin, 1942).

During the years 1872-1875 there was a severe drouth. By 1879 many of
the upland farms had been abandoned and the former cultivated fields al-
lowed to go back to native grass. At this time, the uplands were generally
used for grazing and the bottom lands were used for general farming.

During the decade of the 1880's the region experienced a livestock boom
(Malin, 1942). The use of barbed wire began slowly about 1879 and 1880 and
reached boom proportions by 1883. The bottom lands were fenced from the
uplands first and by 1885 the uplands had been fenced off into separate pas-
tures. This in a sense brought about a sharp line of distinction between the
small farmers in the bottom lands and the cattlemen controlling the uplands
that exists even today (Kollmorgen & Simonett, 1965).

In the first part of the 1880's the major emphasis was placed upon the
expansion and improvement of local cow herds, but by the latter part of the
decade this practice was replaced by the maturing and grazing of transient
cattle. By 1886 cattlemen realized that cattle could be fattened to market
weights on the Kansas Flint Hill prairies in a single summer. The practice of
bringing yearling steers to the area by rail in mid-April, allowing them to
graze on the bluestem prairies through the summer months, and then ship-
ping them by rail to eastern markets by mid-August or, at the latest, mid-
September soon became the common rule. Much outside capital was invested
by southwestern cattlemen and various syndicates organized during the 1880's
to assemble acreages large enough for these extensive cattle operations. To
build large enough acreages required large sums of capital and the small
farmers were never able to compete with the cattlemen in acquiring the
uplands. So, early in the history of the area, absentee ownership of the up-
lands was established. Today, more so than ever before, the uplands are
controlled by absentee owners who are mainly interested in stock raising
rather than general farming. For an example, in 1959 in Chase County, only
45% of the land was locally owned. It is also interesting to note that only 30%
of all the pasture land was locally owned while 81% of the cropland was
locally owned (Kollmorgen & Simonett, 1965).

Due to changes in agricultural technology and economic conditions in the
United States, the number of small farmers in the region has steadily declined
and the large holdings of the cattlemen have continued to increase in size at
the expense of the small farmers and small ranchers. In recent years the prac-
tice of maturing and grazing of transient cattle has given way somewhat to
the practice of maintaining local cow herds. This practice increases grazing
pressures on the prairies. The native grasses of the region begin to grow in
March and continue their development until the first killing frost in early
October. Transient cattle have usually been turned into the prairies by mid-

The Flora of the Kansas Flint Hills 535

April and are then shipped to market around mid-August or by mid-Septem-
ber at the latest. Anderson (1953) feels that the maintenance or relatively good
native prairie in the region has been due to the fact that since the cattlemen
have wanted rapid gains on their steers, light stocking has been practiced.
Early in the history of the area large holdings were built and these owners
could afford the luxury of light stocking. Along with this, cattle often reached
market condition by mid-summer and were removed, thus allowing the native
grasses to recover by allowing carbohydrate reserves to accumulate, flowering,
seed set and finally, reseeding of the prairie.

Overgrazing in the Flint Hill region has occurred often, especially where
both farming and stock raising has been practiced and the smaller ranges have
been more or less continually overstocked. Weaver (1954) stated that light
grazing involves only a shift in number of plants, not in species composition
of the native prairie. Heavier overgrazing practically eliminates the big blue-
stem, while little bluestem, sideoats grama, tall dropseed and buffalo grass
increase. If the overgrazing becomes more severe, there is an invasion of
weeds and many brushy species. In the northern part of the Flint Hills,
Juniperus virginiuna is becoming a serious problem in overgrazed situations.
Throughout the area the brushy species, Comas drummondii, Rhus aro-
matica and Rhus glabra increase during overgrazing. Ranchers are using
selective herbicides, 2,4-D (2, 4-D Dichlorophenoxyacetic acid) and 2, 4, 5-T
(2, 4, 5-Trichlorophenoxyacetic acid), to control these. These selective herbi-
cides kill many dicotyledonous forbs, especially the leguminous plants, which
are important prairie components.

A trend that is becoming more and more evident is that cow herds are now
being maintained in the Flint Hill region. This practice increases the grazing
pressures on the prairies since the cows and calves are on the prairies nine to
twelve months rather than just in the summer months, as with the transient
steers. To emphasize this point, a common rule-of-thumb that ranchers follow
in stocking their pastures holds that a good to superior pasture should carry
a long-yearling or two-year old steer on five acres of pasture for a six month
summer grazing period whereas, a cow or a cow with a small calf requires
at least twelve acres for a nine-month grazing period.

Burning of the prairies in the spring has long been a common practice.
Many ranchers feel that if they burn early each spring they will have better
grass sooner, and others burn in an attempt to control brush. Anderson
(1965) points out that the burning of bluestem range in the Flint Hill region
reduces soil moisture. Burning of prairies in very dry years undoubtedly
causes deterioration of the condition of native prairie. Anderson (1965) sug-
gests the following rules for ranchers in the region: "Burn no more than is
necessary for good range management and good livestock husbandry. If burn-
ing is practiced, do it in late spring rather than earlier, recognizing that soil

536 The University Science Bulletin

moisture levels, and hence herbage yield, will be reduced by burning. The
earlier one burns, the less herbage he will have for harvest by livestock."

"Improvements" that man has made and is making in the area, such as
building roads and impoundments, will have an influence on the flora of the
area. Many acres of native vegetation are destroyed when a new road or
impoundment is built. Disturbances of this sort provide open habitats and
the number of plants of the weedy species of the area will increase. The num-
ber of aquatic plants will increase in the new aquatic habitats produced. A
few new species may become established in these new habitats.

PRESENT VEGETATION

The largest remaining tract of true prairie in Kansas occupies the Flint
Hill region. As previously stated, the region includes approximately 4.5 mil-
lion acres. According to the 1965 Biennial Report of the Kansas State Board
of Agriculture there are about 3 million acres of pasture (prairie) and approx-
imately 1.5 million acres of cropland and woodland in the Flint Hill counties.
In 1965 171,000 acres of native prairie was mowed for hay.

The vegetation of the Flint Hills region has been subjected to disturbances
of various sorts throughout its history. Various degrees of degradation from
the natural state can be found due to these disturbances and it is beyond the
scope of this paper to discuss all of these. The habitats available for plant life
in the Flint Hills are well-defined topographically and are easily recognized
by their characteristic vegetation. These habitats range from aquatic situ-
ations provided by streams and artificial impoundments to the more xeric
upland prairie. The following habitats are typical of the Kansas Flint Hills.

Seepage Areas, Streams and Artificial Impoundments
In the Flint Hill region aquatic plants are found in seepage areas around
springs, in and along streams and artificial impoundments. In relatively deep
water of impoundments and slow-moving streams there may be several spe-
cies of submerged aquatics. Examples of these are Ceratophyllum demersum,
Myriophyllum heterophyllum , M. pinnatum, Potamogeton diversijolius, P.
obtusifolins and P. pectinatus. Nearer the shores in shallow water there are
plants which are rooted in the mud and whose leaves float on the water.
Some of these plants are Bacopa rotundijolia, Jussiaea repens and Nuphar
ad vena. Among these plants are found free-floating plants such as Lemna
minor, L. perpitsilla, Spirodela polyrhiza and Wolffia columbiana. There are
many plants, rooted below the water at the edge of these aquatic situations,
with their foliage above the water. These include Alisma triviale, Echino-
dorus cordijolius, Eleocharis calva, E. compressa, E. macrostachya, E. obtusa,
Heteranthera dubia, H. limosa, H. renijormis, Juncus diffiisissimus, J. in-
terior, f. torreyi, Sagittaria lati folia, S. montevidensis, Scirpus valid us, Spar-

The Flora of the Kansas Flint Hills 537

ganium eurycarpum and Typha latijolia. In the moist soil found along the
margins of these aquatic situations, the vegetation usually consists of smart
weeds, grasses, sedges and occasional shrubs. Examples of these are Poly-
gonum bicorne, P. lapathifolium, P. pennsylvanicum , P. punctatum, Ritmex
altissimus, R. crispits, Echinochloa crusgalli, Eragrostis hypnoides, Glyceria
striata, Leersia oryzoides, L. virginica, Leptochloa fascicularis, Phragmites
communis, Carex crawei, C. crus-corvi, C. emoryi, C. jran\ii, C. hyalinolepsis,
C. hystricina, C. normalis, C. vulpinoidea, Cyperus erythrorhizos, C. escu-
lentus, C. jerruginescens, C. strigosus, Scirpus atrovirens, S. lineatus, Amor-
pha fruticosa, Cephalanthus occidentalis, Salix caroliniana, S. interior and S.
nigra. Outside of the latter zone of vegetation there is an abrupt change to
the surrounding plant community, whether it is a prairie or a woodland
community.

In seepage areas below springs, such plants as Equisetum laevigatum,
Spartina pectinata, Typha latijolia, Carex annectans, C. crawei, C. interior,
C. lanuginosa, C. vulpinoidea, Cyperus jerruginescens, C. strigosus, Eleo-
charis obtusa, E. calva, E. compressa, E. macrostachya, Furiena simplex,
Scirpus atrovirens, S. lineatus, ] uncus torreyi, Nasturtium officinale, Prunella
vulgaris, Veronica angallis-aquatica, Bidens frondosa, B. polylepis, B. cernua
and B. vulgata are found. Just a few feet away from the seepage area very dry
conditions are usually found and a typical prairie community usually exists.

Woodland
In the Kansas Flint Hills limited amounts of woodland are developed
along floodplains, ravines and steep hillsides. In the Flint Hills region one
finds the greatest proportion of woodlands in the northern counties and the
lowest proportion in the southern counties. When one considers the major
stream courses and the resulting erosional patterns of the region it becomes
quite obvious that there are more habitats suitable for the development of
woodlands in the northern and southern parts (see Map II). In the north,
the Kansas River courses from west to east through the region. There are
numerous large (Big Blue, Little Blue and Vermillion) rivers and small
north-south tributaries of the Kansas River that dissect the area to the north.
Each of these north-south tributaries have their smaller tributaries that run
east and west. The result is more floodplains, ravines, and steep hillsides
(hence, more woodlands) in the north as compared to the central and south
portions of the area. The Cottonwood and Neosho Rivers are the only major
streams in the central part of the Flint Hills. The floodplains of these streams
are wooded and they have their small wooded tributaries. However, the cen-
tral portion is less dissected by the stream courses and the resulting valleys are
usually wide with gently sloping sides. In general, the terrain of the central
portion is more rolling, which favors the development of a prairie community

538 The University Science Bulletin

rather than a woodland community. In the southern part of the region the
Walnut River and Grouse Creek dissect the area in a north-south direction
and their tributaries dissect the area in an east to west fashion. The resulting
erosional pattern provides suitable sites for the development of woodlands.

The floodplain woods, as would be expected, are more mesic than the
woods encountered along the ravines and on the hillsides. Ulmus americana
dominates the floodplain woods. Other important tree species found in these
woods are Ulmus rubra, Platanus occidentalis, Populus deltoides, Acer ne-
gundo, A. saccharinum , Fraxinus pennsylvanica var. subintegerrima, Celtis
occidentalis, Salix amygdaloides, S. nigra, S. interior, Morns rubra, and
fuglans nigra. Common shrubs found in the understory of the floodplain
woods are Kibes missouriense , Sambucus canadensis, Asimina triloba and
Euonymus atropurpureus. Woody vines that are found in these woods in-
clude Celastrus scandens, Parthenocissus quinquefolia, Rhus radicans, Smilax
hispida, Vitis cinera, V. riparia and V. vulpina.

The floodplain woods are subject to periodic flooding and this has a pro-
found effect on the herbaceous flora. The herbaceous species found in these
woods vary from year to year or from flood to flood. The herbaceous species
appear as spring, summer and fall aspects. Spring plants include Chaerophyl-
lum procumbens, Dicentra cucullaria, Phlox divaricata var. laphamii, Viola
papilionacea, Galium aparine, Ellisia nyctelea, Oxalis stricta, Parietaria penn-
sylvanica, Carex aniphibola, C. blanda, C. davisii, C. molesta and Ruellia
strepens. Conspicuous summer species are Geum canadensis, Oxalis dillenii,
Laportea canadensis, Urtica dioica, Silene stellata, Thalictrum dasycarpum ,
Chenopodium hybridum and Scrophularia marilandica. Fall plants include
Bid ens bipinnata, Eupatorium rugosum, Polygonum virginianum, Muhlen-
bergia brachyphylla, M. mexicana, M. racemosa, M. schreberi, Sicyos angu-
latus, Verbesina alter ni folia and Ambrosia trifida. The floodplain woods are
relatively uniform in their composition throughout the region.

Progressing from the floodplain woods up the steep hillsides and ravines,
an abrupt change in the vegetation occurs and a more xeric type of woodland
is encountered. At the base of the steep hillsides and ravines Quercus macro-
carpa and Carya cordiformis are common. In the northern part of the Flint
Hills Quercus muehlenbergii becomes the dominant plant further up on the
hillsides. In the southern portion of the region, Quercus muehlenbergii is
replaced by Q. prinoides as the dominant tree high on the hillsides and in the
ravines. Other important tree species found in these drier woods are Ulmus
americana, Celtis occidentalis, Cercis canadensis, Ostrya uirginiana, Madura
pomifera and Gleditisia triacanthos. funiperus virginiana is a common tree
in this type of woodland in the northern part of the Flint Hills. Dense stands
of this tree are found in the northern woodlands but only scattered trees are
found in the central and southern portions of the Flint Hills. Common shrubs
found in the understory of this dry type of woods are Aesculus glabra, Sym-

The Flora of the Kansas Flint Hills 539

phoricarpos orbicitlatus, Rhus aromatica, Cornus drummondii, Staphylea
trifolia, Zanthoxylum americanum, Ritbus allegheniensis, R. flagellaris, R.
occidentalis and R. ostryifolius. Corylus americana and Rhamnus lanceolata
are common shrubs in the northern portion of the Flint Hills but are not
found in the southern portion. Ptelea trijoliata, Bumelia lanuginosa and
Sapindus drummondii are found frequently in the southern hillside wood-
lands but are not found in the northern woodlands.

One of the major differences between the woodlands in the Flint Hill
region as compared with those of the Dissected Till Plains, the Osage Plains,
the Chautauqua Hills and the Cherokee Plain is the reduced number of
herbaceous species present. Very often there is little or no herbaceous cover
in the dry hillside woods of the Flint Hills. The herbaceous species of these
woods can be listed as spring, summer and fall aspects. Spring plants include
Erythronium albidum, Galium aparine, Viola papilionacea, Carex aggregata,
C. bic\nellii, C. brevior, C. cephalophora, C. heliophila, Antennaria planta-
gini folia, Ruellia strepens and Parietaria pennsylvanica. Conspicuous sum-
mer species are Galium circaezans, Sanicula canadensis, Desmodtum glu-
tinosum, Phyrma leptostachya, Oxalis dillenii, Silene stellata, Campanula
americana, Geum canadensis and Ely m us virginicus. Common fall species
are Hachelia virginiana, Agastache nepetoides, Helianthus hirsutus, Eupa-
torium rugosum, Polygonum virginianum, Ambrosia trifida, Solidago nemo-
ralis, S. ulmifolia, Verbena urticifolia, Verbesina alternijolia, Bidens bipin-
nata, Uniola lati folia and Aster drummondii. In the Flint Hills region
woodlands are seldom, if ever, developed on the uplands. As one ascends to
the uplands, a prairie community is abruptly encountered.

A zone of shrubby vegetation is common along the limestone benches on
the escarpments. The shrubs that make up this zone are Rhus aromatica,
Cornus drummondii and Rhus glabra.

Prairie

The prairies found in the lowlands along streams are dominated by
Andropogon gerardi, while Sorghastrum nutans and Panicum virgatum are
common. Scattered colonies of Tripsacum dactyloides may be present and
Spartina pectinata is often dominant in very wet places. Leaving the lowland
prairie and ascending the prairie hillsides, Andropogon gerardi is still the
dominant grass, but A. scoparius and Bouteloua curtipendula increase in
abundance. On the uplands Andropogon scoparius is the dominant and other
important grasses found are A. gerardi, Bouteloua curtipendula, Panicum
virgatum, Sorghastrum nutans, Bouteloua hirsuta, Buchloe dactyloides and
Sporobolus heterolepis.

As a result of his studies of numerous Flint Hill prairies, Weaver (1954)
gives the following average species composition figures for the area: Andro-
pogon gerardi 38.9%, A. scoparius 34.4%, Bouteloua curtipendula 5.3%, B,

540 The University Science Bulletin

hirsuta 4.2%, Sorghastrum nutans 2.9%, Panieum virgatum 2.9%, Buchloe
dactyloides 1.9%, Sporobohts heterolepis 1.7%, and various forbs 2.2%.

Light overgrazing of Flint Hill prairies reduces the abundance of the tall
grasses, Andropogon gerardi and A. scoparius, while Bouteloita ciirtipendula,
B. hirsuta and Buchloe dactyloides become more abundant. At this stage of
degeneration the Andropogon scoparius and Bouteloita ciirtipendula are re-
duced to scattered bunches and Buchloe dactyloides becomes dominant. As
overgrazing becomes severe, Andropogon gerardi is eliminated and A. sco-
parius is greatly reduced in abundance. In cases of severe overgrazing, soil
erosion often results and the native vegetation is nearly destroyed. Aristida
oligantha readily invades these exposed sites and appears as light areas in the
vegetation.

A number of species give the prairie spring, summer and fall aspects. The
most apparent plants in the spring and summer aspects are forbs, and it is
important to remember that forbs account for less than 3% of the prairie
community. Before the dominant native grasses have a chance to grow, much
of the spring aspect becomes apparent due to the colorful flowering species
involved. Important spring species are Anemone caroliniana, Antennaria
neglecta, Carex meadii, Nothoscordum bivale, Lomatiutn joeniculaceum,
Nemastylis gemini flora, Senecio plattensis, Sisyrinchium campestre, Achillea
millefolium , Zigadenus nuttallii, Baptisia leucophaea, Baptisia australis var.
minor, Asclepias viridis, Cacalia tuberosa, Polytaenia nuttallii, Callirhoe al-
caeoides, Erigeron strigosus, Erigeron annuus, Echinacea angustifolia and
Ratibida columnifera.

The summer aspect of the prairie is less striking than that of the spring
because by this time grasses have undergone considerable growth. Common
plants of the summer are Schranl(ia uncinata, Psoralea tenuiflora, P. argo-
phylla, Petalostemon purpureum, P. canidum, Lin um sulcatum, Cassia
fasciculata, Monarda fistulosa, Hedyotis nigricans, Ruellia humilis, Silphium
lacimatum , Asclepias verticillata, Salvia azurea, Liatris aspera and Liatris
mucronata. In the fall the flowering of the dominant tall grass species, An-
dropogon gerardi, A. scoparius, Sorghastrum nutans and Panieum virgatum
gives the prairie its aspect along with other conspicuous plants such as
Euphorbia marginata, Vernonia baldwinii, Solidago missouriensis, Eupa-
torium altissimum, Solidago altissima, Kuhnia eupatorioides, Solidago rigida,
Helianthus laetiflorus, H. maximiliani, H. grosseserratus, H. salicifolius,
Gutierrezia dracunculoides, Artemisia ludoviciana, Aster ericoides, A. ob-
longifolius and Solidago graminifolia.

Along the limestone benches and the summits of the escarpments the soil
is very cherty and shallow. The native grasses do not grow well in these
marginal sites but certain shrubs and forbs do grow well and when these
plants are in flower it is easy to observe these zones. In the spring, blooming

The Flora of the Kansas Flint Hills 541

Amorpha canescens is common, followed in the summer by Monarda fistu-
losa. As Monarda fistulosa goes by, "Euphorbia marginata appears, then is
replaced by Gutierrezia dracunculoides in the fall. In many places Rhus
aromatica, Symphoricarpos orbiculatus, Cornus drummondii and Rhus glabra
are locally abundant.

Roadsides and Railroad Right-of-Ways

A complete list of plants found along the roadsides and railroad right-of-
ways would nearly duplicate the list of plants compiled for the entire region.
The vegetation of these areas consists of the border plants of whatever com-
munity the road or railroad passes.

Thickets are common along roadsides and railway right-of-ways. Com-
mon plants of these thickets are Cornus drumtnondii, Primus americana,
Madura pomifera, Rhus glabra, R. aromatica and Symphoricarpos orbicu-
latus. Some common herbaceous plants along these areas are Cassia jascicu-
lata, Coronilla varia, Delphinium virescens, Euphorbia marginata, E. mis-
surica, Kuhnia eupatorioides, Melilotus alba, M. officinalis, Oenothera
speciosa and Saponaria officinalis. Roadsides and railroad right-of-ways
form convenient avenues of escape for many cultivated species and are likely
spots for plants being introduced into a region. Cattle trucks passing along
the highways afford excellent possibilities for the dispersal for seed along the
route. The same is true of freight cars passing along railroad right-of-ways.
Most of these introduced plants persist for just a season, but undoubtedly
some have become established in the region and others will become estab-
lished in the region in this manner.

Cultivated Fields

Cultivated fields are invaded by a host of plants commonly referred to as
weeds. Most of these are annuals and occur in such large numbers or in such
close proximity with the cultivated crop that they can't all be destroyed;
therefore, a plentiful seed supply for the ensuing year is nearly always as-
sured. The most abundant of these are Thlaspi arvense, Chenopodium album,
Koclua scoparia, Amaranthus tamariscinus, A. retroflexus, Polygonum bi-
corne, P. pensylvanicum , Hibiscus trionum, Digitaria sanguinalis, Setaria
viridis, Panicum capillare, Abutilon theophrasti, Helianthus annuus, Am-
brosia trifida, Lactuca scariola, Tribulus terrestris, Xanthium chinense, X.
ttaltcum,X. pennsylvanicum and X. speciosum.

Abandoned Fields

As previously mentioned, much of the Flint Hills prairie was early plowed

in an attempt to practice general farming. Farming in much of the upland

prairie proved to be marginal at best and many fields were abandoned and

allowed to revert to native grass. These areas are easily recognized today

542 The University Science Bulletin

because of the abundance of some species and the scarcity of others. Aristida
oligantha, Artemisia ludoviciana, Bromus japonicus, B. inermis, Digitaria
sangitinalis, Gutierrezia dracunculoidcs, Panicum oligosanthes var. scrib-
nerianum, Vernonia baldwinii and Verbascum thapsus are common in these
abandoned fields. After a long period of succession, scattered clumps of
Andropogon scoparins, Bouteloua curtipendida, Sporobolus heterolepsis be-
come evident and finally Andropogon gerardi and Sorghastrum nutans will
reappear.

RELATIONSHIP OF THE KANSAS FLINT HILLS FLORA TO
THAT OF ADJACENT AREAS

The present Kansas Flint Hills flora consists of 987 species, of which
approximately 18% have been introduced. The families Aizoaceae, Caryo-
phyllaceae, Chenopodiaceae, Cruciferae, Malvaceae, Papaveraceae, Simarou-
baceae and Zygophyllaceae are represented in the area largely and in some
cases totally by introduced species.

The Kansas Flint Hills area is a transition region where eastern and
western species meet and, at the same time, one that has some northern and
southern floristic affinities. Twenty-five per cent of the native Flint Hills
species in one way or another have the limit of their distribution within the
Hills. There are about 150 species of eastern North America, representing a
wide-range of families, that reach the limit of their westward distribution in
Kansas along the western border of the Flint Hills. Examples from this group
of eastern species are Botrychium virginianum, Erythroniutn albidum, Smi-
lax herbacea, Festuca obtusa, F . paradoxa, Asimina triloba, Anemone cana-
densis, Podophyllum peltatum, Tilia americana, Zanthoxylum americanum ,
Dicentra cucullaria, Dentaria lacinata, Silene stellata, Genhana puberulenta,
Ruellia strepens, Agastache nepetoides, Stachys tenuifolia, Agrimonia pu-
bescens, Circaea quadrisulcata, Ceanothus americanus, Rhamnus lanceolata,
Aesculus glabra, Rhus copallina, Corylus americana, Ostrya virginiana, Quer-
cus borealis, Q. marilandica, 0. muehlenbergii, Q. prinoides, 0. velutina,
Eryngium yuccifolium , Osmorhiza longistylis, Poly taenia nuttallii, Campa-
nula americana, Lobelia spicata, Helianthus hirsutus, H. laetiflorus, Aster
drummondii, Solidago ulmifolia, Cacalia atriplicifolia, C. tuberosa and
Prenanthes aspera. There are about 50 western species that reach their east-
ern limit of distribution in Kansas at the eastern escarpment of the Flint
Hills. The following species are examples of this group: Heteranthera dubia,
Commelina erecta, Furiena simplex, Triplasis purpurea, Chloris virgata,
Ceratophyllum demersum , Callirhoe involucrata, Sphaeralcea coccinea, Cro-
ton texensis, Di taxis mercurialina, Euphorbia missurica, Froelichia florid ana,
F. gracilis, Asclepias speciosa, Dalea aurea, D. enneandra, Oxytropis lam-
bertii, Gaura coccinea, Stenosiphon linij alius, Myriophyllum pinnatum, Cur-

The Flora of the Kansas Flint Hills 543

curbita foetidissima, Thelesperma megapotamicum , Pyrrhopappus scaposa
and Hybanthns verticillatus.

In addition to those species that reach their limit of distribution within the
Flint Hill region, there are 100 species that are limited either in their west-
ward or eastern distribution in Kansas by the Flint Hills but do not occur in
the Hills. Fifty eastern species reach their western limits along the eastern
escarpment of the Flint Hills. Examples of these species are Adiantum
pedatum, Cyperus ovularis, Scleria triglomerata, Digitaria ischaemum, Pani-
cum clandestinum , Isopyrum biternatum, Talinum, parviflorum, Hydrophyl-
lum virginianum ,Pycnccnthemum flexuosum , Tephrosia virginiana, Coreopsis
palmata, Ratibida pinnata and Grindelia lanceolata. Even though the west-
ern border of the Flint Hills is relatively indistinct, another 50 species seem to
be limited in their eastern distribution at the western border of the area. Some
of these species are Zannichellia pulustris, Carex grisea, Cyperus schweinitzii ,
C. fdiculmis, Scirpus paliidosus, Aristida longiseta, Distichlis stricta,
Eragrostis oxylepis, Muhlenbergia aspen folia, Panic um obtusum, Erysium
asperum. Paronychia jamesii, Talinum calycinum. Poly gala alba, Asclepias
pumila, Petalostemon villosum, Hymenopappus tenuijolius, Aster exilis, A.
jendleri and Haplopappus spinulosus.

There are about 50 species whose range, within the Flint Hills, is restricted
to the southern portion, and another group of about 25 species that are re-
stricted to the northern counties. Examples of the northern group are Carex
emoryi, lodanthus pinnatifidus, Dentaria lacinata. Anemone virginiana,
Arisaema triphyllum , Corylus americana, Hystrix patula, Cyperus diandrus,
Clematis virginiana, Mimulus ringens, Impatiens biflora, Viola sororia, Quer-
cus borealis, Juniperus virginiana and Parthenocissus vitacea. The following
species are examples of the group restricted to the southern area: Camassia
angusta, Nemastylis geminiflora, Luzula bulbosa, Acacia angustissma, Mo-
narda citriodora, Cocculus carolinus, Viburnum, rufidulum, Bumelia lanugi-
nosa, Carya illinoensis, Penstemon digitalis, Celtis teniti flora, Ptelea trijoliata,
Rhus copallina, Quercus palustris, Verbesina virginica and Parthenium
hispid um.

While there are no species endemic to the Kansas Flint Hills, a few have
some interesting relationships. Phlox ol(lahomensis is known only from its
type locality in Woodward County, Oklahoma and from Butler, Chautauqua,
Cowley and Elk counties in the southern Flint Hills. Thus this phlox is
largely restricted to the Flint Hills, except for the disjunct colony found on
gypsiferous soils in Oklahoma.

Parthenium hispidum is relatively common in the southern Flint Hills,
extends southward in Oklahoma, and is found again on dry hillsides in the
Missouri Ozarks. A similar distribution is found with Tradescantia tharpii
though it extends to the northern part of the Hills and south into Texas.

544 The University Science Bulletin

Aster sericeus is a common aster in the Flint Hills but becomes infrequent
northward to North Dakota and Michigan. In northeast Kansas and north-
west Missouri, it is found in dry hilltops of the Missouri River Bluffs. In the
Missouri Ozarks and the Cedar Barrens of Tennessee, the species is found in
bald knobs and rocky barrens. It is also known from similar areas in North
Carolina and Georgia.

Stipa spartea is a northern grass common in South Dakota and eastern
Nebraska, and is scattered in Iowa and northern Missouri. It is found in all
the northeastern counties of Kansas, but it extends southward through the
Flint Hills to Oklahoma.

In comparing the 987 species found in the Flint Hills with those found in
other physiographic areas of Kansas, some interesting relationships are re-
vealed. Work done by staff and students at the University of Kansas (unpub-
lished) indicates the small Ozarkian section in extreme southeast Kansas has
approximately 1300 species of vascular plants, of which at least 125 are known
nowhere else in the state. By contrast, the flora of Morton County in extreme
southwest Kansas was found by Richards (unpublished) to consist of 458
species. Morley (1964) reported 594 species for Republic County in north-
central Kansas and Birkholz (1968) lists 560 species for Sumner County in
south-central Kansas. Records accumulated by McGregor (unpublished) for
Douglas County in northeastern Kansas indicates approximately 1100 species
for that county. The flora of the Kansas Flint Hills includes approximately
one half of the total number of vascular plants known to occur in Kansas.

ANNOTATED LIST

The following list includes all the native or naturalized vascular plants
found in the Kansas Flint Hills. Cultivated plants have been omitted. All of
the plants listed have been collected from the field, represented by voucher
specimens in the University of Kansas Herbarium or included on the
authority of Gates (1940) .

The frequency of each species, the habitat of each and the counties in
which they were found are given. For frequency, the terms abundant, com-
mon, occasional and rare have been used. A species is considered to be
abundant if large numbers of individual plants can be found with ease, com-
mon if numerous individuals are found fairly often, occasional if a few scat-
tered plants are found and, finally, rare if the species is seldom found or is
known from only a single or a few records. If the plant is found in the
northern counties, i.e., Marshall, Pottawatomie, Riley, Wabaunsee and Geary
counties, this is designated by the statement, "found in the northern part of
the area." On the other hand, if the plant is found only in the southern
counties, i.e., Cowley, Butler, Chase, Lyon and Marion counties, the statement

The Flora of the Kansas Flint Hills 545

"found in the southern part of the area" is used. When the plant is distributed
throughout the area the designation "found throughout the area" is used.

The families are arranged according to the traditional familiar sequence
of Engler and Prantl. The nomenclature employed follows that of Gleason
and Cronquist (1963) except when the species is outside their range and
where more recent opinions are given in current monographs and revisions.
The genera and species are arranged alphabetically within the families.

EQUISETACEAE

Equisetum arvense L. Rare, on moist streambanks and in low moist prairies. Known from the

northern part.
Equisetum hyemcde L. Occasional, usually found along streambanks and sometimes in moist

prairies. Found in the northern part.
Equisetum laevigatum A. Br. Occasional, in low moist prairies and along drainage and seepage

areas on prairie slopes. Encountered throughout the area.

OPHIOGLOSSACEAE
Botrychium virginianum (L.) Sw. Occasional, in rich wooded canyons and stream valley woods.

Found throughout the area.
Ophioglossum engelmannii Prantl. Rare, in prairie areas on shallow soil above limestone outcrops

and on open wooded slopes. Known from Cowley and Marion counties.

POLYPODIACEAE

Camptosorus rhizophyllus (L.) Link. Rare, grows on calcareous soil and in crevices of shaded

limestone escarpments. Found in Cowlcv County and reported for Riley County by Gates

(1940).
Cystopteris fragilis (L.) Rernh, var. fragilis. Rare, grows in crevices of limestone escarpments in

moist wooded areas. Found in Riley and Cowley counties.
Cystopteris tentiessensis Shaver. Rare, grows in the crevices of limestone escarpments in moist

wooded areas. Found in Cowley and Marshall counties.
Notholaena dealbata (Pursh.) Kunze. Rare, grows in the crevices of limestone escarpments in

moist wooded areas. Found in Chase, Cowley and Riley counties.
Pellaea atropurpwea (L.) Link. Occasional to locally abundant, grows in crevices of limestone

rocks. Known from Cowley, Riley and Wabaunsee counties.
Pellaea glabella Mett. Rare, grows in the crevices of limestone escarpments in moist wooded

areas. Known from Cowley, Geary, Riley and Wabaunsee counties.
Thelypteris palustris Schott. Rare, in marshes and floodplain woods. Reported for Pottawatomie

County by Gates (1940).
Woodsia obtusa (Spreng.) Torr. Occasional, in moist canyon woods. Found in Cowley, Morris,

Pottawatomie and Riley counties.

MARSILEACEAE
Marsilea mucronata A. Br. Rare, grows in roadside ditches and along the edge of ponds and
marshes. Known from Butler, Chase, Cowley and Pottawatomie counties.

CUPRESSACEAE
Jitniperus virginiana L. Common to abundant, in overgrazed upland prairies, on prairie slopes
and in wooded canyons. Often forming dense stands in the northern part to occurring as
scattered individuals in the central and southern parts of the Flint Hills.

TYPHACEAE
Typ/ia latijolia L. Common, in wet places, along streams, impoundments, drainage areas, seep-
age areas and roadside ditches. Occurs throughout the area.

SPARGANIACEAE
Sparganium eurycarpum Engelm. Occasional, along the margins of streams, impoundments and
other low wet places. Occurs in Riley and Wabaunsee counties.

546 The University Science Bulletin

NAJADACEAE
Potamogeton diversijolius Raf. Occasional, in impoundments. Found in Cowley County.
Potamogeton obtusifolius Mert. & Kech. Occasional, in impoundments. Reported for Riley

County by Gates (1940).
Potamogeton pectinattis L. Occasional, in impoundments. Reported for Rilev Countv bv Gates

(1940).

ALISMATACEAE

Alisma triviale Pursh. Rare, grows in wet roadside ditches, along streambeds and margins of

impoundments. Known from Chase, Geary, Morris, Pottawatomie and Riley counties.
Eclnnodorus cordifolius (L.) Griseb. Rare, along muddy margins of impoundments. Found in

Lyon and Riley counties.
Sagittaria ambigua J. G. Sin. Rare, growing in shallow water along the margins of ponds and

lakes. Known from Cowley and Wabaunsee counties.
Sagittaria cuneata Sheldon. Rare, grows in shallow water along margins of ponds and lakes.

Reported for Lyon County by Gates (1940).
Sagittaria latijolia Willd. Occasional, grows in shallow water along the margins of ponds and

lakes. Found throughout the area.
Sagittaria montevidensis Cham. & Schlecht. Rare, grows in shallow water along the margins of

ponds and lakes. Reported for Riley County by Gates (1940).

GRAMINEAE
Aegilops cylindrica Host. Common, along the roadsides. Found throughout the area.
Agropyron smithii Rydb. Common, in moist prairie swales and roadside ditches. Found through-
out the area.
Agrostis alba L. Common, in roadside ditches and overgrazed prairies. Found in the northern

part.
Agrostis hyemalis (Walt.) BSP. Occasional, in roadside ditches. Found throughout the area.
Alopecurus carolinianus Walt. Occasional, in moist prairies. Found in Cowley, Pottawatomie

and Riley counties.
Alopecurus myosuroides Huds. Rare, in waste places and railway banks. Reported for Riley

County by Gates (1940).
Alopecurus pratensis L. Rare, in meadows and waste places. Reported for Riley County bv

Gates (1940).
Andropogon gerardi Vitman. Abundant, in well-managed upland prairies, on north-facing

prairie slopes, in prairie canyons and open lowland woods. Found throughout the area.
Andropogon ischaemum L. Occasional, introduced along prairie roadsides and in overgrazed

prairies. Known from Wabaunsee County.
Andropogon saccharoides Swartz. Occasional to common, along roadsides and sometimes in

overgrazed prairies. Common in the southern part becoming occasional in the northern part

of the regifin.
Andropogon scoparius Michx. Abundant in well-drained upland prairie and on well-drained

and south-facing prairie slopes. Known throughout the area.
Andropogon virginicus L. Occasional, in overgrazed prairies. Reported for Lyon and Riley

counties by Gates (1940).
Aristida basiramea Engelm. Rare, in dry overgrazed prairies. Reported for Pottawatomie and

Riley counties by Gates (1940).
Aristida longiseta Steud. Rare, in dry overgrazed prairies. Reported for Riley Countv by Gates

(1940).
Aristida oligantha Michx. Common, on clay-pan areas and in overgrazed prairies. Found

throughout the area.
Aristida purpurascens Poir. Rare, in dry overgrazed prairie. Reported for Rilev Countv by Gates

(1940).
Arena sativa L. Occasional, escaped or introduced along roadsides and in cultivated fields.

Known from Cowley County.
Bouteloua curtipendula Michx. Abundant, in upland prairies and on prairie slopes. Found

throughout the Flint Hills.
Bouteloua gracilis (HBK.) Lag. Occasional, in overgrazed prairies and along roadsides. Known

throughout the area.
Bouteloua lursuta Lag. Occasional to common, in overgrazed prairies. Known throughout the

area.
Bromus catharticus Vahl. Rare, in overgrazed prairies and waste places. Reported for Riley

County by Gates (1940).

The Flora of the Kansas Flint Hills 547

Bromns inermis Leyss. Common, introduced in roadside ditches, overgrazed prairies and culti-
vated for hay. Known throughout the area.
Bromns japonictis Thunb. Common, along roadsides and in overgrazed prairies. Known

throughout the area.
Bromtis mollis L. Rare, along roadsides and in waste places. Reported for Riley County by

Gates (1940).
Bromtis purgans L. Occasional, in moist rocky woods. Found in the northern part.
Bromtis secalinus L. Occasional, in overgrazed prairies and waste places. Known throughout

the area.
Bromtis tectortim L. Common, in roadside ditches, in cultivated and abandoned fields. Found

throughout the region.
Buchloe datyloides (Nutt.) Engelm. Common to locally abundant, in overgrazed upland prairies.

Found throughout the area.
Calamovilfa gigantea (Nutt.) Lamson-Scribner. Rare, on sand dunes along the Kansas River.

Reported for Riley County by Gates (1940).
Calamovilfa longijolia (Hook.) Scribner. Rare, on sandy areas along streambanks. Reported for

Geary, Pottawatomie and Riley counties by Gates (1940).
Cenchrtis longispinns (Hack.) Fern. Common, in roadside ditches, waste places and severely

overgrazed prairies. Found throughout the area.
Chloris vcrticillata Nutt. Common, in roadside ditches, waste places and overgrazed prairies.

Occurs throughout the area.
Chloris virgata Sw. Occasional, along roadsides and in overgrazed prairies. Found in the south-
ern part.
Cinna arundinacea L. Occasional, in moist lowland woods. Reported tor Pottawatomie and

Riley counties by Gates (1940).
Cynodon dactylon (L.) Pers. Occasional, cultivated in lawns and occasionally escaped to road-
sides and overgrazed prairies. Found throughout the area.
Dactylis glomerata L. Occasional, in cultivated fields, meadows, and waste places. Found in the

northern part.
Danthonia spicata (L.) Beauv. Rare, grows in rocky soil along limestone escarpments. Known

from Marshall County.
Diarrhena americana Beauv. Occasional, in rich stream valley woods. Found throughout the

area.
Digitaria filiformis (L.) Kocl. Rare, on sandy soil along streambeds. Reported for Pottawatomie

County by Gates (1940).
Digitaria ischaemum (Schreb.) Muhl. Occasional, in cultivated fields and waste places. Reported

for Pottawatomie and Riley counties by Gates (1940).
Digitaria sangtiinalis (L.) Scop. Abundant, along roadsides, in cultivated fields and on sandy

streambanks. Found throughout the area.
Distichlis stricta (Torr.) Rydb. Rare, in waste places with alkaline soils. Known throughout the

area.
Echinochloa crusgalli (L.) Beauv. Abundant, along roadsides, streambanks. ponds, reservoirs

and in cultivated fields. Found throughout the area.
Eleusine indica (L.) Gaertn. Common, along roadsides, in cultivated ground and other waste

places. Encountered throughout the area.
Elymtis canadensis L. Common, in roadside ditches, overgrazed prairies, prairie canyons and

along streambanks. Known throughout the area.
Elymtis villostis Muhl. Occasional, in flood plain woods and on streambanks. Found in the

northern part.
Elymtis t'irginictis L. Common, in moist lowland prairies, canyon woods along streambanks.

Encountered throughout the area.
Eragrostis capillaris (L.) Nees. Occasional, in dry overgrazed prairies, open woods and waste

places. Reported for Morris, Riley and Wabaunsee counties by Gates (1940).
Eragrostis cilianensis (All.) Link. Abundant, along roadsides, in cultivated fields and other

waste places. Known throughout the area.
Eragrostis hypnoides (Lam.) BSP. Occasional, dense mats are found growing along moist

streambanks and margins of impoundments. Known from Riley and Pottawatomie counties.
Eragrostis oxylepis (Torr.) Torr. Rare, occurs in dry sandy prairies. Reported for Riley and

Pottawatomie counties by Gates (1940).
Eragrostis pectinacea (Michx.) Nees. Occasional, along impoundments, streambanks, in culti-
vated fields and waste places. Found throughout the area.
Eragrostis pilosa (L.) Beauv. Rare, in cultivated fields and waste places. Reported for Chase

County by Gates (1940).

548 The University Science Bulletin

Eragrostis poaeoides Beauv. Rare, cultivated fields and waste places. Reported for Gearv County

by Gates (1940).
Eragrostis spectabilis (Pursh.) Steud. Common, along roadsides and in overgrazed prairies.

Known throughout the region.
Eragrostis trichodes (Nutt.) Wood. Occasional, in sandy soil along rivers and streambeds. Re-
ported for Geary, Riley and Pottawatomie counties by Gates (1940).
Eriochloa contracta Hitchc. Common, in roadside ditches, cultivated fields and low wet places.

Encountered throughout the area.
Festttca elatior L. Occasional, in prairie hay meadows, open lowland woods and in roadside

ditches. Reported for Marshall and Riley counties by Gates (1940).
Festttca obtusa Biehler. Occasional, in floodplain woods. Found in the northern part.
Festuca octoflora Walt. Common, in overgrazed upland prairies and on prairie slopes. Found

throughout the area.
Festttca paradoxa Desv. Common, in roadside ditches, low moist prairies and open lowland

woods. Known from the northern part.
Glyceria striata (Lam.) Hitchc. Occasional, in low wet areas along impoundments and wet low-
land prairies. Found in the northern part.
Hordeam jubatum L. Common, along roadsides and in overgrazed upland prairies. Known in

the northern part.
Hordeum pusillam Nutt. Common, in overgrazed upland prairies and on prairie slopes. En-
countered throughout the region.
Koeleria cristata (L.) Pers. Common, in well-managed upland prairies. Found throughout the

area.
Leersia oryzoides (L.) Sw. Rare, in wet places along streams and impoundments. Reported for

Pottawatomie and Riley counties by Gates (1940).
Leersia virginica Willd. Common, in wet places along streams and impoundments. Found

throughout the region.
Leptochloa fascicalaris (Lam.) Gray. Occasional, along streambeds, drainage areas and margins

of impoundments. Known from the northern and southern part of the area.
Leptoloma cognation (Schultes) Chase. Common, in rocky roadside ditches and overgrazed

prairies. Found throughout the area.
Lohitm perenne L. Rare, introduced into cultivated areas and waste places. Reported for Marion

and Riley counties by Gates (1940).
Loliutn temulentum L. Rare, introduced into cultivated areas and waste places. Reported for

Riley County by Gates (1940).
Melica nitens Nutt. Rare, in rocky soil of wooded canyons. Known from Riley and Geary

counties.
Mn/ilenbergia brachyphylla Bush. Common, in floodplain woods and along streambanks. Found

throughout the area.
Mn/ilcnbcrgia cttspidata (Torr.) Rydb. Common, on prairie slopes and in overgrazed upland

prairies. Known from throughout the area.
Mahlenbergia frondosa (Poir.) Fern. Rare, in low ground along roadsides, fence rows and other

waste places. Collected once in Marshall County.
Mahlenbergia mexicana (L.) Trin. Common, in floodplain woods and along streambanks. Found

in the northern part.
Mahlenbergia racemosa (Michx.) BSP. Common, in moist prairie canyons and open floodplain

woods. Known from the northern part.
Mahlenbergia schreberi Gmcl. Common, in floodplain woods. Known from Chase, Pottawatomie

and Riley counties.
Mahlenbergia sobolifera (Muhl.) Trin. Rare, in wooded canyons on rocky soil. Reported from

Geary and Riley counties by Gates (1940).
Mahlenbergia sylratica Torr. Occasional, in floodplain woods and along streambanks. Known

from Geary, Pottawatomie and Riley counties.
Panicam capillarc L. Abundant, along roadsides, in cultivated fields, waste places and severely

overgrazed prairies. Found throughout the area.
Panicam clandestinam L. Rare, low moist wooded areas. Reported for Pottawatomie County by

Gates (1940).
Panicam dichotomifloram Michx. Common, along roadsides, in cultivated fields anil in over-
grazed prairies. Found throughout the area.
Panicam lanaginosum Ell. var. jascicalatam (Torr.) Fern. Occasional, in well-managed upland

prairies and in open wooded areas. Known from Cowley, Pottawatomie and Riley counties.
Panicam oligosanthes Schult. var. scribnerianum (Nash.) Fern. Common, in overgrazed upland

prairies, on prairie slopes and in open wooded canyons. Found throughout the area.

The Flora of the Kansas Flint Hills 549

Panicttm perlongum Nash. Common, in well-managed upland prairies. Known from Geary,

Morris and Riley counties.
Panicttm praecocitts Hitchc. & Chase. Rare, in overgrazed upland prairies. Reported from Geary,

Pottawatomie and Riley counties by Gates (1940).
Panicttm sphaerocarpon Ell. Occasional, in well-managed upland prairies. Known from Marshall

and Cowley counties.
Panicttm virgatum L. Abundant, in upland prairies, prairie canyons and often used in seed

mixtures to reseed abandoned fields and overgrazed prairies. Known throughout the area.
Panicttm wilcoxianum Vasey. Rare, in well-managed upland prairies. Reported for Pottawatomie

and Riley counties by Gates (1940).
Paspalttm ciliatifo/itim Michx. Common, along roadsides and in overgrazed upland prairies.

Known from throughout the region.
Paspalttm pubiflorum Rupr. Common, along roadsides and in overgrazed prairies. Found in the

southern part.
Phalaris arttndinacea L. Rare, along streambanks and in low wet places. Reported for Potta-
watomie County by Gates (1940).
Phalaris canadensis L. Rare, introduced overgrazed prairies and in waste places. Known from

Butler, Lyon and Morris counties.
Phleum pratense L. Common, introduced along roadsides, in prairie hay meadows, overgrazed

prairies and waste places. Found throughout the area.
Phragmites communis Trin. Rare, grows in large colonies in swamps, marshes and at the edge

of impoundments. Reported from Pottawatomie County by Gates (1940).
Poa annua L. Rare, in lawns, waste places and cultivated ground. Reported for Pottawatomie

County by Gates (1940).
Poa pratensis L. Common, in overgrazed upland prairies, prairie hay meadows, on prairie slopes

and in open wooded canyons. Known throughout the area.
Poa sylvestris Gray. Occasional, in floodplain woods and sometimes in the drier wooded canyons.

Known from Riley and Pottawatomie counties.
Polypogon monspeliensis (L.) Desf. Occasional, grows in sandy soil along Arkansas River in

Cowley County. Known from Cowley County.
Schedonnardus panictdatus (Nutt.) Trel. Common, in roadside ditches, cultivated fields and

severely overgrazed prairies. Found throughout the area.
Setaria genicttlata (Lam.) Beauv. Common, along roadsides, in cultivated fields, waste places

and severely overgrazed prairies. Known throughout the area.
Setaria jabertt Herrm. Common, in roadside ditches, cultivated fields, waste places and in open

floodplain woods. Found throughout the region.
Setaria ttalica (L.) Beauv. Rare, sometimes cultivated and escaping along roadsides. Reported

for Cowley, Pottawatomie, Riley and Wabaunsee counties by Gates (1940).
Setaria lutescens (Weigel) Hubb. Common, along roadsides, in cultivated fields, waste places

and overgrazed prairies. Found throughout the area.
Setaria verttctllata (L.) Beauv. Rare, in waste places and around dwellings. Reported for Riley

County by Gates (1940).
Setaria virtdts (L.) Beauv. Common, along roadsides, in cultivated fields, waste places, over-
grazed prairies and open wooded canyons. Found throughout the area.
Sorg/iastrttm nutans (L.) Nash. Abundant, in well-managed upland prairies, on moist north-
facing prairie slopes, in prairie canyons and in open wooded canyons. Found throughout

the area.
Sorghum halepense (L.) Pers. Common, in roadside ditches, in cultivated fields and other waste

places. Occurs throughout the region.
Sparttna pectinata Link. Common, in wet roadside ditches, along drainage areas and along the

margins of impoundments. Encountered throughout the area.
Sphenopholis obtttsata (Michx.) Scribn. Occasional, in well-managed upland prairies, prairie hay

meadows and sometimes in overgrazed prairies. Encountered more often in the northern

part becoming very rare in the southern part.
Sporoboltts asper (Michx.) Kunth. Common, on prairie slopes and in open wooded canyons.

Occurs throughout the area.
Sporoboltts clandestintts (Biehler) Hitchc. Rare, in sandy fields, prairies. Reported for Potta-
watomie and Riley counties by Gates (1940).
Sporoboltts cryptandrtts (Torr.) Gray. Common, on roadside banks, in overgrazed upland prairies

and on prairie slopes. Found throughout the area.
Sporoboltts heterolepis Gray. Common, in well-managed upland prairies, prairie canyons and

wooded canyons. Found throughout the area but much more common in the northern part.

550 The University Science Bulletin

Sporobohis neglectus Nash. Occasional, on roadside banks and on prairie slopes. Encountered

throughout the area.
Sporobohis raginiflorns (Torr.) Wood. Rare, on roadside banks and on shallow soil along

limestone escarpments. Known from Butler, Morris, Pottawatomie and Riley counties.
St/pa spartea Trin. Occasional, grows in small patches in well-managed upland prairies and in

prairie hay meadows. Occurs in the northern and southern counties of the Flint Hills but

not in the central counties.
Triodia flava (L.) Smyth. Common, along roadsides, in waste places, overgrazed prairies and

open wooded areas. Found throughout the area.
Tripiasis purpurea (Walt.) Chapm. Rare, occurs in sandy places. Reported for Riley County by

Gates (1940).
Tripsaattn dactyloides L. Common, well-managed upland prairies, prairie hay meadows, prairie

canyons and open lowland woods. Found throughout the area.
Trisetum flarescens (L.) Beauv. Rare, occurs in waste places. Reported for Riley County by

Gates (1940).
Uniola latijolia Michx. Occasional, in floodplain woods. Known Cowley, Geary, Pottawatomie,

Riley and Wabaunsee counties.

CYPERACEAE
Bidbostylis capillaris (L.) C. B. Clarke. Rare, in moist prairie swales and in sandy soils along

streams. Reported for Pottawatomie County by Gates (1940).
Carex aggregata Mackenzie. Occasional, on open wooded hillsides and in open wooded canyons.

Found only in the northern part.
Carex amphibola Steud. Common, in rich woods and on wooded north-facing slopes. Occurs

throughout the area.
Carex annectans Bickn. var. annectans. Common, in wet areas of upland prairies and in prairie

canyons. Known from Lyon County.
Carex annectans Bickn. var. xanthocarpa (Bickn.) Wieg. Occasional, in wet areas of upland

prairies and in prairie canyons. Known from Lyon County.
Carex artitecta Mackenzie. Rare, on wooded canyon slopes. Reported from Pottawatomie and

Riley counties by Gates (1940).
Carex bic\neUii Britt. Occasional, in upland prairie and in prairie canyons. Known from Lyon,

Riley and Wabaunsee counties.
Carex blanda Dewey. Common, in wooded areas. Occurs throughout the area.
Carex brevior (Dewey) Mackenzie. Common, in upland prairies, on prairie slopes and in prairie

canyons. Found throughout the area.
Carex bushii Mackenzie. Common, in upland prairies, in prairie canyons, and sometimes found

on open wooded slopes. Known from Butler, Chase, Lyon and Wabaunsee counties.
Carex cephalophora Muhl. var. cephalophora. Rare, on open wooded slopes, and sometimes

found in open lowland woods. Known from Marshall and Riley counties.
Carex crawei Dewey. Occasional, in wet places in upland prairies and along the margins of

impoundments. Found in Cowley County.
Carex crits-corvi Shuttlew. Rare, in low wet places and in floodplain woods. Known from Lyon

and Wabaunsee counties.
Carex davisii Schwein. & Torr. Rare, in floodplain woods. Reported for Rilev Countv by Gates

(1940).
Carex eleocharis Bailey. Rare, in upland prairie. Reported for Riley County by Gates (1940).
Carex emoryi Dewey. Occasional, on streambanks and in floodplain woods. Found in the

northern part.
Carex jran\ii Kunth. Rare, along wet places in upland prairies, on streambanks and along

margins of impoundments. Known from Chase, Lyon and Morris counties.
Carex gravida Bailey. Common, in upland prairies, on prairie slopes, in prairie canyons and less

commonly in open wooded areas. Found throughout the area.
Carex heliophila Mackenzie. Occasional, on dry wooded slopes. Reported for Pottawatomie,

Riley and Wabaunsee counties by Gates (1940).
Carex hyalinolepsis Steud. Rare, in wet places along streams, impoundments and roadside ditches.

Found in Cowley and Pottawatomie counties.
Carex hystricina Muhl. Common, in wet areas along streams, impoundments and seepage areas.

Found in the northern part.
Carex interior Bailey. Rare, in swampy prairie areas and along seepage areas along limestone

escarpments. Reported from Riley County by Gates (1940).
Carex laeviconica Dewey. Rare, in wet soil of floodplain woods, along streams and in low wet

prairies. Known from Pottawatomie and Riley counties.

The Flora of the Kansas Flint Hills 551

Carex lanuginosa Michx. Common, in wet swales in upland prairies, along seepage areas and
in moist prairie canyons. Found in the northern part.

Carex meadii Dewey. Common, in upland prairies, prairie hay meadows and prairie canyons.
Found throughout the area.

Carex molesta Mackenzie. Occasional, in wet places, in moist upland prairies, on moist open
wooded slopes and in prairie canyons. Found in Chase, Cowley and Marshall counties.

Carex muhlenbergia Schkuhr. var. australis Olney. Common, in upland prairies, and on open
wooded slopes. Encountered throughout the area.

Carex normalis Mackenzie. Common, in wet places, along drainage areas, in moist floodplain
woods and along streambanks. Found throughout the area.

Carex oligocarpa Schkuhr. Rare, in floodplain woods and along streambanks. Found in Wa-
baunsee County.

Carex praegracilis W. Boot. Rare, in upland prairies. Reported for Wabaunsee County by Gates

(1940).

Carex stipata Muhl. Rare, along streambanks, in floodplain woods and in wet lowland prairies.
Reported for Riley County by Gates (1940).

Carex vulpinoidea Michx. Common, in wet places, in upland prairies, along margins of im-
poundments, along streambanks and in floodplain woods. Found throughout the area.

Cyperus acuminatus Torr. & Hook. Common, along wet places, along streambanks, margins of
impoundments, and in wet cultivated ground. Found throughout the area.

Cyperus aristatus Rottb. Common, along wet places in prairies, margins of impoundments,
streams and in low wet cultivated ground. Occurs throughout the area.

Cyperus diandrus Torr. Rare, in wet places along limestone escarpments, along margins of im-
poundments and streambeds. Known from Pottawatomie and Riley counties.

Cyperus erythrorhizos Muhl. Occasional, in wet places, along margins of impoundments and
streambeds. Found in Butler, Lyon and Riley counties.

Cyperus esculentus L. Common, in wet places, in prairie swales, along margins of impound-
ments, streams and in low moist cultivated ground. Found in the northern part.

Cyperus jerruginescens Boeckl. Common, in wet places, in prairie swales, along margins of
impoundments and streambeds. Found throughout the area.

Cyperus filiculmis Vahl. Common, in upland prairies, on gentle-sloping prairie hillsides, in open
woods on prairie slopes. Occurs throughout the area.

Cyperus oiularis (Michx.) Torr. Rare, in prairies, in dry open wooded slopes and along streams.
Found in Cowley County.

Cyperus schweinitzii Torr. Rare, sandy low ground along streams and ponds. Reported for
Riley County by Gates (1940).

Cyperus setigerus Torr. & Hook. Common, in wet places, along the margins of impoundments
and streams. Found in the southern part.

Cyperus strigosus L. Common, in wet places, along the margins of impoundments and streams.
Found in the northern part.

Eleocharis calva Torr. Common, along margins of impoundments, streams, sloughs and in wet
prairies. Found in Marshall, Pottawatomie and Wabaunsee counties.

Eleocharis compressa Sulliv. Common, in wet places, in prairie drainage areas, roadside ditches
and along margins of impoundments. Occurs throughout the area.

Eleocharis macrostachya Britt. Common, along margins of impoundments, streams, sloughs and
in wet prairies. Found in the northern part.

Eleocharis montevidensis Kunth. Rare, in wet sandy places along streams and impoundments.
Known from Cowley County.

Eleocharis obtusa (Willd.) Schultes. Occasional, in muddy areas along streams and impound-
ments. Reported for Morris and Pottawatomie counties by Gates (1940).

Finibristylis puberula (Michx.) Vahl. Common, in upland prairies, prairie hay meadows, prairie
canyons and sometimes in open wooded areas. Collected from Butler, Cowley, Pottawatomie
and Riley counties.

Fuirena simplex Vahl. Rare, grows in wet soil along springs, seepage areas and streams. Found
in Cowley, Pottawatomie and Riley counties.

Hemicarpha micrantha (Vahl.) Pax. Rare, grows in sandy soil. Reported for Pottawatomie
County by Gates (1940).

Scirpus americanus Pers. Rare, grows in sandy soil along the Kansas River. Reported for Riley
County by Gates (1940).

Scirpus atrovirens Willd. Common, grows in wet roadside ditches, along margins of impound-
ments and open streambanks. Found throughout the area.

Scirpus lineatus Michx. Common, in wet soil of prairie drainage areas, roadside ditches, margins
of impoundments and along streambanks. Encountered throughout the area.

552 The University Science Bulletin

Scirpus validus Vahl. Common, in wet soil of roadside ditches and along the margins of im-
poundments. Found throughout the area.

ARACEAE

Acorns calamus L. Rare, sometimes introduced in swampy areas along margins of ponds and
streams. Reported for Pottawatomie and Riley counties by Gates (1940)

Arisacma dracontium (L.) Schott. Rare, in moist stream valley woods. Encountered throughout
the area.

Arisaema triphyllum (L.) Schott. Rare, in moist stream valley woods. Reported for Potta-
watomie County by Gates (1940).

LEMNACEAE
Lemna minor L. Occasional, in impoundments and slow-moving streams. Reported for Geary

County by Gates (1940).
Lemna perpusilla Torr. Occasional, in impoundments and slow-moving streams. Reported for

Riley County by Gates (1940).
Spirodela polyrhiza (L.) Schleid. Occasional, in impoundments and slow-moving streams.

Known from Cowley, Pottawatomie and Riley counties.
Wolffia cohtmbiana Karst. Occasional, in impoundments. Reported for Geary County by Gates

(1940).

COMMELINACEAE
Commelina erecta L. Rare, in stream valley woods and along streambanks. Found in Potta-
watomie and Riley counties.
Commelina virginica L. Common, in wooded canyons, floodplain woods and in thickets along

the roadsides. Found throughout the area.
Tradescantia bracteata Small. Common, in overgrazed upland prairies, prairie hay meadows,

along roadsides and railway right-of-ways. Occurs throughout the area.
Tradescantia ohiensis Raf. Common, in overgrazed upland prairies, prairie hay meadows, along

the roadsides, railway right-of-ways and sometimes in open wooded areas. Encountered

throughout the area.
Tradescantia tharpii Anderson & Woodson. Common, in overgrazed upland prairies, on gentle

prairie slopes and in prairie canyons. Found throughout the Flint Hills.

PONTEDERIACEAE

Heteranthera ditbia (Jacq.) MacM. Occasional, grows at the edge of streams or impoundments,

either in shallow water or on the mud. Reported for Riley County by Gates (1940).
Heteranthera limosa (Sw.) Willd. Occasional, grows at the edge of streams or impoundments,

either in shallow water or on the mud. Reported for Morris and Riley counties by Gates

(1940).
Heteranthera renijormis R. & P. Occasional, grows at the edge of streams or impoundments,

either in shallow water or on the mud. Reported for Pottawatomie and Riley counties by

Gates (1940).

JUNCACEAE
Juncus diffusissimus Buckl. Rare, grows in wet soil in roadside ditches, seepage areas, along

margins of impoundments and streams. Known from Pottawatomie County.
Juncus dudleyi Wieg. Common, in wet prairies, along seepage areas, drainage areas, margins of

impoundments and streams. Encountered throughout the area.
Juncus interior Wieg. Common, in moist prairies, along seepage areas, drainage areas, margins

of impoundments and streams. Occurs throughout the area.
Juncus marginatus Lam. Rare, in moist prairies and prairie hay meadows. Found in Cowley

and Pottawatomie counties.
Juncus torreyi Coville. Common, in wet prairies, along seepage areas, drainage areas, margins

of impoundments and streams. Known from throughout the area.
Luzula bulbosa (Wood) Rydb. Rare, in open wooded areas, on rocky prairie slopes and in over-
grazed upland prairies. Known from Cowley County.

LILIACEAE
Allium canadense L. var. canadense. Occasional, in open floodplain woods, prairie canyons,

prairie hay meadows, upland prairies, along roadsides and railway right-of-ways. Found in

Butler and Cowley counties.
Allium canadense L. var. jraseri. Owenby. Common, in open floodplain woods, prairie canyons,

prairie hay meadows, upland prairies, along roadsides and railway right-of-ways. Found

throughout the area.

The Flora of the Kansas Flint Hills 553

Allium canadense L. var. lavendulare (Bates) Owenby & Ashe. Occasional, in open floodplain

woods, prairie canyons, prairie hay meadows, upland prairies, along roadsides and railroad

right-of-ways. Found in the southern part.
Allium stcllatum Ker. Common, in rocky upland prairies, and on rocky prairie slopes. Found in

the northern part of the area.
Androstephium coeritlem (Scheele) Greene. Occasional, in rocky upland prairies and on rocky,

prairie slopes. Found in Butler, Cowley, Geary and Marion counties.
Asparagus officinalis L. Common, escapes from cultivation into roadside ditches, cultivated

fields and other such waste places. Occurs throughout the area.
Camassia angusta (Engelm. & Gray) Blankinship. Occasional, in rocky upland prairies and on

rocky prairie slopes. Occurs in Butler and Cowley counties.
Camassia scilloides (Raf.) Cory. Occasional, in open wooded areas, prairie canyons, prairie hay

meadows and well-managed upland prairies. Known from Cowley County.
Erythronium albidum Nutt. Common, in floodplain woods and wooded canyons. Found in the

northern part of the area and in Cowley County.
Erythronium mesochoreum Knerr. Common, in rocky upland prairies, on rocky prairie slopes

and in open woods on prairie slopes. Found in Butler, Geary, Lyon, Morris and Wabaunsee

counties.
Not/ioscordum hi vale (L.) Britt. Common, in upland prairies, on prairie slopes and in open

wooded areas on prairie slopes. Found throughout the area.
Polygonatum biflorum (Walt.) Ell. Common, in floodplain woods and on wooded slopes. Found

throughout the area.
Smilacina racemosa (L.) Desf. Rare, in rocky wooded areas. Reported for Riley County by

Gates (1940).
Smilacina stellata (L.) Desf. Rare, in moist floodplain woods. Reported for Riley and Potta-
watomie counties by Gates (1940).
Smilax hcrbacea L. var. lasioneuron (Small) Rydb. Common, in stream valley woods and

wooded canyons. Found throughout the area.
Smilax hispida Muhl. Common, in floodplain woods and wooded canyons. Occurs throughout

the area.
Yucca glauca Nutt. Occasional, in rocky upland prairies, on rocky prairie slopes and on road-
cuts. Occurs throughout the area.
Zygadcnus nuttallii Gray. Occasional, in moist upland prairies, on prairie slopes and in prairie

canyons. Encountered throughout the area.

IRIDACEAE
Belamcanda chinensis (L.) DC. Rare, escaped from cultivation and established in prairies and

along roadsides. Reported for Cowley, Riley and Wabaunsee counties by Gates (1940).
Nemastylis geminiflora Nutt. Rare, in well-managed upland prairies, prairie hay meadows and

open-wooded areas. Known from Cowley County.
Sisyrinchium campestre Bickn. var. campestrc. Common, in upland prairies, prairie hay

meadows, on prairie slopes, in prairie canyons and in open wooded areas. Found throughout

the area.
Sisyrinchium campestre Bickn. var. l(ansanum Bickn. Occasional, in upland prairies, prairie hay

meadows, on prairie slopes, in prairie canyons and in open wooded areas. Occurs throughout

the area.

ORCHIDACEAE
Habenaria leucophaea (Nutt.) Gray. Rare, in wet prairie areas. Reported for Lyon and Riley

counties by Gates (1940).
Liparis loeselii (L.) Richard. Rare, in woods and thickets. Reported for Pottawatomie County

by Gates (1940).
Spiranthes cernua (L.) Rich. Rare, in moist upland prairies, prairie hay meadows, and prairie

canyons. Found in Lyon and Riley counties.
Spiranthes vernalis Engelm. & Gray. Rare, in moist upland prairies and wet prairie hay meadows.

Found in Lyon County and reported for Cowley, Morris and Riley counties by Gates (1940).

SALICACEAE
Populus alba L. Occasional, escapes from cultivation to along streams, fence rows and roadsides.

Collected from Riley and Wabaunsee counties.
Populus deltoides Marsh, (including P. sargentii Dode). Common, in floodplain woods. Occurs

throughout the area.
Salix amygdaloides Anderss. Common, in floodplain woods and along streambanks. Found in

Geary, Marshall, Morris, Pottawatomie, Riley and Wabaunsee counties.

554 The University Science Bulletin

Salix caroliniana Michx. Common, in floodplain woods and along streambanks. Found in Butler,

Chase, Cowley, Geary, and Wabaunsee counties.
Salix eriocephala Michx. Rare, in alluvial soil along streams. Reported for Geary, Morris, Potta-
watomie and Riley counties by Gates (1940).
Salix humilis Marsh, var. microphylla (Anderss.) Fern. Occasional, in upland prairies, on rocky

prairie slopes, in open areas in wooded canyons and along prairie roadsides. Known from

Marshall, Pottawatomie and Riley counties.
Salix interior Rowlee. Common, in floodplain woods and along streambanks. Found throughout

the area.
Salix nigra L. Common, in floodplain woods and along streambanks. Occurs throughout the

area.
Salix rigida Muhl. Occasional, in alluvial soil along streams. Known from Marshall, Morris and

Riley counties.

JUGLANDACEAE

Carya cordijormis (Wang.) K. Koch. Common, in rich woods along streams, drainage areas and

in moist wooded canyons. Found throughout the area.
Carya illinoensis (Wang.) K. Koch. Rare, in floodplain woods and often cultivated in low

places. Known from Cowley County.
Carya laciniosa (Michx.) Loud. Occasional, in rich woods along streams. Found in Cowley

County.
Carya ovata (Mill.) K. Koch. Rare, in rich woods along streams. Reported for Riley County by

Gates (1940).
Juglans nigra L. Common, in rich wooded canyons, rich woods along streams and sometimes

occurring as scattered individuals in upland areas. Found throughout the area.

CORYLACEAE

Coryhts americana Walt. Occasional, in wooded canyons and sometimes in floodplain woods.
Known from Pottawatomie, Riley and Wabaunsee counties.

Ostrya virginiana (Mill.) K. Koch. Occasional, in wooded canyons and along margins of flood-
plain woods. Found in Pottawatomie, Riley and Wabaunsee counties.

FAGACEAE

Quercus borealis Michx. Occasional, on well-drained rocky wooded slopes. Known from Geary,

Pottawatomie and Riley counties.
Qtierctis macrocarpa Michx. Common, on rocky wooded slopes and in wooded canyons. Occurs

throughout the area.
Quercus marilandica Muenchh. Rare, on rocky wooded slopes. Known from Pottawatomie and

Riley counties.
Quercus mttehlenbergii Engclm. Common, on rocky wooded slopes. Found throughout the

area.
Quercus palustris Muenchh. Rare, in floodplain woods along the Arkansas River. Known only

from Cowley County.
Quercus prinoides Willd. Common, on rocky wooded slopes and in wooded canyons. Found

throughout the area.
Quercus shumardii Buckl. Rare, on rocky wooded slopes. Known only from Cowley County.
Quercus stellata Wang. Rare, in sandy soil of upland wooded areas. Reported for Riley County

by Gates (1940).
Quercus velutina Lam. Occasional, on rocky wooded slopes and in well-drained soil along

streams. Known from Cowley, Pottawatomie and Riley counties.

ULMACEAE
Celtis laevigata Willd. var. texana Sarg. Rare, in rocky wooded canyons. Known from Cowley

County.
Celtis occidentalis L. var. canina (Raf.) Sarg. Rare, in rocky wooded canyons. Known from

Lyon County.
Celtis occidentalis L. var. occidentalis. Common in rocky wooded canyons and wooded stream

valleys. Occurs throughout the area.
Celtis occidentalis L. var. pumila (Pursh.) Gray. Occasional, in rocky wooded canyons and

wooded stream valleys. Found throughout the area.
Celtis tcnuijolia Nutt. var. georgiana (Small) Fern. & Schub. Rare, in rocky wooded canyons.

Known from Cowley County.
Ulmus americana L. Common, in rocky wooded canyons and in floodplain woods. Found

throughout the area.

The Flora of the Kansas Flint Hills 555

Ulmiis pumila L. Occasional, escapes from cultivation to along roadsides, in wooded canyons

and floodplain woods. Occurs throughout the area.
Ulmus rubra Muhl. Common, in rocky wooded canyons and in floodplain woods. Found

throughout the area.

MORACEAE
Cannabis sativa L. Common to abundant, in waste places, along moist railroad right-of-ways

and in open floodplain woods. Found in the northern part of the area.
Hamulus lupalus L. Occasional, in waste ground, in open wooded areas, and along railroad

right-of-ways. Occurs throughout the area.
Madura pomifera (Raf.) Scheid. Common, in floodplain woods, margins of woods, as scattered

individuals in grazed prairies, along fence rows and roadsides. Occurs throughout the area.
Morns alba L. Common, in moist woods along streams, along fence rows and grazed prairies.

Known from Lyon, Marshall, Morris, Riley and Wabaunsee counties.
Moms rubra L. Common, in floodplain woods, wooded canyons, along fence rows and grazed

prairies. Found throughout the area.

URTICACEAE
Bochmeria cylindrica (L.) Sw. Rare, in floodplain woods and low wet ground. Reported for

Marshall and Pottawatomie counties by Gates (1940).
Laportea canadensis (L.) Wedd. Occasional, in moist stream valley woods. Found throughout

the area.
Parietaria pennsylvanica Muhl. Occasional to common, in wooded canyons and stream valley

woods. Occurs throughout the area.
Pilea pumila (L.) Gray. Occasional, in moist stream valley woods. Found throughout the area.
Urtica dioica L. var. proccra (Muhl.) Wedd. Common, in moist stream valley woods. Occurs

throughout the area.

SANTALACEAE
Comandra umbcllata (L.) Nutt. subsp. pallida (A. DC) Pic hi. Rare, in rocky prairie areas and

in rocky wooded canyons. Collected once by W. H. Horr in Cowley County.
Comandra umbcllata (L.) Nutt. subsp. umbcllata. Rare, in rocky wooded canyons. Reported

for Pottawatomie and Wabaunsee counties by Gates (1940).

POLYGONACEAE

Eriogonum annunm Nutt. Rare, in prairies on sandy soils. Known from Riley and Cowley
counties.

Polygonum avicularc L. Common, in waste places, lawns, along roadsides, about farmyards and
in severely overgrazed prairies. Found throughout the area.

Polygonum bicorne Raf. Occasional, in wet low ground along drainage areas and streams. Found
throughout the area.

Polygonum coccincum Muhl. Occasional, in water at the edge of impoundments, streams and
roadside ditches. Occurs throughout the area.

Polygonum convolvulus L. Occasional, along roadsides, railway right-of-ways, in cultivated
fields and waste ground. Occurs throughout the area.

Polygonum erectum L. Occasional, in overgrazed prairies, waste ground along roadsides and
railway right-of-ways. Encountered throughout the area.

Polygonum hydropiper L. Rare, in cultivated fields, along prairie drainage areas, margins of
impoundments and streams. Reported for Riley County by Gates (1940).

Polygonum hydropiperoides Michx. Rare, in wet ground along impoundments, seepage areas
and streams. Reported for Riley County by Gates (1940).

Polygonum lapathijolium L. Common, in low wet alluvial soil along drainage areas, impound-
ments and streams. Occurs throughout the area.

Polygonum pennsylvanicum L. Common, in wet alluvial soil along drainage areas, impound-
ments and streams. Found throughout the area.

Polygonum persicaria L. Common, in waste ground, cultivated fields, moist overgrazed prairies
and along streams. Encountered throughout the area.

Polygonum punctatum Ell. Common, in alluvial soil along drainage areas, impoundments and
streams. Occurs throughout the area.

Polygonum ramosissimum Michx. Occasional, in upland and canyon prairies. Known through-
out the area.

Polygonum scandens L. Common, in floodplain woods and wooded canyons. Found throughout
the area.

Polygonum tenue Michx. Occasional, in rocky upland prairie and on rocky prairie slopes. Known
from Chase, Marion, Morris, Pottawatomie and Wabaunsee counties.

556 The University Science Bulletin

Polygonum virginianum L. Common, in moist wooded canyons and floodplain woods. Found

throughout the area.
Rumex acetosella L. Rare, in waste ground and cultivated fields. Reported for Pottawatomie,

Riley and Wabaunsee counties by Gates (1940).
Rumex altissimtts Wood. Common, in alluvial soil along streams, impoundments, in cultivated

fields, waste ground and roadside ditches. Encountered throughout the area.
Rumex crispus L. Common, in cultivated fields, waste ground, along roadsides, railway right-
of-ways, and margins of floodplain woods. Found throughout the area.
Rumex maritimus L. var. fueginus (Phil.) Dusen. Rare, in sandy soil along the Kansas River.

Reported for Riley County by Gates (1940).
Rumex patientia L. Occasional, in waste ground, along roadsides and railway right-of-ways.

Known from the northern part of the area.
Rumex stenophyllus Ledeb. Rare, in alluvial soil along the streams. Found only in the northern

part of the area.
Rumex venosus Pursh. Rare, in prairies on sandy soils. Reported for Geary, Pottawatomie and

Riley counties by Gates (1940).

CHENOPODIACEAE

Atriplex hortensis L. Rare, escapes from cultivation into waste places. Reported for Pottawatomie

County by Gates (1940).
Chenopodium album L. Common, in waste places, cultivated fields, along roadsides and railroad

right-of-ways. Found throughout the area.
Chenopodium ambroisoides L. Rare, in waste places around dwellings and along railroad

right-of-ways. Reported from Morris, Pottawatomie and Riley counties.
Chenopodium berlandieri Moq. var. boscianum (Moq.) Vahl. Occasional, in alluvial woods,

moist wooded canyons, and in moist waste places. Occurs throughout the area.
Chenopodium gigantospermum Aellen. Common, in floodplain woods and wooded canyons.

Found throughout the area.
Chenopodium glaucum L. Rare, in floodplain woods and moist waste places. Found in Chase

County.
Chenopodium leptophyllum Nutt. Occasional, in dry wooded canyons and waste places. Occurs

throughout the area.
Chenopodium standleyanum Aellen. Occasional, in floodplain woods, along streambeds, and in

moist waste places. Encountered throughout the area.
Cycloloma atriplicijolium (Spreng.) Coult. Rare, in sandy areas along streams and sandy waste

places. Found in the northern part and in Cowley County along Arkansas River floodplain.
Kochia scoparia (L.) Schrader. Common, in waste places, along roadsides and railway right-of-
ways. Found throughout the area.
Monolepis nuttalliana (Schult.) Green. Occasional, in waste ground, overgrazed prairies, along

roadsides and railroad right-of-ways. Found throughout the area.
Salsola \ali L. Occasional, in open waste places, along open floodplains, roadsides and railroad

right-of-ways. Reported for Pottawatomie, Riley and Wabaun-ee counties by Gates (1940).

AMARANTHACEAE

Amaranthus graecizans L. Common, in waste places, cultivated fields, along roadsides and rail-
road right-of-ways. Occurs throughout the area.

Amaranthus hybridus L. Common, in floodplain woods, and moist wooded canyons. En-
countered throughout the area.

Amaranthus rettofiexus L. Common, in waste places, cultivated fields and along roadsides.
Found throughout the area.

Amaranthus spinosus L. Occasional, in waste places and cultivated fields. Found in Potta-
watomie and Wabaunsee counties.

Amaranthus tamariscinus (Nutt.) Wood. Common, in waste places, cultivated fields, along
roadsides and railroad right-of-ways. Encountered throughout the area.

Froelichia floridana (Nutt.) Moq. var. campestris (Small) Fern. Occasional, grows in sandy soil
along streambeds. Known from Cowley, Morris, Pottawatomie and Riley counties.

Froelichia gracilis (Hook.) Moq. Occasional, grows in sandy soil along streambeds. Known
from Geary, Pottawatomie and Riley counties.

Tidestromia lanuginosa (Nutt.) Standi. Rare, occurs in waste ground and along railway right-
of-ways. Reported for Riley and Lyon counties by Gates (1940).

NYCTAGINACEAE
Mirabilis albida (Walt.) Heimerl. Occasional, in rocky upland prairies, on rocky prairie slopes,

The Flora of the Kansas Flint Hills 557

margins of wooded canyons and along rocky roadside embankments. Found in Cowley,

Marshall, Morris, Riley and Wabaunsee counties.
Mirabilis linearis (Pursh.) Heimerl. Common, in rocky upland prairies, on rocky prairie slopes,

margins of wooded canyons and along rocky roadsides. Occurs throughout the area.
Mirabilis nyctaginea (Michx.) MacM. Common, in moist rich soil along railway right-of-ways,

roadside ditches, drainage areas, waste places and open floodplain woods. Encountered

throughout the area.

PHYTOLACCACEAE

Phytolacca americana L. Common, in waste ground, farmyards, along roadsides and margins of
wooded areas. Occurs throughout the area.

AIZOACEAE
Mollugo vcrticillata L. Common, along roadsides, railroad right-of-ways, in cultivated fields and
waste places. Occurs throughout the area.

PORTULACACEAE

Claytonia virginica L. Occasional, in upland prairies, prairie canyons, open areas in wooded

canyons and open floodplain woods. Known from Cowley County.
Portnlaca mundula Johnston. Rare, in sandy soil. Reported for Riley County by Gates (1940).
Portulaca oleracea L. Occasional, in waste ground and cultivated fields. Occurs throughout

the area.

CARYOPHYLLACEAE

Agrostemma githago L. Rare, introduced into cultivated fields and waste ground. Reported for

Riley County by Gates (1940).
Arenaria scrpyllijolia L. Rare, in sandy soil, cultivated fields and open waste places. Reported

for Geary and Riley counties by Gates (1940).
Arenaria stricta Michx. var. texana Robins. Rare, on rocky prairie slopes, along prairie ravines

and in low sandy areas. Reported for Cowley County by Gates (1940).
Cerastium brachypodum (Engelm.) Robins. Rare, in open wooded areas, overgrazed prairies,

roadside ditches, cultivated fields and waste places. Known from Wabaunsee County.
Lychnis alba Mill. Rare, introduced into cultivated fields and waste places. Reported for Rilev

County by Gates (1940).
Saponaria officinalis L. Common, escapes cultivation and grows along roads, abandoned farm-
yards and in waste places. Encountered throughout the region.
Saponaria vaccaria L. Rare, introduced along roadsides, railroad right-of-ways and in waste

places. Reported for Riley County by Gates (1940).
Silenc antirrhina L. Common, along roadsides, railway right-of-ways in overgrazed prairies,

cultivated fields and waste places. Occurs throughout the area.
Stlcne caciibalus Wibel. Rare, introduced into cultivated fields and waste places. Reported for

Marion and Riley counties by Gates (1940).
Sdcne noctiflora L. Rare, introduced into cultivated fields and waste places. Reported for Rilev

County by Gates (1940).
Sdene stellata (L.) Ait. Common, in floodplain woods and along the margins of rich wooded

areas. Occurs throughout the area.
Spergitla arvensis L. Rare, introduced into cultivated fields and waste places. Reported for Rilev

County by Gates (1940).
Stellaria media (L.) Cyrill. Common, in lawns, cultivated fields and other such waste places.

Found throughout the area.

CERATOPHYLLACEAE
Ceratophyllum demersum L. Occasional, grows in shallow water of impoundments. Found in

Chase and Riley counties.

NYMPHAEACEAE
Nelumbo lutea (Willd.) Pers. Rare, grows in impoundments. Known from Morris County.
Nuphar advena (Ait.) Ait. Rare, grows in impoundments. Known from Cowley County.

ANNONACEAE

Asimina triloba (L.) Dunal. Common, in stream valley woods. Found throughout the area
except in Chase, Geary, Marion, Morris and Marshall counties.

RANUNCULACEAE
Anemone canadensis L. Occasional, in alluvial soil along wooded floodplains. Reported for
Pottawatomie and Riley counties by Gates (1940).

558 The University Science Bulletin

Anemone caroliniana Walt. Occasional, in upland prairies and prairie canyons. Occurs through-
out the area.
Anemone virginiana L. Occasional, in floodplain woods. Known from Marshall, Pottawatomie,

Riley and Wabaunsee counties.
Aquilegia canadensis L. Occasional, on rocky wooded slopes and in rich lowland woods.

Collected from Cowley and Marshall counties.
Clematis dioscoreijolia Levi. & Vaniot. Occasional, escapes from cultivation to along roadsides,

railway right-of-ways and waste places. Known from Cowley County.
Clematis ligusticijolia Nutt. Rare, in wooded canyons. Reported for Geary and Cowley counties

by Gates (1940).
Clematis pitcheri T. & G. Occasional, in wooded canyons and floodplain woods. Occurs through-
out the area.
Clematis virginiana L. Occasional, in wooded canyons and floodplain woods. Reported for

Marshall, Pottawatomie and Riley counties by Gates (1940).
Delphinium ajacis L. Occasional, escapes from cultivation to along roadsides and in waste places.

Known from Marshall and Cowley counties.
Delphinium tricorne Michx. Occasional, in rich wooded canyons and in floodplain woods. Found

throughout the area.
Delphinium virescens Nutt. Common, in upland prairies, prairie canyons, cultivated fields, along

roadsides and railway right-of-ways. Occurs throughout the area.
Isopyrum biternatum (Raf.) T. & G. Rare, in moist wooded canyons. Reported for Riley County

by Gates (1940).
Myosurus minimus L. Rare, in moist cultivated fields, low alluvial ground and low waste places.

Known from Geary, Pottawatomie and Riley counties.
Ranunculus abortivus L. Common, in wet places in prairies, open areas in woodlands, along

streambanks, drainage areas, roadside ditches and railroad right-of-ways. Found throughout

the area.
Ranunculus acris L. Occasional, in cultivated fields and moist prairie hay meadows. Reported

for Lyon, Pottawatomie and Riley counties by Gates (1940).
Ranunculus cymbalaria Pursh. Common, growing in wet soil along margins of impoundments

and streams. Known from Riley County.
Ranunculus flabellans Raf. Rare, grows in water of impoundments and slow-moving streams.

Reported for Morris and Riley counties by Gates (1940).
Ranunculus longirostris Godron. Rare, grows in water of impoundments and slow-moving

streams. Reported for Riley County by Gates (1940).
Ranunculus sceleratus L. Occasional, in wet prairie areas, open floodplain woods, cultivated

fields, along roadsides and railway right-of-ways. Known from throughout the area.
Tha/ictrum dasycarpum Fisch. and Ave-Lall. Occasional, along margins of moist woods, in

moist wooded canyons and floodplain woods. Occurs throughout the area.

BERBERIDACEAE
Podophyllum pcltatum L. Rare, in stream valley woods. Reported for Pottawatomie and Riley
counties by Gates (1940).

MENISPERMACEAE

Cocculus carolinus (L.) D.C. Occasional, on rocky wooded slopes and wooded canyons. Known
from Cowley County.

Menispcrmum canadense L. Common, in wooded canyons and in stream valley woods. Found
throughout the area.

PAPAVERACEAE

Argemone polyanthcmos (Fedde) Ownb. Occasional, in rocky overgrazed prairies, rocky road-
side embankments and along railroad right-of-ways. Occurs in Geary, Pottawatomie, Riley
and Wabaunsee counties.

Papaver rhoeas L. Occasional, escapes into waste places and moist woods. Reported for Wabaun-
see County by Gates (1940).

FUMARIACEAE

Corydalis micrantha (Engelm.) Gray. Occasional, in rich woods, low cultivated fields and
shaded waste places. Known from Cowley, Geary, Riley and Wabaunsee counties.

Dicentra cucullaria (L.) Bernh. Occasional, in rich woods. Found in Cowley, Marshall, Riley
and Wabaunsee counties.

CRUCIFERAE

Alliaria officinalis Andrz. Rare, in floodplain woods. Found in Riley and Wabaunsee counties.

Arabis canadensis L. Occasional, in floodplain woods and wooded canyons. Known from
Marshall, Morris and Wabaunsee counties.

The Flora of the Kansas Flint Hills 559

Arabis hirsuta (L.) Scop. Occasional, in stream valley woods and wooded canyons. Found in
Marshall, Pottawatomie and Riley counties.

Barbarea vulgaris R. Br. Occasional, in cultivated fields, waste places, open floodplains and
along roadsides. Found in Pottawatomie and Riley counties.

Berteroa incana (L.) DC. Rare, in waste places. Reported for Riley County by Gates (1940).

Brassica campestris L. Rare, cultivated fields and waste places. Reported for Morris, Potta-
watomie, Riley and Wabaunsee counties by Gates (1940).

Brassica juncea (L.) Cross. Rare, in cultivated fields and waste places. Reported for Riley
County by Gates (1940).

Brassica \aber (DC.) L. Occasional, in cultivated fields and waste places. Known from Chase,
Geary, Marshall, Morris and Riley counties.

Brassica nigra (L.) Koch. Occasional, in cultivated fields and waste places. Found throughout
the area.

Camelina microcarpa Andrz. Occasional, in cultivated fields, waste places and along roadsides.
Found throughout the area.

Capsella bursa-pastoris (L.) Medic. Common, in lawns, cultivated fields, waste places and
along roadsides. Occurs throughout the area.

Cardamine bulbosa (Schreb.) BSP. Rare, in wet woods and prairies. Reported for Pottawatomie
County by Gates (1940).

Cardaria draba (L.) Desv. Occasional in waste places, along roadsides and railroad right-of-
ways. Occurs throughout the area.

Conringia orientalis (L.) Dumont. Rare, in waste places, along roadsides and railroad right-
of-ways. Reported for Geary, Pottawatomie, and Riley counties by Gates (1940).

Dcntaria laciniata Muhl. Occasional, in rich floodplain woods and sometimes in wooded can-
yons. Found in Lyon, Riley and Wabaunsee counties.

Descurainia pinnata (Walt.) Rritt. Common, in waste places, overgrazed prairies, along railroad
right-of-ways and roadsides. Occurs throughout the area.

Descurainia sophia (L.) Webb. Occasional, in waste places, along roadsides and railway
right-of-ways. Known from Butler, Cowley and Riley counties.

Draba brachycarpa Nutt. Rare, in rocky prairies, cultivated Holds, along railroad right-of-ways
and roadsides. Reported for Riley County by Gates (1940).

Draba ctineijolia Nutt. Occasional, in rocky prairies, along limestone escarpments, railroad
right-of-ways, and roadsides. Encountered throughout the area.

Draba reptans (Lam.) Fern. Occasional in rocky prairies, cultivated fields, waste places, along
railroad right-of-ways and roadsides. Found in Cowley, Lyon, Pottawatomie, Riley and
Wabaunsee counties.

Eruca sativa Mill. Rare, introduced into waste places. Reported for Riley County by Gates
(1940).

Erysimum asperum DC. Rare, in rocky prairies, along railway right-of-ways and roadsides.
Reported for Riley County by Gates (1940).

Erysimum repandum L. Common, in waste places, cultivated fields, along railroad right-of-
ways and roadsides. Found throughout the area.

Hcsperis matronalis L. Occasional, introduced along roadsides and in waste places. En-
countered throughout the area.

lodanthus pinnatifidus (Michx.) Steud. Rare, in floodplain woods. Reported for Pottawatomie
County by Gates (1940).

Lepidium campestre (L.) R. Br. Rare, in cultivated fields and waste places. Reported for Riley
County by Gates (1940).

Lepidium densiflorum Schrader. Common, in overgrazed prairies, cultivated fields, waste places,
along railway right-of-ways and roadsides. Occurs throughout the area.

Lepidium ramosissimum Nels. Rare, in waste places and cultivated fields. Reported for Riley
County by Gates (1940).

Lepidium virginicutn L. Occasional, in cultivated fields, overgrazed prairies, waste places, along
railway right-of-ways and roadsides. Found throughout the area.

Nasturtium officinale R. Br. Common, along streams and spring seepage areas growing either
in shallow water or in mud at the edge of the water. Occurs throughout the area.

Raphantts sativus L. Occasional, escapes from cultivation into waste places. Reported for Riley
County by Gates (1940).

Rorippa islandica (Oeder) Borbas. Occasional, in wet alluvial soil along streams, drainage areas,
impoundments and in low cultivated fields. Found in Ruder, Cowley, Riley and Wabaunsee
counties.
Rorippa sessiliflora (Nutt.) Hitchc. Common, in wet alluvial soil along streams, impound-
ments, in floodplain woods and low cultivated fields. Occurs throughout the area.

560 The University Science Bulletin

Rorippa sinuata (Nutt.) Hitchc. Common, along floodplains, in low cultivated fields, railway
right-of-ways and roadsides. Known throughout the area.

Sisymbrium altissimum L. Occasional, in cultivated fields, waste places, along railway right-of-
ways and roadsides. Found in Geary and Mar:hall counties.

Sisymbrium officinale (L.) Scop. Occasional, in cultivated fields, farmyards, waste places, along
railroad right-of-ways and roadsides. Found in Marshall, Pottawatomie, Riley and Wabaun-
see counties.

Thlaspi arvense L. Common, in cultivated fields, overgrazed prairies, waste places, along rail-
road right-of-ways and roadsides. Occurs throughout the area.

CAPPARIDACEAE

Cleome scrrulata Pursh. Rare, in sandy soil along streams. Reported for Pottawatomie and

Riley counties by Gates (1940).
Polanisia trachysperma T. &. G. Occasional, in sandy prairies, sandy areas along streams and in

sandy roadside ditches. Known from Cowley, Pottawatomie and Riley counties.

CRASSULACEAE
Pcntlwrum sedoides L. Common, along streambanks, impoundments, drainage areas and road-
side ditches. Found throughout the area.

SAXIFRAGACEAE

Ribes missouriense Nutt. Common, in wooded canyons, along fence rows and wooded road-
sides. Found throughout the area.

Ribes odoratum Wendl. Occasional, in wooded canyons and along fence rows. Known from
Butler and Cowley counties.

PLATANACEAE
Platanus occidentalis L. Common, in floodplain woods and wooded canyons. Occurs through-
out the area.

ROSACEAE
Agrimonia parviflora Ait. Rare, in low wet prairie areas, along margins of drainage areas,

impoundments and streams. Known from Pottawatomie, Riley and Wabaunsee counties.
Agrimonia pubescens Wallr. Occasional, in wooded canyons and floodplain woods. Known from

Geary, Marshall, Morris, Pottawatomie, Riley and Wabaunsee counties.
Crataegus macl<enzii Sarg. Rare, in rocky wooded canyons. Known only from Wabaunsee

County.
Crataegus mollis (T. & G.) Scheele. Occasional, in wooded stream valleys. Found in Chase,

Lyon, Marshall and Riley counties.
Crataegus succulenta Link. Occasional, in rocky wooded canyons and floodplain woods. Known

from Pottawatomie and Wabaunsee counties.
Fragaria virginiana Duchesne. Occasional to common, in upland prairies, on prairie slopes, in

prairie canyons and in open areas of wooded canyons. Found throughout the area.
Geum canadense Jacq. Common, in rich wooded canyons in wooded stream valleys. Encountered

throughout the area.
Potentilla or gut a Pursh. Occasional, in upland prairies, prairie canyons and in prairie railway

right-of-ways. Found in Lyon, Marshall and Pottawatomie counties.
Potentilla nicolletii (S. Wats.) Sheld. Rare, in low ground along streams. Reported for Riley

County by Gates (1940).
Potentilla paradoxa Nutt. Rare, in wet alluvial soil along impoundments and streams. Reported

for Riley County by Gates (1940).
Potentilla recta L. Occasional, in cultivated fields, prairie hay meadows, overgrazed prairies,

waste ground, along roadsides and railroad right-of-ways. Found in Cowley, Lyon, Marshall,

Riley and Wabaunsee counties.
Potentilla rivalis Nutt. Rare, in waste ground and alluvial soil along streams and impound-
ments. Reported for Riley County by Gates (1940).
Potentilla simplex Michx. Occasional, along margins of wooded areas and in rocky wooded

canyons. Found in Pottawatomie and Wabaunsee counties.
Fiu nits americana Marsh. Occasional, in wooded fence rows, wooded canyons and sometimes

overgrazed prairies. Encountered throughout the area.
Primus hortulana Bailey. Occasional, in open wooded canyons, along borders of wooded areas

and along streams. Found in Butler anil Cowley counties.
1'runus mahalcb L. Rare, escapes from cultivation to along roadsides and wooded fence rows.

Known from Lyon, Marshall, Riley and Wabaunsee counties.

The Flora of the Kansas Flint Hills 561

Primus mexicana S. Wats. Occasional, in rocky wooded canyons and along wooded roadsides.

Found in Cowley, Geary, Marshall, Pottawatomie, Riley and Wabaunsee counties.
Primus munsoniana Wight & Hedrick. Occasional along the margins of wooded areas and

streams. Known from Butler and Cowley counties.
Primus serotina Ehrh. Occasional, in wooded canyons and wooded areas along streams. Occurs

in Riley and Wabaunsee counties.
Primus virginiana L. Occasional, in wooded canyons and along streams. Occurs throughout

the area.
Pyrus coronaria L. Rare, in low open woods along streams. Reported for Riley County by Gates

(1940).
Pyrus ioensis (Wood.) Bailey. Occasional, in open areas in wooded areas, along margins of

woods and wooded streambanks. Found in Marshall, Pottawatomie and Riley counties.
Rosa arkjansas Porter var. suffulta (Greene) Cockerell. Common, in upland prairies on prairie

slopes, along roadsides and railroad right-of-ways. Occurs throughout the area.
Rosa blanda Ait. Occasional, in open wooded areas and along roadsides. Found in Marshall,

Pottawatomie and Riley counties.
Rosa mulfiflora Thunb. Occasional, planted along fence rows and roadsides, sometimes escaping

into overgrazed prairies and open ground. Known from Butler, Cowley and Lyon counties.
Rubus allegheniensis Porter. Occasional, in wooded canyons. Found in Cowley County.
Rubus enslenii Tratt. Occasional, in upland prairies, on prairie slopes and along roadsides.

Known from Pottawatomie and Wabaunsee counties.
Rubus flagellars L. Occasional, in rocky wooded canyons and along wooded roadsides. Found

in Cowley, Morris, Pottawatomie and Riley counties.
Rubus occidentalis L. Occasional to common, in rocky wooded canyons. Found throughout

the area.
Rubus orarius Blanchard. Rare, in open wooded areas, along margins of woods and roadsides.

Collected once by H. A. Stephens in Wabaunsee County.
Rubus ostryijolius Rydb. Occasional to common, in wooded canyons, on prairie slopes, along

roadsides and railway right-of-ways. Found throughout the ana.

LEGUMINOSA1.

Acacia angustissima (Mill.) Ktze. Rare, in rocky upland prairies and occasionally along the
roadsides. Known from Cowley County.

Amorpha canescens Pursh. Common to abundant, in upland prairies, prairie hay meadows,
prairie canyons and sometimes in open wooded canyons. Occurs throughout the area.

Amorpha fruticosa L. Common, along streams, drainage areas, impoundments, and along mar-
gins of stream valley woods. Found throughout the area.

Amorpha nana Nutt. Rare, in rocky upland prairies and on rocky prairie slopes. Known from
Chase, Pottawatomie, Riley and Wabaunsee counties.

Amphicarpa bracteata (L.) Fern. Occasional, in wooded canyons and floodplain woods. Found
throughout the area.

Apios americana Medic. Rare, in floodplain woods. Known from Cowley, Marshall, Potta-
watomie and Riley counties.

Astragalus canadensis L. Occasional, in rocky or open lowland area, margins of canyon and
floodplain woods. Found in Cowley, Geary, Marshall, Morris, Pottawatomie and Riley
counties.

Astragalus caryocarpus Ker. Common, in rocks upland prairies, on rocky prairie slopes, in open
canyon woods, along railroad right-of-ways and roadsides. Occurs throughout the area.

Astragalus lotiflortts Hook. Occasional, in rocky upland prairies and on rocky prairie slopes.
Occurs in Geary, Pottawatomie, Riley and Wabaunsee counties.

Astragalus plattcnsis Nutt. Occasional, in rocky upland prairies and on rocky prairie slopes.
Found in Geary, Lyon, Riley and Wabaunsee counties.

Baptisia australis (L.) R. Br. var. minor (Lehm.) Fern. Common, in rocky upland prairies,
prairie hay meadows and prairie canyons. Occurs throughout the area.

Baptisia leucantha T. & G. Occasional, in upland prairies, prairie hay meadows, prairie canyons
and in alluvial soil along streams. Collected from Lyon and Wabaunsee counties.

Baptisia leucophaea Nutt. Common, in rocky upland prairies and prairie hay meadows. En-
countered throughout the area.

Cassia fasciculata Michx. Common, in rocky upland prairies, on rocky prairie slopes, along
railway right-of-ways and roadsides. Found throughout the area.

Cassia marilandica L. Occasional, in open rocky woods, on rocky prairie slopes, in rocky upland
prairies, alluvial soil along roadsides and drainage areas. Occurs throughout the area.

562 The University Science Bulletin

Cercis canadensis L. Common, in open canyon woods, along the margins of lowland, flood-
plain woods and along open rocky streambanks. Found throughout the area.

Coronilla varia L. Occasional, in waste places and often planted along roadside embankments.
Known from Geary, Marshall and Wabaunsee counties.

Crotalaria sagittalis L. Rare, in prairies, on open wooded slopes, in waste places and along
roadsides. Reported for Morris, Pottawatomie, and Riley counties by Gates (1940).

Dalea alopectiroides Willd. Rare, in prairies on sandy soil and along rivers. Reported for
Pottawatomie, Riley and Wabaunsee counties by Gates (1940).

Dalea atirea Nutt. Occasional, in rocky upland prairies, on rocky prairie slopes and along rocky
roadsides. Occurs throughout the area.

Dalea enncandra Nutt. Rare, in rocky upland prairies and along roadsides. Known from
Cowley, Pottawatomie and Riley counties.

Desmanthits illinoensis (Michx.) MacMill. Common, in rocky upland prairies, on rocky prairie
slopes, in open wooded canyons, along railway right-of-ways and roadsides. Found through-
out the area.

Desmoditim canadense (L.) DC. Occasional, in moist upland prairies, moist prairie hay meadows
and prairie canyons. Found throughout the area.

Desmoditim canescens (L.) DC. Occasional, in wooded canyons and in alluvial soil along
streams. Known from Chase, Geary, Marshall, Morris, Pottawatomie, Riley and Wabaunsee
counties.

Desmodiitm cuspidatum (Muhl.) Loud. Occasional, in rocky wooded canyons. Known from
Geary, Lyon, Morris and Riley counties.

Desmoditim glutinosum (Muhl.) Wood. Common, in rocky wooded canyons and in floodplain
woods. Occurs throughout the area.

Desmoditim illinoense Gray. Common, in rocky upland prairies, on rocky prairie slopes and in
prairie canyons. Encountered throughout the area.

Desmodiitm paniculatum (L.) DC. Rare, in wooded canyons and moist woods along streams.
Known from Cowley County.

Desmoditim sessilifolium (Torr.) T. & G. Common, in upland prairies, on rocky prairie slopes
and in prairie canyons. Found throughout the area.

Gleditsia triacanthos L. Common, in floodplain woods and in wooded canyons. Occurs through-
out the area.

Glycyrrhiza lepidota Pursh. Common, in overgrazed upland prairies, prairie hay meadows,
prairie canyons, along railway right-of-ways and roadsides. Found throughout the region.

Gymnocladtts dtotca (L.) K. Koch. Common, in floodplain woods. Occurs throughout the area.

Lespedeza capitata Michx. Common, in upland prairies, on rocky prairie slopes, on open
wooded slopes, in prairie canyons along railway right-of-ways and roadsides. Found through-
out the area.

Lespedeza cuneata (Dumont) G. Don. Rare, in upland prairies and along roadsides. Collected
from Butler County.

Lespedeza intermedia (S. Wats.) Britt. Rare, in open wooded canyons. Known from Butler
County.

Lespedeza nuttallii Dark Rare, in open wooded canyons. Known from Chase County.

Lespedeza repens (L.) Bart. Occasional, in open wooded canyons. Occurs in Geary, Potta-
watomie and Riley counties.

Lespedeza stipulacea Maxin. Common, along roadsides, railroad right-of-ways and planted in
overgrazed prairies. Found throughout the area.

Lespedeza striata (Thunb.) H. & A. Rare, in open, wooded canyons and along roadsides.
Known from Cowley and Pottawatomie counties.

Lespedeza violacea (L.) Pers. Occasional, in open rocky woods and in rocky prairies. Occurs
in Geary, Pottawatomie, Riley and Wabaunsee counties.

Lespedeza virginica (L.) Britt. Occasional, in open rocky woods. Founil in Cowley County.

Lotus corniculatus L. Occasional, in waste ground, along roadsides and sometimes planted in
overgrazed prairies. Known from Pottawatomie County.

Lotus purshianus Clements & Clements. Occasional, in upland prairies, on rocky prairie slopes
and along roadsides. Found in Marshall County.

Medicago lupulina L. Occasional, in lawns, cultivated fields, waste ground, overgrazed prairies,
along roadsides and railroad right-of-ways. Found throughout the area.

Medicago sativa L. Common, cultivated for hay and escapes into waste ground, roadsides and
railway right-of-ways. Occurs throughout the area.

Melilotus alba Desr. Common, along roadsides, railway right-of-ways, in cultivated fields and
other waste places. Encountered throughout the area.

The Flora of the Kansas Flint Hills 563

Melilotus officinalis (L.) Desr. Common to abundant, along roadsides, railroad right-of-ways,
in cultivated fields and other waste places. Occurs throughout the area.

Oxytropis lambertii Pursh. Rare, in upland prairies and on rocky, prairie slopes. Known from
Geary and Riley counties.

Petalostemon candidum (Willd.) Michx. Common, in rocky upland prairies, on rocky prairie
slopes, in prairie canyons, along railroad right-of-ways and roadsides. Found throughout
the area.

Petalostemon multiflorum Nutt. Common, in rocky upland prairies, on rocky prairie slopes, along
railway right-of-ways and roadsides. Occurs throughout the area.

Petalostemon pwpiiretim (Vent.) Rydb. Common, in rocky upland prairies, on rocky prairie
slopes, in prairie canyons, along railroad right-of-ways and roadsides. Encountered through-
out the area.

Psoralea argophylla Pursh. Occasional, in upland prairies, prairie hay meadows and in prairie
canyons. Occurs throughout the area.

Psoralea esctdenta Pursh. Occasional, in upland prairies, prairie hay meadows and prairie can-
yons. Found throughout the area.

Psoralea lanceolata Pursh. Rare, in sandy, prairie areas. Known from Geary and Riley counties.

Psoralea tentiiflora Pursh. var. floribunda (Nutt.) Rydb. Common, in upland prairies, prairie
hay meadows and prairie canyons. Occurs throughout the area.

Psoralea tentiiflora Pursh. var. tenuiflora. Common, in upland prairies, prairie hay meadows
and prairie canyons. Found throughout the area.

Robinia pseudo-acacia L. Occasional, on rocky wooded slopes, in open floodplain woods and
waste ground. Found in Cowley, Geary, Lyon, Marshall, Morris, Pottawatomie, Riley and
Wabaunsee counties.

Schranhja uncinata Willd. Common, in rocky upland prairies, on rocky prairie slopes, along
railway right-of-ways and roadsides. Occurs throughout the area.

Sesbania exaltata (Raf.) Cory. Rare, reported as escaped from cultivation in Rilev County by
Gates (1940).

Strophostyles htivola (L.) Ell. Occasional, in rocky wooded canyons and floodplain woods.
Known from Lyon, Marshall, Pottawatomie and Riley counties.

Strophostyles leiosperma (T. & G.) Piper. Occasional, in rocky upland prairies, on rocky prairie
slopes, cultivated fields, waste ground, along railway right-of-ways and roadside embank-
ments. Known from Cowley, Geary, Marshall, Pottawatomie, Riley and Wabaunsee counties.

Tephrosia virginiana (L.) Pers. Occasional, in open wooded canyons. Reported for Riley and
Pottawatomie counties by Gates (1940).

Tri folium hybridum L. Occasional, in abandoned fields, overgrazed prairies, waste ground,
along railway right-of-ways and roadsides. Found in Cowley and Riley counties.

Trifolitim pratense L. Common, in cultivated fields, overgrazed prairies, waste ground, along
railway right-of-ways and roadsides. Found throughout the area.

Trifolitim procumbens L. Occasional, in cultivated fields, prairies, open woodlands, along rail-
way right-of-ways and roadsides. Known from Cowley County.

Trifolitim repens L. Common, in cultivated fields, overgrazed prairies, along railway right-of-
ways and roadsides. Found throughout the region.

Vicia americana Muhl. Occasional, in open woods and thickets. Found throughout the area.

Vicia sparsijolia Nutt. Occasional, in overgrazed prairies, waste ground, and along roadsides.
Known from Pottawatomie County.

Vicia villosa Roth. Occasional, in cultivated fields, overgrazed prairies, waste ground and along
roadsides. Known from Cowley and Riley counties.

OXALIDACEAE
Oxalis dillenii Jacq. Common, in cultivated fields, rocky prairies, open wooded areas, along

railroad right-of-ways and roadsides. Throughout the area.
Oxalis stricta L. Common, in wooded canyons, wooded floodplains, cultivated fields, waste

ground, along railroad right-of-ways and roadsides. Occurs throughout the region.
Oxalis violacea L. Common, in open wooded canyons, cultivated fields, rocky upland prairies

and along roadsides. Encountered throughout the area.

GERANIACEAE

Erodium cicutarium L'Her. Rare, in waste places, cultivated fields and along margins of im-
poundments. Found in Chase and Geary counties.

Geranium caroliniantim L. Common, in cultivated fields, waste places, overgrazed prairies,
prairie hay meadows, along roadsides and railway right-of-ways. Throughout the area.

564 The University Science Bulletin

LINACEAE

Linum sulcatum Riddell. Common, in upland prairies, prairie hay meadows and prairie can-
yons. Occurs throughout the area.

Linum usitatissimum L. Occasional, along roadsides, railway right-of-ways and in waste places.
Known from Cowley, Marshall and Riley counties.

ZYGOPHYLLACEAE

Tribulus terrestris L. Common, in cultivated fields, waste ground, severely overgrazed prairies,
along roadsides and railroad right-of-ways. Occurs throughout the area.

RUTACEAE

Ptclea trijoliata L. Occasional, along margins of wooded canyons, in wooded canyons and stream
valley woods. Collected in Cowley County and reported for Riley County by Gates (1940).

Zanthoxylum americanum Mill. Occasional, in wooded canyons. Encountered throughout the
area.

SIMARUBACEAE

Ailanthus altissima (Mill.) Swingle. Occasional, escapes from cultivation to waste ground and
along roadsides. Known from Butler, Geary, Riley and Wabaunsee counties.

POLYGALACEAE
Polygala incarnata L. Rare, in sandy soils of prairie areas. Reported for Pottawatomie County

by Gates (1940).
Polygala verticillata L. Common, in open wooded canyons, on rocky prairie slopes and in rocky

upland prairies. Known from Cowley, Geary, Pottawatomie, Riley and Wabaunsee counties.

EUPHORBIACEAE

Acalypha gracilens Gray. Occasional, in rocky upland prairies, on rocky prairie slopes, in open
wooded canyons. Occurs throughout the area.

Acalypha ostryaefolia Riddell. Common, in floodplain woods and in low cultivated ground.
Found throughout the area.

Acalypha rhomboidea Raf. Occasional, in alluvial soil along drainage areas and streams. Re-
ported for Riley County by Gates (1940).

Acalypha virginica L. Common, in upland prairies, prairie canyons, open wooded canyons,
cultivated fields, along railroad right-of-ways and roadsides. Occurs throughout the area.

Croton capitatus Michx. Common, in overgrazed upland prairies, cultivated fields, waste places,
along railway right-of-ways and roadsides. Found throughout the area.

Croton glandulosas L. Rare, in upland prairies, open wooded canyons, cultivated fields, waste
places, along railway right-of-ways and roadsides. Known from Cowley County.

Croton monanthogynous Michx. Common, in rocky upland prairies, along limestone escarp-
ments on rocky slopes and along roadside embankments. Encountered throughout the area.

Croton texensis (Klotzsch) Muell. Arg. Rare, in sandy prairies, alluvial soil along streams, along
railway right-of-ways and roadsides. Reported for Riley County by Gates (1940).

Ditaxis mercurialina (Nutt.) Coult. Occasional, in rocky upland prairies and on rocky prairie
slopes. Found in Butler, Chase and Cowley counties.

Euphorbia commutata Engelm. Rare, in rocky wooded canyons and in floodplain woods. Re-
ported for Geary County by Gates (1940).

Euphorbia corollata L. Common, in upland prairies, prairie canyons, open wooded canyons,
along railway right-of-ways and roadsides. Found throughout the area.

Euphorbia cyparissias L. Occasional, planted in cemeteries and occasionally escaping into culti-
vated fields and roadsides. Known from Pottawatomie, Riley and Wabaunsee counties.

Euphorbia dcntata Michx. Common, in wooded canyons, overgrazed prairies, cultivated fields,
waste places, along railway right-of-ways and roadsides. Encountered throughout the area.

Euphorbia glyptosperma Engelm. Common, in prairie canyons, in alluvial soil along streams and
low waste places. Occurs throughout the area.

Euphorbia heterophylla L. Common, in wooded canyons and floodplain woods and along
wooded drainage areas. Known throughout the area.

Euphorbia hcxagona Nutt. Occasional, in alluvial soil along streams. Known from Cowley,
Geary, Pottawatomie, Riley and Wabaunsee counties.

Euphorbia humistrata Engelm. Rare, in alluvial soil along streams and impoundments. Reported
for Geary County by Gates (1940).

Euphorbia maculata L. Common, in cultivated fields, waste places, along railroad right-of-ways
and roadsides. Found throughout the area.

The Flora of the Kansas Flint Hills 565

Euphorbia marginata Pursh. Common, in cultivated fields, overgrazed prairies, waste places,

along railway right-of-ways and roadsides. Found throughout the area.
Euphorbia missurica Raf. Common, on rocky prairie slopes along limestone escarpments and in

rocky upland prairies. Occurs throughout the area.
Euphorbia nutans Lag. Common, in cultivated fields, waste places, open floodplain woods along

railway right-of-ways and roadsides. Found throughout the area.
Euphorbia podperae Croiz. Rare, in cultivated fields, waste places, along railway right-of-ways

and roadsides. Known from Marion and Marshall counties.
Euphorbia prostrata Ait. Occasional, in overgrazed prairies, waste places, along railway right-
of-ways and roadsides. Found throughout the area.
Euphorbia serpens HRK. Common, in open alluvial soil along streams, drainage areas and

impoundments. Occurs throughout the area.
Euphorbia spathulata Lam. Common, in rocky upland prairies, on rocky prairie slopes, in

wooded canyons and in roadside ditches. Occurs throughout the area.
Euphorbia stictospora Engelm. Common, in overgrazed prairies, cultivated fields, waste places,

along railway right-of-ways and roadsides. Encountered throughout the area.
Tragia betonicijolia Nutt. Occasional, in overgrazed prairies, on rocky prairie slopes anil in

roadside ditches. Found throughout the area.
Tragia ramosa Torr. Rare, in overgrazed prairies, fin rocky prairie slopes and in roadside ditches.

Occurs throughout the area.

CALLITRICHACEAE
Callitrichc heterophylla Pursh. Rare, grows either in the water or at the edge of streams, springs

and impoundments. Reported for Riley County by Gates (1940).

ANACARDIACEAE

Rhus aromatica Ait. Abundant, on limestone escarpments and on prairie slopes. Occurs through-
out the area.

Rhus copallina L. Rare, in prairies and along margins of wooded areas. Known from Cowley
and Riley counties.

Rhus glabra L. Abundant, in upland prairies and along the margins of wooded areas. Found
throughout the area.

Rhus radicans L. Common, in floodplain woods, on rock) wooded slopes, along roadsides, fence-
rows and railroad right-of-ways. Found throughout the ana.

CELASTRACEAE
Celastrus scandens L. Common, in alluvial woods, wooded canyons, on wooded slopes and

along fence rows. Encountered throughout the area.
Euonymus atropurptireus Jacq. Common, in alluvial woods, wooded canyons and on wooded

slopes. Found throughout the area.

STAPH VLEACEAE
Staphvlca trifolia L. Occasional, in wooded canyons. Known from Cowley, Geary, Pottawatomie,
Riley and Wabaunsee counties.

ACERACEAE
Acer negundo L. Common, in floodplain woods and wooded canyons. Found throughout the

area.
Acer saccharinum L. Common, in floodplain woods anil wooded canyons. Occurs throughout
the area.

HIPPOCASTANACEAE

Aesculus glabra Willd. Common, in wooded areas along streams and drainage areas. En-
countered throughout the area.

SAPINDACEAE

Sapindus drummondii Hook. & Arn. Occasional, along the margins of wooded areas developed
on rocky slopes and sometimes in low areas along streams. Found in Chase, Cowley, Geary
and Riley counties.

BALSAMINACEAE

Impatiens capensis Meerb. Rare, in alluvial woods anil along streambanks. Reported for Potta-
watomie and Wabaunsee counties by Gates (1940).

Impatiens pallida Nutt. Rare, in alluvial woods and along streambanks. Reported for Riley
County by Gates (1940).

566 The University Science Bulletin

RHAMNACEAE
Ceanothus americanus L. Occasional, on rocky prairie slopes and in open areas of wooded

canyons. Occurs throughout the area.
Ceanothus ovatus Desf. Common, on rocky prairie slopes. Occurs throughout the area.
Rhamntts lanceolata Pursh. Occasional, in open wooded canyons and open floodplain woods.

Found in the northern part of the area.

VITACEAE
Ampelopsis cordata Michx. Occasional, in rich stream valley woods. Occurs in Cowley, Geary,

Pottawatomie, Riley and Wabaunsee counties.
Parthenocissus quinquefolia (L.) Planch. Common, in rocky wooded canyons and rich stream

valley woods. Found throughout the area.
Parthenocissus vitacea (Knerr.) Hitchc. Occasional, in rocky wooded canyons and rich stream

valley woods. Known from Chase, Geary, Morris, Pottawatomie, Riley and Wabaunsee

counties.
Vitis aestivalis Michx. Rare, in rocky wooded canyons and rich stream valley woods. Found in

Chase and Cowley counties.
Vitis cinera Engelm. Occasional, in rocky wooded canyons and rich stream valley woods. Occurs

in Butler, Cowley, Pottawatomie, Riley and Wabaunsee counties.
Vitis riparia Michx. Occasional, in rocky wooded canyons and rich stream valley woods. Found

throughout the area.
Vitis I'tilpina L. Common, in rocky wooded canyons and rich stream valley woods. Occurs

throughout the area.

TILIACEAE

Tilta americana L. Occasional, in rich wooded canyons and in wooded areas along streams.
Known from Geary, Lyon, Marshall, Pottawatomie, Riley and Wabaunsee counties.

MALVACEAE

Abutilon thcophrasti Medic. Common, in cultivated fields, waste places, overgrazed prairies,
along railroad right-of-ways and roadsides. Occurs throughout the area.

Callirhoe alcacoides (Michx.) Gray. Common, in upland prairies, prairie hay meadows and in
prairie canyons. Found throughout the region.

Callirhoe involacrata (T. & G.) Gray. Occasional, along roadsides, railway right-of-ways and
sometimes in overgrazed prairies. Encountered throughout the region.

Hibiscus militaris Cav. Occasional, along streambanks, impoundments and drainage areas. Re-
ported for Pottawatomie, Riley and Wabaunsee counties by Gates (1940).

Hibiscus trionum L. Common to abundant, in cultivated fields, waste places, along roadsides
and railroad right-of-ways. Occurs throughout the area.

Malva neglecta Wallr. Occasional, in cultivated fields, waste ground, along railway right-of-ways
and roadsides. Found throughout the area.

Malva rotundijolia L. Occasional, in cultivated fields, waste ground, along railway right-of-ways
and roadsides. Known from Cowley, Geary, Pottawatomie and Riley counties.

Sida spinosa L. Common, in cultivated fields, overgrazed prairies, along railroad right-of-ways
and roadsides. Occurs throughout the area.

Sphaeralcea angusta (Gray.) Fern. Rare, in rocky prairies, along rocky streambeds and rocky
roadside ditches. Known from Chase, Geary, Pottawatomie and Riley counties.

Sphaeralcea coccinea (Pursh.) Rydb. Rare, in rocky prairie areas. Known from Cowley and
Riley counties.

HYPERICACEAE

Hypericum mutilum L. Occasional, in moist places in upland prairies, along margins of drain-
age areas and impoundments. Reported for Pottawatomie and Riley counties by Gates
(1940).

Hypericum perforatum L. Occasional, in rocky upland prairies, waste places, along railway
right-of-ways and roadsides. Found in Chase, Marshall, Pottawatomie and Riley counties.

Hypericum punctatum L. Occasional, in cultivated fields, waste places, along railroad right-of-
ways, roadsides and in open wooded areas. Known from Marshall, Pottawatomie and Riley
counties.

Hypericum sphaerocarpon Michx. Occasional, in rocky prairies, drainage ravines, and open
wooded areas. Known from Lyon, Pottawatomie and Wabaunsee counties.

ELATINACEAE
Bergia texana (Hook.) Seubert. Rare, in sandy or alluvia] soil. Reported for Pottawatomie
County by Gates (l'HO).

The Flora of the Kansas Flint Hills 567

CISTACEAE
Helianthemum bic\nellii Fern. Rare, in rocky, prairie areas and in dry, open woods. Reported

for Pottawatomie and Riley counties by Gates (19-10).
Lechea tenuijolia Michx. Rare, in rocky, open, wooded areas. Reported for Pottawatomie
County by Gates (1940).

VIOL ACE AE

Hybanthus verUcillatus (Ortega) Baill. Occasional, in rocky upland prairies and on rocky prairie
slopes. Found throughout the area.

Viola bicolor Pursh. Common, in upland prairies, prairie hay meadows, along roadsides and
railway right-of-ways. Found throughout the area.

Viola eriocarpa Schw. Occasional, in rocky wooded canyons and wooded stream valleys. Found
in Cowley, Lyon and Wabaunsee counties.

Viola missottriensis Greene. Occasional, in wooded canyons and wooded stream valleys. Known
from Cowley, Geary, Lyon and Riley counties.

Viola papilionacea Pursh. Occasional, in wooded canyons and wooded stream valleys. Found
throughout the area.

Viola pedatifida G. Don. Occasional, in upland prairies, on prairie slopes, along roadside em-
bankments and railway right-of-ways. Occurs in Cowley, Geary, Marshall, Morris, Potta-
watomie and Riley counties.

Viola sororia Willd. Occasional, in wooded canyons and wooded stream valleys. Found in
Geary, Riley and Wabaunsee counties.

LOASACEAE
Mentzelia oligosperma Nutt. Common, on rocky prairie slopes and rocky, roadside embank-
ments. Occurs throughout the region.

CACTACEAE

Opuntia compressa (Salisb.) Macbr. Common, in overgrazed upland prairies and on rocky

prairie slopes. Found throughout the area.
Mammillaria missouriensis Swett. Occasional, in upland prairies. Known from Chase, Riley and

Waubaunsee counties.
Mammillaria viviparia (Nutt.) Haw. Rare, in upland prairies. Known from Marion County.

ELAEAGNACEAE
Elaeagnus angustifolia L. Occasional, planted around farmyards and escaping to roadsides anil
other waste ground. Occurs throughout the area.

LYTHRACEAE
Ammannia auriculata Willd. Rare, on muddy margins of impoundments, streams, drainage areas

and roadside ditches. Known from Cowley and Rilcv counties.
Ammannia coccinea Rottb. Common, on muddy margins of impoundments, streams, drainage

areas and roadside ditches. Found throughout the area.
Lythrum alarum Pursh. Common, in wet prairie areas, along drainage areas, impoundments,

streambanks and roadside ditches. Found throughout the area.

ONAGRACEAE
Circaea qaadrisulcata (Maxim.) Franch. & Sav. Occasional, in rich floodplain woods. Known

from Marshall, Pottawatomie anil Riley counties.
Epilobium coloratum Biehler. Occasional, in wet ground along streams, impoundments and

low places. Known from Pottawatomie and Riley counties.
Epilobium leptophyllum Raf. Rare, in wet ground along streams, impoundments and low places.

Reported for Riley County by Gates (1940).
Gatira coccinea Pursh. Rare, in overgrazed prairies, along roadsides and railway right-of-ways.

Known from Geary and Riley counties.
Gatira filiformis Small. Common, in rocky upland prairies, along roadsides and railroad right-
of-ways. Found throughout the region.
Jussiaea repens L. Occasional, in shallow water and on muddy margins of impoundments.

Occurs throughout the area.
Ludwigia altcmijolia L. Rare, in low wet places, along the margins of impoundments and

streams. Known from Butler and Cowley counties.
Oenothera biennis L. Occasional, in cultivated fields, waste places, overgrazed prairies, along

railroad right-of-ways and roadsides. Occurs throughout the area.

568 The University Science Bulletin

Oenothera laciniata Hill. Occasional, in cultivated fields, waste ground, prairie canyons, along
railroad right-of-ways and roadsides. Occurs throughout the area.

Oenothera missouriensis Sims. Common, in rocky upland prairies, on rocky prairie slopes and
along rocky roadside embankments. Occurs throughout the area.

Oenothera rhombipetala Nutt. Rare, in sandy prairies. Reported for Riley County by Gates
(1940).

Oenothera semdata Nutt. Common, in upland prairies, on prairie slopes, along railroad right-
of-ways and roadsides. Encountered throughout the area.

Oenothera speciosa Nutt. Common to abundant, in overgrazed prairies, cultivated fields, waste
places, along railway right-of-ways and roadsides. Found throughout the area.

Oenothera triloba Nutt. Occasional, in upland prairies, on prairie slopes and along roadsides.
Known from Cowley County.

Stenosiphon linifolius (Nutt.) Britt. Common, on rocky prairie slopes and roadside embank-
ments. Occurs throughout the area.

HALORAGIDACEAE
Myriophyllum heterophyllum Michx. Occasional, in impoundments and slow-moving streams.

Reported for Riley County by Gates (1940).
Myriophyllum pinnatum (Walt.) BSP. Occasional in impoundments and slow-moving streams.

Known from Chase, Morris and Riley counties.

UMBELLIFERAE
Berala pusilla (Nutt.) Fern. Rare, along low wet areas and spring seepage areas. Reported for

Pottawatomie County by Gates (1940).
Chaerophyllum procambens (L.) Crantz. var. procumbens. Occasional, in rocky wooded can-
yons and wooded stream valleys. Found in Chase, Cowley, Riley and Wabaunsee counties.
Chaerophyllum tainturieri Hook. var. tainttirieri. Rare, in rocky wooded canyons and wooded

stream valleys. Known in Pottawatomie and Wabaunsee counties.
Chaerophyllum tcxanitm Coult. and Rose. Occasional, in rocky upland prairies, along roadsides

and railroad right-of-ways. Occurs throughout the area.
Cicitta maculata L. var. maculata. Occasional, along prairie drainage areas, spring seepage areas,

roadside ditches, margins of impoundments and streams. Occurs throughout the area.
Coniam maculatum L. Common, in lowland cultivated fields, waste ground, fence rows, road-
side ditches and railroad right-of-ways. Found throughout the area.
Cryptotaenia canadensis (L.) DC. Rare, in rocky wooded canyons and wooded stream valleys.

Reported for Marshall, Riley and Pottawatomie counties by Gates (1940).
Daticus carota L. Rare, in cultivated fields, waste ground and along roadsides. Known from

Morris, Pottawatomie and Riley counties.
Eryngium leavenworthii T. &. G. Occasional to common, on rocky prairie slopes along rocky

roadside embankments and rocky railroad right-of-ways. Found in Butler, Cowley, Chase,

Marion, Morris and Geary counties.
Eryngium yaccifoliam Michx. Occasional in moist prairie hay meadows and in prairie canyons.

Found only in Butler County.
Valcaria sioides (Wibel) Aschers. Rare, in waste places and along roadsides. Reported for

Mashall County by Gates (1940).
Lomatiitm joeniciilaceatn (Nutt.) Coult. & Roe. Common, in rocky upland prairies and on

rocky prairie slopes. Found throughout the area.
Osmorhiza longistylis (Torr.) DC. Occasional, in wooded canyons and rich stream valley woods.

Found in Cowley, Pottawatomie, Riley and Wabaunsee counties.
Polytaenia nuttallii DC. Common, in upland prairies, prairie hay meadows and prairie canyons.

Found throughout the area.
Sanicula canadensis L. Common, in rocky wooded canyons and wooded stream valleys. Found

throughout the area.
Sanicula gregaria Bicknell. Occasional, in rocky wooded canyons and wooded stream valleys.

Known from Lyon and Riley counties.
Sanicula marilandica L. Occasional, in rocky wooded canyons and wooded stream valleys. Occurs

in Lyon, Marshall, Riley and Wabaumee counties.
Spermolepis inermis (Nutt.) Math. & Const. Occasional, in upland prairies, prairie hay meadows,

along roadsides and railroad right-of-ways. Found throughout the area.
Thaspium trifoliatum (L.) Gray. Rare, in wooded canyons and rich stream valley woods. Re-
ported for Wabaunsee County by Gates (1940).
Zizia aitrea (L.) Koch. Occasional, in rocky wooded canyons and rich river valley woods. Occurs

throughout the area.

The Flora of the Kansas Flint Hills 569

cornaceae

Cormts amomum Mill. Rare, on rocky prairie slopes, in wooded canyons and along margins of

floodplain woods. Known from Marshall County.
Cormts drummondii C. A. Meyer. Common, grows in dense stands along limestone escarpments

on rocky slopes, in wooded canyons and along the margins of floodplain woods. Occurs

throughout the area.
Cormts obhqita Raf. Rare, grows along limestone escarpment on rocky slopes. Known from

Morris County.

PRIMULACEAE

Anagallis arvensis L. Rare, in cultivated fields and waste places. Reported from Lyon and Riley
counties by Gates (1940).

Androsace occidentalis Pursh. Occasional, in cultivated fields, rocky upland prairies and open
areas in wooded canyons. Found in Cowley, Marshall, Pottawatomie, Riley and Wabaunsee
counties.

Lysimachia ciliata L. Occasional, in low wet prairie areas, floodplain woods, along impound-
ments and streams. Found in Butler, Geary, Morris, Pottawatomie. Riley and Wabaunsee
counties.

SAPOTACEAE

Bumeliu lanuginosa (Michx.) Pers. Occasional, in wooded canyons. Known from Cowley

County.

EBENACEAE
Diospyros rirginiana L. Occasional, along margins of floodplain woods, in wooded canyons and

sometimes as scattered trees in prairie areas. Grows as a native in Cowley County and is

cultivated in Riley County.

OLEACEAE
Fraxinus americana L. Occasional, in floodplain woods. Found in Butler. Marshall. Morris. Riley

and Wabaunsee counties.
Fraxinus pennsylvanica Marsh var. subintegerrima (Vahl.) Fern. Common, in floodplain woods.

Occurs throughout the area.

GENTIANACEAE
Gentiana puberulenta Pringle. Occasional, in well-managed upland prairies, on prairie slopes and

in prairie canyons. Found in Geary, Marshall, Morris, Pottawatomie. Riley and Wabaunsee

counties.

APOCYNACEAE
Apocymtm cannabinum L. Common, in prairies, open wooded areas on rocky slopes, roadsides

and waste places. Occurs throughout the area.
Apocymtm sibtricitm Jacq. Occasional, in prairie canyons, wet upland prairie areas and in

alluvial soil along streambanks. Known throughout the area.

ASCLEPIADACEAE
Asclepias amplexicaulis Sm. Occasional, in upland prairies, open wooded areas on rocky slopes

and along roadsides. Reported for Geary, Pottawatomie, Riley and Wabaunsee counties by

Gates (1940).
Asclepias aspcrttla (Dene.) Woodson, spp. Capricorn (Woodson) Woodson. Occasional, in up-
land prairies and on rocky, prairie slopes. Known from Butler and Cowley counties.
Asclepias incamata L. Occasional, along the margins of streams and man-made impoundments.

Found in Cowley, Marshall, Pottawatomie and Riley counties.
Asclepias lanuginosa Nutt. Rare, in prairie areas. Reported for Rilev and Wabaunsee counties

by Gates (1940).
Asclepias speciosa Torr. Rare, in prairie canyons. Reported for Geary, Pottawatomie and Riley

counties by Gates (1940).
Asclepias stenophylla Gray. Common, in upland prairies, on rocky prairie slopes and in prairie

canyons. Found throughout the area.
Asclepias sullivantii Engelm. Occasional, in moist upland prairies, prairie hay meadows and in

prairie canyons. Known throughout the area.
Asclepias syriaca L. Common, along roadsides, railway right-of-ways, in overgrazed prairies, and

in open wooded areas. Occurs throughout the area.
Asclepias tuberosa L. Common, in upland prairies, prairie hay meadows, and in prairie canyons.

Encountered throughout the area.
Asclepias verticillata L. Common, in rocky upland prairies, on rocky prairie slopes, open wooded

slopes, along roadsides and railway right-of-ways. Found throughout the area.

570 The University Science Bulletin

Asclepias viridiflora Raf. Common, in rocky upland prairies and on rocky prairie slopes. Found

throughout the area.
Asclepias viridis Walt. Common, in upland prairies and on rocky prairie slopes. Encountered

throughout the area.
Cynanchum laeve (Michx.) Pers. Occasional, in floodplain woods. Occurs throughout the area.

CONVOLVULACEAE
Convolvulus arvensis L. Common, in overgrazed prairies, cultivated fields, waste places, along

ri^'ht-of-ways and roadsides. Occurs throughout the area.
Convolvulus septum L. Common, in moist soil of prairie canyons, margins of impoundments,

prairie hay meadows, wooded fence rows, along railway right-of-ways and roadsides. Found

throughout the area.
Cuscuta cephalanthi Engelm. Occasional, in moist ground on coarse herbs and shrubs. Reported

for Marshall County by Gates (1940).
Cuscuta coryli Engelm. Occasional, in moist ground on coarse herbs and shrubs. Reported for

Pottawatomie County by Gates (1940).
Cuscuta glomerata Choisy. Common, in moist ground on coarse herbs and shrubs. Found

throughout the area.
Cuscuta gronovii Willd. Occasional, in moist ground on coarse herbs and shrubs. Reported for

Marshall County by Gates (1940).
Cuscuta indecora Choisy. Occasional, in moist ground on coarse herbs and shrubs. Reported for

Geary County by Gates (1940).
Cuscuta pentagona Engelm. Occasional, in moist ground on coarse herbs and shrubs. Reported

for Cowley and Riley counties by Gates (1940).
Cuscuta polygonorum Engelm. Occasional, in moist ground on coarse herbs and shrubs. Known

from Geary, Pottawatomie, Riley and Wabaunsee counties.
Evolvulus nuttallianus R. & S. Common, in rocky upland prairies and on rocky, prairie slopes.

Found throughout the area.
Ipomoea hederacea (L.) Jacq. Common, in cultivated fields, waste places, stream valley woods

and along roadsides. Found throughout the area.
Ipomoea lacunosa L. Occasional, in cultivated fields, waste places, stream valley woods and

along roadsides. Known from Cowley, Morris, Pottawatomie and Riley counties.
Ipomoea leptophylla Torr. Rare, in sandy soil along streams and in low prairies. Reported for

Riley County 'by Gates (1940).
Ipomoea pandurata (L.) G. F. W. Meyer. Occasional, in cultivated fields, waste places, rocky

prairie slopes and along roadsides. Found throughout the area.
Ipomoea purpurea (L.) Roth. Common, in cultivated fields, waste places, floodplain woods and

along roadsides. Occurs throughout the area.

POLEMONIACEAE

Phlox bifida Reck. Rare, in wooded canyons. Reported for Pottawatomie County by Gates

(1940).
Phlox divaricata L. var. laphamii Wood. Common, in wooded canyons and floodplain woods.

Found throughout the area.
Phlox o/{lahomensis Wherry. Occasional, in rocky upland prairies, on rocky prairie slopes and

along rocky roadside embankments. Known from Butler and Cowley counties.
Phlox pilosa L. var. jtdgida Wherry. Occasional, in upland prairies, prairie hay meadows and

prairie canyons. Known from Marshall and Riley counties.

HYDROPHYLLACEAE

Ellisia nyctclea L. Common, in floodplain woods, low cultivated ground and low waste places.
Found throughout the area.

Hydrophyllum virginianum L. Occasional, in rich woods along streams. Known from Potta-
watomie County.

BORAGINACEAE

Cynoglossum officinale L. Rare, in overgrazed prairies, rocky waste places, and along roadsides.
Reported for Geary and Riley counties by Gates (1940).

Echium vulgare L. Rare, in waste places and along roadsides. Reported for Pottawatomie and
Riley counties by Gates (1940).

Hac/{elia virginiana (L.) Johnst. Common, on open wooded slopes, in floodplain woods and
lowland waste places. Found commonly in the northern part of the area while only occa-
sionally in the southern part.

The Flora of the Kansas Flint Hills 571

Helioiropium tenellum (Nutt.) Torr. Occasional, in rocky upland prairies, on rocky prairie

slopes and often growing in the crevices of outcropping limestone rocks. Known from Buder

and Cowley counties.
Lappula echinata Gilib. Rare, in waste places and cultivated fields. Reported to be in the Kansas

River valley in Riley County by Gates (1940).
Lappula redowskji (Hornem.) Greene var. occidentalis (Wats.) Rydb. Rare, in waste places,

along railroad right-of-ways, and in cultivated fields. Occurs in Geary, Pottawatomie and

Riley counties.
Lithospermum arvense L. Occasional, in waste places, cultivated fields, overgrazed prairies,

along roadsides and railroad right-of-ways. Found in Cowley and Chase counties.
Lithospermum canescens (Michx.) Lehm. Occasional, in rocky prairies, open wooded areas on

slopes and along roadsides. Known from Chase, Marshall and Riley counties.
Lithospermum incisum Lehm. Common, in rocky prairie, on rocky prairie slopes, open wooded

areas on rocky slopes and along prairie roadsides. Occurs throughout the area.
Myosotis virginica (L.) BSP. Rare, in cultivated fields, open wooded areas, along roadsides and

railroad right-of-ways. Reported for Lyon and Pottawatomie counties by Gates (1940).
Onosmodium molle Michx. Common, in rocky upland prairies, on rocky prairie slopes and in

open wooded areas on rocky slopes. Found throughout the area.

VERRENACEAE

Lippia cutlet folia (Torr.) Steud. Rare, along the margins of impoundments and streams. Re-
ported for Geary, Riley and Wabaunsee counties by Gates (1940).

Lippia lanceolata Michx. Occasional, along drainage areas, margins of impoundments and
streams. Occurs throughout the area.

Verbena bipinnattfida Nutt. Occasional, in rocky upland prairies, along roadsides and railway
right-of-ways. Reported for Covvlev, Geary, Pottawatomie, Riley and Wabaunsee counties by
Gates (1940).

Verbena bracteata Lag. & Rodr. Occasional along roadsides, cultivated fields and railway right-
of-ways. Found throughout the area.

Verbena canadensis (L.) Britt. Common, in rocky upland prairies, on rocky prairie slopes, on
rocky roadside embankments and along railroad right-of-ways. Found throughout the area.

Verbena hastata L. Occasional, along prairie drainage areas, in low prairie canyons, along mar-
gins of impoundments and streams. Occurs throughout the area.

Verbena simplex Lehm. Occasional in rocky upland prairies, on rocky prairie slopes, along
roadsides and railway right-of-ways. Found in Butler, Chase, Cowley, Lyon and Wabaunsee
counties.

Verbena stricta Vent. Common, in upland prairies, on prairie slopes, along roadsides and rail-
road right-of-ways. Occurs throughout the area.

Verbena urtictjolta L. Common, in floodplain woods and on wooded slopes. Found throughout
the area.

LABIATAE

Agastachc nepetoides (L.) Ktze. Occasional, in wooded canyons and rich woods along drainage
areas and streams. Found throughout tin- area.

Glechoma hederacea L. Rare, in waste places and thickets. Reported lor Riley County by Gates
(1940).

Hedeoma hispida Pursh. Common, in rocky upland prairies, on rocky prairie slopes, in prairie
canyons, waste places, along railway right-of-ways and roadsides. Found throughout the area.

Isanthus brachiatus (L.) BSP. Rare, along streambeds, on moist limestone escarpments in
prairie areas and in roadside ditches. Known from Chase, Cowley, Geary, Marshall, Potta-
watomie, Riley and Wabaunsee counties.

Lamtttm amplexicaule L. Common, in lawns, waste places, overgrazed prairies, prairie hay
meadows, along railway right-of-ways and roadsides. Occurs throughout the area.

Leonurus cardiaca L. Occasional, in cultivated fields, waste places, about farm dwellings, in
overgrazed pastures and along roadsides. Found in Marshall, Pottawatomie, Riley and
Wabaunsee counties.

Lycopus americanus Muhl. Common, in rich woods along streams, drainage areas and along
margins of impoundments. Found throughout the area.

Lycopus asper Greene. Rare, in wet soil especially in thickets. Reported for Marshall and Riley
counties by Gates (1940).

Lycopus uniflorus Michx. Rare, in moist soil. Reported for Pottawatomie County by Gates
(1940).

Lycopus t'irginicus L. Rare, in low wet woods and open ground. Reported for Pottawatomie
and Riley counties by Gates (1940).

572 The University Science Bulletin

Marrubium vulgare L. Occasional, in cultivated fields, waste ground, overgrazed prairies, and
around old dwellings. Found throughout the area.

Mentha arvensis L. Occasional, in wet soil along springs, drainage areas, streams and impound-
ments. Found throughout the area.

Monarda citriodora Cerv. Common, in rocky upland prairies and on rocky prairie slopes.
Known from Butler, Chase, Cowley and Lyon counties.

Monarda fisttdosa L. var. mollis (L.) Benth. Common to abundant, in rocky upland prairies,
along limestone escarpments on prairie slopes and sometimes in open wooded canyons.
Found throughout the area.

Nepeta cataria L. Occasional, in cultivated fields, waste places, open woodlands and along road-
sides. Found throughout the area.

Perilla jrutescens (L.) Britt. Rare, in cultivated fields, waste places and in rich woods along
streams. Reported for Riley County by Gates (1940).

Physostegia virginiana (L.) Benth. Rare, in moist places in upland prairies, along railway
right-of-ways and roadsides. Reported for Pottawatomie and Riley counties by Gates (1940).

Prunella vulgaris L. Common, along drainage areas, streambanks, impoundments and in road-
side ditches. Found throughout the area.

Pycnanthemum flexuosum (Walt.) BSP. Rare, in upland prairies, prairie hay meadows, prairie
canyons and sometimes in open woodlands. Reported for Riley County by Gates (1940).

Pycnanthemum virginianum (L.) Durand & Jackson. Rare, in rocky woods and thickets. Re-
ported for Pottawatomie County by Gates (1940).

Salvia azurea Lam. var. grandiflora Benth. Common, in rocky upland prairies, prairie hay
meadows, on rocky prairie slopes, railway right-of-ways and roadsides. Occurs throughout
the area.

Salvia reflexa Hornem. Common, in cultivated fields, overgrazed prairie pastures, waste places
and rocky roadsides. Found throughout the area.

Scutellaria lateriflora L. Rare, along streams and impoundments. Known from Butler, Potta-
watomie and Riley counties.

Scutellaria parvtda Michx. var. australis Fasset. Occasional, in rocky upland prairies and on
rocky prairie slopes. Found throughout the area.

Scutellaria parvula Michx. var. leonardi (Epling) Fern. Occasional, in rocky upland prairies and
on rocky prairie slopes. Found throughout the area.

Stachys tenuijolia Willd. Rare, in floodplain woods and wooded canyons. Reported Pottawatomie
and Riley counties by Gates (1940).

Teucrium canadense L. Common, in prairies, prairie hay meadows, wooded canyons, along
railway right-of-ways and roadsides. Found throughout the area.

SOLANACEAE
Datura metel L. Rare, in waste ground. Reported for Morris and Riley counties by Gates

(1940).
Datura stramonium L. Occasional to common, in farmyards, cultivated fields, waste ground and

open floodplains. Occurs throughout the area.
Lycium halimifolium Mill. Rare, escapes from cultivation to waste ground and along roadsides.

Reported for Marshall, Riley and Wabaunsee counties by Gates (1940).
Physalis angtdata L. var. pendula (Rydb.) Waterfall. Occasional, in open alluvial soil in low-
land areas, cultivated fields, waste ground and along roadsides. Found in Butler, Chase,

Marion, Morris and Riley counties.
Physalis hedcraejolia Gray var. comata (Rydb.) Waterfall. Rare, in prairie areas. Reported for

Geary and Riley counties by Gates (1940).
Physalis hctcrophylla Nees. Occasional, in overgrazed prairies, open areas in wooded canyons,

waste ground and along roadsides. Found throughout the area.
Physalis ixocarpa Brot. Rare, along railroad right-of-ways. Reported for Pottawatomie and

Riley counties by Gates (1940).
Physalis longijolia Nutt. var. suhglabrata (Mackcnz. & Bush) Cronq. Occasional, in overgrazed

prairies, open areas in wooded canyons, waste ground and along roadsides. Encountered

throughout the area.
Physalis missouriensis Mackenz. & Bush. Rare, in rich stream valley woods. Reported for

Geary, Marshall, Pottawatomie and Riley counties by Gates (1940).
Physalis pubescens L. Rare, in rocky wooded canyons, cultivated fields, waste ground and along

roadsides. Known from Chase, Geary and Riley counties.
Physalis pumila Nutt. Occasional, in rocky upland prairies, along roadsides and railway right-
of-ways. Found throughout the area.

The Flora of the Kansas Flint Hills 573

Physalis virginiana Mill. Occasional, in rocky wooded canyons, waste ground and along road-
sides. Occurs throughout the area.

Solatium carolinense L. Occasional, in waste ground, cultivated fields, overgrazed prairies and
along roadsides. Occurs throughout the area.

Solatium elaeagnifolium Cav. Occasional, in waste ground, cultivated fields, overgrazed prairies
and along roadsides. Found throughout the area.

Solatium rostratum Dunal. Common, in farmyards, severely overgrazed prairies, waste ground,
cultivated fields and along the roadsides. Occurs throughout the area.

Solatium torreyi Gray. Rare, in rocky or sandy open ground. Reported for Cowley County by
Gates (1940).

Solatium triflorum Nutt. Rare, in waste ground, cultivated fields and along roadsides. Reported
for Riley County by Gates (1940).

SCHROPHULARIACEAE

Bacopa rotundijolia (Michx.) Wcttst. Occasional, in alluvial soil along the margins of im-
poundments and streams. Known from Geary, Morris, Pottawatomie and Riley counties.

Castilleja sessiliflora Pursh. Rare, in rocky upland prairies and on rocky prairie slopes. Known
from Cowley, Pottawatomie and Riley counties.

Conobea multiftda Michx. Occasional, in alluvial soil along impoundments, streams, roadside
ditches and in open areas of lowland woods. Occurs throughout the area.

Dasistoma macrophylla (Nutt.) Raf. Rare, in wooded canyons and rich stream valley woods.
Reported for Pottawatomie, Riley and Wabaunsee counties by Gates (1940).

Gerardia aspera Dougl. Occasional, in moist upland prairies. Known from Butler, Chase, Geary.
Morris, Pottawatomie and Rilcv counties.

Gerardia dcnsiflora Renth. Rare, in rocky upland prairie anil on rocky prairie slopes. Known
from Chase and Lyon counties.

Gerardia gattingcri Small. Rare, in alluvial soil in roadside ditches along the margins of im-
poundments and streams. Found in Wabaunsee County.

Gerardia tcnuijlora Vahl. Occasional, in alluvial soil in roadside ditches, drainage areas, along
margins of impoundments and streams. Occurs in Morris, Pottawatomie, Riley and Wa-
baunsee counties.

Linaria vulgaris Hill. Occasional, in cultivated fields, overgrazed pastures, waste ground and
along roadsides. Known from Pottawatomie and Rilcv counties.

l.indcrnia unagallidea (Michx.) Pennell. Occasional, in alluvial soil along drainage areas, mar-
gins of impoundments and streams. Found in Rutler and Cowley counties.

Lindernia dubia (L.) Pennell. Occasional, in alluvial soil alone drainage areas, margins of im-
poundments and streams. Known from Cowley, Morris, Pottawatomie and Riley counties.

Mimulus alatus Ait. Rare, in alluvial soil along drainage areas, margins of impoundments and
streams. Collected from Rutler and Pottawatomie countie .

Mimulus glabratus HRK. Occasional, in alluvial soil along margins of impoundments and
streams. Found in Butler, Geary. Pottawatomie. Riley anil Wabaunsee counties.

Mimulus ringens L. Occasional, in alluvial -oil along drainage areas, margins ot impoundments
and streams. Encountered in Pottawatomie, Riley and Wabaunsee counties.

Pcnstemon cobaea Nutt. Common, in rocky upland prairies, on rocky prairie slopes and in open
areas of wooded canyons. Found throughout the area.

Penstemon digitalis Nutt. Rare, in upland prairies and on rocky prairie slopes. Known from
Rutler County.

Penstemon grandiflorus Nutt. Occasional, in rocky upland prairies, on rocky prairie slopes and
in open areas of wooded canyons. Occurs throughout the area.

Penstemon tubaeflorus Nutt. Occasional, in upland prairies, prairie hay meadows and in prairie
canyons. Found in Cowley, Lyon, Morris and Riley counties.

Scrophularia lanceolata Pursh. Rare, in rich floodplain woods. Reported for Riley County by
Gates (1940).

Scrophularia marilandica L. Common, in floodplain woods. Occurs throughout the area.

Verbascum blattaria L. Occasional, in cultivated fields, waste ground, severely overgrazed
prairies and along roadsides. Encountered throughout the area.

Verbascum thapsus L. Common to abundant, along cultivated fields, in waste ground, severely
overgrazed prairies and along roadsides. Found throughout the area.

Veronica anagallis-aquatica L. Rare, in alluvial soil along the margins of impoundments and
streams. Known from Cowley County.

Veronica arvensis L. Common, in cultivated fields, waste ground, severely overgrazed prairies
and along roadsides. Occurs throughout the area.

574 The University Science Bulletin

Veronica catena/a Pcnnell. Rare, in water and wet places. Reported for Morris and Riley coun-
ties by Gates (1940).

Veronica peregrina L. Common, in cultivated fields, waste ground, severely overgrazed prairies
and along the roadsides. Encountered throughout the area.

Veronica polita Fries. Rare, in cultivated fields, waste ground, severely overgrazed prairies and
along the roadsides. Reported for Geary County by Gates (1940).

Veronica triphylos L. Rare, in waste ground. Reported for Geary County by Gates (1940).

BIGNONIACEAE

Campsis radicans (L.) Seem. Occasional, in floodplain woods, and around abandoned farm-
yards. Often cultivated. Found throughout the area.

Justicia americana (L.) Vahl. Common, grows at the margins of streams or impoundments,
either in shallow water or mud. Found throughout the area.

Ruellia humilis Nutt. Common, in upland prairies, prairie hay meadows and prairie canyons.
Found throughout the area.

Ruellia strepens L. Common, in dense floodplain woods. Found throughout the area.

PHRYMACEAE

Phryma leptostachya L. Occasional, in wooded canyons and stream valley woods. Throughout
the area.

PLANTAGINACEAE

Plant ago aristata Michx. Occasional, in overgrazed prairies, open wooded canyons, along road-
sides and railroad right-of-ways. Found in Butler, Lyon, Morris, Riley and Wabaunsee
counties.

Plant ago lanceolata L. Occasional, in lawns, cultivated fields, waste ground and along roadsides.
Known from Lyon, Riley and Wabaunsee counties.

Plant ago patagonica Jacq. Common, in overgrazed prairies, waste places and along the road-
sides. Throughout the area.

Plantago rhodosperma Dene. Occasional, in overgrazed prairies and waste ground. Occurs
throughout the area.

Plantago rugelhi Dene. Common, in lawns, cultivated fields, overgrazed prairies, waste places
and along the roadsides. Throughout the area.

Plantago virginica L. Occasional, in overgrazed prairies and waste ground. Encountered
throughout the area.

RUBIACEAE

Cephalanthus occidentals L. Common, in floodplain woods, along the margins of impoundments
and streams. Occurs throughout the area.

Diodia teres Walt. Rare, in rocky soil of prairie drainage areas, alluvial soil along impound-
ments and streams. Known from Butler and Wabaunsee counties.

Galium aparine L. Common, in wooded canyons and in stream valley woods. Found through-
out the area.

Galium circaezans Michx. Common, in rocky wooded canyons and in stream valley woods.
Occurs throughout the area.

Galium conanum T. & G. Rare, in wooded canyons and moist stream valley woods. Known
from Wabaunsee County.

Galium triflorum Michx. Rare, in wooded canyons. Known from Wabaunsee County.

Galium virgatum Nutt. Occasional, in rocky upland prairie areas, on rocky prairie slopes and
on rocky roadside embankments. Known from Cowley County.

Hedyotis nigricans (Lam.) Fosberg. Common, in rocky upland prairies, on rocky prairie slopes,
in prairie hay meadows and along prairie railway right-of-ways. Occurs throughout the area.

CAPRIFOLIACEAE

Sambucus canadensis L. Common, in moist floodplain woods, wooded canyons, along stream-
banks, fence rows, roadsides and railway right-of-ways. Occurs throughout the area.

Symphoricarpos orbiculatus Moench. Common, in floodplain woods, wooded canyons, wooded
slopes and overgrazed prairies. Encountered throughout the area.

Triosteum perfoliatum L. Occasional, in wooded canyons and on wooded slopes. Known from
Cowley, Morris and Wabaunsee counties.

Viburnum rufidulum Raf. Occasional, in stream valley woods and in wooded canyons. Known
from Cowlev Countv.

CUCURBITACEAE

Cucurbita joetidissma HBK. Occasional, in overgrazed prairies, along railway right-of-ways
and roadsides. Found throughout the area.

The Flora of the Kansas Flint Hills 575

Echinocystis lobata (Michx.) T. & G. Occasional, in floodplain woods along drainage areas and
streams, along roadsides and in low waste places. Occurs throughout the area.

Melothria pendula L. Rare, in floodplain wooded areas. Known from Cowley County.

Sicyos angulatus L. Common, in stream valley woods, along drainage areas, in wooded canyons,
along roadsides and railroad right-of-ways. Encountered throughout the area.

CAMPANULACEAE
Campanula americana L. Common, in floodplain woods, moist wooded canyons and along the

margins of wooded areas. Occurs throughout the area.
Triodanis leptocarpa (Nutt.) Nieuwl. Common, in moist upland prairies, along prairie drainage

areas and in prairie canyons. Found throughout the area.
Triodanis perjoliata (L.) Nieuwl. Common, in moist upland prairies, along prairie drainage
areas, in prairie canyons, open alluvial soil along streams and along roadsides. Encountered
throughout the area.

LOBELIACEAE

Lobelia cardinalis L. Occasional, along wet streambanks, seepage areas and in wet roadside

ditches. Known from Butler, Chase, Cowley and Pottawatomie counties.
Lobelia siphilitica L. Occasional, along wet streambanks, margins of impoundments, seepage

areas, drainage areas and roadside ditches. Found in Marshall, Pottawatomie and Wabaunsee

counties.
Lobelia spicata Lam. var. hirtella Gray. Common, in moist upland prairies, along drainage areas

and in prairie canyons. Occurs throughout the area.
Lobelia spicata Lam. var. leptostachys (A. DC.) Mackenz. & Bush. Common, in moist upland

prairies, along drainage areas and in prairie canyons. Found throughout the area.

COMPOSITAE
Achillea millefolium L. Common, in upland prairies, abandoned fields, waste places and along

roadsides. Found throughout the area.
Ambrosia artemisiijolia L. Common to abundant, in cultivated fields, abandoned fields, severely

overgra7.ed pastures, waste places, along roadsides and railway right-of-ways. Occurs

throughout the area.
Ambrosia bidentata Michx. Occasional, in roadside ditches, overgrazed prairies, waste places and

along the margins of impoundments. Known from Cowley County.
Ambrosia psilostachya DC. Occasional, in roadside ditches and overgrazed prairies. Found

throughout the area.
Ambrosia trifida L. Common to abundant, in waste places, low wooded areas, fallow fields,

cultivated fields, along streams, roadsides and railroad right-of-ways. Occurs throughout

the area.
Antennaria neglecta Greene. Common, in rocky upland prairies, prairie hay meadows, on rocky

prairie slopes and in dry, open wooded areas. Found throughout the area.
Antennaria plant aginijolia (L.) Richards. Occasional, in rocky wooded canyons. Found in

Pottawatomie, Riley and Wabaunsee counties.
Anthemis cotula L. Occasional, introduced in overgrazed prairies, farmyards, cultivated fields,

waste ground, roadsides and railroad right-of-ways. Scattered throughout the area.
Arctium minus Schk. Occasional, in waste places, farmyards, disturbed wooded areas, and along

railroad right-of-ways. Known from the northern part of the area.
Artemisia biennis Willd. Rare, in alluvial soil of floodplain woods. Reported for Marshall,

Pottawatomie and Riley counties by Gates (1940).
Artemisia ludoviciana Nutt. Common, in rocky upland prairies, on rocky prairie slopes, in open

canyon woods, along roadsides and railway right-of-ways. Found throughout the area.
Aster azureus Lindl. Occasional, in rocky prairies, along prairie roadsides and in open wooded

canyons. Occurs in Pottawatomie, Riley and Wabaunsee counties.
Aster drummondii Lindl. Common, in floodplain woods, wooded canyons and along wooded

fence rows. Occurs throughout the area.
Aster ericoides L. Common to abundant, in rocky upland prairies, on rocky prairie slopes, in

prairie canyons, along roadsides, railway right-of-ways and sometimes in rocky, open

wooded areas on the slopes. Found throughout the area.
Aster laevis L. Occasional, in rocky upland prairies, on dry wooded, rocky slopes and along

roadsides. Occurs throughout the area.
Aster novae-angliae L. Rare, in low, alluvial ground along streams, and in low moist prairies.

Reported for Riley County by Gates (1940).
Aster oblongifolius Nutt. Common, in rocky upland prairies, on rocky prairie slopes, along

roadsides and railroad right-of-ways. Encountered throughout the area.

576 The University Science Bulletin

Aster pilosus Wilkl. Occasional, in cultivated fields, abandoned fields, prairie hay meadows,
disturbed rocky prairies and along roadsides. Found in Pottawatomie and Marion counties.

Aster pracaltus Poir. Common, in moist canyon prairies, along prairie drainage areas, seepage
areas and in moist roadside ditches. Found in the northern part in Cowley County.

Aster sericetis Vent. Common, in rocky upland prairies and on rocky prairie slopes. Occurs
throughout the area.

Aster simplex Wilkl. Common, in moist prairie areas, along drainage areas, seepage areas, fence
rows, margins of floodplain woods and moist roadside ditches. Encountered throughout
the area.

Aster vitninetis Lam. Rare, in moist prairie canyons, and alluvial soil along drainage areas,
streams and man-made impoundments. Known from Chase County.

Bidens bipinnata L. Occasional, in open wooded areas, moist areas in prairies, along roadsides
and railway right-of-ways. Known from Cowley and Pottawatomie counties.

Bidens cernua L. Occasional, wet places along streams, drainage areas and impoundments.
Found in Riley and Marshall counties.

Bidens comosa (Gray) Wieg. Occasional, wet places along drainage areas, streams and im-
poundments. Collected from Pottawatomie, Riley and Wabaunsee counties.

Bidens jrondosa L. Common, in floodplain woods, wet lowland prairies, along drainage areas,
streams, impoundments and roadside ditches. Found throughout the area.

Bidens polylcpis Blake. Common, in alluvial soil along streams, drainage areas, impoundments,
in cultivated fields, waste ground and roadside ditches. Found throughout the area.

Bidens vulgata Greene. Common, in alluvial soil along drainage areas, streams, impoundments,
waste places and roadside ditches. Found in the northern part.

Boltonia asteroides (L.) L'Her var. latisquama (Gray) Cronq. Occasional, in wet prairie can-
yons, along drainage areas, streams and roadside ditches. Known from throughout the area.

Cacalia atriplicifolia L. Occasional, in floodplain woods, along streambanks, drainage areas, in
open canyon woods and in wet roadside ditches. Found in the northern part and in Cowley
County.

Cacalia tuberosa Nutt. Common, in upland prairies, prairie hay meadows, and in prairie can-
yons. Occurs throughout the area.

Cardnns nutans L. Common to abundant, in severely overgrazed prairies, on rocky prairie
slopes, in cultivated fields, abandoned fields and other waste places. A common noxious weed
found in the northern part of the area, that is spreading southward.

Centaurea americana Nutt. Rare, in rocky prairies, open wooded areas, along roadsides and
railway right-of-ways. Reported for Butler County by Gates (1940).

Centaurea cyanus L. Rare, escapes from cultivation into roadsides, overgrazed prairies, cultivated
fields and other waste places. Reported for Riley County by Gates (1940).

Centaurea rcpens L. Rare, in waste places and cultivated ground. Reported for Cowley and
Riley counties by Gates (1940).

Chrysanthemum halsamita L. Rare, escapes from cultivation into waste places. Reported for
Riley County by Gates (1940).

Chrysanthemum leueanthemum L. Common, in cultivated fields, prairie hay meadows, grazed
prairies, open wooded areas, along railroad right-of-ways and roadsides. Found throughout
the area.

Chrysopsis villosa (Pursh.) Nutt. Rare, in rocky prairies and prairie roadsides. Found in
Butler County.

Cichorium intybus L. Occasional, in cultivated fields, along roadsides, railway right-of-ways and
other waste places. Known from Morris, Pottawatomie and Riley counties.

Cirsium altissimum (L.) Spreng. Common, in wooded canyons, stream valley woods, wet places
along roadsides and railroad right-of-ways. Occurs throughout the area.

Cirsium undtdatum (Nutt.) Spreng. Common, in upland prairies, prairie canyons, along road-
sides and railroad right-of-ways. Found throughout the area.

Cirsium vulgare (Savi) Tenore. Rare, in cultivated fields, grazed prairies, waste places and
along roadsides. Reported for Wabaunsee County by Gates (1940).

Conyza canadensis (L.) Cronq. Common, in cultivated fields, abandoned fields, overgrazed
prairies, waste places, along roadsides and railroad right-of-ways. Found throughout the area.

Conyza ramosissima Cronq. Occasional, in cultivated fields, rocky overgrazed prairies, waste
places, along railroad right-of-ways and roadsides. Occurs throughout the area.

Coreopsis grandiflora Hogg. Occasional, in upland prairies, prairie canyons and in open wooded
areas. Known from Cowley County.

Coreopsis tinctoria Nutt. Occasional, in sandy soil along streams, cultivated fields, along road-
sides and railroad right-of-ways. Found throughout the area.

The Flora of the Kansas Flint Hills 577

Dyssodia papposa (Vent.) Rydb. Common, in cultivated fields, overgrazed prairies, along rail-
road right-of-ways and roadsides. Found throughout the area.
Echinacea angustijolia DC. Common, in rocky upland prairies, on rocky prairie slopes and

along prairie roadsides. Occurs throughout the area.
Echinacea atrorubens Nutt. Occasional, in upland prairies, prairie hay meadows along prairie

roadsides and railroad right-of-ways. Found in Butler, Cowley, Lyon and Wabaunsee

counties.
Echinacea pallida Nutt. Common, in rocky upland prairies, on rocky prairie slopes along prairie

roadsides and railroad right-of-ways. Found in the southern part.
Eclipta alba (L.) Hassk. Occasional, along the margin of streams, impoundments and in wet

roadside ditches. Known from Butler, Cowley and Lyon counties.
Erechtites hieracijola (L.) Raf. Rare, along streambanks, in open woods and waste places.

Reported for Pottawatomie and Riley counties by Gates (1940).
Erigeron anntms (L.) Pers. Common, in upland prairies, prairie hay meadows, prairie canyons,

waste places, along railroad right-of-ways and roadsides. Found throughout the area.
Erigeron philadelphicus L. Common, in cultivated fields, stream valley woods, moist prairie

canyons, prairie hay meadows, along railway right-of-ways and roadsides. Occurs throughout

the area.
Erigeron strigosus Muhl. Common, in rocky upland prairies, on rocky prairie slopes, along

railroad right-of-ways and roadsides. Found throughout the area.
Eupatorium altis.fi/niim L. Common, in upland prairies, along prairie roadsides, in wooded

canyons and stream valley woods. Found throughout the area.
Eupatorium maculatum L. Rare, creek valley woods. Reported tor Pottawatomie County by

Gates (1940).
Eupatorium perjoliatum L. Rare, in rich woods along streams and in wooded canyons. Re-
ported for Pottawatomie and Riley counties by Gates (1940).
Eupatorium rugosum Houtt. Common, in rich woods along streams and wooded canyons.

Found throughout the area.
Evax prolijera Nutt. Rare, in rocky upland prairies and on rocky prairie slopes. Known from

Cowley County.
Franseria tomentosa A. Gray. Rare, in prairie canyons and moist places in upland prairies.

Reported for Marion County by Gates (1940).
Gaillardia pulchclla Foug. Rare, in sandy soil along the Kansas and Arkansas Rivers and along

railroad right-of-ways in Cowley and Riley counties.
Galinsoga ciliata (Raf.) Blake. Rare, a weed of gardens, yards and waste places. Reported for

Pottawatomie County by Gates (1940).
Galinsoga parrij/ora Cav. Rare, in waste places. Reported for Riley County by Gates (1940).
Gnaphalium obtusi folium L. Occasional, in abandoned fields, overgrazed prairies, open wood-
lands, along railway right-of-ways and roadsides. Found throughout the area.
Grindelia lanceolata Nutt. Rare, in overgrazed, rocky prairies, along railroad right-of-ways and

roadsides. Known from Cowley County.
Grindelia squarrosa (Pursh.) Dunal. Common, in waste places, overgrazed prairies, open stream

valley woods, along railroad right-of-ways and roadsides. Known from all counties except,

Cowley and Butler counties.
Gutierrezia dracunculoid.es (DC.) Blake. Common, in overgrazed prairies, cultivated fields,

waste places, along railroad right-of-ways and roadsides. Occurs throughout the area.
Haplopappus ciliatus (Nutt.) D.C. Occasional, in upland prairies, prairie canyons, open flood-
plains and low waste places. Found in Butler, Chase, Cowley, Lyon and Riley counties.
Helenium am arum (Raf.) Rock. Rare, in cultivated fields, waste grounds, along railroad right-
of-ways and roadsides. Known from Butler, Pottawatomie and Wabaunsee counties.
Hclianthus annuiis L. Common, in cultivated fields, abandoned fields, waste places, overgrazed

lowland prairies, along railway right-of-ways and roadsides. Encountered throughout the

area.
Hclianthus grosseserratus Martens. Common, in upland prairies, prairie canyons, along fence

rows, margins of impoundments, railroad right-of-ways and roadsides. Occurs throughout

the area.
Hclianthus hirsutus Raf. Common, in wooded canyons, stream valley woods, and along wooded

roadsides. Known from throughout the area.
Helianthus lactiflorus Pers. var. rigid us (Cass.) Fern. Common, in rocky upland prairies, on

rocky prairie slopes, along railroad right-of-ways and roadsides. Occurs throughout the area.
Hclianthus maximiliani Schrader. Common, in rocky upland prairies, on rocky prairie slopes,

in waste places, along railway right-of-ways and roadsides. Occurs throughout the area.

578 The University Science Bulletin

Helianthus mollis Lam. Rare, in prairie areas, thickets and barren ground. Reported for Riley

County by Gates (1940).
Helianthus petiolaris Nutt. Occasional, in overgrazed prairies, open areas along streams, waste

places, along railway right-of-ways and roadsides. Known from Chase, Cowley, Riley and

Wabaunsee counties.
Helianthus salicijolius A. Dietr. Common, in rocky upland prairies, along the limestone es-
carpments on the prairie slopes, along railway right-of-ways and roadsides. Found through-
out the area.
Helianthus tuberosus L. Common, in wooded areas along streams and in wooded canyons.

Found throughout the area.
Heliopsis helianthoides (L.) Sweet. Common, in stream valley woods, wooded canyons, along

wooded fence rows, railway right-of-ways and roadsides. Encountered throughout the area.
Heterotheca latijolia Buckl. Rare, in sandy alluvial soil along the Arkansas River floodplain.

Known from Cowley County.
Hieracium longipilum Torr. Occasional, in moist upland prairies and in prairie lowlands.

Occurs throughout the area.
Hymenopappus scabiosaeus L'Her. Common, in rocky upland prairies, along limestone escarp-
ments and in roadside ditches. Encountered throughout the area.
Iva annua L. Occasional to common, in alluvial soil along streams, impoundments, in prairie

canyons, roadside ditches and railroad right-of-ways. Found throughout the area.
Iva xanthifolia Nutt. Rare, in waste places and along streams. Reported for Riley County by

Gates (1940).
Krigia oppositijolia Raf. Rare, in prairies, thickets and open sandy ground. Reported for Morris,

Riley and Wabaunsee counties by Gates (1940).
Kuhnia eupatorioides L. Common, in overgrazed upland prairies, on rocky prairie slopes, in

open wooded canyons, and along rocky roadside embankments. Occurs throughout the area.
Lactuca canadensis L. Common, in wooded canyons, along impoundments, streams, lowland

fields and roadsides. Encountered throughout the area.
Lactuca floridana (L.) Gaertn. Occasional, in rich woods along streams, wet lowland areas and

along roadsides. Found throughout the area.
Lactuca ludoviciana (Nutt.) DC. Common, in rich wooded areas along drainage areas and

streams. Found throughout the area.
Lactuca scariola L. Common, in cultivated fields, abandoned fields, waste places, along roadside

ditches and railway right-of-ways. Occurs throughout the area.
Liatris aspera Michx. Common, in upland prairies, prairie meadows, sometimes in prairie

openings in wooded areas and along railway right-of-ways. Encountered throughout the area.
Liatris mucronata DC. Common, in shallow soil above limestone outcrops in prairie areas.

Known from throughout the area.
Liatris punctata Hook. Occasional, in upland prairies and along prairie roadsides. Found

throughout the area.
Liatris pycnostachya Michx. Occasional, in moist places in upland prairies, low prairie hay

meadows, along moist railway right-of-ways and roadsides. Known from throughout the

area.
Liatris squarrosa (L.) Michx. var. glahrata (Rydb.) Gaiser. Rare, in upland prairies and open

wooded canyons. Reported for Pottawatomie and Riley counties by Gates (1940).
Liatris squarrosa (L.) Michx. var. hirsuta (Rydb.) Gaiser. Rare, in upland prairies and open

wooded canyons. Collected in Marshall County and reported for Marshall, Pottawatomie and

Riley counties by Gates (1940).
Lygodesmia juncea (Pursh) D. Don. Rare, in sandy prairie areas. Reported for Geary and

Riley counties by Gates (1940).
Microseris cuspidata (Pursh) Schulta-Bip. Occasional, in rocky upland prairies and along the

roadsides. Found in Cowley, Geary, Lyon, Morris, Pottawatomie, Riley and Wabaunsee

counties.
Parthenium hispidum Raf. Common, in rocky upland prairies, on rocky prairie slopes and along

roadsides. Known from Butler and Cowley counties.
Prenanthes aspera Michx. Common, in open wooded canyons and rocky prairie areas. Found

throughout the area.
Pyrrhopappus carolinianus (Walt.) DC. Rare, in cultivated fields, open floodplain woods, moist

railway right-of-ways and roadsides. Known from Pottawatomie and Butler counties.
Pyrrhopappus scaposus DC. Common, in upland prairies, prairie canyons, and along prairie

railway right-of-ways. Found in Butler, Cowley and Riley counties.
Ratibida columnijera (Nutt.) Woot. & Standi. Common, in upland prairies, on rocky prairie

slopes, in prairie canyons, along railroad right-of-ways and roadsides. Occurs throughout

the area.

The Flora of the Kansas Flint Hills 579

Ratibida pinnata (Vent.) Barnh. Rare, in wooded areas along streams, wooded canyons and

along the margins of wooded areas. Known from Lyon, Marshall and Wabaunsee counties.
Rudbeckja amplexicaulis Vahl. Rare, in moist prairies, sandy soil along Arkansas River flood-
plain, and along roadsides. Known from Cowley County.
Rudbeckia hirta L. Occasional to common, in open wooded canyons, prairie canyons, prairie hay

meadows and along railway right-of-ways. Occurs throughout the area.
Rudbeckia laciniata L. Rare, in low open woods and wooded areas along streams. Reported for

Marshall, Pottawatomie and Riley counties by Gates (1940).
Rudbeckia triloba L. Rare, in wooded plains along streams and in canyon wooded areas. Known

from Lyon County.
Senecio plattensis Nutt. Common, in upland prairies, prairie hay meadows, prairie canyons and

along railway right-of-ways. Encountered throughout the area.
Silphium integrifolium Michx. Common, in open places of wooded canyons, rocky prairie

canyons and along railway right-of-ways. Encountered throughout the area.
Silphium laciniatum L. Common, in well-managed upland prairies, prairie hay meadows and

prairie canyons. Occurs throughout the area.
Silphium perfoliatum L. Occasional, in wooded areas along streams, along impoundments and

moist railroad right-of-ways. Known from Marshall, Pottawatomie and Riley counties.
Silphium speciosum Nutt. Occasional to common, in upland prairies, on rocky prairie slopes, in

prairie canyons, along railway right-of-ways and roadside ditches. Found throughout the

area.
Solidago altissima L. Common, in waste places, cultivated fields, abandoned fields, overgrazed

prairies, open wooded canyons, along railroad right-of-ways and roadsides. Found through-
out the area.
Solidago canadensis L. var. gilvocanescens Rydb. Occasional, in cultivated fields, abandoned

fields and open stream valley woods. Occurs throughout the area.
Solidago gigantea Ait. var. serotina (Kuntze.) Cronq. Occasional, in floodplain woods, margins

of impoundments and in wet places of prairie canyons. Encountered throughout the area.
Solidago gramintjolia (L.) Salisb. Common, in overgrazed, upland prairies, cultivated fields

and along railroad right-of-ways. Found throughout the area.
Solidago gymnospermoides (Greene) Fern. Occasional, in overgrazed prairies and along railway

right-of-ways. Known from Chase, Pottawatomie, Riley and Wabaunsee counties.
Solidago missouriensis Nutt. var. fasciculata Holzinger. Common, in rocky upland prairies, on

rocky prairie slopes and on rocky roadside embankments. Occurs throughout the area.
Solidago nemoralis Ait. Occasional, in open canyon woods, on rocky prairie slopes and along

railroad right-of-ways and roadsides. Known from the northern part of the area.
Solidago petiolaris Ait. Occasional to common, in rocky upland prairies, on prairie slopes, along

railroad right-of-ways and roadsides. Occurs commonly in the northern part of the area,

while only occasionally in the southern part.
Solidago rigida L. var. rigida. Common, in overgrazed prairies, cultivated fields, abandoned

fields, along railroad right-of-ways and roadsides. Found throughout the area.
Solidago rigida L. var. humidis Porter. Occasional, in overgrazed prairies, cultivated fields,

abandoned fields, along railroad right-of-ways and roadsides. Known from Lyon and Riley

counties.
Solidago speciosa Nutt. var. rigidiusula T. & G. Occasional, in rocky upland prairies, on rocky

prairie slopes and along roadsides. Found throughout the area.
Solidago ulmtjolia Muhl. Common, in wooded canyons and in wooded areas along streams.

Known from Cowley County.
Sonchus asper (L.) Hill. Occasional, in cultivated fields, abandoned fields, waste places, farm-
yards, along railway right-of-ways and roadside ditches. Occurs throughout the area.
Sonchus oleraceus L. Rare, in overgrazed prairies, cultivated fields, waste places, along railway

right-of-ways and roadsides. Known from Chase County.
Tanacctum vulgare L. Rare, escapes from cultivation into waste places, farmyards, prairie hay

meadows, cultivated fields and along roadsides. Reported for Marshall, Morris and Potta-
watomie counties.
Taraxacum erythrospermum Andrz. Common, in lawns, gardens, cultivated fields, overgrazed

prairies, along railroad right-of-ways and roadsides. Occurs throughout the area.
Taraxcum officinale Wiggers. Common to abundant, in lawns, gardens, cultivated fields, over-
grazed prairies, railway right-of-ways and roadside ditches. Encountered throughout the

area.
Thelesperma filijolium (Hook.) A. Gray. var. intermedium (Rydb.) Melchert. Occasional, in

rocky upland prairies and on rocky prairie slopes. Known from Butler, Chase, Cowley and

Marion counties.

580 The University Science Bulletin

Thelesperma megapotamicam (Spreng.) Kuntzc. Occasional to common, in rocky upland
prairies and on rocky prairie slopes. Found throughout the area.

Tragopogon ditbius Scop. Common, in waste places, cultivated fields, along railroad right-of-
ways and roadsides. Found throughout the area.

Tragopogon porrijolius L. Occasional, in cultivated fields, abandoned fields, waste places and
along roadsides. Known from Cowley, Riley and Wabaunsee counties.

Verbesina alternifolia (L.) Britt. Common, in wooded areas along streams and drainage areas.
Found throughout the area.

Verbesina encelioides (Cav.) Benth. & Hook. Rare, on rocky prairie slopes and in prairie can-
yons. Collected once from Butler County by Dr. R. L. McGregor and reported for Riley
County by Gates (1940).

1 'erbesina virginica L. Rare, in wooded canyons and in floodplain woods. Known from Cowley
County.

Vernonia baldwini Torr. var. interior (Small) Schubert. Common, in upland prairies, prairie hay
meadows, rocky prairie slopes, waste places, along railway right-of-ways and roadsides.
Found throughout the area.

Vernonia jascicidata Michx. Occasional, in upland prairies, prairie hay meadows, prairie can-
yons, along railway right-of-ways and roadsides. Known from Chase, Marshall, Potta-
watomie, Riley and Wabaunsee counties.

Xanthium chinense Mill. Occasional in cultivated fields, waste places, open floodplains and
roadside ditches. Known from Geary, Marshall, Pottawatomie and Riley counties.

Xanthium itdlicum Moretti. Occasional, in cultivated fields, waste places and along roadsides.
Reported for Riley County by Gates (1940).

Xanthium pennsylvanicum Wallr. Occasional, in cultivated fields, waste places and along road-
sides. Known from Geary, Marshall, Morris, Pottawatomie, Riley and Wabaunsee counties.

Xanthium speciosum Kearney. Occasional in cultivated fields, waste places and along roadsides.
Known from Chase and Morris counties.

The Flora of the Kansas Flint Hills

581

STATISTICAL SUMMARY
A. Tabular list of Kansas Flint Hill plant families:

Fa m Hies

Genera

Species

Native Introduced

Pteridophytes:

Conifers:

Monocots:

Dicots

Equisetaceae 1

Ophioglossaceae 2

Polypodiaceae 6

Marsileaceae 1

Totals J0_

Cupressaceae 1_

Totals 1_

Typhaceae 1

Sparganiaceae 1

Najadaceae 1

Alistmataceae 3

Gramineae 55

Cyperaceae 8

Araceae 2

Lemnaceae 3

Commelinaceae 2

Pontederiaceae 1

Juncaceae 2

Liliaceae 1 1

Iridaceae 3

Orchidaceae 3

Totals - 96

Salicaceae 2

Juglandaceae 2

Corylaceae 2

Fagaceae 1

Ulmaceae 2

Moraceae 4

Urticaceae 5

Santalaceae 1

Polygonaceae 3

Chenopodiaceae 6

Amaranthaceae 3

Nyctaginaceae 1

Phytolaccaceae 1

Aizoaceae 1

Portulacaceae 2

Caryophyllaceae 8

Ceratophyllaceae 1

Nymphaceae 2

Annonaceae 1

Ranunculaceae 8

Rerberidaceae 1

Menispermaceae 2

Papaveraceae 2

Fumariaceae 2

Cruciferae 22

Capparidaceae 2

Crassulaceae 1

Saxifragaceae 1

Platanaceae 1

Rosaceae 9

Leguminosae 3 1

Oxalidaceae 1

3
2

3

6

88

55

2

4

5

3

6

15

2

_4

195

8

5
2
9
5
3
5
2
16
8
5
3
1

2

4
1

2
1
17
1
2

2

17
2
1
2
1
29
54
3

35
1

38
1

1
2

7
4
3

2
20

2
15

582 The University Science Bulletin

Species
Families Genera Native Introduced

Geraniaceae 2

Linaceae 1

Zygophyllaceae 1

Rutaceae 2

Simarubaceae 1

Polygalaceae 1

Euphorbiaceae 5

Callitrichaceae 1

Anacardiaceae 1

Celastraceae 2

Staphyleaceae 1

Aceraceae 1

Hippocastanaceae 1

Sapindaceae 1

Balsaminaceae 1

Rhamnaceae 2

Vitaceae 3

Tiliaceae 1

Malvaceae 6

Hypericaceae 1

Elatinaceae 1

Cistaceae 2

Violaceae 2

Loasaceae 1

Cactaceae 2

Elaeagnaceae 1

Lythraceae 2

Onagraceae 7

Haloragidaceae 1

Umbelliferae 1 5

Cornaceae 1

Primulaceae 3

Sapotaceae 1

Ebenaceae 1

Oleaceae 1

Gentianaceae 1

Apocynaceae 1

Aslepidaceae 2

Convolvulaceae 4

Polemoniaceae 1

Hydrophyllaceae 2

Roraginaceae 8

Verbenaceae 2

Labiatae 19

Solanaceae 4

Schrophulariaceae 12

Rignoniaceae 2

Orobanchaceae 1

Lentibulariaceae 1

Martyniaceae 1

Acanthaceae 3

Phrymaceae 1

Plantaginaceae 1

Rubiaceae 4

Caprifoliaceae 4

Cucurbitaceae 4

Campanulaceae 2

Lobeliaceae 1

Compositae 61

Totals 340"

1

1

1

1

..

1

2

1

2

..

26

2

1

4

..

2

..

1

..

2

1

1

2

3

..

7

1

4

6

3

1

1

..

2

6

1

1

..

3

..

1

..

3

15

2

17

3

3

2

1

1

..

1

..

2

1

..

2

13

12

3

4

2

7

4

9

20

7

14

3

22

6

1

1

2

1

I

..

4

1

5

1

8

4

4

..

3

3

124

20

The Flora of the Kansas Flint Hills

583

B. Components of the Flora of the Kansas Flint Hills:

Major Groups of Tracheophyta

Species

Genera

Families

Pteridophytes 14 10

Conifers 1 1

Monocots 233 96

Dicots 739_ 349

Totals 987 456

4

1

14

91

110

C. Largest families with total number of species in each:

Compositae 144

Gramineae 123

Leguminosae 69

Cyperaceae 55

Cruciferae 37

Rosaceae 31

Euphorbiaceae 28

Schrophulariaceae 28

Labiatae 27

Polygonaceae 23

Umbelliferae 20

Ranunculaceae 20

Solanaceae 17

Liliaceae 16

Onagraceae 15

Convolvulaceae 15

Asclepidaceae 13

Caryophyllaceae 13

Chenopodiaceae 12

Roraginaceae 11

LITERATURE CITED

Adams, G. I. 1903. Physiographic divisions of Kansas. Trans. Kansas Acad. Sci. 18:109-123.
Agrelius, F. U. G. 1945. Botanical notes. Trans. Kansas Acad. Sci. 49:107-108.
Albertson, F. W., and G. W. Tomanek. 1965. Vegetation changes during a 30-year period on
grassland communities near Hays, Kansas. Ecology 46:714-720.

Anderson, K. L. 1951. The effects of grazing management and site conditions on Flint Hills

Bluestem Pastures in Kansas. Ph.D. thesis. University of Nebraska, Lincoln, Nebraska,

pp. 1-2.
. 1953. Utilization of grasslands in the Flint Hills of Kansas. Jour. Range Management

6:86-93.
. 1965. Time of burning as it affects soil moisture in an ordinary upland bluestem prairie

in the Flint Hills. Jour. Range Management 18:311-316.
, and C. L. Fly. 1955. Vegetation-soil relationships in Flint Hills Bluestem pastures.

Jour. Range Management 8:164-169.

Bidwell, O. W. 1966. The Flint Hills range sites. Trans. Kansas Acad. Sci. 69:205-213.

Birkholz, D. W. 1968. Flora of Sumner County, Kansas, unpublished Master's Thesis, Kansas
State Teachers College, Emporia, Kansas, 67 pp.

Caldwell, M. B. 1937. The southern Kansas boundary survey. Kansas Historical Quarterly

6:339-377.
Carruth, J. H. 1877. Centennial catalogue of the plants of Kansas. Trans. Kansas Acad. Sci.

5:40-59.

Fenneman, N. M. 1939. Physiography of Eastern United States. McGraw-Hill, New York.
714 pp.

Fish, M. C. 1953. Differential effect of presence or absence of glaciation upon the pteropsida of
three adjacent Kansas counties. Trans. Kansas Acad. Sci. 56:203-219.

Flora, S. D. 1948. Climate of Kansas. Rpt. of the Kansas St. Bd. Agric. 67:1-320.

Fremont, J. C. 1845. Report of the Exploring Expedition to the Rocky Mountains in the Year

1842 and to Oregon and North California in the Years 1843-1844. Gales and Seaton,

Printers, Washington, D.C. 693 pp.

Frye, J. C, and A. B. Leonard. 1952. Pleistocene geology of Kansas. Kansas Geol. Survey
Bull., 230 pp.

584 The University Science Bulletin

. 1957. Ecological interpretations of pliocene and pleistocene stratigraphy in the Great

Plains Region. Amer. Jour. Sci. 255:1-11.
Gates, F. C. 1940. Flora of Kansas. Contrib. no. 291, Dept. of Botany, Kansas State College,

Manhattan, Kansas, 266 pp.
. 1945. Amp hiac hyris dracunatloides as a poisonous plant. Trans. Kansas Acad. Sci.

48:87-89.
Gibson, E. S. 1963. Vascular flora of Crawford County, Kansas. Trans. Kansas Acad. Sci.

66:685-726.
Gier, H. T. 1967. Vertebrates of the Flint Hills. Trans. Kansas Acad. Sci. 70:51-59.
Gleason, H. A., and A. Cronquist. 1963. Manual of Vascular Plants of Northeastern United

States and Adjacent Canada. D. Van Nostrand, New York. 810 pp.
Hitchcock, A. S. 1899. Flora of Kansas. Kans. State Agr. College, Manhattan, Kansas, 25 pp.
Hoyle, W. L. 1938. Drought in Cowley County, Kansas. Trans. Kansas Acad. Sci. 40:211-214.
Jewett, J. M. 1941. The geology of Riley and Geary counties, Kansas. Kansas Geol. Surv. Bull.

39, 164 pp.
Kollmorgen, W. M., and D. S. Simonett. 1965. Grazing operations in the Flint Hills-Blue-

stem pastures of Chase County, Kansas. Ann. Assoc. Amer. Geographers 55:260-290.
Kuchler, A. W. 1967. Some geographic features of the Kansas prairies. Trans. Kansas Acad.

Sci. 70:388-401.
Lathrop, E. W. 1958. The flora and ecology of the Chautauqua Hills in Kansas. Univ. Kansas

Sci. Bull. 39:97-209.
Malin, J. C. 1942. An introduction to the history of the Bluestem-Pasture Region of Kansas.

Kansas Historical Quarterly 1 1:3-28.
. 1967. The Grassland of North America Prolegoma to its History with Addenda and

Postscript. Peter Smith, Gloucester, Mass. 490 pp.
Maus, P. M. 1929. Flora of Wabaunsee County, Kansas. Trans. Kansas Acad Sci, 32:88-104.
McGregor, R. L. 1955. Taxonomy and ecology of Kansas Hepaticae. Univ. Kansas Sci. Bull.

37:55-141.
Miller, N. H. 1931. Surveying the southern boundary line of Kansas. Kansas Historical

Quarterly 1:104-139.
Morley, G. E. 1964. A floristic study of Republic County, Kansas. Trans. Kansas Acad. Sci,

67:716-746.
Schoewe, W. H. 1949. The geography of Kansas: part II, physical geography. Trans. Kansas

Acad. Sci. 52:261-333.
Stevens, O. A. 1917. Plants of Manhattan and Blue Rapids, Kansas; with dates of flowering.

Amer. Mid. Nat. 5:71-116.
Taft, R. 1948. Over the Santa Fe Trail through Kansas in 1858. Kansas Historical Quarterly

16:337-380.
Weaver, J. E. 1954. North American Prairie. Johnsen Publishing Company, Lincoln,

Nebraska. 347 pp.
Wedel, W. R. 195^. An Introduction to Kansas Archaeology. Smithsonian Institution Bu-
reau of American Ethnology Bull. 174. United States Government Printing Office, Wash-
ington, D.C. 723 pp.

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLVIII Pages 585-687 October 31, 1969 No. 15

Skulls of Gymnophiona and Their Significance
in the Taxonomy of the Group1

Edward Harrison Taylor2

TABLE OF CONTENTS

PAGE FIGURE

Abstract 587

Introduction 587

The Bones of the Skulls of Gymnophiona 589 1

Description, of Skulls 591

Ichthyophidae Taylor 591

Ichthyophis Fitzinger 592

glutinosus (Linnaeus) 592 2

beddomei Peters 593 3

singaporensis Taylor 593 4

\ohtaoensis Taylor 593 5-6

mindanaoensis Taylor 595 7

Caudacaecilia Taylor 595

nigroflava (Taylor) 595 8

asplenia Taylor 596 9

larutensis (Taylor) 596 10

iveberi (Taylor) 596 1 1

Typhlonectidae Taylor 597

Typhlonectes Peters 598

compressicaudus Dumeril and Bibron 598 12

natans Fischer, in Peters 599 13

Potamotyphlus Taylor 599

kaupii (Berthokl) 600 14

Scolecomorphidae Taylor 601

Scolecomorphus Boulenger 602

vittatus Boulenger 602 15

1 This study was pursued under a National Science Foundation Grant, GB 4510.

2 Research Associate, Kansas University Museum of Natural History.

586 The University Science Bulletin

PACE FIGURE

\ir\ii Boulenger 603 16

Caeciliidae Gray 603

Subfamily Caeciliinae Taylor 604

Caecilia Linnaeus 604

nigricans Boulenger 604 17

tentaculata Linnaeus 605 18

degenerata Dunn 605 19

disossea Taylor 607 20

volcani Taylor 607 21

orientalis Taylor 607 22

albiventris (Daudin) 607 23

Oscaecilia Taylor 608

ochrocephala (Cope) 608 24-25

bassleri (Dunn) 609 26

Subfamily Dermophinae Taylor 610

Dermophis Peters 610

eburatus Taylor 611 27

costaricensis Taylor 611 28

balboai Taylor 612 29

parviceps (Dunn) 612 30

glandnlosus Taylor 612 31

occidentcdis Taylor 612 32

mexicanus mexicanus Dumeril and Bibron 614 33

Gymnopis Peters 614

multiplicata Peters 615 34

proximo. (Cope) 616 35

Siphonops Wagler 616

annulatus (Mikan) 616 36-38

paulensis Boettger 617 39

Geotrypetes Peters 617

seraphini seraphini (A. Dumeril) 618 40

seraphini occidentalis Parker 619 41

Herpele Peters 619

squalostoma (Stutchbury) 620 42

Uraeotyphlus Peters 620

oxyurus (Dumeril and Bibron) 620 43

Schistometopum Parker 622

gregorii (Boulenger) 622 44-45

Hypogeophis Peters 622

rostratus rostratus (Cuvier) 623 46

Grandisonia Taylor 623

alternans (Stejneger) 624 47

Skulls of Gymnophiona and Their Significance 587

sechellensis (Boulenger) 624 48

Idiocranium Parker 625

russeli Parker 625 49

Boulengeruh Tornier 627

boulengen Tornier 627 50

Ajrocaecilia Taylor 627

uluguruensis (Barbour and Loveridge) 628 51

taitana (Loveridge) 628 52

changamwensis (Loveridge) 628 53

Gegeneophis Peters 629

ramaswamii Taylor 629 54-55

ABSTRACT

The skulls of 48 species and subspecies of Caecilians are reviewed and three
photographs of each are presented. Significant differences arc demonstrated in
the skulls, which would warrant the taxonomic arrangement proposed. Four
families are recognized.

INTRODUCTION

Although considerable anatomical study has been done on the skulls of
certain genera of the Gymnophiona, skulls of numerous genera are as yet
unknown, owing to a dearth of specimens available for such study. I have,
with the help of certain museums and individuals, brought together the
skulls of nearly four dozen species representing 4 families and 20 genera.
For a number of years we have been wont to regard the caecilians as repre-
senting a single family and in such case perhaps committing ourselves to
the premise that the evolution of the now Gymnophiona has all taken place
since they arrived at their present form and habitat. With an examination of
the literature descriptions of known species, and the skull studied here, one
is convinced that, of the present members of the group, doubtless several had
undergone a considerable evolution prior to becoming refugees in subter-
ranean food habitats where competition for food was less severe, and where
the loss of limbs and girdles was an adaptation essential to a subterranean
existence. Some forms seemingly left the subterranean habitat for a com-
pletely aquatic one. It is likewise true that some anatomical studies on the
skulls, though excellent, have not been directed in the immediate direction
of taxonomy, since the workers were seemingly less concerned with the
taxonomic implications than with other aspects.

In my study of caecilians (Taylor, 1968), having utilized such characters
as were apparent from preserved specimens (such as body shape, size, folds,
scales, proportions and dentition, and, to a lesser extent, reproductive organs,

588 The University Science Bulletin

gills, gill slits, eye position, character and position of the tentacle and snout
projection, etc.), it became obvious to me that evolution had differentiated
the Gymnophiona into several groups of greater significance than genera.
Three such groups were considered as families, the Ichthyophidae, Typhlo-
nectidae and Caeciliidae, while a fourth, still little known, likewise appeared
that it might be worthy of admission to such a category.

Since this publication appeared, I have proposed that this fourth group
be recognized as a new family, the Scolecomorphidae. I have also proposed
that the family Caeciliidae, sensu strictu, be divided into two subfamilies,
the Dermophinae and Caeciliinae (Taylor, 1969b).

The present study was undertaken to ascertain to what extent the gross
anatomy of the skull displayed the evolution that was evident in other
anatomical areas. Rather than write detailed descriptions of conditions ob-
taining in each of the skulls, I offer figures, which, in a way, serve as a
universal language.

The following abbreviations are used for the names of various collections
containing caecilians:

AMNH : American Museum of Natural History, New York, New York.

BMNH: British Museum (Natural History), London, England.

EHT-HMS: Edward H. Taylor-Hobert M. Smith Herpetological Col-
lection, Lawrence, Kansas.

KUMNH: Kansas University Museum of Natural History, Lawrence,
Kansas.

MCZ: Museum of Comparative Zoology, Harvard College, Cambridge,
Massachusetts.

DSBM: Division of Systematic Biology, Museum, Stanford University,
Stanford, California.

UIM: University of Illinois Museum, Urbana, Illinois.

Skulls of Gymnophiona and Their Significance 589

THE BONES OF THE SKULLS OF GYMNOPHIONA

Figure 1 is a diagrammatic drawing of a representative gymnophionan
skull showing the names of the bones as referred to in this paper.

The skulls of caecilians seem to have undergone a reduction in numbers
of bones when compared with many other primitive amphibians, a reduc-
tion largely brought about by fusion, or in some cases perhaps by complete
loss. For instance the skulls of some generic groups have certain dermal
bones which in others are lacking. The usual interpretation is that the
missing bones have been fused to surrounding or contiguous bones or that
they have "dropped out." One may sometimes trace a bone through a series
of related genera to find it diminishing in size until it no longer appears, or,
as in the case of the splenial of certain salamanders, it seems as if the bone,
with its numerous teeth, is resorbed or eliminated completely during the
lifetime of the individual animal.

One common observation of paleontologists and others who treat of skull
anatomy is that the earlier geological forms of many groups are very likely
to have the greater number of bones; that later forms, usually considered
more specialized, have fewer bones. By such a criterion one might consider
the family Ichthyophidae as a more primitive group (i.e., retaining more
of the primitive characters). However, this would not necessarily bespeak a
greater age.

In the Gymnophiona one of the recognized family groups, Ichthyophidae,
has pairs of separate dorsal dermal bones, the prefrontals, orbitals (or post-
frontals), septomaxillae, premaxillae, and nasals. Two other groups recog-
nized as families, Typhlonectidae and Caeciliidae, lack this series of separate
bones. The first three bones are fused to neighboring elements or have
dropped out completely. In the case of the nasals and premaxillae, these
obviously become fused to form a pair of nasopremaxillae.

The remaining bones of the skull likewise show obvious fusions in all
four recognized families. The occipital bones, the otic bones (except stapes)
and the otic capsules fuse with the sphenoid bones to form a compound
element which is here designated the basisphenoid and serves as the major
part of the braincase.

The stapes are present in three of the recognized family groups but are
absent in the fourth (Scolecomorphidae).

The maxillary bones are invariably fused to the palatines, and are here
called the maxillopalatines, the palatine portion being occasionally referred
to as the palatine shelf.

The family Scolecomorphidae has retained the septomaxillae, separate
nasals and premaxillae but seeming invariably lacks the prefrontals, the
oculars and, strangely enough, the stapes.

590 The University Science Bulletin

The pterygoids are sometimes discernible as separate elements, sometimes
they seem to be more or less solidly fused to the quadrate. A tiny ectoptery-
goid may or may not be present. The quadrate is invariably present, sutured
to the squamosal and presenting a surface for the articulation of the lower
jaw, as well as a surface or a notch for contact with the stapes (except in
the Scolecomorphidae).

The bones of the various groups are fitted together by an overlapping
suture or they may be narrowly separated and held together by a cartilage.
Sometimes these regions of contact become still more widened by the inter-
vention of musculature. These separations are described as diastemata. One
occurs frequently between the squamosal and the parietal. Similarly, the
same may happen between the basisphenoid and pterygoid with the palatine
shelf of the maxillary. There may be typical sutures, overlapping or inter-
digitating cartilage covered openings, or there may be very wide diastemata,
so that the posterior side walls of the basisphenoids are exposed and the
orbitosphenoids, usually not seen, are largely exposed (Scolecomorphidae).

Another variable element is the mesethmoid, which may be completely
covered by the dorsal dermal bones or may appear on the dorsal midline of
the skull, exposing little or much surface, and serving, often, to separate the
frontals, or the nasopremaxillae. Occasionally the anterior part of the pari-
etals may likewise be separated for a distance.

In making the comparisons, I have chosen only a part of the characters.
Many others might be mentioned as, for instance, the blood and nerve fene-
strae, shape and size of teeth, bone sculpturing, and the internal skull struc-
tures. Most of the characters I have used are visible in the cleaned skull
without a dissection or dismemberment of the cranium.

Unfortunately, skulls of many of the species are unknown to me or
known through only a single skull. Where several skulls of a single species
were available, the variation was very meager.

In the Figures, often it is a bit difficult to follow the skull sutures. Per-
haps the greatest disappointment is in the teeth. Being at different levels in
the skull, rarely are all the teeth in focus at the same time with the magnifi-
cation used. Moreover, in many of the skulls, many of the teeth are absent,
having been lost in the preparation of the skull. Usually the numbers can
be counted and their relative size discerned in the skull itself. In the Figures
this is not true to the same extent. The reason for this condition is that the
teeth are jointed and the distal portions may be lost.

It will be noted that I have not utilized the lower jaws of these specimens
in this discussion. The jaws were photographed and studied. The material
was on the whole very unsatisfactory inasmuch as the fragile jaws were
almost invariably broken in the specimens that had been skeletonized. This
was perhaps due to the opening of the closed jaw for the original identifica-

Skulls of Gymnophiona and Their Significance 591

tion of the species of the preserved specimen. Occasionally, the specimens
had been damaged before preservation. The teeth on the lower jaw were
often destroyed by the opening of the jaw.

These elements do contribute characters for the differentiation of certain
genera and families, but the illustrations, which are most necessary, do not
show the characters adequately or at least not to the satisfaction of the author,
hence the omission.

DESCRIPTIONS OF SKULLS
Family Ichthyophidae

The family characteristics as displayed by the skulls of Ichthyophis and
Caudacaecilia examined are as follows: the number of separate bones is
greater than in other known families of the Gymnophiona. Beginning dor-
sally on the snout tip are the large paired nasals, separated anteriorly by a
small wedge formed on the upper tips of the premaxillae; a pair of frontals
follow the nasals, and these in turn are followed by a pair of parietals,
usually the largest dorsal bones; on each side of the median series are
(anteriorly) a septomaxilla bordering the anterior nares, a prefrontal, and an
ocular which may be completely free or partially fused to the squamosal;
the latter is a large lateral bone. This is preceded by the maxillopalatine,
that borders the mouth laterally and reaches forward to contact the pre-
maxillaries, which border the anterior part of the mouth. The posterior part
of the skull roof is formed from the sides of the posterior brain case (basi-
sphenoid) which surrounds the large foramen magnum, but the continuity
is broken dorsally by a suture in the bone between the foramen and the
parietals. Behind the squamosal, and attached to it, is the quadrate. Just
posterior to the quadrate is the stapes, whose base serves as part of the brain
case wall and sends forward a projection that forms a suture with the
quadrate.

The brain case is largely composed of a single compound bone, here des-
ignated the basisphenoid, which has incorporated the various occipital,
sphenoid and otic bones and possible certain others. It forms a large part of
the floor and sides of the brain case and sometimes, in certain families,
reaches forward to the prevomerine tooth series. Ventrally there is a con-
striction of the basisphenoid preceding the otic capsules, anterior to which
the bone forms two more or less distinct lateral "wings" which are in contact
with the quadrates and the pterygoids. The anterior border of the ventral
surface of the skull is formed by the premaxillae and the maxillopalatines.
The premaxillae are followed by paired prevomers which are relatively very
large in this family, often extending farther back than the internal nares and
usually if not always forming a part of their border. The major part of the

592 The University Science Bulletin

border of the nares is the palatine shelf of the maxillopalatine. Posteriorly
the shelf contacts the pterygoid, which posteriorly joins the quadrate either
by suture or in certain cases seemingly by fusion.

The tooth-bearing bones of the skull proper are the premaxillae, the
maxillary and palatine sections of the maxillopalatines, and the prevomers.
(In certain families where an ectopterygoid is present, it may also bear
teeth.) The lower jaws each consist of two compound bones, the dentary,
which has incorporated the splenial, and the articulare. In this family this
bone bears a series of dentary teeth (usually the largest of all in size) and
usually a splenial series (sometimes called the inner mandibular series).
This latter may be absent completely in certain families while in other
families, including the Ichthyophidae, it may be absent in only some of
the genera.

Ichthyophis Fitzinger

Ichthyophis Fitzinger, Neue Classification der Rcptilien. . . . Wien, 1826, pp. 36, 63. Type
species Coealia hypocyanea van Hasselt.

Fitzinger assigned a single species to this genus. Peters (1879) recognized
three species, Ichthyophis glutinosus (Linnaeus), Ichthyophis heddomei
Peters and Epicriitm monochroum Bleeker. Boulenger, who reviewed the
caecilians in 1882, as one of his maiden efforts, placed all Asiatic forms of
the genus having a lateral stripe in the species glutinosus (Linnaeus) ; all
lacking the stripe in the species monochrous (Bleeker).

The genus as now constituted has a large number of forms recognized as
species which are confined to southern and southeastern Asia and adjacent
island groups. The cranial skeleton is known to me from the skulls of
several forms which I recognize as species. This genus differs from the
Caudacaecilia chiefly in the absence of the splenial teeth in the adult.

Ichthyophis glutinosus (Linnaeus)
(Fig. 2)
Caecilia glutinosa Linnaeus, Systema Naturae . . . , 10th Ed., vol. 2, 1758, p. 229. Type-
locality "Habitat in Indiis." It appears to be most probable that the tvpe-locality is Ceylon
(see Taylor, 1965).

Data here given are from the skull of KUMNH No. 31291, Tonacumbe
Estate, Numunukula, Ceylon.

The general characters agree with those listed for the family and genus.
The skull bones show a rather strong overlap on the dorsal surface, the
anterior elements overlap the bones behind them and the frontals have two
narrowed processes pushing forward between the prefrontals and the nasals.
The parietals maintain the same width until they narrow posteriorly. The
parietals and squamosals have a very narrow separation, bridged by carti-
lage; the ocular bone is separate and complete (the bone of this skull figured
is broken partially on one side). The septomaxillae border the anterior nasal

Skulls of Gymnophiona and Their Significance 593

openings. The tentacular foramen is separate from the orbit but close to it,
directed outward and downward rather than forward. The prevomers are
not narrowed and not or scarcely separated posteriorly by the spine of the
basisphenoid. The pterygoid, which is not fused to the quadrate, sends a
narrow process forward that reaches to the internal nares. I do not discern
an ectopterygoid and believe it normally absent. The stapes, which has a
fenestra through the bone, projects forward into a notch in the quadrate.
The "wings" of the basisphenoid are rather poorly developed, and the otic
capsules are inflated very little. The tooth rows are subparallel. See Table 1.

Ichthyophis beddomei Peters
(Fig. 3)

Ichthyophis beddomei Peters, Monatsb. Akad. Wiss. Berlin, 1879, p. 932, fig. 4. Type-locality,
Nilgherrie Hills, India.

Data given here are from the skull of EHT-HMS No. 3186, Kotegehar,
India.

The skull resembles /. glutinosus in most characters. The tentacular
aperture is somewhat farther forward or, rather, the groove continues farther
forward, so that when the tentacle reaches the surface, it is distinctly farther
from the eye. The stapes is fenestrated. There are two anterior fenestrae in
the prevomers instead of one as in glutinosus. See Table 1.

Ichthyophis singaporensis Taylor
(Fig. 4)
Ichthyophis singaporensis Taylor, Univ. Kansas Sci. Bull., I960, vol. 40, pp. 55-58, figs. 5, 6.
Type-locality, Singapore Island.

Data are from the skull of the type-specimen, BMNH No. RR 1959.1. 2.43.
(figured in Taylor, loc. cit., fig. 6) .

The ocular bone actually is circular, but on one side the bone tends
to fuse to the maxillopalatine. The parietals are notched, constricted
near their anterior end, behind which they attain their greatest width. The
skull is more compact than in /. glutinosus, not or scarcely narrowing
anteriorly. The anterior process of the pterygoid is narrower and shorter,
not reaching to the level of the internal nares; the squamosals are shorter
and wider than in the preceding species. See Table 1.

Ichthyophis \ohtaoensis Taylor

(Figs. 5-6)

Ichthyophis \ohtaoensis Taylor, Univ. Kansas Sci. Bull., 1960, vol. 40, pp. 110-113, fig. 38.
Type-locality, Koh Tao Island, Gulf of Siam.

Data are from the skull of EHT-HMS No. 3935, from about 10 mi. north
of Chiang Doi, N. Thailand.

The skull tends to narrow anteriorly as in /. glutinosus. The orbit is not
completely surrounded by the ocular, and the orbit and the tentacular groove
are continuous. The fenestrae, posterior to the internal nares and separating

594

The University Science Bulletin

O

3

J4
in

J3

o
u

o

o

-a

G
M

s

a

3

CO

<

~c

On -L- <g

to

« -a

k ^ n c

<lj LT\ T1 '— ' -^

O ^ i-V bcrG

"o ?t 1c H

W(j

bo
C

c/3

ft MD

C/3

f- >. C

IS"

w

5 "1 J G

V C

a 3

3 a

3 C

in -r o
^ i< vo

q n x rj o
n f 't i- «i

00

—

^r

m

t~~» ^ t^ 1^

in _

—

no

CM

CnI

in

oroxqi/imtNiNin
fnooNOr/->t^minr<-)(NO

1^ o no no r^

3C On ~^ — < — h (vj .— 1
o „ ^ ^ ^

~ no'

in O

t- in

t^ in -t- r-^

in j^ — on —1 rN .-h

t- o t^. i-o in o o 00 r^

Ot^NOONOinmr/lco

r^

X -< (O rt

On

ro

in f>

1-

N ^O X Ml

O

m

o o

r-5 {<

in q X m o N tNl
in o no t in t~ ro

"1

t^.

in

O

in

n-

n..

J^ 00
in

00

-1-

—

-t-

-f

-1-

qq«i(Ni*NiAiAm

CN)

in in >^i

OO ■— ' OO -H (N)

N IT ft \t n.

a

o -c

u

0

--t-i

o

>- _y

o

G
V

-c :
a p

"w C

" V. ^J .

'3 ki Vi 9

~
O

—

c
o

c _e

.2 2

i <*

C I-

a

•g a 5

c_ •£

bo
C
u

>

rt

W)

c

O

«

«

"o

G

IS

0

u

~

_G

Si

-£

-s

So

0

a

£

c
0
-

1

^

c

g g g:

r* CJ . — X ttj G

W).

G

~ '-* 73 U ' —

H a *s a a

2 '5

g J^;

^ a

1—, t/)

Skulls of Gymnophiona and Their Significance 595

the pterygoid and basisphenoid, are proportionally larger, and the prevomers
are proportionally longer than in glutinosits. Table 1 gives data on this and
one other skull.

Ichthyophis mindanaoensis Taylor

(Fig. 7)

Ichthyophis mindanaoensis Taylor, Univ. Kansas Sci. Bull., 1960, vol. 40, pp. 69-74, figs. 13-15.

Type-locality, Todaya, Mt. Apo, Davao, Mindanao, Philippine Islands. Elev. 2800 ft.

Data are from the skull of DSBM No. 20926, from 11 km. SE of Buena
Suerte, on side of Dapitan Peak, Mindanao, P.I. Elev. 3700 ft.

This is similar to the preceding /. \ohtaoensis, but in this the ocular bone
is seemingly fused to the squamosal for much of its region of contact, and
the tentacular aperture is separated from the orbit. When the cartilage of the
squamosal-parietal suture is removed, there is a very narrow diastema be-
tween these bones. The prevomers are shorter than in /. \ohtaoensis. See
Table 1.

Caudacaecilia Taylor

Caudacaecilia Taylor, Caecilians of the World, 1968, p. 165. Type of genus, Ichthyophis
nigroflarns Taylor.

Taylor referred five species to this genus: Ichthyophis nigroflewus, which
was designated as type, /. paucidentulus Taylor, /. weberi Taylor, /. laruten-
sis Taylor and Caudicaecilia asplenia Taylor. The range of this series of
species is spread from the western Philippine Islands through Borneo,
Sumatra, and the Malay Peninsula. All differ from Ichthyophis in the
absence of splenial teeth.

These have the general skull characters of the family Ichthyophidae;
that is, the presence of prefrontals, septomaxillae, oculars and separate pre-
maxillae and nasals. The oculars may be partially fused.

The maxillaries and palatines are fused to form the maxillopalatines, and
the basisphenoid has incorporated the various occipital and otic bones and
the otic capsules to form the basisphenoid that serves as the major part of the
brain case. The squamosals, frontals, parietals, quadrates and stapes appear
dorsally or laterally; ventrally there are prevomers and pterygoids. The
mesethmoid is not visible dorsally.

Skulls are available for four of the five known species as follows:

Caudacaecilia nigroflava (Taylor)
(Fig. 8)

Ichthyophis nigroflavus Taylor, Univ. Kansas Sci. Bull., I960, vol. 40, pp. 101-103. Type-
locality, near Kuala Lumpur ("within 20 miles"), Selangor, Malaya.

The data are from the skull of EHT-HMS No. 1734, Bukit Lagong
Forest Reserve, Selangor, Malaya.

The skull is relatively broad, not tapered or tapering but very little
anteriorly; the premaxillae are visible dorsally between the nasal openings;

596 The University Science Bulletin

the nasals are narrowed posteriorly; the frontals are nearly as long as the
parietals, which widen somewhat posterior to the squamosals. The pre-
frontals form relatively broad sutures with the septomaxillae. The stapes,
partly hidden from above by a shelf of the dorsal part of the basisphenoid,
push forward into a rather deep notch in the quadrate. The pterygoids
reach to near the edge of the internal nares; no ectopterygoid is present. The
orbital rim of eye and the tentacular aperture and groove are continuous.
The ocular bones, if considered present, are fused to the squamosals which
largely surround the eye. The diastemata following the internal nares are
large. The otic capsules are not noticeably inflated. The tooth rows are
subparallel, with the inner series extending farther posteriorly than the
maxillae. See Table 2.

Candacaecilia asplenia Taylor
(Fig. 9)

lchthyophis asplenias Taylor, Univ. Kansas Sci. Bull., 1965, vol. 46, pp. 278-283, figs. 14-15.
Type-locality, "Boven Mahakkam," Borneo.

Data are from the skull of EHT-HMS No. 1373, La Doo Tin Mine, Yala
Province, Thailand.

This differs from the preceding species in having the orbit almost com-
pletely separated from the squamosal by a separate ocular bone, the tentacular
aperture distinctly separate from the orbit, and the frontals distinctly shorter
than the parietals. The maxillopalatine borders the orbit for a short distance.
The skull is slenderer, narrowing more anteriorly than in C. nigroflava. On
the left pterygoid there are two small teeth (anomaly?). See Table 2.

Caudacaecilia larutensis (Taylor)
(Fig. 10)
lchthyophis larutensis Taylor, Univ. of Kansas Sci. Bull., 1960, vol. 40, p. 44, figs. 1,2. Type-
locality, Maxwells Hill, Larut Hills, 3380 ft. elev. Perak, Malaya.

Data are taken from the skull of EHT-HMS No. 3359, Topotype.

The ocular, if present, is fused more or less completely with the squamosal.
The maxillopalatine is excluded from the orbit and the tentacular aperture is
not continuous with the orbit. Dorsally the frontals are as long as or longer
than the parietals, with the latter scarcely widening behind the posterior level
of the squamosal. The skull tends to be narrower than that of nigroflava,
especially anteriorly. See Table 2.

Caudacaecilia weberi (Taylor)
(Fig. 11)

lchthyophis icc/'cri Taylor, Philippine Jour. Sci., 1920, vol. 16, p. 227; type-locality, Malatgan
River, Palawan, Philippines. (Type destroyed.)

Data are taken from the skull of DSBM No. 21764. (Paratype of neo-
type.)

The prevomers are slightly wider posteriorly and do not extend beyond
the posterior level of the internal nares; the orbit and the tentacular aper-

Skulls of Gymnophiona and Their Significance

597

Table 2. Measurements in mm and tooth counts in skulls of Caudacaecilia.

Species nigroflara asplenia larutensis ueberi

Number 1734 1373 3359 21764

Museum EHT-HMS EHT-HMS EHT-HMS DSBM

Locality Selangor, Yala, Perak Palawan,

Malaya Thailand Malaya Philippines

Total length of skull 12.0 9.5 11.0 9.0

Greatest width 8.0 6.3 6.8 5.4

Width at orbits 7.4 4.2 6.0 5.0

Length of jaws 12.4 10.5 11.1 8.0

Length of basisphenoid 7.4 6.7 6.2 5.4

Width at "wings" 5.4 4.0 4.2 3.9

Width at otic capsules 5.2 4.3 5.0 4.4

Length of prevomers 4.3 3.3 3.9 3.3

Combined width prevomers .. 4.4 3.5 3.4 3.15

Front edge of internal

nares to condyle 7.9 6.2 7.0 6.0

Length of specimen 385 238 252 209

Premaxillary teeth 11-10 7-8 9-8 8-10

Maxillary teeth 17-15 16-16 19-20 17-17

Prevomerine teeth 13-12 10-10 11-10 11-11

Palatine teeth 15-15 14-15 18-18 13-13

Dentary teeth 18-? 21-21 20-20 20-21

Splenial teeth 0-0 0-0 0-0 0-0

ture are seemingly connected by a narrow groove, and the ocular if present
is fused completely to the squamosal. The maxillopalatines border the orbits
narrowly. The frontals are shorter than the parietals. The general skull
proportions seem to be more similar to C. nigroflcwa than to the other two
species. See Table 2.

Family Typhlonectidae

The general characteristics of the skull in the Typhlonectidae based on
the type genus and Potamotyphhts are as follows: The premaxillary and
nasal bones are fused and seemingly also have incorporated the septomaxilla.
The frontals are largely in contact their entire length, the mesethmoid not
visible. The parietals are in contact their entire length. The posterior dorsal
portion of the skull is formed by the dorsal part of the basisphenoid, which
is sutured above the foramen magnum.

Laterally the maxillopalatine has an anterior opening for the tentacle and
posteriorly it surrounds the eye socket for about three fourths of its circum-
ference. This is followed by the squamosal, separated from the parietal by
a wide diastema, and bordered posteriorly by the quadrate. There is a small
external stapes (this element has been accidentally lost on one side) on the
skull of Potamotyphhts.

Seen from the ventral surface, the anterior portion of the skull is occupied

598 The University Science Bulletin

by the nasopremaxillae. Following this are the two prevomers, bearing the
prevomerine teeth on their anterior borders, then narrowing as they extend
backward between the huge internal nares, bordering them on their upper
anterior edges for about one fifth to one third of their circumference or
perhaps more in Potamotyphlus.

The lateral portion of the ventral surface is occupied by the maxillopala-
tine, the palatine portion largely surrounding the internal nares (choanae)
and posteriorly separated by an irregular rather narrow diastema from the
basisphenoid (partly cartilage covered) or without a noticeable diastema in
Potamotyphlus. The pterygoid is at least partially fused to the quadrate.
The basisphenoid is large with distinct lateral "wings" bordering the ptery-
goid; it runs forward, and anteriorly as a narrow point, tends to separate the
prevomers for a greater or lesser distance (in T. nutans separated nearly to
the prevomerine teeth).

Typhlonectes Peters

Typhlonectes Peters, Monatsb. Akad. Wiss. Berlin, 1879, pp. 930, 940. Type of genus, Caecilia
compressicauda Dumeril and Bibron.

Peters (1879) originally referred four species to the genus: Caecilia com-
pressicauda Dumeril and Bibron, Caecilia natans Fischer, in Peters, Siphon-
ops syntremus Cope, and Caecilia dorsalis Peters.

Of these the first two, T. compressicauda and T. natans are now recog-
nized in the genus. Caecilia dorsalis, a synonym of Caecilia Itaupii Berthold,
is placed in a different genus, Potamotyphlus; and Siphonops syntremus has
likewise been placed in a new genus, Copeotyphlinus. Formerly it had been
placed in Typhlonectes with a ? by Peters, in Dermophis by Cope (1866),
and treated as incertae sedis by Dunn (1942).

Peters recognized the most salient generic features: body lacking scales;
eye in an orbit, not covered by bone; tentacular aperture opening close to
nostril; splenial teeth present. Both lungs well-developed; foetus with
bladder-shaped gills. Aquatic.

Certain other species have been added to the genus: Thyphlonectes [sic]
venzuelense Fuhrman, Chthonerpeton microcephalum Miranda-Ribeiro, Ty-
phlonectes eiselti Taylor, Typhlonectes obesus Taylor, and Typhlonectes
anguillajormis Taylor. On the authority of Dr. Joseph Bailey, in Dunn
(1942), Chthonerpeton microcephalum is a synonym of Typhlonectes
kaupii=Potamotyphlus \aupii (Berthold).

The skulls of two species of Typhlonectes are known to me, T. com-
pressicauda and T . natans.

Typhlonectes compressicaudus (Dumeril and Bibron)

(Fig. 12)
Caecilia compressicauda (Dumeril and Bibron), Erpetologie Generate, . . . 1841, vol. 8, p.
278 (type-locality, Cayenne).

Skulls of Gymnophiona and Their Significance 599

The skull is from a specimen, EHT-HMS No. 1731, Belem, Brasil. Since
the type came from a different drainage system, one cannot be absolutely
certain that this Amazonian population agrees in detail with that from
the Guianas.

The characters agree with those of the family: the skulls bones are
reduced in number, the prefrontals, septomaxillae, and oculars do not
appear, while the premaxillae and nasals are fused to form the two naso-
premaxillae. As in all gymnophiones the occipitals, otics, and the otic
capsule and sphenoids are fused with the basisphenoid in making the brain
case. The palatine and the maxillary are fused; small stapes are present.
The pterygoid may be free or fused at least partially to the quadrate. A very
small ectopterygoid may be present. The orbits are cut chiefly in the maxil-
lopalatines but are bordered posteriorly by the squamosals; the tentacular
aperture notches the anterior part of the same bone, with the groove con-
tinuing along the nasopremaxilla. The external narial openings are pro-
portionally large. A wide diastema is present between the parietal and
squamosal. The internal nares are of very large size, surrounded, except
anteriorly, by the palatine shelf of the maxillary. The prevomers, which
border the inner anterior edge, are rather narrowed posteriorly but are
separated for about a fourth of their length by the anterior point of the
basisphenoid. Seemingly the pterygoid is at least partly fused to the quad-
rate. The wings of the basisphenoid are distinct. See Table 3.

Typhlonectes nutans Fischer, in Peters
(Fig. 13)
Typhlonectes nutans Fischer, in Peters, Monatsb. Akad. Wiss. Berlin, 1879, p. 9-11. Type-
locality, Rio Cauca, Colombia.

Data are from the skull of MCZ No. 24524 from "Cucuta," Colombia.

Compared with the previously described skull, T. natans differs in having
a proportionately much narrower pair of frontals and still narrower pari-
etals which are less wide than the frontals; the very large diastema between
the squamosal and the parietal is as wide as the parietal itself, the greater
part of its length. The prevomers have a slight projection anterior to the
prevomerine teeth, much narrowed posteriorly and terminating in points
posterior to the back level of the nares. These are separated for about two
thirds of their length. The pterygoid seemingly is not fused to the quadrate.
See Table 3.

Potamotyphlus Taylor

Pqtamotyphlus Taylor, Caecilians of the World. 1968, pp. 256-257. Type of genus, Caecilia
kaupii Berthold.

This genus comprises two known species that have the following charac-
ters: no scales, no secondaries, head relatively small with a long slender
neck, a dorsal skinfold, eyes in sockets, tentacular aperture very close to large

600 The University Science Bulletin

Table 3. Measurements in mm and tooth counts of the skulls of Typhlonectes

and Potamotyphlus.

Species T. compressicaudus T. natans P. /iaupii

Number 1731 24524 787

Museum EHT-HMS MCZ UIM

Locality Belem, Brasil Cucuta, Col. (Uncertain)

Total length of skull 16.5 22.4 11.1

Greatest width 11.2 14.8 5.5

Width at middle of orbits 10 12 5.05

Length of jaw 14.2 21.5 ? broken

Length of basisphenoid 11.1 15.5 7.2

Width at "wings" 7.7 8.9 3.5

Width at otic capsules 7.2 8.3 5.0

Length of prevomers 4.8 6.8 3.0

Greatest width of prevomers 4.0 5.0 1.9

Diameter of internal nares 1.8 2.0 2.5

Length of internal nares 3.2 3.2 1.3

Anterior edge of nares to condyle 11.0 15.0 7.2

Premaxillary teeth 9-10 9-9 12-10

Maxillary teeth 15-15 20-20 17-17

Prevomerine teeth 7-5 5-8 4-4

Palatine teeth 19-19 23-20 16-17

Dentary teeth 20-21 21-? 27-27

Splenial teeth 6-6 6-? 6-6

Total length of specimen 432 650 ?

nostril, strongly modified body terminus in male serving as a clasper, com-
pletely aquatic.

Of the two species known, P. \aitpii and P. melanochrus Taylor (1968),
I have been able to study the skull of only the former.

Potamotyphlus haupii (Berthold)
(Rg. 14)

Caecilia kaupii Berthold Nacht. Gesel. Gottingen, 1859, p. 181. Type-locality, "Angostura"=
Ciudad Bolivar, Venezuela.

Data from the skull of UIM No. 787 (locality uncertain).

The diastema between the parietals and squamosals is strong, also sep-
arating the squamosal and the frontals for half the length of the latter. The
eye socket is between the squamosal and the maxillopalatine, cut chiefly in
the latter. The upper edge of the maxillopalatine, bordering the frontal and
the nasopremaxilla, is very short, less than half its lower border. In this
skull the stapes are absent as they have fallen out while the skull was being
prepared, thus accounting for the large opening where the stapes would
normally occur. The internal nares are enormously large proportionally,
surrounded largely by the palatine shelf of the maxillary. The prevomers are
greatly narrowed, with the total distance between the nares about one third
of their transverse diameter. The pterygoid is moderate and the ectoptery-

Skulls of Gymnophiona and Their Significance 601

goids, if present, are very small. Minute diastemata are present posterior to
the nares between the palatal shelf and the basisphenoid, the latter remain-
ing wide to near the prevomers but sending forward a spine to separate the
prevomers for less than half their length. The suture separating the pre-
maxillary area from the prevomers is transverse, anterior to the prevomerine
tooth series. The basisphenoid is but little widened at the poorly developed
"wings." The nostrils are large and the large tentacular aperture notches
the maxillopalatine. The tentacular groove continues along the nasopre-
maxilla. See Table 3.

Family Scolecomorphidae Taylor

This recently proposed family of the Gymnophiona comprises a single
genus with six forms, recognized as species.

In studying the cranial antomy, skulls of two species have been available,
Scolecomorphits vittatus and S. I{irkji. Most of the data here recorded are
from the former species.

This family retains the premaxillae and the nasals as separate elements,
the latter bending down between the nostrils, meeting the premaxillae on
the ventral surface of the snout. Of three separate paired bones retained by
the Ichthyophidae (prefrontals, oculars, and septomaxillae), only the latter
are retained. These are short bones partly bordering the nostrils. The maxil-
lae and palatines are fused together as in all Gymnophiona, and are visible
both laterally and ventrally. The upper anterior part of the maxillopalatine
has an area resembling a prefrontal, but in skulls examined, this area, while
of a somewhat different color, is solidly fused with the maxillopalatine.

The nasal bones are large, truncate posteriorly. The frontals form a
narrow median suture while their lateral edges are 25 times the length of
the median suture. The parietals are elongate, sloping downwards in their
posterior parts and failing to enter the rim of the foramen magnum by a
fraction of a millimeter (0.15 mm). The squamosal and parietal are widely
separate, the squamosal attached lightly, and extending farther back than
is usual. No stapes are present. The squamosals are lightly attached to the
maxillopalatine.

The otic capsules are inflated and solidly fused with the basisphenoid.
The prevomers are narrow, slightly widened at the prevomerine tooth series,
extending a greater distance in front of the teeth than they do behind them.
No eye sockets are present. The tentacular apertures are large, opening in
the maxillopalatines, and extending forward, the grooves not covered by
bone, widening considerably anteriorly. The internal nares are surrounded
by the orbitosphenoid and partially by the palatine shelf of the maxillo-
palatine. This latter shelf, bearing the palatine teeth, is directed diagonally
outwards so that the back part of this dental series comes to lie in a line

602

The University Science Bulletin

behind the maxillary teeth, and thus widely separated from the basisphenoid.
A very wide diastema is present between the basisphenoid and the palatine
shelf, thus exposing the orbitosphenoid, which tends to reach up to the
dorsal skull surface. The basisphenoid forms the lateral part of the brain
case. There is a wide diastema between the prevomerine teeth and the
palatine teeth. I have not discerned ectopterygoids. The pterygoids seem-
ingly are fused to the quadrates. The splenial teeth are absent.

Scolecomorphus (Boulenger)

Scolecomorp litis Boulenger, Ann. Mag. Nat. Hist., ser. 5, vol. 11, 1883, p. 48. Type of genus
Scolecomorphus \irkii by monotypy. The type-locality thought to be probably "Lake
Tanganyika."

Scolecomorphus vittatus (Boulenger)
(Fig. 15)

Bdellophis vittatus Boulenger, Proc. Zool. Soc, London, 1895, p. 412, pi. 24, fig. 4. Type-
locality, Usambara Mountains, Tanganyika (Tanzania).

Data are taken from the skull of EHT-HMS 4642, from Nyange, Kenya.
For the general skull description, see the preceding family description
and Table 4.

The genus Bdellophis was erected by Boulenger for a caecilian presum-
ably differing from Scolecomorphus in having the eye not covered by bone

Table 4. Measurements in mm and tooth counts of skulls of Scolecomorphus.

Species vittatus k}rkii kirkii kjrkii

Number 4642 27120 27106 27116

Museum EHT-HMS MCZ MCZ MCZ

Locality Nyange, Cholo Mts., Cholo Mts., Cholo Mts.,

Vituri, Tanzania Malawi Malawi Malawi

Greatest length of skull 7.25 7.9 7.2+ 7.4

Greatest width 3.7 4.05 4.0+ ?

Jaw length 5.9 6.0 6.0 ?

Length of basisphenoid 4.8 5.35 broken 5.3

Width at otic capsules 2.5 3.0 3.+ 3.1

Length of the prevomers 1.75 1.9 1.85 2.1

before prevomerine teeth .. 0.9 1.0 1.0 1.2

behind prevomerine teeth .. 0.55 0.7 .65 .09

Combined width, greatest 1.0 1.15 1.1 1.2

Length, anterior part of

internal nares to condyle .. 4.8 5.3 5.3

Length of specimen, total 230± 270 ?

Premaxillary teeth 2-3 3-3 3-2 i-3

Maxillary teeth 6-6 8-7 8-?

Prevomerine teeth 2-3 4-3 3-3 2-3

Palatine teeth 7-6 7-7

Dcntary teeth 13-? 12-? ?-13

Splenial teeth 0-0 0-0 0-0 0-0

Skulls of Gymnophiona and Their Significance 603

(but not in a socket). Barbour and Loveridge (1928), seemingly not examin-
ing the skull carefully, state "An examination of the skull of a topotypic
specimen of B. vittatits shows that the character of 'eyes distinct,' which
separates the genus from Scolecomorphns whose eyes are below the cranial
bones, is a sign of youth." I have not been able to verify this statement.

I did examine the type in London and from notes taken I read "The eye
is very far forward, not under bone, nor in an orbit, but carried bv the
tentacle from under the bone into the tentacular trough, which for much of
its length is not bone-covered." In many specimens the eye may be seen
dimly through the skin at the base of the tentacular aperture or just under
the nostril.

Scolecomorphus \ir\ii Boulenger
(Fig. 16)

Scolecomorphus t(ir^ii Boulenger, Ann. Mag. Nat. Hist., ser. 5, vol. 11, 1883, p. 48. Type-
locality uncertain; thought to be presumably in the region of Lake Tanganyika (Tanzania).

Data are from the skulls of MCZ 27120, MCZ 27106 and MCZ 27116,
all from the Cholo Mountains, Malawi.

The specimens of the skulls available and the photographs leave much to
be desired. They are fragile and the differences between S. vittatits and
S. kjrl{ii as displayed in the Figures are not great. They seem to agree in all
family characters. In two of the specimens the eyes were discovered im-
bedded in the tentacles in the anterior part of the tentacular grooves. A
detailed description of the skulls may be found in Tavlor (1969). See
Table 4.

Family Caeciliidae

The prefrontals, oculars and septomaxillae are absent as separate bones
while the nasals and premaxillae are fused to form the two nasopremaxillae.
Usually the frontals are completely separated by the mesethmoid in the sub-
family Caeciliinae (one exception may be C. tetJtacitlata), but it is not
generally true of the Dermophinae. Thus the genus Dermophis has at least
one species with the frontals separated; some species with the frontals partly
separated and some without an external trace of the mesethmoid. Geo-
trypetes has at least one species with the frontals completely separated and
one incompletely separated. The skulls of the species seen of Uraeotyphlus,
Gegeneophis, Ajrocaecilia and Grandisonia have no external trace of the
mesethmoid on the dorsal surface of the skull. In Boitlengerula, Schisto-
metopum and Siphonops the mesethmoid completely separates the frontals.
(The other three families show no trace of the mesethmoid dorsally in the
species here studied.)

This family is treated under the two subfamilies, Caeciliinae and
Dermophinae.

604 The University Science Bulletin

Subfamily Caeciliinae

The significant characters of the two genera included in this subfamily
are the longer and heavier teeth, and the anterior processes on the prevomers
notching the nasopremaxillaries. Also, one group of species has a well-
defined eye socket, the other group has the eye solidly roofed over with bone.

In this subfamily the tentacular opening appears almost directly below
the nostril, usually about the same distance from the nostril and edge of the
lip and concealed by the snout so as not to be visible from above.

Caecilia Linnaeus

Caecilia Linnaeus, Systema Naturae, 1758, Eel. 10, vol. 1, p. 229. Type of genus, Caecilia
tentacidata Linnaeus (by monotypy).

This genus now has a large number of species all confined to southern
Central America and South America.

Skulls of seven species have been available. The skull characters are
given in greater detail for the species C. nigricans, than for the others. This
is one of the largest members of the genus and in fact of all Gymnophiona.

Caecilia nigricans Boulenger
(Fig. 17)

Caecilia nigricans Boulenger, Ann. Mag. Nat. Hist., ser. 7, 191)2, vol. 9, p. 51. Type-locality,
Rio Lita, N. W. Ecuador or southwest Colombia (3000 feet).

Data here recorded are chiefly from the skull of KUMNH No. 94377
from Darien, Panama.

Two other forms, described by Boulenger (1913), Caecilia intermedia and
Caecilia palmeri, from Ecuador and Colombia respectively, have been placed
as synonyms of this species.

This skull displays the family characters of the reduced number of dorsal
dermal bones with the usual fusions. The parietals are sharply declivous
posteriorly, while anteriorly two median processes push forward separating
the posterior parts of the frontals and contacting the narrow mesethmoid
which is between the frontals. The very small orbits of the eyes are cut in
the maxillopalatines anterior to the squamosal border; the squamosal bones
are wider but shorter than the maxillopalatines; the tentacular aperture is
at the anterior end of the maxillopalatine, with the groove extending along
the side of the nasopremaxilla for some distance. No diastemata are present
between the parietals and squamosals and the stapes are seemingly more
posterior than is usual.

On the ventral surface of the skull the prevomers, which have flattened
projections anterior to the prevomerine tooth series, are longer than their
combined width, and are separated posterior to the tooth series for four
fifths of their length by a narrow pointed spine of the basisphenoid. The
premaxillary tooth series is widely separated from the prevomerine teeth.

Skulls of Gymnophiona and Their Significance 605

The internal nares are bordered on their inner edges by the prevomers, else-
where by the palatine shelf of the maxillopalatine. A diastema exists be-
tween the basisphenoid, pterygoid process, and the palatine shelf, the opening
being about double the size of a naris. The "wings" of the basisphenoid are
prominent, tending to bend down, leaving the basisphenoid somewhat
domed as seen from below; the otic capsules are not or are scarcely inflated;
no ectopterygoid is discernible, and the pterygoids do not reach to the
internal nares. The transverse bosses below the otic capsules are prominent.
See Table 5.

Caecilia tentaculata Linnaeus

(Fig. 18)
Caecilia tentaculata Linnaeus, Systema Naturae, Ed. 10, vol. 1, 1758, p. 229. Type-locality
"America"=Surinam (see the earlier description by Linnaeus which preceded the 10th
edition).

Data are chiefly from the skull of KUMNH No. 10443S, Santa Cecilia,
Napo-Pastaza, Ecuador.

Caecilia ithmica, described by Cope (1887) from Darien, appears to be a
synonym of this species. Amphiumophis andicola Werner from Chancha-
mayo, Peru has been referred to this species also, but of this I am not certain,
not having yet been able to examine the type. Caecilia albiventris Daudin
also has been synonymized with C. tentaculata, but after examination of a
skull I regard this as a legitimate species (see later discussion).

Although this specimen is about 200 mm shorter than the C. nigricans
described, the skull is higher, longer and tapers more toward the snout. The
eye is cut into the maxillopalatine farther from its posterior end. The pari-
etals are shaped much the same but are more narrowed at the posterior end
and the mesethmoid does not complete the separation of the frontals and
consequently does not separate the posterior parts of the nasopremaxillae.
The pterygoid (seemingly fused to the quadrate) is shorter and wider than
that in C. nigricans. See Table 5.

Caecilia degenerata Dunn

(Fig. 19)
Caecilia degenerata Dunn, Bull. Mus. Comp. Zool., Harvard College, 19-12, vol. 91, pp. 505-
508. Type-locality, Garagoa, Cundinamarca, Colombia.

Data are from the skull of a paratype, AMNH No. 23354, "Colombia,"
without a specific locality.

All the bones of this specimen are greenish, not impossibly caused by the
method of preservation. This is a much smaller species than tentaculata but
the arrangement of the dorsal bones of the skull are much the same. The
orbit is proportionally even larger; the maxillopalatine is slightly longer
proportionally. See Table 5.

606

The University Science Bulletin

O

3

C

O
u

o
o

c

a

w
J

CO

<

h

^ CD

R NO
^ ON

.5 on

S o°-2

"s ** o
* f^ t* -a

*-» O

~5

Sj rv-j

"<3

hU
X o

H 15
£ E

< °

« „ .2 n

ii ° D 2 °
2 ~ ^ c £-

W3 £

s

a s a s

C 3 3 ^

in

m on r^-

O U"N 1-

o

o o

in

no oo

CnI

C~) CO "« *»-■

I I I I

cv> CO ON ro

in in

CDCNr*")— 'NONOOt-«rN

o

o

-1- ON f — <

— , (xi

CNlAtNlAf^ttNN

no t^

ro oo ro on

CD (N

in

oo in

in

_ ^ O NO

l/-\ l/N

ia m fi t\i

ITN

«1 *0 I- in r< rH

-3- no «■> ir\ — i i— i

ON CO CD ON (X) 1- o o
NOroroTfc^POCN

-r no -r
co

NO

i- no on cnj

PO NO ON CNl

CO t-v O c<1 CD t^. CD rn

UN NT «' VD 1- KN r«i (N

—

1-

Tt-

ro

t^

—|

rN

NO

o

ITN

rn

xr

m

r<->

t^

o

CN

ooomt^oO"— j r <

-r no' <vi o in o on on in
cn — i ■— ■ c-4 ■— i — i

-1- -1-

CO rn ON — i

rN

-f- CD or,

i— i on

X m N ^

CN

—

(V)

NO

o

CD

t^

On

co

fNl

in

-r

ON

-1-

co

rn

m

CN

"NO

, |

, 1

in

oo

ON

NO

NO

CN

-1-
O

rO

ON

T

oo

CM

PO

3 >

n] ri

o Si

O

o

c

OJ

c

3
Q.

O
u -r"

a"0

O, -C

be

C

a j=

a " a s ■? cr a

« c ° «

-5 — r -£

o

So
c

to

C D3

c

Ecu

J U J

2 3 X ^

r «J es 4J

C &.

c
C

Skulls of Gymnophiona and Their Significance 607

Caealia disossea Taylor

(Fig. 20)
Caecilia disossea Taylor, Caecilians of the World, 1968, pp. 374-378, figs. 196-197. Type-
locality, mouth of the Rio Santiago, Peru (a river rising in southern Ecuador, flowing into
the Rio Maranon).

Data are taken from the skull of EHT-HMS No. 1808, from "Alto
Cararey Napo-Pastaza, Ecuador."

This is a very elongate, slender species. The mesethmoid separates the
frontals but does not enter between the posterior parts of the nasopre-
maxillae. The orbit of the eye is well-defined. The median points of the
parietals separate the posterior part of the frontals slightly. See Table 5.

Caecilia volcani Taylor
(Fig. 21)

Caecilia volcani Taylor, Univ. Kansas Sci. Bull., 1969, vol. 48, pp. 315-323, figs. 1-4. Type-
locality, Valle de Anton, Panama.

Data are recorded from the skull of EHT-HMS 4696, topotypic paratype.

The mesethmoid separates the frontals and pushes forward between the
nasopremaxillae slightly; the parietals have two short median prongs that
touch the mesethmoid. There is a fine suture from the orbit to the frontal
and one from the orbit to the tentacular aperture. Thus the eye is between
a large upper portion and a lower portion of the maxillopalatine. This may
be an anomalous condition since it is the only known member of this family
that I have seen having such an arrangement. The two parts seem to be
fused solidly with the maxillopalatine. See Table 5.

Caecilia orientalis Taylor

(Fig. 22)
Caecilia orientalis Taylor, Caecilians of the World, 1968, pp. 417-425, figs. 220-224, 224a.
Type-locality, La Bonita, Napo-Pastazo prov., elev. 6300 ft. Ecuador.

Data are from the skull of EHT-HMS No. 4677, from "Ecuador."
The mesethmoid is minute, separating the frontals completely; the pre-
vomers reach back to the posterior level of the internal nares; the pterygoids
reach forward to practically the same level. See Table 5.

Caecilia albiventris Daudin
(Fig. 23)

Caecilia albiventris Daudin, Histoire naturelle, generale et particuliere des reptiles. . . . 1803,
pp. 423-426, pi. 92, fig. 1 (not fig. 2 as stated). Type-locality, "Surinam."

Data are from the skull of AMNH No. 49960, "Bogota," Colombia.

I am resurrecting this name from the synonymy of Caecilia tentaculata
Linnaeus. The skull of the species differs from that of C. tentaculata in
being longer and slenderer proportionally; the eye socket is larger, and the
stapes is somewhat differently shaped; seemingly the nostril is proportionally
much larger. See Table 5.

608 The University Science Bulletin

Oscaecilia Taylor

Oscaecilia Taylor, Caecilians of the World, 1968, p. 598. Type by designation, Caecilia
ochrocephala Cope.

Taylor referred six species to this genus as follows: Caecilia ochrocephala
Cope (1866); C. polyzona Fischer in Peters (1879); C. elongata Dunn
(1942); C. bassleri Dunn (1942); Oscaecilia hypereumeces Taylor (1968);
and 0. zweifeli Taylor (1968). Of these, 0. polyzona, known from three
specimens, is a close relative of O. ochrocephala. C. elongata is problematical
since the poorly described type and typotypic paratype have been destroyed,
and one fragmentary specimen, also referred to the form, is seemingly not
a member of the species. It unfortunately lacks the head and the anterior
part of the body; the posterior part differs from the type in certain recorded
characters.

These species appear to be more closely related to Caecilia and they differ
so much from the remainder of the genera recognized in the family Caecili-
idae that I have proposed that these two genera be placed in a distinct
subfamily.

Oscaecilia ochrocephala (Cope)

(Figs. 2-1-25)
Caecilia ochrocephala Cope, Proc. Acad. Nat. Sci. Philadelphia, 1866, vol. 18, p. 132. Type-
locality, Atlantic side, Isthmus of Darien, Panama.

The data here recorded are from the skull of MCZ No. 14817, from
Ancon, Canal Zone, Panama.

For a considerable time this species was regarded a member of the
African genus Herpele, which association seemingly is wholly unjustified.

The narrow mesethmoid separates the frontals completely and runs for-
ward to separate the posterior parts of the nasomaxillae and posteriorly to
separate the anterior prongs of the parietals. The squamosal is short (2.6
mm), the maxillopalatine being much longer (4.6 mm). No diastemata are
present between the parietals and squamosals.

The snout projects considerably beyond the mouth, its anterior edge pre-
senting a sharp transverse edge; the under surface of the projection is smooth
and flat. The teeth of the premaxillaries and anterior maxillaries are large
with very thick bases. The prevomers have an anterior flattened area pre-
ceding the prevomerine teeth; posteriorly they reach considerably behind
the posterior level of the posterior nares, and are separated by a spine from
the basisphenoid, which overlays the inner edges of the prevomers. The
nares are largely surrounded by the palatine shelves of the maxillaries, the
remainder by the prevomers.

The pterygoids are narrow and elongate, seemingly fused to the quad-
rates posteriorly. No ectopterygoids are apparent. The "wings" of the basi-
sphenoid tend to turn down, leaving the posterior ventral surface somewhat
domed (seen from below). A strong transverse boss is evident below the

Skulls of Gymnophiona and Their Significance

609

Table 6. Measurements in mm and tooth counts of skulls of Oscaecilia.

Species ochrocephala ochrocephala bassleri

Number 14817 41092 4675

Museum MCZ UIM EHT-HMS

Locality Ancon, Gatun, Ecuador

Canal Zone Canal Zone

Total length 12.1 8.9 9.0

Greatest width 6.6 5.2 4.8

Jaw length 9.5 6.25 broken

Length of basisphenoid 8.6 5.7 5.6

Width at wings 3.55 33 3.2

Width at otic capsules 4.0 3.5 3.4

Length of prevomers 4.2 2.5 3.0

Combined width anterior border

internal nares to condyle 7.2 5.0 5.15

Total length of specimen 542 ± 890±

Premaxillary teeth 3-2 2-3 4-3

Maxillary teeth 5-6 5-4 9-9

Prevomerine teeth 5-4 4-3 4-3

Palatine teeth 8-7 7-8 8-8

Dentary teeth 11-12 10-11 11-11

Splenial teeth 4-4 3-3 3-3

otic capsule. This capsule is scarcely inflated. The stapes makes a rather
widened contact with the quadrate. The tentacular apertures are visible
from a ventral view of the skull.

A second skull, EHT-HMS No. 1810, from Gatun, Canal Zone, Panama
differs in no essential detail, though it is smaller. See Table 6.

Oscaecilia bassleri (Dunn)
(Fig. 26)
Caccilia bassleri Dunn, Bull. Mus. Comp. Zool. Harvard College, 1942, vol. 91, p. 518. Type-
locality, Rio Pastaza, Ecuador.

The data here recorded are from the skull of EHT-HMS No. 4675, from
Ecuador.

Unfortunately the type series of C. bassleri contains specimens of two
species, Oscaecilia bassleri and Caecilia disossea. This has led to much
confusion.

The character of the mesethmoid is quite similar to that of O. ochroceph-
ala save that the forward prongs of the parietals are scarcely developed;
there is no orbit, but the position of the eye is dimly visible through the
semitransparent bone of the maxillopalatine. The area of contact of the
stapes with the quadrate is smaller proportionally. The tentacular apertures
are visible from the ventral face of the skull. The spine of the basisphenoid
overlays the edges of the posterior projections of the prevomers. The for-
ward projection of the snout has a rounded rather than a sharp tip in lateral
view. See Table 6.

610 The University Science Bulletin

Subfamily Dermophinae

The Dermophinae have the general characteristics of the Caeciliidae but
lack the characteristic distinguishing features of the Scolecomorphidae,
Typhlonectidae and the Ichthyophidae. From the subfamily Caeciliinae,
they differ in lacking the relatively large teeth, the tentacular aperture is
never as far forward as the nostril, and there is a strong reduction in the
range of the numbers of vertebrae. The tentacular aperture is in the maxil-
lopalatine and seemingly never emerges from under the anterior edge of
the bone.

The genera included in this subfamily show a considerable differentia-
tion on the generic level, and future studies may provide data that would
suggest the presence of other subfamily groups. Probably the most aberrant
genus is Idiocranium, comprising only a single known species.

Dermophis Peters

Dermophis Peters, Monatsb. Akad. Wiss. Berlin, 1879, p. 937, fig. 6. Type of genus: Siphonops
mexicanus Dumeril and Bibron.

Peters (1879) assigned a series of species to this genus without designat-
ing a type. One, however, was named by G. N. Noble (1924) who desig-
nated D. mexicanus as the type. The species originally placed in the genus
were Siphonops mexicanus Dumeril and Bibron, S. brevirostris Peters, and
four questioned species: S. thomensis Barboza du Bocage, S. brasihensis
Ltitken, S. proximus Cope and S. simus Cope.

Of these species only the first is retained in the genus; the others are
treated under other genera. Thus S. brevirostris and S. thomensis are recog-
nized in Schistometopum Parker; S. proximus in Gymnopis Peters; S. simus
in Cryptopsophis Boulenger; and S. brasiliensis in Luet\enotyphlus Taylor.

Several other forms, species or subspecies, have, however, been added to
the genus Dermophis by Taylor (1968). These are Gymnopis oaxacae Mer-
tens, Dermophis occidentalis Taylor, Gymnopis parviceps Dunn, Dermophis
balboai Taylor, Dermophis glandulosus Taylor, Dermophis eburatus Taylor,
Dermophis septentrionalis Taylor, Gymnopis gracilior Giinther; and Gym-
nophis [sic] clarkj Barbour as a subspecies of Dermophis mexicanus.

Dermophis is readily separated from Gymnopis by two well-defined
generic characters: the eye appears in an open socket, externally visible, and
the splenial teeth are lacking. Gymnopis, on the other hand, has the splenial
teeth and there is no orbit, the eye being covered by bone. There are certain
other differences as well.

Both of these genera are largely confined to Central America, only two
known forms of Dermophis entering South America.

The species of the genus have the following characters: secondary folds
present; scales present; three series of teeth (the splenials absent); parietals

Skulls of Gymnophiona and Their Significance 611

and squamosals forming sutures; tentacle closer to eye than to nostril; eye
in a socket, not covered by bone. Large and small species occur.

Skulls of several species are illustrated. That of Dermophis eburatus is
described more at length; the other species illustrated are compared to it to
point out differences that exist.

Dermophis eburatus Taylor
(Fig. 27)

Dermophis eburatus Taylor, Caecilians of the World, 1968, pp. 473-475, figs. 252, a, b, c, d, e;
type-locality, "Nicaragua."

The data are taken from the skull of MCZ No. 12121, from Guatemala,
C.A.

This species agrees in the basic family characters, lacking separate pre-
frontals, septomaxillae and orbitals, and having the premaxillae fused to the
nasals to form nasopremaxillae. The frontals are normal but posteriorly they
are partially separated by the mesethmoid, which is very narrow.

The squamosal is widened posteriorly, thus narrowing the parietal near
its middle, but it widens considerably posterior to the squamosal. The outer
posterior edges of the squamosal are slightly elevated above the jaw condyle,
leaving a cavelike overhang for muscle attachment. The stapes, much
widened, is widely overhung by a free edge of the dorsal basisphenoid,
which conceals much of the bone from a dorsal view. The orbit is largely
in the anterior part of the squamosal but is bordered anteriorly by the
maxillopalatine; the tentacular aperture is relatively large near the anterior
end of the maxillopalatine. not forming an anterior groove. The two upper
dental series are subparallel. The compound basisphenoid bone has a sharp
anterior spine that separates the prevomers for most of their length; the
internal nares are largely surrounded by the palatine shelf; however, the
prevomers border them on the anterior inner border for a short distance.
The nares are elevated above the remainder of the palate (the skull seen
from its ventral face); the pterygoid, seemingly, is partly fused to the quad-
rate, and an ectopterygoid is not discernible. The otic capsules are not
swollen, but below the area two strongly developed outer transverse bosses
are evident. See Table 7.

(All the bones, skull, jaws, vertebrae, and ribs of this specimen are a
dark reddish brown. I presume that this is due to some chemical used in
preservation.)

Dermophis costaricensis Taylor

(Fig. 28)
Dermophis costaricense Taylor, Univ. Sci. Bull., 1955, vol. 37, pp. 506-509, fig. 2, Photo.
(Type locality, Cinchona [Isla Bonita], Heredia Province, C. R. at near 4000 ft.)

Data are from the skull of KUMNH No. 66805.

This resembles the preceding skull of D. eburatus but with the following
differences: the mesethmoid is evident between the anterior part of the

612 The University Science Bulletin

frontals and is shorter. The parietals are but slightly constricted mesially,
and the squamosals are less widened posteriorly. The free edge of the dorsal
part of the basisphenoid hides much less of the stapes, while the free outer
posterior edge of the squamosal makes a very slight cavelike indentation
above the jaw condyle. The otic capsules are slightly more inflated. See
Table 7.

Dermophis balboai Taylor
(Fig. 29)
Dermophis balboai Taylor, Caecilians of the World, 1968, pp. 461-467, figs. 244-248. Type-
locality, Tacarcuna, Darien, Panama.

Data are taken from the skull of KUMNH No. 108935.

This is similar to D. eburatus but the mesethmoid is visible only as a
narrow mesial line the entire length of the frontals, widening slightly an-
teriorly; the maxillopalatines border on the orbits for a greater distance; the
stapes resemble those of D. eburatus more than those of D. costaricensis.
The parietals are narrowed and the posterior part of the squamosals are as
wide as in D. eburatus. See Table 7.

Dermophis parviceps (Dunn)
(Fig. 30)
Siphonops parviceps Dunn, Occ. Papers, Boston. Soc. Nat. Hist., 1924, vol. 5, pp. 93-94. Type-
locality, La Loma (1200 ft.), Boco del Toro Province, Panama.

Data are taken from the skull of KUMNH No. 36276.

There is no evidence of a mesethmoid. On the ventral surface the most
striking character is that the palatine shelf of the maxillopalatine completely
surrounds the nares, the rim of which is unbroken; the separation of the
prevomers is marked by a rather strong ridge, the most elevated part being
the spine of the basisphenoid. The parietals are only slightly constricted
mesially. Only the inner edge of the stapes is hidden when seen from above.
See Table 7.

Dermophis glandulosus Taylor

Dermophis glandularis Taylor, Univ. Kansas Sci. Bull., 1955, vol. 37, pp. 509-511, fig. 3.
Type-locality, San Isidro del General, San Jose Prov., Costa Rica.

Data are from the skull of KUMNH No. 56070, San Isidro del General,
San Jose, C.R.

This skull has the mesethmoid well developed, completely separating
the frontals. The parietals are only slightly narrowed. In contrast to D.
parviceps, the posterior part of the prevomers is wider and with a concom-
mitant widened space between the internal nares. The dorsal bones of the
skull show a strong overlap anteriorly. See Table 7.

Dermophis occidentalis Taylor

(Fig. 32)
Dermophis occidentalis Taylor, Univ. Kansas Sci. Bull., 1955, vol. 37, pp. 503-506, fig. 1.

Data are from the skull of KUMNH No. 36296, topotypic paratype.

Skulls of Gymnophiona and Their Significance

613

ex.

3

C
3
O
u

O

o

4— '

c

c

<u

3

W

pa
<

S on _, a 8
:~ oc £ n x

no

ntali.<:

96

NH

e

rt
CO

c

o

; re

~« ^c ~
"C ^ H

<u

u,

-- rt

5

V3

0 c

r

CO

"So X<*

S. E n o ™

•g cm S g 2 «

5 "1 J *; ™ c

ON .^ .—

V.

in

o
oo

<a ^

«5 no

c .a

u. ~ re

u

S E a S

O. 3 3 o

r^NOI-rOOOOinCNJON

Noo'ocTrONONOTM-

— i o

oc —i —i O

~h oc ~- -r o

U"N

tN o

ri — • o —i oo in On
"l in 1A (N <N « —

IA OO VO N in o

no oo no oc o o

oo

oc NO rt-

q ia in

IN fN rn

■—I CnI

lt r^ o — — o

in O NO OC
OC

NO — . <VJ o

oo ir>

in
CM

in

O

O

O NO — ' NO

— r-j

CN

in "I no oo no

ON

— PN]

O in on — _ in rv; o ""> o

in

no — . in «

— I o

ON NO

in
m

no r~i no in

t- o

in in ~

"i x in <n.

(N| I tN N -

« 1" IA "1 ^

o

ON

O NO
» (N - -

O
©

-

in

0

in

in

K

-r

in

in

NO

On

r*1

ON

rr\

oo

o

o

■d-

rvi

oo

NO

IN

oc

in

NO

r~j

ON

00

rr,

On

1-

oc

—

o
a

3

a.

-S "2

c

o ^

u

re w
o t

.« *

J3
O

OC

c

re
c

■S u, c

C .3 JJ

"S a •£ 2

a i-

t: t: j= .s t:

c
u

c

V

be n

C 33

J3 O

_C J= J3

t_ — ' ^ rt J—

c

<

s s a

H £ S

2 '5
c ^

oj

iyj

614 The University Science Bulletin

The constriction of the parietals anterior to their middle is rather pro-
nounced. The frontals are proportionally shorter than in previous species.
The maxillopalatines form as much of the orbit as the squamosal. The
posterior part of the prevomers is narrower. See Table 7.

Dermophis m. mexicanus (Dumeril and Bibron)

(Fig. ii)
Siphonops mexicanus Dumeril and Bibron, Erpetologie generate. . . . 1836, pp. 284-285. Type-
locality "Mexico."

Data are from the skull of UIM No. 66889, Chiapas, Mexico.

The narrow mesethmoid separates the rather elongate frontals for about
half their length. The posterior parts of the prevomers are not especially
narrowed. The stapes is proportionally less covered by the dorsolateral edge
of the dorsal part of the basisphenoid; the transverse bosses below the otic
capsules are prominent. See Table 7.

Gymnopis Peters

Gymnopis Peters, Monatsb. Akad. Wiss. Berlin, 1874, p. 616, pi. 1, fig. 1. Type of the genus,
Gyt7inopis multiplicata Peters.

Peters described this genus very briefly as follows: "Augen nicht von der
Haut iiberzogen, frei, keine Gesichtsgruber.,, Only the type was included.
Other distinguishing characters are: splenial teeth, secondaries, and scales
present. Eye covered by bone. The tentacle is nearer to the commissure of
the jaws than to the nostril.

A number of other species have been treated in the genus by Dunn
(1942) as follows: Siphonops proximus Cope, Rhinatrema unicolor A.
Dumeril, Siphonops oligozonus Cope, Gymnopis multiplicata oaxacae, Mer-
ters, Gymnophis [sic] nicefori Barbour, Dermophis albiceps Boulenger,
Gymnophis [sic] clarkj Barbour, Dermophis gracilior Gunther, and Si-
phonops parviceps Dunn. Dunn later (1944, 1945) described Gymnopis
pricei and Gymnopis braziliensis. Gorham (1962) also referred Dermophis
costaricensis Taylor, Dermophis glandulosus Taylor and Dermophis occi-
dentalis Taylor to the genus Gymnopis!

Of this list of species only Gymnopis multiplicata, G. proxima and G.
oligozona agree in having the essential characters of the genus, and only
these are recognized by Taylor (1968). The other species mentioned are
considered as belonging to other genera.

The skull characters are known to me by three specimens: one of G.
proxima and two of G. multiplicata. (Since these two rather closely related
forms appear to be separated without direct evidence of integradation, I am
treating them as species rather than subspecies.)

Skulls of Gymnophiona and Their Significance 615

Gymnopis multipltcata Peters

(Fig. 34)
Gymnopis multipltcata Peters, Monatsb. Akad. Wiss. Berlin 1874, p. 616, pi. 1, fig. 1. Type-
locality, Veragua, Panama.

Data are recorded from two skulls from Guanacaste, C.R., EHT-HMS
No. 4702, and KUMNH 117457, but mostly from the latter.

The prefrontals, septomaxillae, and oculars do not appear as separate
bones and the premaxillae and nasals are fused to form the nasopremaxillae.
The mesethmoid is visible mesially at the junction of the nasopremaxillae
and the frontals. The eye is without socket, bone covered, but is visible
under the anterior part of the squamosal. The aperture of the tentacle is
subcircular, cut in the maxillary, and not forming a forward-directed open
groove. The stapes is present, large, and the bone seemingly not fenestrated.
Ventrally the internal nares are somewhat narrowed, definitely diagonally
placed and surrounded almost completely by the palatine portion of the
maxillopalatine, although the prevomer borders the inner anterior edge for
a very short distance. The prevomers are narrowly separated for more than
four fifths of their length by a forward-projecting spine of the basisphenoid.
A very narrow diastema follows the nares, between the basisphenoid and
the palatine shelf. The tooth rows are nearly parallel. An ectopterygoid is
present! The second skull differs in no significant point from the one
described. See Table 8.

Table 8. Measurements in mm and tooth counts of the skulls of Gymnopis.

Species multiplicata multiplicata proxima

Number 4702 117457 4712

Museum EHT-HMS KUMNH EHT-HMS

Locality Guanacaste, Guanacaste, Turrialba,

C.R. C.R. C.R.

Total length of skull 16.0 14.5 16.85

Greatest width 10.8 9.7 11.6

Length of jaw 16.0 13.7 15.7

Length of basisphenoid 13.65 12.6 11.6

Width at wings 5.9 5.1 6.3

Width greatest posteriorly 7.0 6.0 6.7

Length of prevomers 4.7 4.0 4.5

Combined width, greatest 5.0 4.1 5.0

Anterior edge of choanae to condyle 11.0 9.2 11.8

Premaxillary teeth 10-10 11-8 10-9

Maxillary teeth 14-15 12-12 16-16

Prevomerine teeth 10-10 7-8 9-9

Palatine teeth 15-16 14-13 20-19

Dentary teeth 19-20 16-16 20-20

Splenial teeth 1-1 1-1 1-1

Length of preserved animal 430 459

616 The University Science Bulletin

Gymnopis proxima Cope

(F.g. 35)
Siphonops proximits Cope, Proc. Amer. Phil. Soc. 1877, vol. 17, p. 90. Type-locality, "Eastern
Costa Rica."

Data are from the skull of EHT-HMS 4712, from Turrialba, C.R.

This skull is similar to that of the preceding species but the following
characters obtain: the prevomers do not border the internal nares, they are
separated by the basisphenoid for about two fifths of their length only. The
diastema (fenestra) between the basisphenoid and the palatine shelf of the
maxillary is a little larger. The anterior narrowed portion of the stapes is
fenestrated near to the point of contact with the quadrate. The eye is visible
under the anterior edge of the squamosal. The mesethmoid is not visible
dorsally. See Table 8.

Siphonops Wagler

Siphonops Wagler, Isis, von Oken, Leipzig, 1828, p. 742, pi. 10, figs. 1, 2. The type of the
genus is Caealia anmdata Mikan, by monotypy.

In an article, Isis, von Oken, 1828, "Ansziige aus seinem Systema Am-
phibiorum" Tab. X., descriptions and comments on several genera are given
by Wagler. He includes a single species, annitlatus, in Siphonops, seemingly
attributing it to himself, citing Wagler's "Serpentes brasiliens" pi. 26, fig. 1.
In his Naturliches System der Amphibien, 1830, p. 198, the single species
Caecilia annulata is properly attributed to Mikan.

The genus as now understood includes six species: the type Siphonops
annulatus Mikan, S. paulensis Boettger, S. hardyi Boulenger, S. insulanus
Ihering, S. confusionis Taylor, and S. leucoderus Taylor. One species that
has been associated with the genus, Siphonops brasiliensis, I regard as be-
longing to a genus Luetkenotyphlus Taylor.

Siphonops has been characterized as lacking secondary folds, being with-
out scales, the tentacle much nearer to the eye than to the nostrils, the eye
in a socket that is continuous with the tentacular aperture, and having a
very narrow diastema between the squamosal and parietal. The splenial
teeth are lacking, and there is no terminal sucking disc. There is a well-
defined unsegmented terminal "shield."

Skulls of two species, Siphonops annulatus and S. paulensis have been
available.

Siphonops annulatus (Mikan)

(Figs. 36-38)

Caealia annulata Mikan, Delectus florae et faunae brasiliensis, Vklabonae, 1820, folio, pi. 11.
Type-locality, Sebastianopolis ( = Rio de Janeiro), Brasil.

Data here given are principally from the skull of MCZ No. 19407, from
"Rio Pastazo to Maranon," Ecuador.

This species lacks prefrontals, septomaxillae, and oculars, and has the
premaxillae and nasals fused. The mesethmoid is moderately large, com-

Skulls of Gymnophiona and Their Significance 617

pletely separating the frontals. The latter are greatly narrowed on their
inner margins so that their contact with the mesethmoid is less than one fifth
of their length, measured on their outer margins. The parietals, on the other
hand, have their greatest length on their inner margins, sometimes more
than double the length of their outer margins.

The orbits and the tentacular apertures are continuous with each other,
cut between the squamosals and the maxillopalatin.es. The stapes are present
but relatively small and slender. The prevomers are relatively large, extend-
ing back a little beyond the internal nares. Following the nares are large
diastemata between the basisphenoid and the palatal shelves of the maxillary;
the internal nares are almost completely surrounded by the palatine shelves,
the prevomers touching them only on their anterior borders for a short
distance. The pterygoids are widened, and seemingly fused to the quadrate;
ectopterygoids are present.

A second skull, UIM No. 56668, agrees very well with that described.
In this it is likewise difficult to interpret the relationship between the
pterygoid and quadrate. Anteriorly the pterygoid is attached to the palatine
shelf by thin cartilage rather than by overlap.

A third skull, EHT-HMS No. 1848, purporting to be of this species, is
from Teresopolis, Guanabara, Brasil. It seems to differ in that if the ptery-
goid is present, it is seemingly fused solidly with the quadrate. If an
ectopterygoid was present, it has been lost. The shape of the stapes is slightly
different. See Table 9.

Siphonops paulensis Boettger

(Fi^. 39)

Siphonops paulensis Boettger, (Catalog der Batrachicr Sammlung im Museum <ler Sencken-
bergischen Naturforschenden Gesellschaft in Frankfurt am Main, 1892, p. 62. Type-
locality, Sao Paulo, Brasil.

The data here given are taken from the skull of AMNH No. 23433,
"Brasil." (Skull partly broken.)

This differs from the preceding species in having the stapes differently
shaped. Also in this species the palatine shelves completely surround the
internal nares. Ectopterygoids are present and each seemingly has one tooth.
See Table 9.

Wiedersheim, in his "Anatomie der Gymnophionen" Jena, 1879, pi. 1,
figs. 1-12, gives figures of "S. annulatus." These however may well be
S. paulensis, since this species was not recognized until three years later.
The external appearance of the two species in preservatives is very similar.

Geotrypetes Peters

Geotrypetes Peters, Sitzb. Ges. Naturf. Freunde, Berlin, 1882, p. 55. The type-species is
Caecilia seraphini A. Dumeril, by monotypy.

Peters (1879) originally placed the type species in the genus Hypogeophis.

618

The University Science Bulletin

Table 9. Measurements in mm and tooth counts of skulls of Siphonops.

Species
Number
Museum
Locality

Total length of skull

Greatest width

Width at orbit

Length of jaws

Length of basisphenoid ...

Width at "wings"

Width at otic capsules

Length of prevomers

Combined width of prevomers
Anterior border internal

nares to condyle

Premaxillary teeth

Maxillary teeth

Prevomerine teeth

Palatine teeth

Dentary teeth

Splenial teeth

Total length of specimen

anniilatus
19407
MCZ

Ecuador

anntilatiis

56668

UIM

Limon

Cocha,

Ecuador

anniilatus

1848

EHT-HMS

Teresopolis,

Guanabara,

Brasil

paiilensis

23433

AMNH

"Brasil"

. 15.4

15.4

11.2

9.3

15.1

10.4

6.8

6.4

5.2

5.0

10.4
10-10

8-7

8-6

10-10

15-15

0-0

397

13.0

9.4
7.9
12.7
8.3
5.2
6.0
3.5
3.9

8.55
8.7
7-7
6-6
8-8
11-11
0-0
?

13.7

11.0

9.55

9.1

8.0

. 14.8

12.7

9.9

8.5

. 6.2

5.7

6.8

6.2

. 4.2

4.2

! 5.0

10.2
. 12-11

3.8

9.6

8+-?

7-7

11-?

7-7

7-?

10-9

11-?

. 16-16

15-?

. 0-0

0-0

. 356

453

Since 1880, Parker (1936) has described Geotrypetes angeli and recog-
nized a new subspecies G. seraphim occidentalis. Taylor (1968) described
G. congoensis and G. pseud oangeli.

Geotrypetes seraphini seraphini (A. Dumeril)

(Fig. 40)
Caecilia seraphini A. Dumeril, Archiv. Mus. Nat. Hist., 1859, vol. 10, p. 222. Type-locality,
Gaboon.

Two skulls are at hand, AMNH No. 23466, South Cameroons and MCZ
No. 3424, Metet, Cameroons. The data here presented are from the second.

This skull is quite remarkable in having teeth on the ectopterygoid,
which lies below the pterygoid. Dorsally the mesethmoid separates the
frontals for nearly half their length. Diastemata are present between the
parietals and squamosals which run forward into the orbit. The tentacular
apertures emerge at the orbits and, as deep, wide-open grooves, run forward
to a point on the nasomaxillae. On the ventral surface the prevomers do
not reach as far back as the middle of the internal nares. The prevomers
each have a large fenestra and a median forward projection anterior to the
prevomerine teeth. They are separated for most of the length behind the
teeth by a spine of the basisphenoid. The maxillopalatines completely sur-

Skulls of Gymnophiona and Their Significance 619

round the internal nares, behind which are enormous diastemata between
the basisphenoid and palatine shelves, and the pterygoids. See Table 10.

The "wings" of the basisphenoid are greatly narrowed as compared with
skulls of other genera. The skulls of Geotrypetes as far as known, differ
from most of the other genera of the subfamily Dermophinae of the
Caeciliidae in that they have ectopterygoid teeth and have the diastemata
between the squamosals and parietals penetrating the orbital rim. The skulls
are seemingly more fragile than is usual in other genera of the family.

Geotrypetes seraphini occidentalis Parker

(Fig. 41)
Geotrypetes seraphini occidentalis Parker, Zool. Meded. Rijks. Mus. Nat. Hist., Leiden, 1936,
pp. 99-100. Type-locality, French Guinea and the Gold Coast Region (Ghana).

Data here given are taken from the skull of EHT-HMS No. 4653, Tafo,
Ghana.

Dorsally the mesethmoid separates the frontals, runs forward separating
the nasopremaxillae for a distance, and then appears again at their anterior
parts; posteriorly the anterior prongs of the parietals arc likewise separated
by the mesethmoid. The parietals narrow mesially with their posterior parts
declivous. Broad diastemata are present between the squamosals and pari-
etals; also, the suture between the anterior part of the squamosal and parietal
and the squamosal and frontal runs forward, breaking the continuity of the
large orbit of the eye. The orbit is continuous with the proximal part of
the tentacular groove. The tentacular groove is open for a short distance,
then is covered over by bone of the maxillopalatine, but later emerges from
under the anterior part of this same bone, the groove continuing forward to
a point below but near the back edge of the nostril. The orbital opening is
large and somewhat pear-shaped. Two processes of the prevomer run for-
ward anterior to the prevomerine teeth. Posteriorly the prevomers are
separated by the short spine of the basisphenoid for a distance, with their
posterior ends not reaching to the level of the middle of the internal nares.
The nares are completely surrounded by the maxillopalatine. There are two
very large openings posterior to the nares. The ectopterygoids bear teeth
(not visible in the Figure); the pterygoid touches the palatine shelf; the
wings of the basisphenoid are much narrowed. See Table 10.

Herpele Peters

Herpelc Peters, Monatsb. Akad. Wiss. Berlin, 1879, p. 939. Type of genus, Caecilia squalo-
stoma Stutchbury, by monotypy.

Peters denned this genus on the basis of the following characters: eye
under the skull bones; tentacular aperture circular, behind and below nostril;
similar to Hypogeophis rostratus. Scales in folds; two tooth rows in lower
jaw.

620 The University Science Bulletin

Herpele squalostoma (Stutchbury)

(Fig. 42)
Caecilia squalostoma Stutchbury, Trans. Linnean Soc. London (1), vol. 17, 1837, p. 362. Type-
locality, Gaboon.

Data here recorded are from a skull of EHT-HMS No. 3412 from Metet,
Cameroons, Africa.

The mesethmoid appears minutely at the anterior and posterior ends of
the common suture between the frontals; in neither case is the visible part
more than 0.5 mm long. The eye is under the squamosal bone. This bone
is rather elongate but not as long as the maxillopalatine. The tentacular
aperture is at anterior end of the maxillopalatine, with the groove continued
slightly along the nasopremaxilla, the apertures barely visible from the
ventral face of skull. The prevomers are short and separated by the spine of
the basisphenoid for less than half their length, their posterior extentions
being somewhat widened and lying below the basisphenoid instead of above
it as in Oscaecilia, their forward extensions reaching anterior to the pre-
vomerine teeth. There is a diastema between the basisphenoid and the
palatine shelves, the small forward-extending part of the pterygoid is above
the palatine shelf rather than on its inner edge. Small ectopterygoids are
present, touching the pterygoid. The otic capsules are somewhat inflated.
See Table 10.

Uraeotyphlus Peters

Uraeotyphlus Peters, Monatsb. Akatl. Wiss. Berlin, 1879, p. 933. Type-species, Caecilia oxyura
Dumeril and Bibron, by subsequent designation.

Peters (1879) assigned two species to the genus, the type and Caecilia
malabarica Beddome. Boulenger (1882) later described Uraeotyphlus ajri-
cans (now transferred to Geotrypetes seraphini seraphini). Annandale
(1913) described Uraeotyphlus menoni and Seshachar (1939) described U.
narayani.

Now four species are recognized in the genus. All are Indian in dis-
tribution. Peters reports oxyurus from the Seychelles Islands but this doubt-
less is in error.

Uraeotyphlus oxyurus (Dumeril and Bibron)

(Fig. 43)

Coecilia oxyurus Dumeril and Bibron, Erpetologie generale, 1841, vol. 8, p. 280. Type-locality,
Malabar, India.

Data here recorded are from the skull of MCZ 9484, Taliparabamba,
Travancore, India.

The mesethmoid does not appear on the dorsal skull surface. The pre-
frontals and septomaxillae are not present and the premaxillae and nasals are
fused, forming the large nasopremaxillae. The eye is in an orbit that is con-
tinuous with the deep open groove of the tentacle. This extends forward to
a point below the back edge of the nostril. The orbits are bordered for more

Skulls of Gymnophiona and Their Significance

621

-s:
<o

-a

c

C3

o

<3

3j

^>
ex.

O

3
v>

O

(/>

C
3
O

u

o
o

c

c

£
u

<u

W

<

N u

<u — O -i rt c
io^, piii-lu^

£>2£ O a G

<*>

^ * 2 ^ o

« ? C m a

DC

S

U

raEl

c

~ NO l « r-

O

w

N

ft.

c
J c

<0

U

» i; F >.

S E H«

a 3 3 o

M Z ^ J

in

<v-| — NO PO —

»n t-^ m

O t^ O* NO
t^ l^. o o

c: in — ■ o

NO
in

(N-i in i—< oo in <N on

O C\ N f<i ^r (N (N

O fN - —
VG N CO IA - - — —

>s 'O » in - 1 — >— • o

NO

3C ro

in
o r^ r<~i -i-

r^ -r

: 1A (N "1 M

in u-\ —

"1

r^

,~H

CN

3C

~

-

r^

o

CM

OC

o

in

o no

K -r

eg orj

cs) in — r^-

'^ -r- rsj no

O 1^
in
r*N

«■> {-^

on r-i

T O 3C

— t^

: oo in

« fl (N

r-* in ng tt o

— i — in
o n m

in

p p p

ON NO 3C

X ts Ji — S
VO «1 ri (N N

r^j rsi on o

»3- — rf i— < i— < — i in

in N in in ic c m

3

O

c

D,

E -

—
K o

o
-S •-

be £

C

2 ^

o J=

AJ Ui »> r! <U

s a * .2

•^ r—

— -^ I-

o i t j: j:

"3 — i-j- hn ^ •--

c K

5C be
c c
u

J J ^ £

C 3

u o

3 C p

^2 C c Jg .2 a

3

o

c

ii c

y a.

622 The University Science Bulletin

than half of their circumference by free, narrow, curved ocular bones (post-
frontals), while below they are bordered by the maxillopalatines. The stapes
are present, touching the quadrates. The prevomers extend forward a little
beyond the prevomerine teeth, while posteriorly they reach to the posterior
level of the the internal nares, the spine of the basisphenoid separating them
for about one fifth of their total length. They border the inner sides of the
internal nares but for the greater part of their circumference the nares are
bordered by the palatine shelves of the maxillae. See Table 10.

The skull is defective, broken, and parts missing, so that certain charac-
ters cannot be examined and recorded.

Schistometopum Parker

Schistometopum Parker, Ann. Mag. Nat. Hist., scr. 11, vol. 7, p. 17, fig. 4. Type-species,
Dennophis gregoni Boulenger, by designation.

Two species were originally assigned to the genus, S. gregoni, and S.
thomensis (Bocage). Peters described Siphonops brevirostris in 1874 which
has been regarded a synonym of S. thomensis by certain authors. Taylor
(1964) described Schistometopum ephele.

Schistometopum gregorii (Boulenger)

(Figs. 44-45)

Dennophis gregorii Boulenger, Proc. Zool. Soc. London, 1894, p. 646, pi. 40, fig. 4. Type-
locality, Ngatana, Tana River, Kenya.

Data recorded are from the skull of MCZ No. 20117, Lake Peccatoni,
Kenya.

This species has been redescribed by Nieden (1912) as Boulenger ula
den hard ti from the Tana Region, Kenya and by Boettger (1913) as Bdello-
phis unicolor in Voeltzkow from Lake Peccatoni, Wituland, Kenya.

The slender mesethmoid separates the frontals and pushes in slightly
between the nasopremaxillae. The eye orbits are cut between the squamosal
and the maxillopalatine, but lie chiefly in the former. The tentacular aper-
tures appear near the anterior end of the maxillopalatines, which are
definitely shorter than the squamosals. The stapes are much constricted
behind the point of contact with the quadrates.

The prevomers are very short, projecting forward beyond the prevomer-
ine teeth for a short distance, but not reaching back as far as the middle of
the internal nares. The diastemata between the basisphenoid and palatal
shelves are small. Seemingly small ectopterygoids are present (broken in
this specimen).

A second skull, MCZ 20146, agrees in all essential details with the pre-
ceding. See Table 10.

Hypogeophis Peters

Hypogeophis Peters, Monatsb. Akad. Wiss. Berlin, 1879, p. 936. Type-species Caecilia roslrata
Cuvicr.

Skulls of Gymnophiona and Their Significance 623

Peters (1879) considered two species as belonging to the genus, the type,
and Caecilia seraphini A. Dumeril. In 1880 he proposed Geotrypetes as a
new genus for the second species.

Parker (1941, 1958), who reviewed the species of Hypogeophis, concluded
that there was a single species H. rostratns, with three subspecies: H. r.
rostratus, H. r. gitentheri Boulenger, and H. r. praslini Parker. All are con-
fined to the Seychelles Islands in the Indian Ocean, but each has its own
range, on different islands. Taylor (1969a) has added a fourth subspecies,
H. r. lionneti.

Hypogeophis rostratus rostratus (Cuvier)

(Fig. 46)
Caecilia rostrata Cuvier, Regne Animal, 2nd ed. 1829, vol. 2, p. 100. Type-locality, Mahe,
Seychelles (most probably).

Data here recorded are taken from the skull, MCZ No. 48935.

The prefrontals, septomaxillae, and oculars are absent or fused to sur-
rounding bones and the premaxillae and nasals are fused to form the naso-
premaxillae. The stapes are present. The mesethmoid does not appear on
the dorsal surface of the skull. The basisphenoid is fused with the various
occipitals and otic elements to form a compound bone (as in all gymno-
phiones).

The eye socket is cut in the squamosal. The tentacular apertures are
separate from the orbits, opening far forward at the end of the maxillo-
palatines, the grooves continuing forward along the nasopremaxillae, the
openings partially visible in a ventral view. No diastemata are present
between the squamosals and parietals. The internal nares are largely sur-
rounded by the maxillopalatines, but the prevomers border them on the
anterior inner edge.

There are very small diastemata behind the internal nares between the
basisphenoid and palatal shelves of the maxillopalatines. The basisphenoid
separates only the posterior tips of the prevomers and is strongly constricted
in front of the otic capsules. The pterygoids seemingly are fused to the
quadrate, and relatively narrow. A small ectopterygoid is present. There
appears to be a small element separated from the palatine portion of the
maxillopalatine which, if it touches the pterygoid, does so below the lateral
wings of the basisphenoid. The snout projects considerably beyond the
premaxillary teeth. See Table 11.

Grandisonia Taylor

Grandisonia Taylor, Caecilians of the World, 1968, p. 749, rig. 409. Type of genus, Hypo-
geophis alternans Stejneger.

Taylor referred three species to the genus. These were Hypogeophis
alternans Stejneger, Dermophis sechellensis Boulenger, and Dermophis lar-
vata Ahl. In addition, two other species were referred with some doubt,
Hypogeophis brevis Boulenger and a new species Grandisonia diminutiva

624 The University Science Bulletin

Taylor, since neither the anatomy nor the life histories of these diminutive
species are known. They are the smallest species of the Gymnophiona.

Grandisonia alternans (Stejneger)
(Fig. 47)

Hypogcophis alternans Stejneger, Proc. U.S. Nat. Mus. Washington, 1893, vol. 16, p. 739;
type-locality, Mahe, Seychelles Islands.

Skull data here recorded are of EHT-HMS No. 4647, from Mahe, Sey-
chelles.

The skull has the typical reduced number of bones of the family Cae-
ciliidae. The mesethmoid does not appear on the dorsal surface of the skull.
There are no distinct diastemata between the squamosals and the pari-
etals; the suture, however, is cartilage covered and when this is removed
there is a slight break between them. The rims of the orbits, between the
squamosals and the maxillopalatines, are continuous with the tentacular
apertures and the tentacular grooves which run forward to the anterior end
of the maxillopalatines. The stapes are prominent.

On the ventral surface of the skull the basisphenoid is strongly con-
stricted between the otic capsules and the "wings," and tapers anteriorly to
a spine, which, running forward, separates the prevomers for more than half
their length. The prevomers are wide, anteriorly they scarcely extend beyond
the prevomerine teeth; posteriorly they do not reach the middle of the internal
nares, but border the nares for a short distance. An ectopterygoid is present;
the pterygoid is seemingly fused to the quadrate. There are small diastemata
between the basisphenoid and the palatine shelves of the maxillopalatines.
The separation between the internal nares is greater than usual because the
part of the basisphenoid extending between the prevomers is widened, thus
greatly narrowing the prevomers posteriorly. See Table 11.

Grandisonia sechellensis (Boulenger)

(Fig. 48)
Dermophis sechellensis Boulenger, Trans. Linnaean Soc. London; Zool., l'Ml, vol. 14, p. 376,

fig. 3. Type-locality, Mahe, Seychelles.
Hypogeophis sechellensis Parker, Am. Mag. Nat. Hist. (1941), ser. 11, vol. 7, p. 16, fig. 2.

The data here given are from a damaged skull, AMNH 23673, "Sey-
chelles Islands," with no specific locality.

Mr. Boulenger and Dr. Parker studied the caecilians of the Seychelles.
Boulenger (1909) referred some specimens of this species to his Cryptopso-
phis midtiplicata, and two years later described them as a new species,
Dermophis sechellensis. Parker first regarded this species as a Dermophis
(1941). Ahl (1926) had described it as Dermophis flaviventer. Parker later
(1958) referred it to Hypogeophis sechellensis and still later seemingly
regarded it as a member of the genus Praslinia Boulenger. Taylor (1968)
placed it in his genus Grandisonia.

Skulls of Gymnophiona and Their Significance 625

This skull differs very little from Parker's figure (loc. cit.), the nasopre-
maxillary is proportionally longer with no trace of a mesethmoid dorsally.
(A slight exposure of this element shows in Parker's figure.) The outer
edges of the frontals are longer, and the parietals are less constricted anterior
to their middle.

On the ventral surface of the skull the prevomers are short, reaching to
the anterior part of the internal nares, and are separated for about half their
length by the spine of the basisphenoid. There is a strong constriction of
the basisphenoid just anterior to the otic capsules. The eye orbit is cut
chiefly in the squamosal but its border is continuous with the tentacular
aperture. A pterygoid process and an ectopterygoid are present. See Table
11.

Idiocranium Parker

Idiocranium Parker, Proc. Zool. Soc. London, 1936, pp. 160-163. Type of genus Idiocranium
russeli.

This genus has certain characters which are unique in the Gymno-
phiona. These were recognized by Dr. Parker who has given an excellent
description of this diminutive creature.

Idiocranium russeli Parker
(Fig. 49)

Idiocranium russeli Parker, Proc. Zool. Soc. London, 1936, pp. 160-163, figs. 6, 8. Type-
locality, Makumunu, Asumbo, Mamfe Division, Cameroons (now included in .Nigeria?).

The data recorded here are taken from EHT-HMS No. 4687. Paratype.

The large dorsal nasopremaxillae are separated their entire length by a
large, somewhat triangular mesethmoid which runs forward, terminating in
a point. Following this on the dorsal surface of the skull is a pair of small
subquadrangular frontals which are separated from the maxillopalatines by
the contact of the nasopremaxillary with the squamosal, and having an area
less than one third that of the parietals. The orbits of the eyes are between
the elongate squamosals and the maxillopalatines. The larger tentacular
apertures emerge from the anterior ends of the maxillopalatines and the
tentacular groove continues forward some distance beyond the opening,
along the side of the nasomaxillae. The stapes is well defined, the "neck"
joining the quadrate being very short. The otic capsule is distinctly inflated.

The internal nares are surrounded by the maxillopalatines and the pre-
vomers, the latter bones bordering their inner edges for a short distance,
then extending back to the posterior level of the internal nares; there are no
diastemata between the basisphenoid and the palatine shelves of the maxil-
lae. The pterygoid reaches forward one third of the way to the internal
nares. No ectopterygoid is present. See Table 11.

The greatly enlarged mesethmoid, the separation of the nasopremaxillae,
the reduction of the size of the prefrontals and the separation of the frontals

626

The University Science Bulletin

a

o

C

.O

~<3

6

"Si

o

be

o

o

V)

3

C
3
O

o
o

C
rt

C

<u

3

W
-J

pa
<

41

^u o

CQ

a!

c

*■ f £ -*

^1 =
-5 ^ £ -^

^^^ C'
. <*>

s»

<n

2

t^

DC

^3

1-

-

S

O

w

~ in ^. G -3

*: ¥ S ~

C/0

a:

u C sj «

u c « g

a 3 3 c

CO -r

in

^ >r n O; on in in »o

00 00 o
oo 00 o

00

fN] — 1

in

1- (N M

CO •— ' fN < Of^lOflNOrHH(N]

in

— 1 QO

in
cni 00 c-i cn

q in N

1 ^^NNXrflMN

iy~\

fN OO VO t^

-1- in o

in a -n x r. n

O OO ■*- Tt-

<Ni(NiinaNinoovoo

in

N£>

00 in

in 00 t^ '-'")

(NNMNOHOOrtrtKl

co co r

in

o
.£

So
a

OJ

X>

O

a

o

a

u

a

l_

bo

«J

c

-

0

:

u

>

<L)

rt

1 — ' 1— ^

Q.

a .2 * -5

G

IS

<= -S

C w tJ S

c

o

u

Ih

a,

c

c

u

a

>. a

~Z1 ti *-' Kr hf *-^ *-J T tl T r~ • —

o

J hJ

bo

#£* qj u qj

Q C
C 'C

c

u -S
ho

B

P -= o -s ~

a a

1/^ H

Skulls of Gymnophiona and Their Significance 627

from the maxillopalatine easily distinguish this genus from all other Gymno-
phiona.

Boulengerula Tornier

Boulengerula Tornier, Kricchthicre Deutsch-Ost-Afrikas, Beitrage zur Systematik unci De-
scendenzlehre. Berlin, 1897, p. 16-1 (type of the genus, B. boulengeri from Usambara Mts.,
Tanganyika [Tanzania] Africa).

Generic characters indicated by Tornier are: parietal and squamosal
forming a suture without diastema; no splenial teeth present; eyes roofed
over by bone; no scales present; tentacle conical and extrusible, equally dis-
tant from nostril and symphysis of jaws. Only the type species is known.

Boulengerula boulengeri Tornier
(Fig. 50)
Botdengeruld boulengeri Tornier, Kriechthiere Deutsch-Ost-Afrikas. Beitrage zur Systematik
und Descendenzlehre. Berlin, 1897, p. 164. Type-locality Usambara Mts., Tanganyika
(Tanzania), Africa.

Data are from a defective skull, MCZ No. 12309, Amani, Usambara Mts.,
Tanganyika (Tanzania), Africa.

The skull agrees with family characters in reduction of the number of
dorsal skull bones; the mesethmoid appears narrowly on the median dorsal
line, completely separating the prefrontals and forming a very slight wedge
between the posterior ends of the nasopremaxillae, and a larger wedge be-
tween the anterior ends of the parietals. The median length of the frontals
is about half the median length of the parietals.

There is no eye socket; the eye, if present, is bone covered. The tentacular
aperture is cut in the maxillopalatine, its groove not extending beyond
these. The parietals and squamosals form sutures without diastemata, the
parietals widening very slightly behind the posterior level of the squamosals.
The snout projects beyond the mouth. The basisphenoid is relatively very
wide anteriorly at the point of contact with the prevomers, except for a
spine from the basisphenoid that separates the greatly narrowed prevomers
for half of their length. The prevomers reach posteriorly behind the pos-
terior level of the internal nares.

A short break occurs in the continuity of the prevomeropalatine tooth
series. The pterygoids are present and fused to the quadrate. The otic
capsules are obviously inflated. See Table 11.

Afrocaecilia Taylor

Ajrocaecilia Taylor, Caecilians of the World, 1968, p. 321. Type of the genus, Boulengerula
taitanus Loveridge.

Ajrocaecilia is characterized as follows: eye under bone; tongue free

laterally and anteriorly; no scales; no secondaries; splenial teeth present;

tentacle small, distanct from nostril; no tail; an unsegmented terminal

"shield."

628 The University Science Bulletin

Taylor referred three species to the genus: the type, Boulengentla tai-
tanns, B. uluguruensis and B. changamwensis. All are small species, east
African in distribution. The members of the genus may readily be differ-
entiated from Boulengentla by the presence of splenial teeth.

Afrocaecilia uluguruensis (Barbour and Loveridge)

(Fig. 51)

Barbourula uluguruensis Barbour and Loveridge, Mem. Mus. Comp. Zoo!., Harvard College,
vol. 50, 1928, pp. 183-184. Type-locality, Vituri (2000 ft. elev.), Uluguru Mountains,
Northeastern Tanganyika (Tanzania).

One skull from a topotypic paratype has been available for study, EHT-
HMS No. 4649. The characteristics are as follows:

The eye if present, is concealed under the squamosal bone. The tentacu-
lar aperture is at or near the anterior end of the maxillopalatine, which is
relatively small. The nasopremaxilla is relatively large, the mesethmoid not
being visible on the dorsal surface of skull. The frontals are quadrangular,
their common suture about equaling half of the common parietal suture
length.

The prevomers are separated by the spine of the basisphenoid for more
than two thirds of their length. Very narrow diastemata posterior to the in-
ternal nares lie between the maxillopalatine shelf and the basisphenoid. The
basisphenoid is widened where it first meets the prevomers. The pterygoid
appears to be fused with the quadrate. The lateral sutures of the maxillo-
palatine are difficult to follow. See Table 12.

Afrocaecilia taitana (Loveridge)

(Fig. 52)

Boulengerula taitanus Loveridge, Bull. Mus. Comp. Zool., Harvard College, vol. 79, 1935, p.
16. Type-locality, Mt. Mbololo (4800 ft. elev.), Teila Hills, Kenya.

Data here are recorded from the skull of MCZ No. 20021, topotype.

The characters differing from A. uluguruensis are: the skull is longer
and perhaps slenderer, the prevomers may reach back only to the posterior
level of the internal nares (not distinctly farther as in A. uluguruensis), and
an ectopterygoid is present (seemingly absent in the two other species). It
has been injured during the lifetime of the animal, as one of the bones on
the head has been broken and partially mended. See Table 12.

Afrocaecilia changamwensis (Loveridge)

(Fig. 53)
Boulengerula changamwensis Loveridge, Bull. Mus. Comp. Zool., Harvard College, 1932, vol.
72, p. 381. Type-locality, Changamwe (192 ft. elev.) near Mombasa, Kenya, Africa.

Data are from the skull of a topotypic paratype, EHT-HMS No. 4651.

The differences of this skull from the above two are small. The prevom-
ers are somewhat differently shaped as is the anterior spine of the basi-

Skulls of Gymnophiona and Their Significance

629

Table 12. Measurements in mm and tooth counts of skulls of Afrocaecilia.

Species uluguruensis taitana changamwensis

Number 4649 20021 4651

Museum EHT-HMS MCZ EHT-HMS

Locality Uluguru Teita Hills, Changamwe,

Mts. Kenya Kenya

Total length of skull 5.8 7.3 5.7

Greatest width 3.2 3.6 3.25

Length of lower jaw 5.2 6.3 5.1

Length of basisphenokl 4.4 3.5 4.2

Width at otic capsules 2.0 2.6 2.05

Width at "wings" 1.35 2.35 1.8

Length of prevomers 1.4 1.7 1.7

Combined width, greatest 1.5 1.85 1.5

Anterior border of internal

nares to end of skull 3.55 4.7 3.7

Premaxillary teeth 6-8 ?-? 6-7

Maxillary teeth =-8 12-11 8-?

Prevomerine teeth 5-4 4-5 5-4

Palatine teeth 6-7 9-8 MO

Dentary teeth 10-10 12-12 10-10

Splenial teeth 3-2 1-1 2-2

Total length of specimen 232 234

sphenoid. The fused? pterygoid is present hut I am not wholly certain that
there is an ectopterygoid. The squamosa] is elongate as in the other two
skulls. See Table 12.

Gegeneophis Peters

Gegenes Giinther, Proc. Zool. Soc. London, 1875, p. 577, type of genus, Epicrium carnosum
Beddome (preoccupied by Hubner, 1816, for a genus of Lepidoptera) .

Gegeneophis Peters, Monatsb. Akad. Wiss. Berlin, 1879, p. 932 (type of genus, Epicrium
carnosum Beddome) .

Gegenophis Boulenger (error or emendation), Catalogue of the Batrachia Gradientia s. Caudata
and Batrachia Apoda in the Collection of the British Museum, 2nd Ed. 1882, p. 101.

This genus is, so far as known, confined to India. It comprises three
species, one of which, G. fulleri, may be doubtfully associated. The other
two, G. ramaswamii and G. carnosus, seemingly belong to the same genus.

The type species, G. carnosus, is small, and none has been available for
an examination of the skull.

The combination of external characters that serve to define the genus
are: eye without orbit, solidly covered by bone; splenial teeth present; no
diastema between prevomerine and palatine teeth; tentacle small, behind
and below nostril; secondary folds present posteriorly; scales are present, two
to four rows posteriorly; tongue with two narial plugs, vent transverse.

Gegeneophis ramaswamii Taylor

(Figs. 54, 55)

Gegeneophis ramaswamii Taylor, Senck. Biol., Frankfurt am Main, 1964, Bd. 45, Heft 3/5,

630 The University Science Bulletin

Table 13. Measurements in mm and tooth counts of skulls of Gegeneophis

ramaswamii (topotypic paratypes).

Number 29452 29454 29456

Museum MCZ MCZ MCZ

Locality Kerala, Kerala, Kerala,

India India India

Total length of .skull 10.0 9.05 8.5

Greatest width 6.0 4.95 5.9

Length of lower jaw 10.0 .... 8.3

Length of basisphenoid 6.5 6.0 5.7

Width (at wings) 3.2 2.75 2.85

Width posterior to constriction 4.3 3.4 3.55

Length of prevomers 2.6 2.0 2.00

Combined width of prevomers (greatest) 3.0 2.35 2.3

Anterior border of choana to condyle 6.7 6.0 5.8

Premaxillary teeth 4-5 4-6 5-5

Maxillary teeth 14-15 11-10 10-11

Prevomers teeth 5-5 5-4 5-5

Palatine teeth 12-13 10-12 12-11

Dentary teeth 13-13 .... 13-13

Splenial teeth 3-3 .... 3-3

Total length of specimen 305 263 242

Dec. 1, 1964, pp. 227-231, text figs. 1, 2 (type-locality Tenmalai Forest, Kerala [state],
southern India); Caecilians of the World, Lawrence, Kansas, 1968, pp. 739-746, figs.
402-406.

Data recorded here are from the skull of MCZ No. 29452.

The prefrontals, septomaxillae, and oculars do not appear; the premaxil-
lae and nasals are united. The mesethmoid is not visible dorsally. The
posterior part of the skull is declivous, rather abruptly so from the parietal
border. The stapes are relatively very large. There is no eye socket. The
internal nares are almost surrounded by the maxillopalatines but are bor-
dered for a short distance on their inner side by the prevomer. A small
diastema is present between the basisphenoid and pterygoid following the
nares.

Two other skulls available do not differ in essential details. See Table 13.

LITERATURE CITED

Ahl, E. 1926. Neue Eidechsen und Amphibicn. Zool. Anz., Leipzig, Band 67, Heft 7/8, pp.

186-192.
Annandale, N. 1913. Some new and interesting Batrachia and lizards from India, Ceylon,

and Borneo. Rec. Indian Mus. Calcutta, vol. 9. pt. 5, no. 20, Dec. 1913, pp. 301-310,

pi. 15, text fig. a.
Barbour, T., and A. Loveridge. 1928. A comparative study of the herpetological faunae of

the Uluguru and Usambara Mountains, Tanganyika Territory, with descriptions of

new species. Mem. Mus. Comp. Zool., Harvard College, vol. 50, no. 2, Dec, pp.

87-265, pis. 1-4.

Skulls of Gymnophiona and Their Significance 631

Boettger, O. 1913. Reptilien unci Amphibien von Madagascar, den Inseln und dem Festland
Ostafrikas. In Voeltzkow, Reise in Ostafhka in den Jahren 1889-1895, 1903-1905.
Stuttgart. Band 3, Heft 4, pp. 269-375, 8 pis.

Boulenger, G. A. 1882. Catalogue of the Batrachia Gradientia s. Caudata and Batrachia
Apoda in the collection of the British Museum (Natural History), pp. i-viii, 1-127,
pis. 1-9.

. 1883. Descriptions of new species of reptiles and batrachians in the British Museum.

Ann. Mag. Nat. Hist., ser. 5, vol. 12, pp. 161-167, pi. 5.

. 1909. A list of the freshwater fishes, batrachians, and reptiles obtained by Mr. J.

Stanley Gardiner's Expedition to the Indian Ocean. Trans. Linn. Soc. London, Ser. 2,
vol. 12, 1909, Zool. (1) pp. 291-300, pi. 40.

Cope, E. D. 1866. Fourth contribution to the herpetologv of tropical America. Proc. Acad.
Nat. Sci. Philadelphia, vol. 18, pp. 123-132.

. 1877. Tenth contribution to the herpetology of tropical America. Proc. Amer. Philos.

Soc, vol. 17, Aug. 15, pp. 85-98.

. 1887. Catalogue of the batrachians and reptiles of Central America and Mexico.

Bull. U.S. Nat. Mus., no. 32, pp. 1-98.

Daudin, F. M. 1802-03. Histoire naturelle, generale et particuliere des reptiles; ouvrage
faisant suite, a l'histoire naturelle, generale et particuliere composee par Leclerc de
BufTon, et redigee par S. C. Sonnini, Paris. Tomes 1-8 (tome 7, pp. 411-429, pi. 92).

Dumeril, A. 1863-1864. Catalogue methodique de la collection des batraciens du Museum
d'histoire Naturelle de Paris. Mem. Soc. Imp. Sci. Nat. Cherbourg, tome 9, 1863,
pp. 1-12 and tome 9, 1864, pp. 307-321, pi. 1.

Dumeril, A. M. C, G. Bibron, and A. Dumeril. 1834-54. Erpetologie generale ou histoire
naturelle complete des reptiles, vols. 1-9 (Caecilians, vol. 9, pp. 390-391 under
Famille, Peromeles).

Dunn, E. R. 1942. The American Caecilians. Bull. Mus. Comp. Zool. Harvard College,
vol. 91, no. 6, pp. 439-540.

. 1944. Notes in Colombian Herpetology, III. A new caecilian of the genus Gymnopis.

Caldasia, Bogota, vol. 2, no. 10, pp. 47-48.

. 1945. A new caecilian of the genus Gymnopis from Brazil. Amer. Mus. Novitates,

no. 1278, p. 1.

Fuhrmann, O. 1914. Le Genre Thyphlonectes in Voyage d'exploration scientifique in Colom-
bie by Dr. O. Fuhrmann et Eng Mayor. Mem. Soc. Sci. Nat. Neuchatel, vol. 5, pp.
112-138, text figs. 1-16.

Gorham, S. W. 1962. Gymnophiona, Das Tierreich, Lief. 78, pp. 1-25.

Mirando-Ribeiro, A. de. 1937. Alguns batracios novas das colleccoes do Museu Nacional.
O Campo, May, pp. 66-69.

Nieden, F. 1912. Ubersicht iiber die afrikanischen Schleichenlurche (Amphibia, Apoda) mit
einer Bestimmungstabelle. Sitz.-Ber. Ges. natur. Fr., 1912 (Jan.), no. 3, pp. 186-214.

Noble, G. K. L>24. Contributions to the herpetology of the Belgian Congo. . . . Part III,
Amphibia. Bull. Amer. Mus. Nat. Hist., vol. 49, May 19, 1924, pp. 147-347, pis.
23-42, text figs. 1-8.

Parker, H. W. 1936. Amphibians from Liberia and the Gold Coast. Zool. Meded. Leiden,
vol. 19, pp. 87-102.

. 1941. The caecilians of the Seychelles Islands. Ann. Mag. Nat. Hist., ser. 11, vol. 7,

pp. 1-17, figs. 1-5.

. 1958. Caecilians of the Seychelles Islands, with descriptions of a new subspecies.

Copeia, June 18, no. 2, pp. 71-76.

Peters, W. C. H. 1874. Uber neue Amphibien (Gymnopis, Siphonops, Polypedates, Rha-
cophorus, Hyla, Cydodus, Euprepes, Clemmys). Monatsb. Akad. Berlin, pp. 617-618,
pi. 1, fig. 2.

. 1879 (1880). Uber die Eintheilung der Caecilien und inbesondere iiber die Gattungen

Rhinatrema and Gymnopis. Monatsb. Akad. Wiss. Berlin, 1880 (1879), pp. 924-943,
pi., figs. 1-8.

. 1880. Eine Mittheilung iiber neue oder weniger begannte Amphibien des Berliner

Zoologischen Museums. . . . 1880, pp. 217-224, 1 pi., figs. 1-4.

632 The University Science Bulletin

Seshachar, B. R. 1939. On a new species of Uraeotyphlus from south India. Proc. Ind.
Acad. Sci., vol. 9, sec. B, Apr., pp. 224-229, pi. 24, figs. 1-3, and text figs. 1-2.

Taylor, E. H. 1964. A new species of caecilian from India (Amphibia: Gymnophiona).

Senck. Biol., Band 45, Heft 3/5, pp. 227-231, figs. 1,2.
. 1965. New Asiatic and African caecilians with redescriptions of certain other species.

Univ. Kansas Sci. Bull., vol. 46, no. 6, pp. 256-302, figs. 1-28.

. 1968. The Caecilians of the World: a Taxonomic Review. Univ. Kansas Press, Law-
rence, Kansas, 848 pp.

. 1969a. Miscellaneous notes and descriptions of new forms of Caecilians. Univ. Kansas

Sci. Bull., 1969, vol. 48, no. 9, pp. 281-296, figs. 1-9.

. 1969b. A new family of African Gymnophiona. Univ. Kansas Sci. Bull., vol. 48,

no. 10, pp. 297-305, figs. 1-5.

Wiederscheim, R. 1879. Die Anatomie der Gymnophionen. Jena. 1879, pp. 1-101, 9 plates.

Skulls of Gymnophiona and Their Significance

633

c\j m r< ^ id ^ oo <r>

o

C\J 00

3

■a j

c *

- ~* r-

~ C

•->

o -

U "

n) «

C 0

u C
Q. 0

3 So

— 3

u 2
« c

H U

C3

3 ON

cr— '

_ _2 2

— TJ

o •- .Ss a"
N * =

1? c a00

c
u

a

-IT <j
« 3

rj « S >
- 2 3 «

634

The University Science Bulletin

Skulls of Gymnophiona and Their Significance

635

3

60
C

oo

60

c

o

60

u

o

oo

Z u

J Efl

"O U

H §

W w
u

t/5 xi

■I E

1 £

§,3

«1 —

g
E

636

The University Science Bulletin

3

tt,

C

a

o

So
c

o

c

O

o.

fee
c

c/3

Q.

. c

pa ^

F 5
lE

O O

-s: -o

Skulls of Gymnophiona and Their Significance

637

u
O

Z

u •

.5 S

i. -^

2 a-
— ■—

3 £

be -

51

c -
i- c

1 s

. JO

s o

o c

3
-*

_3^
< *

. 3

o »
o

CO

_,_ ...

w £
. E

Ui

O Os

->> =

a m
H -

c

"> o

—

638

The University Science Bulletin

3

O

So

a

Mi
C

H

H

c

C 0

— : o

— I—

o _n
U _

H S

W o

c
i- o
o S

H

I £

-> 3

la ^

o

Skulls of Gymnophiona and Their Significance

639

o
o

>
V

u

3

on

3
03

w

GO

£ £
« S

C "3"

'E-_-

" 3
— ^

•Si

o

c jz

. £

vd
(vj ;-'

ON c

O c
to ^

Q S

t-M

u °

« -

^3 0.1

S a

u

o
H

640

The University Science Bulletin

2

3

a.

£
a

3

O

c

u

3i

4_,

to

u

L.

o

Vu

tc

c

o

W)

n

J

t .

_/

c

3

E

CQ

~

t-

od

rr)

„

i — i

3

J<

«

2

0

— ■

_G

3

~

u

J

C/0

5

2

E

DC

H

DC

o

UJ

(N

^*

'

3

C9

H

0

J=

«

So

«

c

t^

u

0

'

.bo

PO

s

„

c

jj

C

a
O

u

a

x

od

<-t-l

0

^

lU

So

Skulls of Gymnophiona and Their Significance

641

E
£

«■>

CN

C

u

a.

c

o

H

c

CO

H E
S~5

— 3

■ 2 =

5 J3

11

£ E

S £

3J

642

The University Science Bulletin

Csl

be

C

X

3

Skulls of Gymnophiona and Their Significance

643

u

a

c

H

c

z

ZM

>> -

wo

•~ ON

C __-
V "5

& j=
2 «
|J
5 ...
<J £

— ON

. o

644

The University Science Bulletin

so
c

u

o

-c
Kb

c

o
H

m

O

2:

o
U

DC
E

a

c
a

Q

g E

<3 H

£ •-

o

*■ £
>> £

3

Skulls of Gymnophiona and Their Significance

645

CM
CM

3

C

So
c

o

M
C
u

o
H

IS

o
O

3
*3

u

CM
ITS

O

z

u

>3
♦a
"3

O OO
3

646

The University Science Bulletin

■a

3
^4

O
C

C

E

in

Skulls of Gymnophiona and Their Significance

647

03
O

c

c

03

c

3
tao
3

3
>
V

ba

c

$*

■ 3

•-t-f

— o

ta 3

a m

3 m

o o

3 2

S E

S E

t+l

£ °

s ™

o ~

2 £

. a

P-,

M

c
u

648

The University Science Bulletin

o

W>

c
o

o

rt

T3

C

—

o

-C

U

CD
O

_0

o

o
rN

t~^ :

6
Z'

c £
u

o t^

5 °

■H -S
5^ so
5 «

& E

"o E

CO o

c

|
u

ft

Skulls of Gymnophiona and Their Significance

649

o

3
J4

O
C

O

o

B

u

a

be
c

o
H

c
*c

n

Q

c

ro

z

D

bo

C
jy

3
O

P9

.b«

<o — -

3

£^

~a

650

The University Science Bulletin

o

C

o
o.

to
c

O

a

3

Skulls of Gymnophiona and Their Significance

651

3

fcsO
C

3J

r*1

u

c

o

V3

u-\

<

c
c

3

\> _
&0 —

o -a

ON

~E
2 E

o

652

The University Science Bulletin

■a

C

oo

c

o/j
C

H

w

o

o
<

oo
o
oo

u

_0 oo

."3 ±L

4l

3
-a!

Skulls of Gymnophiona and Their Significance

653

o

ti

c

o

ON

d

c

o
H

U

<

MO

Z

X
H

£ =

654

The University Science Bulletin

o

-S

o

ON

3

bo
C

Ml

a

bo

a

o

o

u

W

6

2

o

0

GO

s

E

I

H

X
w

>>

.*»

o

G

5J r--

3

Skulls of Gymnophiona and Their Significance

655

-a
'0

o"

3

C
o

oo

o

_c

to

c

o

o

bo

o

a

ON

d
Z

X

z

<
a

3

S 3

656

The University Science Bulletin

3

O
.£

So
c

U

II

D.

St

C

H

c
o

N

C

O

C
u

c

o

S3

O

u

a

o
U

O so

O-

cm -q

Uh

Skulls of Gymnophiona and Their Significance

657

l/N

On

00

3

be
c
v

£

C

iT

c

0
N

c
U

3

ON
O

o

s

1—1

D

Q.
O

U

_2

<:

to

o

G

0

658

The University Science Bulletin

J3

~a

o

3

M

c

+1

O

oo

C

o

M
C
u

o
H

o
~
rt

3
<J

W

o

o
U

s

H
X

c
c

3

Q

fc^

Skulls of Gymnophiona and Their Significance

659

o

-3

o
rsi

3

C
V

C
v

o

M
C

O

s

u

a

3

o

c*4

(V)

o
Z

NJ

U

o

Ql

3

660

The University Science Bulletin

o

-£

c
u

r<1

a

b/j
C

O

es

o
U

Ih

O

O
oc
VO

6

a:
z

s

D

R oc

3

tt.-

*1

Skulls of Gymnophiona and Their Significance

661

3

SO

c

so

c

ON

z

D

H
o

o

S a

662

The University Science Bulletin

3

C
C

C

H

rA

o
U

0
bo

u

X
D

o

3

Skulls of Gymnophiona and Their Significance

663

E
£
o

C

0

e

'S.

c

3

c
n

5

Z

■^ P

>^ 3

S c

11
^ .„

5« p

o 3C

Q-a

IN"! ■

C
u

664

The University Science Bulletin

£

c*

o
H

a

c
c

I)

JO

ON

6

z
z

D

H

2 r^:

D —

. ^

c

Skulls of Gymnophiona and Their Significance

665

o
c

0

c

-C
U

D

« =
0

Q 5

rv, i

. ^°

6 — '

666

The University Science Bulletin

q
u

O

So
c

£

c£

o
0

a

C

3

o

H

o

os

o

WD

o
■*■

d

H_3

.8 £

.'-> 3

S.-S

c

0 £

o

Skulls of Gymnophiona and Their Significance

667

c

I

'u
u

c

o
c

s

c
U

o

U

3

c

OS

CM

c a
Z 5

UJ -S

*3 is*

■*5 ^>

O ^

2 0

c
m — '

(V) .-

ON

in

668

The University Science Bulletin

3

C

c

u

u

a

to

c
o

3

c
o

N

a,

o

o

S3

o
5 'o

"S.-S

3 3

Skulls of Gymnophiona and Their Significance

669

3

c

C

u

o

w

Q

u

c

CO

vo
6

z

s
5

3

72

670

The University Science Bulletin

3

3

It

c

Ih

m

c

of

3

a

o

U

o
o

3C

"n

Skulls of Gymnophiona and Their Significance

671

o

3

SO

c
o

*<1

so

c

o

z

^ c

s .2

"3 -1

2 E

672

The University Science Bulletin

£

E

o

3
J*

C

si
c

£

E

oo

o

o

m

6
Z

N
U
2

P

3

G

a.

o =

3

Skulls of Gymnophiona and Their Significance

673

o

.e
to

c

it.

C

o

H

O

(75

H
DC

S E
o

3 r^!

&1

2 «

s -5
o

Sj •-

rO E

3

674

The University Science Bulletin

3

C

o

e

m
o
H

U

u
u

-r
d

o
U

3

J3
'J

a

W3

| in

"s — *
§•"3

ft

o"

Skulls of Gymnophiona and Their Significanxe

675

3

3

n:

O

c

JO

■ft.
o

<3

S+l

676

The University Science Bulletin

3

O

M

C

c

V

u

a

o

_G

So
c

o

c
u
fe4

a
o

o
0

z

o

bo

c

o

3
O

o
bo

So

S B

3 J

£ —
E
E

Skulls of Gymnophiona and Their Significance

677

o

3

V4-(
0

C

c

M
C

o
H

c
u

B
0

u
^

o

CN

6
Z

N

U

v

C

u

"5
o

CQ

o

fc C

s .

o =
3.8

678

The University Science Bulletin

c

V

be
c

H

6

C/J

in

on

N

p

'-4-J

5 —

o «

a: m

u, -

m

(N

Skulls of Gymnophiona and Their Significance

679

c

a
u

a

C
u

O

U

s

X
H

i2 l"~-

■2 c

UE

u. =

680

The University Science Bulletin

3
J4

be
C

4J

o

u

a.

be
C

o
H

u

CO

CI

CO

d
Z

s

u
M
C
u

"3
o

03

R E

-5: OG

■§•3

^ °

.'%

T S

2 £

Skulls of Gymnophiona and Their Significance

681

3

O
M

C

v

o

be
C

o

c
o

U

o

O

so

U

« -

o -6.

On -£

£ • -

682

The University Science Bulletin

00

c

V

Oh

be
C

o
H

N
C

a

be
c

rs

3

D

<
o

Z E

O rv-i

H o

ty)_'

o

. . t

cq

o

.

-=

o

in

Wl

B

o

u

—

■-

Skulls of Gymnophiona and Their Significance

683

e
«
be

c

3
3

Z E

_. e

(J ro

-< 3

£ °

W S

M ...

3X.

c =

rs 3
u ^
3 "

i'c
^ c

b«rM

- C

•2 o

^ o

Kg

* OJ

o —

£

684

The University Science Bulletin

3

O

_c

bo
C

cm

a

o

be
c

o

c
u

H

o

a
o

H

cm

CD
O
CN

d
2

N
U

S

u

-3

O

I £

3

MD

'r* 3

£^
-5

cm £

2 I

Skulls of Gymnophiona and Their Significance

685

CM

a

u

C

o

OS

c

V

C

U

o
Q

CO

2
E
h

X

^

"3 o

E-fl

bo —

2 =

u-\ be
S

fc

686

The University Science Bulletin

be
C

V

o,

o

C
u

o
H

-a
c

3
O
C/3

M

rt

a!

5

u
H

ON

fN

6

N

U

^3

Skulls of Gymnophiona and Their Significance

687

a

c

o

c

c
u

3

U

s

Q.

O

Q.
O

O ~

£^

NJ =
U 3

. ,-^

u. O

■E-s

H-5

2 °.
5 o

53 —1

<=: 3

o M

c
in ■«

fr,

in

o
"1

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLVIII Pages 689-719 November 14, 1969 No. 16

Biology of the Bee Genus Agapostemon
(Hymenoptera : Halictidae) .'

Radclyffe B. Run I RTS3

ABSTRACT

Insectary observations on life history, foraging, mating, sleeping and nesting
behavior of Agapostemon radiatus, A. splendens and A. texanus are presented.
The biology of the entire genus, insofar as known, is summarized. Lack of
sociality and the presence of a single cell at the end of each of the long lateral
branches of the main nest burrow support the conclusion that Agapostemon is
more closely allied with a South and Central American group of genera than
with the genera allied with either Augochlora or Halictus. Unlike many bees,
none of the species of Agapostemon studied in the insectary lines its cells with
materials other than those secreted by the bee constructing the cell and none
constructs a spiraled cell closure. Field and insectary observations indicate that
./. radiatus and A. texanus nest in loam and ./. splendens nests in sand. Naive
insectary-reared bees gather pollen without hesitation or ineptitude from flowers
with tubular anthers, demonstrating that the complex and peculiar behavioral
responses of bees to these flowers are innate and not learned. In the course of
the study, a lighter and less bulky type of observation nest-box was developed
tor use in the insectary.

INTRODUCTION

Initiated as an adjunct to a revision of the genus Agapostemon. this study
of the life history, nest architecture and nesting biology is in some respects
preliminary. Nevertheless, biological information on this group of halictines
is so scarce as to warrant such a study.

1 Contribution number 1414 from the Department of Entomology, University of Kansas,
Lawrence, Kansas.

2 This study was supported by National Science Foundation Grant GR 3151 to the
University of Kansas (C. O. Michener, principal investigator).

3 Present address: Department of Entomology, Oregon State University, Corvallis, Oregon
97330.

690 The University Science Bulletin

According to G. C. Eickvvort (1969), the genera of American Halictinae
can be placed in three groups: (1) the typically non-metallic group in-
cluding Halictus, Lasioglossum, etc.; (2) the typically metallic group
including Augochlora, Augochlorella, Augochloropsis, etc.; and (3) the
typically metallic group including Agapostemon, Paragapostemon, Pseud-
agapostemon , etc. The nesting biology of representatives of the first two
groups has been extensively studied both in the field and in the insectary
by various authors (recent references are Batra, 1966; Ordway, 1966; Stock-
hammer, 1966). Biology of representatives of the Agapostemon group is
poorly known. Detailed studies have been published on the nests of the
South American Pseudagapostemon (Michener and Lange, 1958a) and
Ruizantheda (Claude-Joseph, 1926) but despite its abundance, observations
on the nesting biology of Agapostemon are few and sketchy.

Despite efforts to obtain data on field nests of Agapostemon from 1962
until 1967, 1 was able to find only a single nest, and it could not be excavated.
Thus all my investigations on nesting biology were conducted in insectaries
at the University of Kansas. Field observations on Agapostemon nesting
made by others are compared with insectary findings. Certain published
observations on "Agapostemon splendens" are, in large part, erroneous and
are largely disregarded (see Appendix).

The species most common in the vicinity of Lawrence, Kansas are
Agapostemon texanus, A. radiatus and A. splendens. All three were studied
in the insectary, and the observations on nesting biology are based on 65
nests (28 A. texanus, 23 A. radiatus and 14 A. splendens).

ACKNOWLEDGMENTS

I am grateful to the following individuals for contributing both information
and valuable suggestions used in this study: S. W. Batra, G. E. Bohart, G. C.
Eickwort, K. Eickwort, W. B. Kerfoot, W. E. LaBerge, R. J. Lavigne, C. D.
Michener, D. W. Ribble, W. P. Stephen and K. A. Stockhammer. Special
thanks are extended to Dr. G. W. Byers and Dr. C. D. Michener for their
helpful advice and careful editing of the manuscript, and to my wife, Guinnevere,
for her aid in the preparation of the manuscript.

MATERIALS AND METHODS

The insectary in which the observations on nesting behavior were made consists of
several small rooms with fluorescent lighting ami air-conditioning. The rooms differ only
slightly in size and intensity of illumination. The room used tor most of the investigations
of nesting behavior has been described by Batra (196-1).

The fluorescent lights were operated by two electric timers. The majority of the lights
were on one circuit and were turned on for 15 hours each day. The remaining lights were
set to go on 30 minutes before the main lights and to remain on until 30 minutes after the
main lights had gone off. The purpose of this two circuit system was to warn the bees of
impending darkness, thus giving them an opportunity to enter their nests or find sleeping

Biology of the Bee Genus Agapostemon 691

sites. A two step increase in light intensity used in the morning was probably unnecessary.
The bees were kept on a 16 hour light regime in order to avoid the possibility of inducing
ovarian diapause under "short day" conditions. More for the comfort of the investigator than
for the bees, the air-conditioner was set to maintain a room temperature of about 82° F
(27.7° C) during the summer months. At this temperature the relative humidity varied from
70-90%. During the fall and winter months the temperature dropped to about 70° F
(21.1° C) and the relative humidity sometimes dropped as low as 50%.

In the field, female and male Agapostemon were captured at flowers. The bees were
removed from the net with an aspirator consisting of copper tubing about 20 cm long and
7 mm in diameter. The copper tubing was attached to a piece of plastic (Tvgon) tubing
which was held in the mouth. The bees were sucked into the copper tubing but were prevented
from entering the plastic tubing by a piece of fine mesh copper screening. They were retained
in the copper tubing by covering the opening with a finger. The finger was removed as the
tip of the aspirator was inserted through an X-shaped cut in the polyethylene lid of a "two
pound" coffee can. A sharp puff expelled the bees from the aspirator and the X-shaped cut
closed as the aspirator was withdrawn from the can. The polyethylene lid of the can was
perforated for ventilation by melting numerous small holes with a heated wire. The can
contained several crumpled sheets of dam]) paper toweling which provided the bees with a
substrate to which they could cling, a cushion to prevent injury when they were expelled
from the aspirator and a humid atmosphere to prevent dehydration. If the bees flew against
the lid of the can or if the can became too warm, mortality rose- rapidly to 100%. When the
outside of the can was covered with aluminum foil to reflect heat and exclude light, the bees
rested quietly on the toweling and could be carried in the field all morning in hot weather
with little or no mortality.

In the insectary the bees were fed with both honey and tresh flowers. Honey was placed
in the centers of plastic flowers and diluted with sufficient water (about one part water to
two parts honey) to reduce its viscosity to the point where the bees could drink it. Excessive
dilution reduces the osmotic pressure so much that \ easts and imperfect fungi become a prob-
lem. Even when sufficient flowers with nectaries are available, the bees frequently utilize the
honey on the plastic flowers (Fig. 18). The availability of this honey prevented mass starva-
tion on those occasions when fresh flowers were not available. It was found that female
Agapostemon could be maintained on honey alone during late fall and winter whin flowers
were not available.

Flowers for feeding the bees were selected on the basis of the following criteria: abundance
or obtainability, amount of pollen produced and longevity alter being cut. In chronological
order of their blooming, those used were: Salix, Prunus, Mains, Cirsium, Amorpha, Ratibida,
Sdphuim, Chamaecrista, Heliantlnts and Aster. These flowers were also those most frequently
visited by Agapostemon in the field. Other flowers frequently visited by Agapostemon
(Opuntia, Convolvulus, and Cucurbita) wilt soon after cutting and were not utilized in
the insectary.

In order to learn more about the behavior of Agapostemon, a large observation nest-box
was constructed (Fig. 3). Modeled after those used by batra (196-1) in her study of Lasio-
glossum, its essentials were two sheets of plate glass (0.6 x 91.4 x 91.4 cm) separated by
a 4 mm laser of soil, a wooden frame to hold the glass in place, a V-shaped trough of soil
above the glass and black oilcloth over the glass to exclude light. As the weight of the layer
of soil was sufficient to bow the glass outward, external braces were employed to maintain
a constant distance of approximately 4 mm between the sheets of glass so that the burrowing
bees would not be concealed by the soil. The main burrows, laterals, and cells of the
Agapostemon species studied were all at least 4 mm wide so that the bees were visible through
the glass. Only the cell entrances were slightly less than 4 mm wide and thus obscured by
soil. The nest-box was none too large, for some individuals of A. texanus, A. radiatus and
A. splendens dug all the way to the bottom. It was found that if damp soil were used in
the initial construction of so large a nest-box, it would remain damp for at least six months,
therefore no special means lor replacing moisture were necessary. Desiccation of the top few
centimeters of soil was prevented by lightly sprinkling the soil surface with water. It is easy
to over-water this type of nest-box and pre-imaginal mortality is the result. While this large

692 The University Science Bulletin

nest-box permitted simultaneous observation of as many as ten nests, it was satisfactory only
for general observations inasmuch as the nests and their contents were inaccessible.

In order to permit access to larvae, two additional nest-boxes were constructed of sheets of
transparent plexiglass 0.25 x 30 x 60 cm (Fii^s. 1-2). Holes were drilled in the sheets at
about 10 cm intervals and the two sheets, spaced apart by washers, were bolted together.
Wing nuts were used in order to facilitate subsequent dismantling. A single length of
plastic (Tygon) tubing was forced between the two sheets along the sides and bottom to act
as a seal . This technique proved very satisfactory and resulted in a nest-box which was
unbreakable and lighter, stronger and more compact than glass nest-boxes. As the plexi-
glass nest-boxes required no bulky frame, they were easy to move and manipulate. The
principal disadvantage of plexiglass is that it is easily scratched and has a greater tendency
to bow outward, hence the necessity of bolting it at 10 cm intervals. The Tygon tubing seal
on the sides and bottom of the nest-box is easily penetrated by a hypodermic needle, thus
facilitating the addition of water or other fluids to the soil. In order to open the nest-box for
the removal of larvae or to manipulate the contents of the nests, the nest-box is laid flat and
the wing nuts removed. The top sheet of plexiglass is then removed, the soil not adhering
to it as it does to glass. Later the plexiglass may be replaced anil the nest-box set back in
place with no damage to the contained nests.

Three soil-filled wooden boxes 30 x 45 x 30 cm were used in addition to the observation
nest-boxes. These wooden nest-boxes accommodated nests of those bees not inhabiting the
observation nest-boxes.

OBSERVATIONS
Seasonal Cycle

In the vicinity of Lawrence, Kansas, overwintering females of Agapos-
temon texanus, A. radiatus and A. virescens emerge from hibernation in
April, when they are commonly found foraging on Salix spp. and Primus
americana. In May and early June, when Salix and Prunus have finished
blooming and before most of the Compositae bloom, females of Agaposte-
mon are rarely seen. In June flowers become abundant in the open areas
and female Agapostemon are common on Cirsium. As the season progresses
the number of females remains almost constant, but males become in-
creasingly abundant. In June males are rarely found, but in late September
they become extremely abundant and females become scarce.

Probably the three species of Agapostemon studied have only two gen-
erations per year in Kansas. One generation overwinters as fertilized
females and, as evidenced by a few males collected in early April, about
0.1% of the males survive the winter. Overwintered females nest in April,
and evidence from the insectary indicates that most of them die by mid-
May. The summer generation begins to emerge in early June and consists
almost entirely of females. These females begin nesting immediately and,
judging by the appearance of numerous males in July and August, most of
their offspring are males. The females which emerge in August could not
be induced to nest in the insectary and are presumably the overwintering
generation. These females may be distinguished from the summer genera-
tion by their unworn wings and mandibles and by their slender ovaries.
These newly emerged females visit flowers for nectar and perhaps pollen

Biology of the Bee Genus Agapostemon 693

(presumably to build fat reserves) and are fertilized by the very abundant
males. Both males and females disappear after the first hard freeze (usually
in late October). In the insectary these diapausing females spend most of
their time in old burrows, emerging only occasionally to feed. Unlike
nesting females, several of the diapausing females may crowd into a single
burrow, or hibernaculum (Fig. 14).

This two-generation-per-year life cycle seems to be very efficient. Males
are abundant only in the late summer and fall when there is a super-
abundance of food (Silphium, Helianthus, and other Compositae). The
overwintering females are already fertilized when they emerge in the spring.
They nest immediately and utilize the abundant and relatively localized
pollen of fruit trees and willows in provisioning the cells which produce
the summer generation. Agapostemon adults are essentially absent in May
and early June, the period of heaviest rains, when the trees have finished
blooming and before most of the pollen-rich Compositae bloom. Excluding
wind pollinated species, the number of species of plants blooming in
unwooded areas (Agapostemon is seldom found in woodlands) during May
nearly doubles during June. In the aculeate Hymenoptera, males are
normally produced from unfertilized eggs, and females from fertilized eggs.
Presumably because their overwintered mothers were fertilized, the majority
of Agapostemon emerging in June are females. The shortage of males in
June means that many of the summer generation females lay unfertilized
eggs, thus the majority of their offspring are males. The abundance of
males in the late summer and fall ensures that most, if not all, of the
females will be fertilized before hibernation.

Nest Orientation

Individual species of Agapostemon are known to nest both in banks and
in level ground, whereas many halictines nest in either one or the other.
Nest entrances of the following species have been found in banks: A.
radiatus (Rau, 1934), A. nasutus (Daly and Wille, in Sakagami and Mich-
ener, 1962; G. C. Eickwort, personal communication), A. spendens (Stevens,
1921) and Agapostemon sp. (probably A. radiatus or A. texanus, S. W.
Batra, personal communication). Nest entrances of the following species
have been found in horizontal or only slightly sloping surfaces: A. radiatus
(LaBerge and Ribble, 1966; K. A. Stockhammer, personal communication),
A. splendens (LaBerge and Ribble, 1966), A. virescens (Felt, 1928; W. E.
LaBerge, D. H. Janzen and G. C. Eickwort, personal communications),
A. texanus or A. angelicus (probably A. angelicus, personal observation),
and Agapostemon sp. (W. B. Kerfoot, personal communication).

In the insectary A. texanus, A. radiatus and A. splendens were presented
with horizontal soil surfaces. All three species readily entered these, but

694 The University Science Bulletin

equally significant are nests (one of each species) which were entered
through small apertures in vertical sides of the nest-boxes.

Sakagami and Michener (1962) included A. splendens in a list of "species
whose nests are known only or principally from flat ground" hut in the only
reference cited (Stevens, 1921), the nest was in the side, or bank, of a sand
"blowout, " or wind-carved depression. In the light of my insectary observa-
tions and field observations of LaBerge and Ribble (1%6) this species,
together with A. texanus and A. radiatus, should be included in the list of
species known to nest in both horizontal and vertical surfaces.

Aggregative Tendencies

There is some question as to whether the species of Agapostemon usually
nest in aggregations or are usually solitary. Felt (1928) reported an aggregation
of A. virescens (given as HaUctus virescens) damaging a lawn in Catskill,
New York: "Bare spots were reported here and there in the lawn and the
bees were said to be burrowing more or less over its entire surface although
there were areas where they were decidedly more abundant." W. P. Stephen
(personal communication) has seen nesting aggregations of A. virescens in
a lawn in Corvallis, Oregon and D. H. Janzen (personal communication)
has found numerous nests of A. virescens in a field near Eugene, Oregon.
Nesting aggregations of A. nasutus were found in Costa Rica by Daly and
Wille (Sakagami and Michener, 1962) and by G. C. Eickwort (personal
communication). Nests of A. radiatus may also be locally abundant, as
Rati (1934) reported: "Dozens of these bees were at work burrowing
vertically in a bank of very fine sand at Creve Coeur Lake, Mo., August 8,
1922." Stevens ( 1(>21 ) and LaBerge and Ribble (1966) reported finding
nesting aggregations of A. splendens.

On the other hand, LaBerge and Ribble (1966) and K. A. Stockhammer
(personal communication) have seen isolated nests of A. radiatus; K. A.
Stockhammer, S. W. Batra and W. B. Kerfoot (personal communications)
have seen isolated nests of Agapostemon spp., and I have seen an isolated
nest of A. texanus or A. angelicas (probably A. angelicus). Insectary ob-
servations show that A. texanus, A. radiatus and A. splendens nest as readily
in isolated containers as they do within a few centimeters of each other.

It appears that nests of Agapostemon species, unlike those of many of
the halictines, may be either aggregated or solitary. The tendency to nest in
aggregations seems to vary among the species. The species which are known
only to nest in aggregations, A. virescens (W. E. LaBerge, personal com-
munication) and A. nasutus (G. C. Eickwort, personal communication),
are those in which numerous females often utilize a common burrow. Pos-
sibly the tendency of A. virescens and A. nasutus to nest in aggregations is
linked to this mutual tolerance among females or possibly to some tendency

Biology of the Bee Genus Agapostemon 695

for females to nest near their birthplace. The females of A. texanus, A.
radiatus and A. splendens have not been seen to nest in natural aggrega-
tions as dense as those of A. uirescens and A. nasutus, and in the insectary
do not tolerate the presence of other females in their nests.

Edaphic Preferences

Agapostemon splendens has been found nesting in the sand hills of
North Dakota (Stevens, 1921) and in the sand hills of Nebraska (LaBerge
and Ribble, 1966). In the insectary the bees were supplied with a choice of
sand or loam in which to nest. All but one of the 22 nests of A. splendens
were in sand, whereas A. texanus and A. radiatus always nested in the loam.
Although the bees do exhibit edaphic preferences in the insectary, one must
be careful in extrapolating from laboratory observations, as Rau (1934) and
LaBerge and Ribble (1966) report A. radiatus nesting in sand.

The insectary and field observations on the edaphic preference of A.
splendens are corroborated by its range and local distribution. This species
is common only in the eastern three fifths of the United States and probably
does not occur in very arid regions. Although I have seen several specimens
from the arid Southwest, it is reasonable to assume either that they are
mislabeled, were recent accidental introductions, or represent local popula-
tions in limited areas of moist sand. A. splendens is most common along
the Gulf and Atlantic ('oasts, the margins of the Great Lakes, and the sand
hill regions of Kansas, Nebraska, the Dakotas, and Manitoba. In all ol
these regions sand is the principal soil type. Although I have collected
Agapostemon extensively in the vicinity of Lawrence, Kansas, I have rarely
seen A. splendens outside the sandy floodplain of the Kansas River.

The biological significance of the edaphic preference of A. splendens is
obscure. From insectary observations it is apparent that excessive moisture
and the resulting increase in the growth of fungi is correlated with, il not
the cause of, high pre-imaginal mortality. As A. splendens nests in regions
where the summer rains are frequently heavy, it is possible that a well
drained soil (e.g., sand) with relatively few microorganisms would mini-
mize the mortality of the pre-imaginal bees. Discounting the several speci-
mens from the Southwest mentioned above, it is interesting that this species
is not found in the desert or other areas of low precipitation even though
sand is common in such regions.

Nest Location

When first released in the insectary, the females of A. texanus, A. splen-
dens and A. radiatus exhibit an "escape" reaction by flying up towards the
lights and bumping against the transparent ceiling for about 15 minutes.

696 The University Science Bulletin

When the bees calm down, they begin visiting the flowers for nectar but not
pollen. After some minutes of feeding and resting on flowers, some females
begin what appears to be a "searching" or "investigating" flight. Almost
all females captured in May exhibit this behavior but those captured later
in the summer are increasingly refractory and by late August none of the
females introduced into the insectary search for nest sites.

During this slow, wavering, "searching" flight the female flies about the
room, 3-10 cm from the floor or walls. When flying above the floor she
lands occasionally and walks about for a few seconds before resuming
flight. When flying near the wall the bee always faces the wall, landing
occasionally to investigate nail heads (which happened to be about the
diameter of an Agapostemon burrow) or other dark spots on the straw-
colored walls. This "searching" behavior is often interrupted by rests and
visits to flowers for nectar.

The bees seem to be attracted to dark spots and shadows. When holes
are poked in the soil of the nest-boxes with a pencil, the bees often enter
the holes and begin nest construction. They also enter the nests of other
bees and, unless ejected, begin further excavation.

Small stones and assorted plastic leaves were scattered on the surfaces of
the soil in nest-boxes and, unless presented with the ready-made holes men-
tioned above, the bees made their nest entrances under the stones and
leaves. When a stone is so embedded that the bees are unable to crawl
under it, they begin digging at its margin. In the same manner, they dig at
the intersection of the soil surface and the side of the nest-box. Although
the bees initiate their nests under the stones and leaves, they do not abandon
their nests if the stones or leaves are removed.

K. A. Stockhammer and S. W. Batra (personal communications) have
seen Agapostemon nesting in the Lawrence area and every nest was either
under a leaf or small stone, or at the intersection of vertical and horizontal
soil surfaces. I have seen a nest of A. te.xanus or A. angelicas (probably the
latter) concealed beneath a prostrate stem of Opuntia near Coaldale, Colo-
rado. Neither Dr. Stockhammer, Dr. Batra nor I would have found the
nests had we not been watching the returning females. Perhaps this com-
mon habit of secreting the nest entrance is one reason the nests of Agapo-
stemon are seldom found. One can only speculate on the biological signifi-
cance of this habit. It may conceal the nest from the eyes of parasites such
as beeflies (Bombyliidae) or cuckoo wasps (Chrysididae). It may also serve
to protect the nest entrance from rain.

Nest Architecture

The nests of Agapostemon consist of a main burrow and laterals with
cells. The long axis of the main burrow is normally perpendicular to the

Biology of the Bee Genus Agapostemon 697

soil surface. The laterals normally slant slightly downward from the main
burrow and are never at the end of the burrow. Cells are nearly horizontal,
with the slightly constricted cell entrance slightly higher than the rear of
the ellipsoidal cell. The undisturbed tumulus is a low symmetrical cratered
cone of unconsolidated soil granules.

The nests of A. nasutus excavated by Daly and Wille and reported by
Sakagami and Michener (1962) are unusual in that they had two end-to-end
cells in most laterals with the older, terminal cell producing a female and
the younger, sub-terminal cell producing a male.

Although the two end-to-end cells per lateral observed by Daly and Wille
could not be confirmed by G. C. Eickwort and K. Eickwort (personal com-
munication), it is interesting that the only other species with end-to-end
cells in the laterals are South American halictines (Pseudagapostemon di-
varicatus, P. perzonatus and Ruizantheda mutabilis) closely allied to
Agapostemon (cf. Sakagami and Michener, 1962). Field observations on
A. splendens (Stevens, 1921) and A. radiatus (LaBerge and Ribble, 1966)
as well as my own insectary observations on A. texanus, A. radiatus and
A. splendens indicate that these species build only i single cell at the end
of each lateral.

The deepest nests (1.5 m) and the longest laterals (30 cm) reported for
Agapostemon are those of A. splendens, one of the largest species, excavated
by Stevens (1921). In the insectary some females of A. texanus, A. radiatus
and A. splendens dug all the way to the bottom of the largest nest-box
(nearly 1 m in depth) where they constructed cells. While A. splendens
females usually dug to the bottom of the nest-box, those of the smaller A.
texanus and A. radiatus usually confined their excavations to depths of
20 to 60 cm.

The laterals of A. texanus, A. radiatus and A. splendens were narrower
than the main burrow, as are the laterals of other halictines. The laterals of
A. texanus and A. radiatus in nest-boxes were 5-15 cm long while those
of A. splendens were from 8-20 cm long. K. A. Stockhammer (personal
communication) has seen a nest of A. radiatus in which one of the cells was
joined directly to the main burrow, although the others had laterals of
various lengths. Although the laterals observed by me were filled after cell
closure with soil taken from the sides of the main burrow, there were no
obvious pits left in the walls. The largest number of cells and laterals con-
structed in a single nest was 14 in an insectary nest of A. texanus. The
usual number of cells and laterals constructed by A. texanus, A. radiatus and
A. splendens in the insectary was six to ten, with some females constructing
as few as five.

698 The University Science Bulletin

Nest Construction

Aside from the edaphic preferences mentioned previously, no significant
specific behavioral differences were noted in observations on A. texanus,
A. radiatus and A. splendens. The nests of A. splendens were often deeper
than those of A. texanus and A. radiatus (cf. Nest Architecture) but this
difference is proportional to the larger size of A. splendens. The following
is a composite picture summarizing insectary observations on 28 nests of
A. texanus (see Daily Activities for remarks on timing of the following
activities).

After a female selects a site for her nest, she begins digging immediately.
The forelegs and mandibles are used in loosening the soil, which is then
passed beneath the body to the middle legs. Cupping the soil between the
hind legs and the underside of the metasoma, she backs up about half her
length and drops the soil. As she deepens the vertical burrow, the bee
disappears beneath the tumulus (out of sight in about 15 minutes), making
no attempt to keep the entrance clear of soil. Continued digging can only
be inferred from dirt which is periodically pushed up from below until the
bee appears between the glass or plastic sides of an observation nest. The
tumulus is allowed to accumulate around the nest entrance and, if not
disturbed or blown away, forms (on level ground) a symmetrical cratered
cone 3-5 cm in diameter and 1-3 cm high. Perhaps because the bee does not
remove the soil from the nest entrance until she leaves to forage, no turret is
ever formed. The bee continues to deepen the nearly vertical burrow. If the
bee encounters an obstacle, she deviates laterally until she can resume her
downward course. At a depth of 10-40 cm the bee constructs her first
lateral tunnel, which may or may not be at the bottom of the main burrow.

The lateral is slightly narrower than the vertical shaft and, while the
bee can turn around in the main shaft and the cells by doubling back on
herself in a forward somersault, she cannot turn around in a lateral. The
soil excavated from the lateral is carried between the hind legs and the
underside of the metasoma while the bee backs out of the lateral and up to
the top of the main burrow, where she pushes it from the entrance. Some-
times the dirt gets away from her and drops to the bottom of the burrow.
When this happens she often just leaves it until she is ready to deepen the
burrow. The completed lateral slopes downward at about 10° (sometimes
as much as 45°) from horizontal and is usually about 8 cm long, but may
be as short as 5 cm or as long as 15 cm.

The female finishes the lateral by excavating an enlarged cavity, or cell,
at its apex. The long axis of the cell is usually approximately continuous
with that of the lateral and always slopes downwards at about 10-15°. Like
the main burrow and the laterals, the cell cavity is excavated with the fore-

Biology of the Bee Genus Agapostemon 699

legs and mandibles. After excavation, the interior is tamped smooth by the
pygidium. Although the bees were not seen to line the interior of their
cells with dirt (Fig. 7) as do many other halictines, it is possible that I over-
looked such activity. Next the female begins to "lick" the interior of the
cell. In this process she turns around and around in the cell and, using
the glossa as a brush, applies a liquid to the inside of the cell with short,
rapid, forward motions of the partially extended mouthparts. As she licks
and moves about the cell, she moves her hind legs with short, rapid, lateral
motions, brushing the interior of the cells with the penicillus, a brush of
hairs at the apex of each hind basitarsus. The liquid applied to the interior
of the cell soaks into the soil and, in pale dry soil, may be seen as a dark
zone around the cavity 1-2 mm thick (Fig. 5). After slightly over an hour
of this licking and brushing, the female begins to tap the opened apex of
the metasoma about the interior of the cell while wiping the interior of the
cell with the penicilli. However, the hind legs are now moved so as to
brush the interior of the cell with slower and more "deliberate" forward
(rather than lateral) motions. I was not able to determine what, if any-
thing, is added to the interior of the cell walls during this metasomal tapping.
The tapping-brushing process lasts about two minutes before the bee leaves
the cell to begin foraging. During most of the licking and metasomal
tapping processes the antennae are usually flexed downward so as to touch
the interior of the cell.

Completion of the new cell is followed by provisioning and oviposition.
Immediately after the egg is deposited, the bee starts to leave the cell with-
out pausing or turning around. Halting with the posterior half of her
metasoma just inside the neck of the cell, the bee begins to rotate (about
two seconds per revolution) about her long axis and wipes the open apex of
her metasoma around the neck of the cell. After 10-20 revolutions the bee
leaves the cell completely and begins scraping soil from the interior of the
lateral adjacent to the cell with her forelegs and mandibles. This soil is
passed rearward beneath the body by means of the middle and hind legs
and is packed into the cell entrance with the pygidial region of the meta-
soma. Cell closure is completed in about two minutes. The female does
not rotate in the lateral during cell closure and the resulting cell plug is
unlike those of many bees in that it does not show a spiral construction
pattern.

Having closed her cell, the bee leaves the lateral and proceeds up the
main burrow for a few centimeters. She begins to dig at the walls of the
shaft with her mandibles and forelegs, and the falling soil particles are
caught between the wall and the hind legs and metasomal sterna. When
she has accumulated a load of soil, she backs down the shaft and into the
lateral. The load of soil is deposited at the rear of the lateral next to the

700 The University Science Bulletin

closed cell, where it is tamped into place by the hind legs and pygidium.
This procedure is continued for the 45-90 minutes necessary to completely
fill the lateral.

Sometimes a load of soil is seemingly accidentally dropped to the bottom
of the shaft. Attempts to retrieve the dropped soil are seldom successful,
as the bee does not turn around and go after it head first. Instead, she
backs down to the bottom of the shaft and tries to bring up the soil in the
head-up position. This is usually unsuccessful because the soil particles
catch on the wall of the shaft and roll out from between the metasomal
sterna and the wall of the shaft, falling back to the bottom.

Some time after a lateral has been filled with soil, the bee begins to
excavate a new lateral. This new lateral, with its terminal cell, may be
either above or below the one just completed (Fig. 6). As the vertical shaft
is periodically deepened, the bee is more likely to begin new laterals at
progressively deeper levels, but this tendency is not pronounced. Although
the length of laterals is quite variable, there is no apparent trend toward
longer or shorter laterals as the nest is deepened. As each lateral is filled
following cell closure, only one lateral and cell is open at a time.

The bees occasionally constructed abortive laterals. Sometimes the
abandonment of a lateral was seemimilv induced by observation, but in
other instances abandonment of a lateral appeared to be spontaneous. Such
laterals might be abandoned at any stage of construction, even after the
terminal cell had been excavated. The aborted laterals were usually filled
with soil in the same manner as completed laterals.

None of the females observed in the insectary was ever seen to begin
a second nest unless (as was the case with two A. radiatus and one A.
texanus) she was prevented from entering her original nest. Ordinarily a
bee worked in her nest until her death. Most females died outside their
nests, thus leaving them open; some (less than 20%) died inside their nests,
which were sometimes plugged and sometimes open. Those females which
died in their nests were soon covered by fungus but it was not determined
whether the fungi were pathogens or saprophytes.

A female of A. texanus nesting in one of the wooden boxes was observed
to close her nest in a unique manner. She emerged from the nest at 8:22
a.m. and began filling the burrow with earth from the tumulus. Biting at
the earth with her mandibles, she raked it toward herself with the forelegs
(used alternately) and, passing it backwards beneath her body, pushed the
earth into her open burrow. The burrow was filled by 8:30 a.m. but the bee
kept biting at the partially consolidated tumulus. She scattered the earth
with lateral movements of her hind legs and forelegs. At 8:40 a.m., having
leveled 80% of the tumulus and completely concealing the filled burrow,
she flew away and never returned. As the bee was unmarked it is not

Biology of the Bee Genus Agapostemon 701

known whether she started another nest. I can offer no explanation for this
peculiar phenomenon, but as the entire sequence seemed well coordinated,
such behavior may be normal under some circumstances.

Provisioning and Oviposition

As a bee does not have to go more than a few feet from her nest to find
pollen and nectar in the insectary, the foraging trips may be short in dura-
tion. When pollen is abundant, the bee may gather a full load and return
to the nest in as little as five minutes. The returning bee lands at the nest
entrance, enters, and proceeds rapidly to the open cell. Entering the cell
head first, she immediately turns around and removes the pollen load from
the hind legs and metasomal sterna by rubbing the hind legs together and
against the sterna while helping with the middle legs. The function of the
large spatulate teeth on the posterior hind tibial spurs is unknown, but it is
possible that they are used to help comb the pollen from the scopa. The bee
takes slightly less than two minutes (1.75 minutes minimum observed) to
enter her nest, deposit her pollen load in the cell, and leave on another
foraging trip, providing she engages in no other activities.

The pollen from the first two or three loads is simply deposited as a dry,
loose mass on the floor of the cell (Fig. S). On the third or fourth foraging
trip, the bee first takes nectar and, after spending some minutes at this, she
collects a load of pollen and returns to the nest. After removing the pollen
in the usual manner, the bee turns around once more and begins to mouth
the pollen while scraping it together with her forelegs. Presumably nectar
is regurgitated as the bee mouths the pollen mass. The mouthing and
scraping together last 3-4 minutes, during which time the bee may turn
around in the cell several times. Having thus moistened the pollen and
shaped it into a crude loaf, the bee leaves on another foraging trip. Three or
four more loads of pollen are added to the loaf. With each addition of
pollen, the loaf is further worked with the mandibles and forelegs until it
assumes the form of a slightly flattened ball. During the final stages of its
formation, the ball is rolled back and forth in the cell while the bee works
it constantly with mandibles and forelegs, all the while tapping the ball
with the ventral apex of her metasoma and tapping rapidly on the floor of
the cell with her flexed antennae.

Facing the cell entrance and grasping the finished pollen ball with hind
tarsi only, the bee applies the apex of her metasoma to the rear of the upper
half of the ball and sticks the apex of the emerging egg to the pollen. Then
the hind legs are slowly straightened, pushing the bee forward until the
anterior end of the egg drops on the pollen ball. In three observations of
oviposition, extrusion of the egg took 30-35 seconds. The sausage-shaped

702 The University Science Bulletin

egg rests on the upper rear of the pollen ball, supported only by the two
ends of the egg (Fig. 9).

Daily Activities

The females of the three species of Agapostemon observed in the in-
sectary spend the night in plugged burrows. When they are ready to forage
in the morning, the plug is removed (I never saw this take place) and the
bee usually spends several minutes with her head at the entrance before
leaving the nest. In social halictines, resting with the head at the entrance
is usually interpreted as "guarding" (Fig. 4). In Agapostemon (and per-
haps in other solitary bees), it is doubtful that the female engages in nest
defense while "guarding." If disturbed, the females of Agapostemon in the
insectary would simply drop down their burrows. Even the slightest motion
on my part was often sufficient to cause them to disappear for several
minutes. If a slender blade of grass was poked down the burrow the female
would sometimes try to push it out, but she would never come all the way
up to the entrance.

The lights in the insectary were set to turn on at 7:00 and 7:30 a.m. Most
bees begin foraging at about 9:00 a.m. and provisioning is usually completed
by about noon, with oviposition taking place shortly thereafter. The cell is
closed and the lateral filled with earth. The bee may then leave the nest to
feed or she may remain inactive. Later in the afternoon she often begins
deepening the burrow or constructing a lateral. The soil is ejected from the
burrow and some of it plugs the entrance. During the afternoon or night a
new lateral and cell may be constructed. Sometimes the new cell is not
finished until mid-morning. This schedule is highly variable in the insectary
and must be much more variable in the field where provisioning may be
interrupted by cloudy or rainy weather and where flowers may be few and
distant. In the insectary some individuals spend so much time "dawdling"
on the flowers between provisioning trips that it takes as long as two days
to provision a cell. During periods when pollen is scarce, provisioning
might take as long as three days.

The reasons for abandoning a new lateral or cell are not always apparent,
but if a bee in the process of deepening the vertical shaft or digging a lateral
is exposed to too much light during observation, she soon becomes agitated
and fills the excavations with soil. She may then leave the nest or simply
cease all activity. As construction of a cell progresses, the bee becomes less
and less sensitive to disturbance and once licking of the cell walls begins,
she usually does not react to light by filling the cell with soil. After the first
load of pollen is brought in, she can rarely be induced to abandon the cell.
Unlike Augochlora pura (Stockhammer, 1966), cells of Agapostemon are
usually not constructed late at night.

Biology of the Bee Genus Agapostemon 703

My impression is that the behavior of the female is controlled by some
physiological factors (probably closely correlated with the maturation of
oocytes) which are minimal following oviposition, cell closure and filling
of the lateral. After she has filled the lateral, the actions of the female are
unpredictable. She may peer out of the nest entrance for a while, work
"half-heartedly" at deepening the burrow or simply rest in the burrow. As
the hours pass she becomes more active, her activities seem increasingly
"directed" toward completion of a new cell and she seems less susceptible
to disturbances or distractions.

Pre-Imaginal Development

The following is a summary of insectary observations on 38 cells of A.
texanus during May and June. All but three of the 21 bees reaching ma-
turity were males. Excessive watering of the nest-box was followed by
fungal growth in 17 of the 38 cells. The temperature of the insectary fluc-
tuated between 78-86° F (25.6-30.0° C). As this temperature is probably
at least 20° F higher than the outside soil temperature, it is reasonable to
assume that the development time of bees in the field is somewhat longer
than that observed in the insectary.

It is very difficult to distinguish between an egg and a newly hatched
larva, but it appears that the egg hatches about two days after oviposition.
The larva eats constantly, except when moulting, and growth is very rapid.
Within 4 to 5 days after hatching, the larva has consumed the entire pollen
ball. Some of the large larvae were able to balance the partially eaten pollen
mass on their abdominal venters (Fig. 10) until it was completely consumed.
For about two days after the last pollen has been eaten, the mature larva
turns around and around in its cell while making continuous feeding
motions with the mandibles. Then it defecates on the floor and walls of
the cell (Fig. 11), and lies on its back for 5-6 days as a prepupa (Fig. 12)
before becoming a pupa (Fig. 13). The eyes turn pink on about the third
day after pupation and turn dark brown after another three days. Next the
body begins to darken in the head and mesosomal regions. The metasoma
and appendages finally darken and, after approximately 16 days as a pupa,
the bee moults for the last time. The adult remains in the cell for one or
two days, presumably while the cuticle tans. Then it begins to dig its way
out through the lateral and emerges from the nest one half to one day later.
The elapsed time from oviposition to emergence from the nest was 32± 4
days for the males and about one day longer for the females, but this differ-
ence is not statistically significant.

Adult Longevity
In order to gauge the longevity of Agapostemon in the insectary, it was

704 The University Science Bulletin

necessary to mark each bee in a distinctive manner. The most successful
technique was to put small spots of various colors of quick drying model
airplane paint on the mesonotum and metasomal terga. Mortality (about
30%) due to handling and painting was the most immediate disadvantage,
but a second problem was the loss of the spots. On some individuals the
paint spots remained for more than two months, on others they came off
within 48 hours.

In the insectary the cause of death was often spiders concealed in
the flowers, but frequently bees were accidentally crushed under foot or
drowned in the honey on artificial flowers. The following represent the
maximum number of days from marking to death (the total adult life span
is undoubtedly somewhat longer): A. texanus £ -22" days, 9 -67+ days;
A. splendens 6 -22+ days, 9 -58+ days; and A. radiatus 9 -63+ days. The
data listed above are based on observations of 70 marked bees (A. texanus 53;
A. radiatus 12; A. splendens 5). Of course these data are for active bees, as
overwintering females must live as long as 6 to 7 months.

Mating

Successful copulation was never observed in the insectary or in the field.
However, attempts at copulation were seen repeatedly in the insectary and
in the field. Once the males of Agapostemon leave their nests, they never
return. Sleeping, eating, searching for mates and mating all take place on
or about flowers. The mate-searching flight of the males is a very rapid
oscillating flight which proceeds quickly from flower to flower. When the
bee is flying in this manner it is difficult to follow with one's eye. The
function of the oscillations can be only surmised, but it could serve to
confuse predators which hunt by sight such as birds, robber flies and
philanthine wasps, or it could serve to increase the depth perception of the
searching male. The male approaches to 2-6 cm of the flowers but does not
land except occasionally to take nectar or to rest. When the male locates
a female on a flower, it ceases its oscillating flight and hovers about 5 cm
from her. The male then situates itself so as to face the female from slightly
above and either directly behind or directly in front, still at a distance of
about 5 cm. After hovering for one half to one second in this position, the
male darts at the female and grasps her with his legs. In each of my many
observations of this action, the female successfully dislodged the male with
her wings and occasionally her legs. Contact was never maintained for
more than one second. The rejected male often repeats the hover-dart-grasp
sequence two or three times before resuming its mate-searching flight.
Apparently the males are poor at discriminating between the anterior and
posterior aspects of the females, as they would approach from the front for
almost half the attempts. It is obviously inefficient to be incorrectly oriented

Biology of the Bee Genus Agapostemon 705

in 50% of the mating attempts, but consecutive attempts to mate with an
unresponsive female would be advantageous if the previous rejection were a
result of incorrect orientation of the male. All three of the species studied
have females which are wholly green from above. However, in many of
the species the abdomen of the female is conspicuously different in colora-
tion from the head and thorax. It is possible that the differently colored
abdomen in these species serves as a visual cue to the proper orientation
of males. However, this may be offset by the similar appearance of males
and females of these species.

Bohart (1950) describes mating in Agapostemon femoratus (given as A.
cockerelli). Although the female of this species is wholly green when seen
from above, the males approached the females from behind. Out of 32
attempts to mate with females of its own species, the males of A. femoratus
were successful in 3 instances. Copulation lasted about 10 seconds before
the male, which had clasped the female from above and slightly behind, was
dislodged by the female. It seems probable that a female Agapostemon
mates soon after emerging and never mates again. This could explain why
mating of Agapostemon is so rarely observed.

Although males and females of Agapostemon frequently were seen to
forage simultaneously on the same flower (Figs. 16-17) they never attempted
copulation, and the usual interaction (if any) was for the female to drive
off the male by lunging and biting as she would at another female.

The males of Agapostemon in the insectary will attempt to mate with
nearly any insect on a flower. Bohart (1950) reported almost 30% of the
mating attempts of A. cockerel// males were with females of Halictus ligatus,
H. jannosus, H. rubicundus and Lasioglossum sp. In the flight chamber,
males could even be induced to pounce on crude wood or clay models of
females.

Sleeping

Agapostemon are quite flexible in their sleeping habits. A. angelicas has
been observed (Linsley, 1962) sleeping on the dried flower heads of a clump
of Heterotheca subaxillaris (Compositae) in southeastern Arizona. Numer-
ous males and occasional females were seen on these plants on 26 of 28
consecutive nights. According to Linsley, the females always slept alone but
the males tolerated the presence of other male A. angelic its and as many as
six males could be seen crowded onto one flower head. These males also
tolerated the presence of wasps of the genus Stenodyncrus (Hymenoptera;
Vespidae). Grasping the plant with their legs, the male bees slept with
their bodies extended, wings folded, and antennae held forward and pressed
together or only slightly divergent.

On the morning of August 21, 1966, I observed males of A. angelicas
sleeping on sunflowers (Helianthus sp.) in a pasture in eastern Colorado.

706 The University Science Bulletin

The bees slept on the tops of the flowers and in the morning both the bees
and the flowers were covered with a heavy dew. As the morning was cold,
the bees did not dry out and fly away until about 9:00 a.m. The sleeping
posture was similar to that described by Linsley (1962).

A male of Agapostemon jemoratus was observed by Hicks (1936) enter-
ing a hole in a sand bank at 5:00 p.m. When the tunnel was excavated, a
female as well as the male was found in the hole. As no cells were found,
the burrow was probably not being used as a nest.

On June 7, 1916, in Manati, Puerto Rico, G. N. Wolcott (1948) observed,
"A cluster of twenty or thirty of these bees | males of A. viequesensis
given as A. portoricensis] was noted on a few grapefruit leaves." Although
he makes no mention of the time of day, this may have been a sleeping
aggregation.

Insectary observations showed that males of A. texanus, A. radiatus and
A. splendens sleep most frequently on flowers. They are usually found on
top of the flowers, although they can often be found on the underside. They
occasionally sleep in abandoned burrows, but are chased out of any burrow
containing a female. A block of wood about 20 cm long, 5 cm wide and
1.5 cm thick was hung from the ceiling of the insectary about 1.75 m from
the floor. Males frequently slept in the 5 mm holes which had been drilled
through the block. Females normally spent the night in their nests, but
sometimes spent the night beneath a flower head or clinging to the ceiling
of the insectary. Both males and females of the species studied slept with
the body extended and never were seen to hang by the mandibles as do
numerous Hymenoptera, including the Nomada which sometimes shared
the insectary with Agapostemon.

Foraging

In my experience (corroborated by the flower records from pinned
specimens and the literature), the species of Agapostemon are polylectic and
may be found at almost any time of day at almost any type of flower
physically available to them (realizing their morphological limitation).
Of course under a particular set of conditions the bees often prefer one
flower species to another, but as conditions change (e.g., as different flowers
appear) these preferences also change. In the insectary a female sometimes
gathers pollen from more than one flower species during a single foraging
trip. On any one trip a bee more commonly gathers pollen from flowers
similar in color and shape (e.g., Silphium and Grindelia) than from very
dissimilar species (e.g., Silphium and Cirsium). Proximity of the blossoms
is an additional factor, as the bees have a tendency to work with a clump
of flowers which are, more often than not, conspecific. In any event, only

Biology of the Bee Genus Agapostemon 707

an intensive behavioral study would yield useful information on the flower
preferences or constancy of these bees.

A label which simply states, "on Helianthus annuus" is not very in-
formative. If the bee were a male, it might have been taking nectar, looking
for a mate, sleeping, or merely resting. If it were a female, it may have
been taking nectar for its own use or for its offspring, collecting pollen,
mating, resting, or sleeping. Certainly these are widely divergent activities,
if resting and sleeping can be so classified, and should not be lumped under
the headings "Flower Records" or "Flower Preferences." Although there
are reliable reports on the activities of Agapostemon at flowers, they are
written primarily from the viewpoint of floral biology.

Both males and females of Agapostemon can be found obtaining nectar
or pollen at flowers at any time during the daylight hours. In Kansas, the
best time to collect Agapostemon is from about 9:00 to noon, CST. While
the abundance of females on flowers decreases rapidly in the afternoon,
males remain at flowers throughout the day. Insectary observations show
that provisioning females usually seal themselves in their nests in the after-
noon, having finished provisioning a cell in the morning. However, females
not provisioning nests spend most of their time during the day on or about
flowers.

It is obvious that the females arc able to adapt their pollen collecting
behavior to coincide with the availability of pollen, and the males to the
availability of females. Agapostemon angelicas has been reported (Linsley
and Hurd, 1959) gathering pollen from Mentzelia pumila, which flowers
in late afternoon, as late as 5:40 p.m. (sunset). The same species has also
been reported (Linsley, I960) collecting pollen from the matinal flowering
Cucurbita joetidissima at 5:35 a.m. when the air temperature was 52-54° F
(11.1-12.2° C) and a heavy overcast was threatening rain. In the same
paper, Linsley reports A. melliventris collecting pollen from Datura sp. at
4:56 a.m., when it was so dark that a flashlight was used to examine the
flowers.

The versatility of foraging Agapostemon is further demonstrated by A.
texanus, A. radiatus and A. splendens, which I have seen collecting pollen
from the horse nettle, Solatium carolinense, and the buffalo burr, S. rostra-
turn (Solanaceae), as well as the showy partridge pea, Chamaecrista jascicu-
lata (Cassiaceae), in Lawrence, Kansas. On these flowers the female bee
grasps the cluster of tubular anthers with her legs and covers the apex of
the anthers with the ventral surface of the thorax. Her head and abdomen
flexed and wings folded, she grasps one of the pollen tubes with her
mandibles (Fig. 15) and vibrates her thorax in several rapid bursts about
1/2 second in duration so as to emit a series of audible buzzes. The bee
then transfers the pollen from the underside of her thorax to her hind legs.

708 The University Science Bulletin

She flies to another flower and repeats the performance. Michener (1962)
and Wille (1963) have observed this buzzing behavior on tropical species of
Solarium and Cassia. A wide variety of bees, including Agapostemon, were
reported to exhibit this behavior. Michener speculated that grasping the
pollen tubes with the mandibles serves to loosen the pollen within the tubes.
While this may be true, insectary observations lead me to believe that the
bees grasp one of the tubes during the "buzz" primarily to prevent being
dislodged by their own vibration. Despite their efforts to cling to their
precarious perch atop the anthers, the bees often slip and fall from the
flower during "buzzing." No attempts to chew at the anthers were observed
and the bees only grasp with the mandibles at the time of "buzzing." I
concur with Michener in his contention that this "buzzing" behavior is
inherited and not learned because inexperienced bees reared in the insectary
went to flowers of Solatium carolinense, S. rostratum and Chamaecrista
jasaculata and began "buzzing" without hesitation.

On most flowers the bees rake pollen from the anthers with the fore-
legs, the mandibles and possibly the labrum. Pollen accumulates in the
thick brush of hairs on the flattened fore-basitarsus. The foreleg is swung
to the rear and clamped between the base of the femur and the apex of the
tibia of the middle leg. There is a longitudinal row of stout bristles on
the distal half of the ventral surface of the middle tibia and a similar row on
the basal half of the ventral surface of the middle femur. The bristles of
these combs are directed slightly rearward. Pollen is removed from the
basitarsal brush of the foreleg as it is drawn forward through the opposed
combs of the flexed middle leg. The flexed middle leg simultaneously
moves rearward and presses the mass of pollen into the scopa on the hind
femur. This passage of pollen from foreleg to hind leg is so rapid as to be
very difficult to observe. However, the positions of the legs can be seen in
my high speed (1/1000 sec.) color photographs. Pollen is passed by the legs
on one side at a time, presumably because simultaneous transfer by both
right and left legs would leave the bee in the awkward position of having
to stand on only two legs. As pollen accumulates on the hind legs, the bee
pauses to press the hind legs and ventral surface of her abdomen together,
thus packing the pollen. When pollen is abundant, significant amounts are
carried on an abdominal scopa. It is possible that pollen is transferred to
the abdominal scopa during the packing process. The ventral surface of
the hind tibia is concave and glabrous, but with still bristles laterally. When
the hind leg is flexed, the ventral surface of the hind tibia compacts and
helps hold the pollen in the femoral scopa.

Owing to their long flight season and to their relative abundance, the
species of Agapostemon are doubtless among our more important pollinators
of native plants. They can also play an important role in the pollination of

Biology of the Bee Genus Agapostemon 709

crops. Linsley (1946) and Linsley and MacSwain (1947) report that the
individuals of A. melliventris are more effective pollinators of alfalfa than
are individual honeybees. Bohart (1957) shows that females of A. jemoratus
(listed as A. cockerelli) trip 6-10 flowers per minute, those of A. virescens,
S-10 flowers per minute, and males of both species 1-5 flowers per minute.
The investigations of LaBerge, Isakson and Kehr (1965) indicate that
A. texanus is not an effective pollinator of alfalfa in screen cubicles 3 feet on
a side. Although they neglect to mention the sex of the bees used, I assume
they were females. If the bees were not nesting in the cages, it is unlikely
they would collect pollen. Thus the results of this study seem irrelevant
to what one might expect of uncaged bees.

From my own observations and the observations of others, Agapostemon
ranks close behind Apis, Bombus, Peponapis and Xenoglossa in abundance
on the flowers of Cucurbitaceae. In the desert Southwest, Linsley (1960)
reports A. melliventris and A. angelicus as common on the flowers of Cucur-
bita. In Kansas A. splendens, A. texanus and A. radiatus are common on
Cucurbita and, judging by the collections of squash pollinators from the
University of California at Berkeley, A. nasutus is the species of Agaposte-
mon most abundant on tropical squashes. While this information indicates
the importance of Agapostemon as a pollinator, it should not be construed
as host preference.

As Batra (1966) was able to induce Lasioglossitm zephyrum to accept
frozen pollen, attempts were made to induce Agapostemon to do like-
wise. Frozen pollen from Alnus (the same sample used by Batra) and
Mains was offered to the bees on a wide variety of substrates including
various plastic flowers and living flowers. Although efforts were made to
present the pollen as naturally as possible, all was to no avail. The bees
would walk through the pollen to obtain honey or nectar but if any pollen
adhered to them, it was removed by grooming. Although Lasioglossitm
would collect even pollen spilled on the floor, Agapostemon texanus, A.
radiatus and A. splendens ignored frozen pollen. If Agapostemon could
have been induced to accept frozen pollen, it would have been possible to
maintain them in the insectary throughout the year.

In the insectary foraging females (never males) were frequently seen
"evaporating" nectar (Figs. 19-21). A bee engaged in this process rests on
her middle and hind legs with her forelegs drawn up under the thorax. A
droplet of nectar is regurgitated, and about once each second the mandibles
are opened, the labrum and proboscis extended, and then the labrum and
proboscis retracted and the mandibles closed. The droplet of nectar is
extended on the base of the extended proboscis and when the proboscis is
retracted into the proboscidial fossa the nectar clings to the underside of
the genal area. Rhythmic movement of the mouthparts lasts for 5-15

710 The University Science Bulletin

minutes. Undoubtedly some water in the nectar evaporates in the process,
and it seems reasonable to assume that the function of this activity is to
concentrate or otherwise condition the nectar prior to adding it to the
provisions in the cell.

Predators and Parasites*

No attempt was made to compile a list of organisms preying on species
of Agapostemon. To do so would be difficult and the resulting list would
undoubtedly be very incomplete and of dubious significance. It is apparent
from the literature and from my observations that the following families of
arthropods are important enemies of Agapostemon: crab spiders (Araneida;
Thomisidae), ambush bugs (Hemiptera; Phymatidae), philanthine wasps
(Hymenoptera; Sphecidae) and robber flies (Diptera; Asilidae). Owing to
their cryptic coloration, crab spiders (most commonly Alisumenoides formo-
sipes) and ambush bugs were sometimes overlooked on cut flowers sup-
plied to the insectary, and a single spider would often kill several bees
before it was discovered.

The absence of published records of nest parasites of Agapostemon re-
flects the paucity of field data on the nests of these bees. The only positive
evidence of parasitism is the single coarctate larva of a blister beetle (Coleop-
tera; Meloidae) recovered from a cell of A. texanus in the insectary. As the
phoretic larvae of blister beetles were found on numerous pinned specimens
of Agapostemon, it seems reasonable to assume that the bees are commonly
parasitized by these beetles. The phoretic larvae of rhipiphorid beetles, also
found on the pinned bees, may parasitize Agapostemon. Conspicuous by
their absence were the Strepsiptera, common parasites of other halictid and
andrenid bees. I examined about 50,000 specimens of Agapostemon without
seeing a single strepsipteran. Mites are commonly found on both sexes of
Agapostemon but identification was not attempted. Many of the mites are
phoretic hypopi and are often symmetrically arranged in particular positions
on the bees. It is probable that the life cycle of some of the mites is closely
linked with that of the bees. Mites were not observed on the bees or in
their nests in the insectary.

Probable parasites of Agapostemon are bee flies (Diptera; P>ombyliidae),
velvet ants (Hymenoptera; Mutillidae), and parasitic bees in the genera
Sphecodes (Hymenoptera; Halictidae) and Nomad a (Hymenoptera; An-
thophoridae). Linsley, MacSwain and Smith (1954) report a female
Nomada around a burrow of A. texanus and I have observed large numbers
of Agapostemon and Nomada flying together. In late July, 1967, about 50
Nomada sp. were released in the insectary with nesting Agapostemon.
Although these Nomada flew about, took nectar from, and mated on the

* Parasite is used here in a broad sense t<> include int]uilines and parasitoids.

Biology of the Bee Genus Agapostemon 711

flowers, they were never seen to search for or exhibit interest in the Aga-
postemon nests.

DISCUSSION

The behavior of the three species of Agapostemon observed in the in-
sectary does not differ significantly from that described for Lasioglossum
zephyrum (Batra, 1964), Augochlorella spp. (Ordway, 1966), and Augo-
c/ilora pura (Stockhammer, 1966), excepting those activities directly asso-
ciated with differences in nest architecture and with a lack of sociality.
Unlike the smaller Halictinae (including those mentioned above), females of
Agapostemon are not known to lick objects when away from their nests.

As yet there is no evidence of true sociality, with the concomitant castes
and division of labor in the genus Agapostemon and its allies. Some of the
species of Agapostemon (e.g., A. texanus) are known to be solitary while
other species (e.g., A. virescens) probably live in communal nests, in which
two or more egg-laying females occupy a single burrow without cooperating
in construction of laterals or provisioning of cells. Species of Pseudagapo-
stemon (Michener and Lange, 1958a) and Ruizantheda (Halictus mutabilis;
Claude-Joseph, 1926), closely related to Agapostemon, live in communal
nests but do not exhibit sociality. The augochlorine and Halutus-Lasio-
glossitm groups of genera contain both solitary and social species but. in
contrast to Agapostemon and its relatives, are not known to occupy com-
munal nests. As so few species have been investigated, it seems premature
to conclude that sociality does not occur in members of the genus Agapo-
stemon and its allies.

The nests of Agapostemon and its relatives, Pseudagapostemon, Habra-
lictits and Ruizantheda, are easily distinguished from those of most other
Halictinae by their relatively long laterals (1-30 cm). Excluding the mem-
bers of the Agapostemon group of genera, the only species with relatively
long laterals are Lasioglossum leucozonum and several other species of
Lasioglossum* The three nearctic species of Agapostemon whose nesting
biology has been investigated excavate a single cell at the end of each
lateral, but it seems that the neotropical A. nasutus sometimes constructs a
pair of contiguous cells at the end of the laterals. This habit of building
two or more cells end-to-end in the laterals is known in the Halictinae only
for Agapostemon and its allies. Conversely, the Agapostemon group of
genera are not known to build the nests without laterals or with the cell

* Nests of Caenaugochlora curticeps described by Michener and Lange (1958b) also had
relatively long (1-4 cm) laterals. The authors noted that the nests were very different from
those of other Caenaugochlora species and very similar to tho-.e of Pseudagapostemon and
Habralictus. As suggested by Michener and Lange, Caenaugochlora curticeps probably belongs
in another genus. On the basis of its nest architecture, C. curticeps is closely related to
Agapostemon and should not be placed with the augochlorine genera.

712 The University Science Bulletin

clusters common to both the augochlorine and Halictus-Lasioglossum
groups. Agapostemon and its relatives retain the usual apoid habit of
constructing a cell, provisioning it, laying an egg, and closing it before
going on to start the next cell. In the augochlorine and the Halictus-
Lasioglossum groups this sequence often breaks down, several cells being in
various stages of construction and provisioning at the same time even in
nests inhabited by only a single adult female. None of the species in the
Agapostemon group are known to construct a turret at the nest entrance,
unlike many species in the augochlorine and Halictus-Lasioglossum groups.
Many of the Halictinae (e.g., Halictus farinosus) have a strong tendency to
excavate each new cell below the preceding cell, but in the insectary about
40% of the new cells of Agapostemon were excavated above older cells.

APPENDIX

The cells and larvae described as A. splendens by LaBerge and Ribble (1%6) were so
anomalous as to lead me to conclude that they were not made by Agapostemon splendens or
any other member of the Halictinae.

The cells were vertical and spaced at intervals along a single lateral. This type of cell
arrangement is unknown in the Halictinae but is common in the North American Nomia of
the subgenera Epinomia and Dieunomia (Cross and Bohart, I960). Furthermore, the de-
scribed cells are very elongate, unlike the oval cells of Agapostemon constructed in the
insectary.

The authors comment on the larvae found in the vertical cells, stating that: "The mor-
phology of the larvae is generally similar to that of other halictid bees as described by
Michener (1953). However, the strongly bidentate mandibles of splendens seem unique."
In point of fact, strongly bidentate mandibles occur in the larvae of some of the Halictidae
(Michener, 1953) and therein lies the explanation for this apparently anomalous nest.
Although none of the Halictinae have bidentate mandibles in the larvae, those of the sub-
family Nomiinae are characterized by strongly bidentate mandibles.

Only six species of Nomia are known from Nebraska: Nomia nortoni, N. universitatis ,
N. apaeha, N. heteropoda, N. neradensis and N. triangitlijera. The cells were not made by
A', heteropoda or N. neradensis because they were too small for the former anil too large for
die latter. Nomia nortoni makes a cluster of cells rather than a linear series (Ribble, 1965)
and nothing is known of the biology of the uncommon IV. apaeha and A', universitatis. The
species most likely to have constructed the cells is Nomia triangitlijera, a very common
Nebraskan species which nests in sandy areas like that described by LaBerge and Ribble.
The dimensions of the cells of N. triangitlijera given by Cross and Bohart in 196(1 (cell width
8.5 mm, cell length 20-22 mm, and cell neck 8-10 mm) arc nearly identical with those
described by LaBerge and Ribble (cell width 9 mm, cell length 20 mm, and cell neck 10 mm).

Seemingly contradictory is the fact that Nomia triangitlijera docs not begin to nest until
the second week of August, whereas LaBerge and Ribble excavated the nests between the
ninth anil fifteenth of June. The "larva" figured by LaBerge and Ribble is "7" shaped and
has very prominent tubercles. Both the angular appearance and the tubercles are characteristic
oi prepupae, feeding larvae being "C" shaped and only slightly tuberculate. The "pollen" in
the cells presumably must have been the undigested pollen shells voided by the defecating
larva. Feces of N. triangitlijera are deposited on the bottom of the cell and could be contused
with pollen. The last cell in the series of four was open and contained pollen but no larva.
It is possible that the female was in the process of provisioning this cell when she died. That
she expired prematurely is also indicated by the small number (4) of cells — N. triangitlijera
normally making 6-13 cells along a lateral.

Biology of the Bee Genus Agapostemon 713

The fact remains that A. splendens females were found in some of the burrows. As it was
early in the season, it is possible that many of these females were still excavating their vertical
shafts but had not yet begun to excavate laterals and cells. This interpretation is supported
by the fact that LaBerge and Ribble reported finding only six cells, although 22 nests were
excavated. Two of the six cells were not visibly connected with a vertical shaft but were
assumed to have been constructed by the same bee which had constructed the main burrow
some 3 cm distant. As these cells contained the only "mature larvae" recovered, one must
conclude that the "mature larva" of their figures 3-8 was not found in the nest illustrated
in Figure 1.

Prepupae of N. triangulifera were compared with the immatures of "A. splendens" col-
lected by LaBerge and Ribble (kindly lent by LaBerge) and were found to be indistinguishable.

Insectary observations of Agapostemon show that the females and sometimes the males
will take up residence in any unoccupied hole or burrow. This may explain their presence in
what seem to be burrows of Nomia triangulifera.

LITERATURE CITED

Hatha, S. W. T. 1964. Behavior of the social bee, Lasioglossum zephyrum, within the nest

(Hymenoptera: Halictidae). Insectes Sociaux 11:159-186.
. 1966. The life cycle and behavior of the primitively social bee. Lasioglossum zephyrum

(Halictidae). Univ. Kansas Sci. Bull. 46:359-423.
Bohart, G. E. 1950. Observations on the mating habits of halictid bees. I'.tn-Paeific Ent.

26:34-35.

. 1957. Pollination of alfalfa and rvA clover. Ann. Rev. Ent. 2:355-380.

Claude-Joseph, F. 1926. Recherches biologiques sur les Hymenopteres du Chili (Melliferes) .

Ann. Sci. Nat. Paris. 9:113-268.
Cross, E. A., and G. E. Bohart. 1960. The biology of Nomia (Epinomia) triangulifera with

comparative notes on other species of Nomia. Univ. Kansas Sci. Bull. 41:761-792.
Eickwort, G. C. 1969. A comparative morphological study and generic revision of the

Augochlorine bees (Hymenoptera: Halictidae). Univ. K.in-as Sci. Bull. 4^:325-524.
Felt, E. P. 1928. Observations and notes on injurious and other insects of New York state.

New York State Mus. Bull. 274:145-176.
Hicks, C. H. 1936. Nesting habits of certain western bees. Canadian Entomol. 68:47-52.
LaBerge, W. E., O. W. Isakson and \V. R. Kehr. 1965. Native insects as pollinators of

caged alfalfa clones and seedling performance of the progeny. Jour. Econ. Ent.

58:63-66.
LaBerge, W. E., and I). W. Ribble. 1966. The nests and larvae of two species of Agaposte-
mon (Hymenoptera: Halictidae). Jour. Kansas Ent. Soc. 3l»:4(>7-472.
Linsley, E. G. 1946. Insect pollinators of alfalfa in California. Jour. Econ. Ent. 39:18-29.
. I960. Observations on some matinal bees at flowers of Curcurbita, Impomoea and

Datma in desert areas of New Mexico and southeastern Arizona. Jour. New York

Ent. Soc. 68:13-20.
. 1962. Sleeping aggregations of aculeate Hymenoptera-II. Ann. Ent. Soc. Amer.

55:148-164.
Linsley, E. G., and P. D. Hurd. 1959. Etiological observations on some bees of southeastern

Arizona and New Mexico (Hymenoptera: Apoidea). Entomol. News 70:63-68.
Linsley, E. G., and J. W. MacSwain. 1947. Factors influencing the effectiveness of insect

pollinators of alfalfa in California. Jour. Econ. Ent. 40:349-357.
Linsley, E. G., J. W. MacSwain and R. F. Smith. 1954. A note on the nesting habits of

Exomalopsis solani Cockered (Hymenoptera, Anthophoridae). Pan-Pacific Ent.

30:263-264.

714

The University Science Bulletin

Michener, C. D. 1953. Comparative morphological and systematic studies of bee larvae with
a key to the families of hymenopterous larvae. Univ. Kansas Sci. Bull. 35:987-1102.

. 1962. An interesting method of pollen collecting from flowers with tubular anthers.

Rev. Biol. Trop. [San Jose, Costa Rica] 10:167-175.

Michener, C. I)., and R. B. Lange. 1958a. Observations on the behavior of Brazilian halictid
bees (Hymenoptcra, Apoidea) I. Ann. Ent. Soc. Amer. 51:155-164.

. 1958b. Observations on the behavior of Brazilian halictid bees III. Univ. Kansas Sci.

Bull. 39:473-505.

Ordway, E. 1966. The bionomics of Augochlorella striata and A. persimilis in eastern Kan-
sas. Jour. Kansas Ent. Soc. 39:270-313.

Rau, P. 1934. Notes on the behavior of certain solitary and social bees. Trans. Acad. Sci.
St. Louis 28:219-224.

Ribblh, I). W. 1965. A revision of the banded subgenera of Nomia in America (Hymenop-
tcra: Halictidae). Univ. Kansas Sci. Bull. 45:277-359.

Sakagami, S., and C. D. Michener. 1962. The Nest Architecture of the Sweat Bees. Univ.
Kansas Press, Lawrence, 135 pp.

Stevens, O. A. 1921. Further notes on evening flowers, panurgine and halictine bees. Cana-
dian Entomol. 53:65-69.

Stockhammer, K. A. 1966. Nesting habits and life cycle of a sweat bee, Augochlora pura
(Hymenoptera: Halictidae). Jour. Kansas Ent. Soc. 39:157-192.

Wille, A. 1963. Behavioral adaptations of bees for pollen collecting from Cassia flowers.
Rev. Biol. Trop. [San Jose, Costa Rica | 11:205-210.

Wolcott, G. N. 1948. Insects of Puerto Rico: Apoidea. Jour. Agric. Univ. Puerto Rico
32:865.

Fig. 1. Insectary room showing arrangement of plexiglass nest-boxes (covered with black
oilcloth) and masses ol fresh flowers necessary to maintain Agapostemon. Top of nests are
against wall (upper nest) and horizontal shelf (lower nest) in order to take advantage of the
tendency of Agapostemon females to search for nests along vertical and horizontal surfaces.

Biology of the Bee Genus Agapostemon

715

Fig. 2. Plexiglass observation nest-box with cells and hibernacula oi Agapostemon texantis.
Fig. 3. Plate glass observation nest-box (1 meter square) with black oilcloth. Fig. 4. ./.
texantis female "guarding" burrow in observation nest-box. Fig. 5. Cell of A. splendens with
liquid lining soaking into surrounding sand. Fig. 6. Cells of A. texanus in loam showing
youngest cell (middle) with pollen ball, older cell with mature larva (top) and oldest cell
with defecating larva (bottom).

716

The University Science Bulletin

Fig. 7. View of inner wall of cell fragment of Agapostemon splendens showing how
liquid cell lining (white) is used to coat and cement sand grains (cell is not lined with sand
or loam taken from burrow walls).

Biology of the Bee Genus Agapostemon

717

*mmmirmm$

Figs. 8-13. Cells of Agapostemon in observation nest-boxes (background in cells blackened
to increase contrast). Fig. 8. Loose pollen in partially provisioned cell of Agapostemon texanus.
Fig. 9. Pollen ball with egg in cell of Agapostemon splendens with arrow indicating piece of
cellophane-like cell lining detached from glass of observation nest-box. Fig. 10. Feeding larva
of A. texanus balancing pollen ball on abdomen. Fig. 11. Defecating larva of A. texanus with
arrow indicating feces smeared on cell wall. Fig. 12. Prepupa of A. texanus with charac-
teristic mesosomal tubercles. Fig. 13. Pupa of A. texanus with arrows indicating moldy feces.

718

The University Science Bulletin

•mgmmmmgmg/m^

Fig. 14. Two diapausing female Agapostemon splendens sharing hibernaculum in observa-
tion nest-box. Note abnormal anastomosis of two burrows. Fig. 15. Female A. splendens
"buzzing" to obtain pollen from tubular anthers of Solatium rostratum. Wings remain folded,
mandibles and legs grasp anthers, and body is flexed while vibrating mesosoma causes pollen
to spray from apical pores of anthers and adhere to underside of mesosoma. Fig. 16. Male
and female of A. texantis foraging together on Silphium flower. Fig. 17. Two female A.
texanus foraging together on Silphium flower.

Biology of the Bee Genus Agapostemon

719

Figs. 18-21. Agapostemon texanus females. Fig. IS. Imbibing honey-water on plastic
flower. Figs. 19-21. Extending mouthparts (extend and retract once a second) to evaporate
water from drop of nectar suspended by surface tension between genal region, basal half of
proboscis, and underside of labrum. All four bees are in characteristic position with forelegs
drawn together under thorax.

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLVIII Pages 721-765 November 14, 1969 No. 17

The Skeleto-Muscular System of Mecoptera: The Head1,

H. Randolph Hepburn3

ABSTRACT

This is a morphological study of the skeleto-muscular system of the head of
adult Mecoptera and is hased on 18 genera representing 9 families. The suhgena
has become ventrally displaced along the lateral margin of the clypeus, with the
formation of a clypeogenal suture. Most of the cranial sutures are suppressed,
resulting in poorly defined cranial regions. The occipital and postoccipital
regions have moved dorsally and cephalad. The maxillae are typical of chewing
insects. The maxillary musculature is also of the general type but is modified in
two families through a secondary shifting of the tentoriocardinal and tentorio-
stipital muscles. The tergolacinial muscles are bifurcate in origin but insert on a
common apodeme. The stipitopalpal muscles are unique to Mecoptera among
the orders of the Panorpoid Complex. There is an overall reduction in the
number of maxillary muscles. The mandibles are variable in shape but constant
in articulation. The articulations include an intimate connection between the
hypocondyle and subgenal wall and between the epicondyle and the Apodemal-
walze, which is unique to Mecoptera. The number of mandibular muscles is
reduced. There is no gula in Mecoptera but there is a well developed postgenal
bridge which results in a reduced labial base. The retention of a membranous
postmentum and a sclerotized mental plate is regarded as secondarily derived.
There are no ligular elements in Mecoptera. The labial musculature is extremely
variable but the trend is toward reduction. The hypopharynx is closely asso-
ciated with the labium and the pharyngeal trough and is not musculated. The
antennae are of the flagellar type and are only unusual in the possession of a
rotator muscle of the scape in some families and an extensive antacorium. The
tentorium is composed of five elements, of which only the corporotentorium is
variable, and in some families it completely divides the occipital foramen into

'Contribution number 1419, Dept. of Entomology, University of Kansas.

2 A portion of this work was supported by NSF Grant GB-7045X to Dr. George W. Byers,
Dept. of Entomology, The University of Kansas.

Address as of January, 1970: Dept. of Zoology, The University of Bristol, Bristol, England
BS-8.

722 The University Science Bulletin

two foramina. There is no mesal fusion of the tentorial arms. Compared to
generalized insects, the modifications of the mecopteran head involve a reduction
of the gnathal elements into a new kind of sucking organ; otherwise, many
primitive features are retained. All of the structures present can be homologized
with similar structures in other orders of insects.

INTRODUCTION

Very little has been done on the anatomy of Mecoptera probably because
there was not available material of all families of Mecoptera. What was
available was examined in detail, particularly by German morphologists
during the 1930's. Heddergott, Bierbrodt, Hasken, and Grell studied Panor-
pidae extensively. The Japanese workers, Miyake and Issiki, contributed
large works on Panorpidae of their region. As a consequence of these
studies, a great body of anatomical information was developed on Panorpi-
dae. Unfortunately, as will be repeatedly shown in this paper, the Panor-
pidae are exceptionally specialized and do not represent the order Mecoptera
as a whole.

At its inception, the principal aim of this study was to gather evidence
with which to reconsider the position of Mecoptera within the Panorpoid
Complex. In the course of the study it became apparent that there were
basic problems in the anatomy of Mecoptera that had to be solved before
the Mecoptera as an order could be compared with the other orders in the
Panorpoid Complex. The point of emphasis shifted to a comparative in-
vestigation of the order as a whole, with the hope that there was a basic
pattern within the skeleto-muscular system of the order. Presumably, such
a pattern could then be compared with those of other orders at a later date.
It also seemed desirable to have a basic knowledge of the order as a cohesive
unit, regardless of the Panorpoid Complex arguments.

Before undertaking a comparative study, it was necessary to determine
whether there is in fact a "genus blueprint" and a "family blueprint" — that
is, whether all of the members of a given genus are anatomically equivalent
in terms of their skeleto-muscular system and then whether all of the genera
of a given family are equivalent in the same way. This task began with the
Bittacidae, which contains the largest number of genera. The findings of
this study demonstrated that a given species would adequately serve as a
representative of the family as a whole and not simply as a representative of
itself. The remaining families are small, being usually monogeneric or
digeneric. The assumption that any one species would serve as a family
representative was carried over from the bittacid study. Family recognition
was based on a recently proposed classification by Byers (1%5) which splits
the order into the largest number of families (nine) to date. At the present
this classification is acceptable and is based on a variety of evidence includ-

The Skeleto-Muscular System of Mecoptera: The Head

723

ing morphological, ethological as well as ecological (most of which is still
unpublished). Species on which this study is based are as follows:

Panorpidae

Panorpa ntiptiahs
Panorpa communis

(from the literature)
Neopanorpa harmandi

Bittacidae

Bittactts chlorostigma
Bittactts pilicornis

(from the literature)
Apterobittacus apterus
Harpohittaciis australis
Harpohittacits till yard i
Pazius obtusus
Anabhtacus iridipennis
Kalo hit t act is m it rot cms
Austro bittactts anomalus
Ncobittactis blancheti

Apteropanorpidae

Apteropanorpa tasmanica
Choristidae

Chorista australis
Meropeidae

Meropc tuber
Boreidae

Boretis unicolor

Borcus californicus
Nannochoristidae

Nannochorista dipteroidi s
Notiothaumidae

Notiot/iauma recdi
Panorpodidae

Panorpodes paradoxa
(from the literature)

Brachypanorpa < aro/incnsis

The investigation treats the order Mecoptera as a whole in terms of its
skeleto-muscular system. In the course of interpreting structure, certain
problems arose, most of which stem from older problems in general insect
morphology and are not specific to Mecoptera. For example, the details of
the preoral cavity of insects have never been adequately treated morpho-
logically, and so it is not surprising that taxonomists have retained mor-
phologically unsatisfactory terminology in this regard. In other cases it
seemed that a functional point should take precedence over the homological,
at least for the present, and in those instances references are included to
sources which consider arguments in the classical manner.

An obvious characteristic of the head of Mecoptera is the modified feed-
ing apparatus or proboscis, which is customarily termed the rostrum. Un-
fortunately, not enough is known about the feeding habits of the order to
discuss feeding in any detail, but the rostrum is considered in a different
way. One objective has been to consider the rostrum as a modification of a
generalized insect head, and I looked for the changes that would have been
necessary to convert a generalized head into a mecopteran head in the most
efficacious manner. There is no intention of proposing a phytogeny here
but merely an argument as to how a generalized head (real or imagined)
could be so rearranged.

The problem of terminology usually involves a choice between traditional
names (even though perhaps morphologically incorrect) and names based
on function, which are often just as subjective. Two courses have been fol-
lowed here: the names of the skeletal elements are based on the designations
of Snodgrass (1935, etc.) while those of musculature are based primarily on

724 The University Science Bulletin

the designations of Matsuda (1965), the most recent comprehensive and
comparative treatment of head musculature in insects.

The list of structures illustrated may appear to favor one taxonomic
group or another. They were chosen most often with an eye to making a
point; other times, certain groups were illustrated simply because they were
not previously illustrated and they demonstrate variations in certain features.

ACKNOWLEDGMENTS

This study could not have been done at all, save for the kindness and interest
shown me by Dr. George W. Byers of The University of Kansas. Along with
Dr. C. D. Michener and Dr. R. W. Lichtwardt, also of The University of Kansas,
he critically reviewed the manuscript. In addition, he materially supported this
study by way of NSF Grant GB-7045X. I also express thanks to Dr. C. W.
O'Brien and Dr. L. B. O'Brien who collected a great deal of material for me.

METHODS AND MATERIALS

Most specimens used had been preserved in a variety of the common fixatives such as
70% ethanol, Bouin's solution, Kahle's solution, or Dietrich's solution, depending on the
source of the material. In some instances pinned museum specimens were used for examina-
tion of external features.

Fluid-preserved specimens were dissected in 70% ethanol, with razor blade fragments and
jewelers' forceps, under a standard dissecting microscope. Occasionally the tissues required
histological staining to accentuate certain features of skeleton and musculature. After initial
dissection, specimens were placed in a 0.5% solution of Congo red and 70% ethanol for
30 minutes, washed in 70% ethanol and then examined. The residual effect is that the stain
partially washes out of the muscles but is retained by the endoskeleton. This greatly facilitated
locating origins and insertions of muscles. A solution of 0.5% methylene blue in 70% ethanol
was also used for high-lighting muscles, but this stain has no residual effect.

The highest magnification of the dissecting microscope (90 x) was occasionally insufficient
tor the positive identification of certain muscles; for example, the rotator of the scape is quite
small (length about 300 fi) and is surrounded by deposits of fatty tissue. Examination of
muscles of this magnitude required that sample tissues be temporarily slide-mounted for
examination under a compound microscope, so that the muscle fibers could be clearly
identified.

The examination of the musculature of most of the gnathal appendages requireil careful
dissection under a dissecting microscope, followed by temporary mounts and examination
under a compound microscope. In these cases the specimens were stained previously with
Congo red for ease in muscle identification.

The external skeletal morphology was examined in several ways. In some instances pinned
specimens were sufficient (and in some cases were the only ones available) to yield gross
information. Fine details of external structure and the details of internal skeletal structure-
were usually elucidated by boiling the specimen in a 10% KOH solution anil washing in 70%
ethanol. After such treatment the specimens are often too hyaline to discern fine elements,
and the specimen must then be stained. Again, Congo red proved useful. If a specimen is
first stained in Congo red, washed, and then boiled in a 10% KOH solution, the soft tissues
are rendered while the cuticle retains the stain perfectly. In those specimens which were not
boiled but simply stained tor muscle work, the residual stain was particularly concentrated
along apodemes and apophyses, presumably owing to chemical differences in these areas.
No attempt was made to identify the cuticular layers, but this reaction is useful in following
apodemes.

The Skeleto-Muscular System of Mecoptera: The Head

725

Where possible, drawings were made after the examination of several individuals to avoid
being misled by teratological specimens. The illustrations were made by use of an ocular
grid and grid paper. The techniques used in such a study are not easily described in formula
fashion since each group has its own tricks which the student of morphology must stumble
through before mastering; for example, the amount of pressure that one can apply to a
muscle band before it separates from its origin or insertion, or the care needed in dissecting
specimens that have been preserved for 20 or 30 years.

Perhaps the greatest difficulty was obtaining material. This has undoubtedly been the
reason why earlier morphologists did not treat Mecoptera in more detail. The difficulty in
getting representatives of each family occasionally led to only a single specimen being used;
for example, all of the information presented here on Notiothaumidae is based on a single
preserved specimen.

Perhaps it should be noted that, in the ensuing account, each feature of Mecoptera is
compared with insects in general. While this has led to the inclusion of information available
in sources such as Snodgrass (1935, etc.), it seemed justifiable to include this information so
that a clearer picture of Mecoptera could emerge and that the Mecoptera as a whole could
be seen in the context of general insect morphology.

LIST OF ABBREVIATIONS

abt

abductor tendon of mandible

hyp

act

antacorium

i

adt

adductor tendon of mandible

i

af

alaforamen

k

aj

antennarium

1

an

antenna

lc

anf

antennifer

let

ans

antennal suture

1cm

apm

anterior dilator of the |

iharynx

1.

aw

Apodemalwalze

IP

c

clypeus

Ipg

ccm

clypeocardinal muscle

Ipm

cdm

clypeoapodemal adductoi

Ism

Cgs

clypeogenal suture

lr

CO

compound eye

mbm

cos

circumocular suture

mdm

cr

cardo

mlc

csm

clypeostipital muscle

mn

ct

corporotentorium

mp

dfm

distal flexor of the palp

mx

dpm

depressor of the pedicel

!ll\t

dsm

depressor of the scape

mxp

eci

epicondyle

nf

eel

ecdysial cleavage line

oc

eph

cpipharynx

occ

epm

extensor of the labial pa

IP

oi

es

epistomal suture

pe

f

frons

Pg

fl

flagellum

Pgb

fo

foramen magnum

pha

fpm

flexor of the palp

pht

g

gena

pk

gl

galea

Pi

he

hypocondyle

po

hy

hypostoma

pos

hvb

hypostomal bridge

ppm

hypopharynx

lacinial bridge

lateral process of anterior arm

stipital process

labium

[acinia

lacinial tendon

labroepi pharyngeal compressors

lateral ocellus

labial palp

labial palpiger

levator of the pedicel

levator of the scape

labrum

tergal abductor of mandible

tergal adductor ot mandible

maxillo'labial complex (zygostipes)

mandible

mental plate

maxilla

maxillary palpifer

maxillary palp

neuraforamen

ocular sclerite

occipital condyle

median ocellus

pedicel

postgena

genaponta

pharyngeal aperture

pharyngeal trough

postmentum

pleurostoma

postocciput

postoccipital suture

posterior dilator of the pharynx

726

The University Science Bulletin

q

anterior tentorial arm

St

stipes

qt

anterior tentorial pits

tcm

tentoriocardinal muscle

r

rostrum

tlm

tergolacinial muscle

ra

retractor apotleme

tna

tentorium

rk

prementum

ts

temporal suture

mm

premental retractor muscle

tsm

tentoriostipital muscle

rsm

rotator of the scape

ua

galeal apodeme

sc

scape

V

dorsal tentorial arm

sg

subgena

vx

vertex

sgs

subrenal suture

w

posterior tentorial arm

spin

stipitopalpal muscle

y

lacinial apodeme

ss

subantennal suture

CRANIAL SUTURES— Introduction

There has been a great deal of controversy in the literature over the true
meaning of the term "suture" and the criteria for recognizing one. The
difficulties attending this problem stem mainly from differences of inter-
pretation of various stages during ontogenetic development. Most recently,
Snodgrass (I960) distinguished three basically different kinds of sutures as
distinct from suture-like lines termed sulci. As a purist, he restricted the
term suture in insect morphology to lines of fusion between sclerites, prefer-
ably intersegmental ones. On what appear to be only etymological grounds,
he used the term sulci for all lines of inflection other than intersegmental
ones.

Matsuda (1965) summarized the history of these terms as well as the
contributions of Hinton, DuPorte, and others in resolving these difficulties.
He presented arguments based on comparative anatomy and embryology
advanced by various authors to support various interpretations of these lines
of inflection. Matsuda most notably pointed out that there can be several
different origins of sutures in an unrestricted sense.

There are still differing homological and functional interpretations of
various sutures. If a primarily topographical viewpoint is taken, then
homological significance of these sutures is largely irrelevant, and the same
may be said for the functional interpretation. I have taken a combined
functional and topographical viewpoint and will make no terminological
distinction between sutures and sulci in the sense of Snodgrass. They are
often functionally the same regardless of the ontogenetic and historical
sequences by which they arose.

Ecdysial Cleavage Line

The ecdysial cleavage line (epicranial, frontal, or coronal suture of vari-
ous authors) was shown by DuPorte (1946) not to be a suture of any kind
but only a line of weakness in the larval cranium to facilitate ecdysis, and
having essentially no morphological significance in the adult cranium.

The Skeleto-Muscular System of Mecoptera: The Head 727

DuPorte's interpretations have been subsequently corroborated by Snodgrass
(1947, 1960) and Matsuda (1965). This line lacks the essential element of
any suture, an internally defined ridge or apodeme. There is a tendency for
this line to become suppressed in the adult cranium of pterygote insects
(Snodgrass, 1935, 1947, 1960).

This trend is borne out in adult Mecoptera where the ecdysial cleavage
line is either entirely suppressed or greatly reduced and confined to the
ocellar region (Figs. 21-29). The line is present in the adult cranium of
Panorpidae, Panorpodidae, and Choristidae. Issiki (1933) incorrectly re-
corded it as absent from the Panorpidae. Heddergott (1938) correctly
observed its presence in Panorpidae and noted that it lacks an internal
apodeme. In the Panorpodidae (Fig. 25), the ecdysial cleavage line (eel) is
more extensive than in Panorpidae but it is still greatly reduced from the
larval condition. In the Choristidae (Fig. 29), the line is faint and occurs as
a light furrow in the ocellar region. There is no trace of an ecdysial
cleavage line in the adult cranium of any other families of Mecoptera
although it is present in the larvae and pupae of those families which have
been checked (Applegarth, 1939, the Bittacidae; Steiner, 1930, Yie, 1951,
Bierbrodt, 1942, the Panorpidae and Panorpodidae). Setty (1940) noted the
absence of the ecdysial cleavage line in adult Bittacidae.

In the only comparative study of the head of adult Mecoptera, Otanes
(1922) alleged that the ecdysial cleavage line (which he termed the
epicranial suture) is "undoubtedly the most important landmark on the
head capsule of insects." This error may be excused since it predated
DuPorte's (1946) developmental study of ecdysis and his elucidation of the
significance of this line. Otanes noted the state of this line in the adult
cranium of Meropeidae, Panorpidae, Panorpodidae, Bittacidae, and Boreidae
but did not advance any argument to support his claim for the importance
of the ecdysial cleavage line.

Epistomal Suture

The epistomal or frontoclypeal suture is possibly a primary formation
during ontogeny that indicates the posterior border of the acronal clypeo-
labrum (Matsuda, 1965). If this interpretation is correct, it would qualify
this line of inflection as a suture in the sense of Snodgrass (1960). The
developmental basis for this interpretation has been discussed by Matsuda.
The epistomal suture (es) traverses the lower region of the face between
the anterior tentorial pits (qt) and joins the upper ends of the subgenal
sutures (Figs. 21-29, sgs). Internally, it forms a distinct ridge or apodeme
and is the boundary between the clypeus (c) and the frons (f). The func-
tional significance of this suture in Mecoptera lies in its close association

728 The University Science Bulletin

with the anterior tentorial pits, where it reinforces the cranial capsule.
Whether this is primitively so has been disputed by DuPorte (1946).

Morphologists have generally used three criteria for distinguishing the
frons and clypeus, none of which is constant for the insects (Matsuda, 1965).
The disparity between ontogenetic interpretations and arguments based on
final products is apparent here. As mentioned earlier, an ontogenetic inter-
pretation suggests a clypeo-labral relationship, but in the completed cranium
this is not apparent. Matsuda (1965) discussed the arguments pertinent to
the recognition of the frons and clypeus and accepted Snodgrass' (1947)
interpretation, which is that the frontal ganglion is a convenient landmark
between these two areas. While this definition is consistently operative, it is
questionable since it rests on the grounds that it is extremely unlikely that
the position of the frontal ganglion would have changed during the course
of ontogeny or phylogeny. Since this suture is often suppressed in higher
insects, a distinction between frons and clypeus is often only theoretical
(Matsuda, 1965).

In his studies on Mecoptera, Miyake (1913) distinguished the frons and
clypeus correctly. Otanes (1922) reverted to the position of a single fusion
product, the frontoclypeus, on the grounds that the epistomal suture is
absent in the forms he studied, which is not true. That the epistomal suture
is present in all Mecoptera (at least in apodemal form) and that it is a
reliable indication of the limits of a distinct frons and clypeus was demon-
strated by Hetrick (1935). He showed that the epistomal suture is always
at the level of the frontal ganglion in Mecoptera and that the pharyngeal
dilators and buccal muscles are respectively above and below this line,
which is characteristic of insects in general (Snodgrass, 1935, 1947). The
anterior tentorial arms pass between the ventral buccal muscles (which
generally arise on the clypeus) and the dorsal pharyngeal dilators (which
generally arise on the frons), with the frontal ganglion and the epistomal
suture disposed between them.

The epistomal suture of Mecoptera is variously developed. It is present
both as a suture and an apodeme in Panorpidae, Apteropanorpidae, Nanno-
choristidae, Notiothaumidae, and Choristidae. The suture is barely retained
in Bittacidae and Panorpodidae, but its apodeme is clearly defined. In
Boreidae and Meropeidae the epistomal suture is present only as an apo-
deme, with no indication of an external suture. When externally visible,
the suture varies from almost straight to strongly curved (Figs. 21-29). The
extreme arching is correlated with the mesal migration of the tentorial arms,
without shortening of the suture. In fact, the suture is longer in those forms
in which the tentorial arms have moved closer together than in those in
which it is straight.

The Skeleto-Muscular System of Mecoptera: The Head 729

Heddergott (1938) recognized the correct relationship between frons and
clypeus in Panorpidae; Setty (1940) in Bittacidae; and Evans (1942) in
Nannochoristidae. I have corroborated their findings as well as established
this relationship in the other families.

Subgenal Suture

The subgenal suture is a lateral continuation of the epistomal suture on
both sides of the cranium. It is considered to be of secondary origin, in the
Snodgrass sense, in response to structural need (Matsuda, 1965). Its internal
ridge reinforces the area above the mandibles and sets off a subgenal area
below it. The sclerotized portion of the head capsule directly above the base
of the mandible is termed the pleurostoma, that behind, the hypostoma.
These designations are entirely arbitrary since there is no suture between
them. There is a tendency for this suture to be suppressed in the higher
Pterygota (Matsuda, 1965). The status of the subgenal suture in Mecoptera
is complicated by the ventral elongation of the head, since the normal rela-
tionship between the gena and clypeus has changed by the elongation of
the subgena along the lateral margin of the clypeus (a situation which
appears to be unique to Mecoptera).

The subgenal suture is present both externally as a line and internally as
an apodeme in all Mecoptera except Apteropanorpidae and Notiothaumidae
(Figs. 21-26, 29, sgs). When present, this suture is not easily defined ex-
ternally (but is visible), but its internal apodeme is distinct. The situation
is obscure in only one family, the Panorpodidae, because the rostrum is
shortened and so rugose that it is not possible to distinguish the wrinkles
from the sutures in this area.

Clypeogenal Suture

There is a new suture, formed along the fusion of the clypeus and sub-
gena, which appears to be identical to a similar suture described by DuPorte
and Bigelow (1953) and Bigelow (1954) in Hymenoptera. This is termed
the clypeogenal suture. According to these authors the genal area (of
Hymenoptera) has descended along the lateral margin of the clypeus and
the new clypeogenal suture represents the fusion line. The same suture
occurs in some of the lower Pterygota.

This suture is distinct and should not be confused with the subgenal
suture. Miyake (1913) and Issiki (1933) both realized the distinctness of
these two sutures, discussed and illustrated both, but did not name the
clypeogenal suture. They did correctly observe that the clypeogenal suture
extends latero-ventrad of the anterior tentorial pits and that it terminates
in the area between the two mandibular articulations on either side of the

730 The University Science Bulletin

head. The clypeogenal suture is present in all Mecoptera (Figs. 21-29, cgs).
Some of the more recent workers (Heddergott, 1938; Ferris and Rees, 1939;
Evans, 1942; Setty, 1940) either ignored or misunderstood this suture.
Heddergott, for example, confused the clypeogenal and subgenal suture in
Panorpidae. Setty (1940) ignored both, although they are present in Bitta-
cidae. Evans (1942) described the subgenal suture in Nannochoristidae
but missed the clypeogenal suture.

Frontogenal Suture

Another suture that commonly occurs in insects is the frontogenal suture,
which extends from the anterior mandibular articulation to the ventrolateral
angle of the antennae. The suture is unusual in that there has been a
tendency for a high degree of development in the higher Pterygota (Mat-
suda, 1965). Its morphological significance lies in that it might have been
the primitive site of the anterior tentorial arms (DuPorte, 1946), although
this interpretation has not been universally accepted (Snodgrass, 1960).
Although this suture does not occur in any mecopteran adults, it is present
in other orders of the Panorpoid Complex (Crichton, 1957, in Trichoptera) ;
its absence is a major deviation from the pterygote tendency to preserve and
develop the extent of this suture. I cannot correlate the absence of this suture
with any particular modification of the mecopteran head.

Clypeolabral Suture

In many insects there is a clypeolabral suture separating the clypeus and
labrum. This suture is present in the larvae and pupae of Mecoptera (Bier-
brodt, 1942) but is absent from the adults, as correctly reported by Otanes
(1922), Hetrick (1935), Issiki (1933), Heddergott (1938), and Setty (1940).
I have found it absent also from those families which the cited authors did
not investigate. There are few criteria for the recognition of the labrum
as distinct from the clypeus. These are the presence of the clypeolabral
suture and the labral depressors. Although Mecoptera lack a clypeolabral
suture, the labral depressors are present.

Postoccipital Suture

The postoccipital suture is generally regarded as an intersegmental suture
between the maxillary and labial segments (Matsuda, 1965) and satisfies
Snodgrass' criteria for a true suture. This suture is functionally important
because it gives rise to the posterior tentorial arms and also serves as a site
for muscle attachment. It is present in all Mecoptera (Figs. 30-38, pos).
Posteriorly it delimits the postocciput and is discussed with that feature and
the tentorium. Intimately associated with the postoccipial suture is the

The Skeleto-Muscular System of Mecoptera: The Head 731

occipital suture which primitively separates the maxillary and mandibular
segments. Matsuda (1965) noted that this suture becomes suppressed in
the higher Pterygota and there is no indication of an occipital suture in
Mecoptera. Evans (1942) misinterpreted an unusually well developed post-
occiput as the occiput, in Nannochoristidae, and confused the postoccipital
suture as an occipital suture. That this suture is not the occipital suture in
Nannochoristidae is indicated by the fact that the posterior tentorial arms
arise from it.

ClRCUMOCULAR SUTURE

The periphery of the compound eyes is indicated by the presence of a
circumocular suture which sets off an ocular sclerite surrounding the eye
(Figs. 28, 41). This suture is present in all Mecoptera and is highly devel-
oped both externally as a line and internally by the ridge it forms. The only
unusual modification of this area is that in Boreidae the ocular sclerite is
anteromedially incised.

Temporal Suture

At the dorsolateral angles of the postoccipital suture there is a short
suture (labeled "loc" by Issiki) which extends anteriorly on either side
approximately parallel to the edge of the compound eye (Figs. 39, 41-43).
This unusual suture occurs in Mecoptera (except Nannochoristidae and
Bittacidae) and was first described in the Panorpidae by Issiki (1933). Issiki
homologized it with similar sutures on the posterior part of the cranium
that terminate on the occipital suture (absent from all Mecoptera) and
laterally separate the occiput from the postgena in two primitive groups, the
Micropterygidae (Lepidoptera) and the Rhyacophilidae (Trichoptera).
Issiki argued that the area partially enclosed by these sutures represented
the occiput of other insects, which would normally occur in the same place
on the back of the cranium. This implies that the occiput of Mecoptera has
migrated from a primitively posterior position to a dorsal position. I am in
complete agreement with this interpretation.

Although there is no occipital suture in Mecoptera, the postoccipital
suture has retained its normal position. In addition, in the process of
rostrum formation (as will be discussed later), there has been a complete
antero-ventral reorganization of the mecopteran cranium, and it is reason-
able to expect the occiput (though not clearly defined) to have participated
in this rearrangement. Homologous sutures have been reported in Lepidop-
tera by Ehrlich (1958) and in Mantodea by Levreault (1936).

SUBANTENNAL SUTURE

In Boreidae, Panorpidae, Panorpodidae, and Meropeidae there is a fine
subantennal suture present internally as a ridge extending from the latero-

732 The University Science Bulletin

ventral side of the base of the antenna to the subgenal suture (Figs. 24, 26).
The apodeme formed by this suture would appear to contribute to the
strengthening of the head capsule against vertical stresses on the area be-
tween the antennal bases and the compound eyes, as the antennae are rather
large in Mecoptera and the antennal muscles fairly well developd. This
apodeme should also contribute a modicum of support for the cranium
against the lateral pressures exerted by the contraction of the mandibular
muscles.

CRANIAL REGIONS— Introduction

It is necessary to have a consistent account of the sutures in a cranium
before cranial sclerites or regions can be defined. In defining areas of the
cranium, one should realize what criteria can be used for their recognition
and, secondly, what significance (real or imagined) can be attached to these
areas. A functional interpretation will not necessitate rigid accounting for
missing parts. In a more classical approach, the various cranial areas are
always defined by sutures (Snodgrass, 1935), and when these sutures are
suppressed, as is often the case, the limits of otherwise well defined cranial
areas become obscured and for the most part entirely arbitrary. Unfor-
tunately, too much argument has been devoted to the recognition of arbi-
trary cranial regions in the heads of insects (such as Mecoptera) in which
these sutures have been suppressed. The important functional aspects of
these sutures are their internal apodemes, which provide muscle attachment
and structural reinforcement.

There are a variety of ways in which the insect cranium can be sub-
divided. Since most are arbitrary, I have selected one that is more function-
ally oriented than the others, that of Snodgrass (1935). According to his
system, the functional areas of the insect cranium are the frontoclypeal area,
the lateral parietal regions, the occipital arch, the postoccipital area, and the
subgenal areas basad of the gnathal appendages. Of these regions, only the
occipital arch is absent from Mecoptera.

Frontoclypeal Region

The frontoclypeal region in insects is situated between the frontal sutures
(when present) or the compound eyes or antennae and ventrally bounded
by the base of the labrum (when distinct). If an epistomal suture is present,
the area is further reduced to a dorsal frons and a ventral clypeus (Snodgrass,
1935). There are no frontal sutures in Mecoptera and the antennae have
moved ventrally. In some instances only the apodeme of the epistomal
suture is present in Mecoptera (Figs. 21-29). The frons usually bears the
median ocellus and is the point of origin for the labral muscles (Snodgrass,
1935). In Mecoptera the same relationship exists except for the two families

The Skeleto-Muscular System of Mecoptera: The Head 733

which lack ocelli (Meropeidae and Apteropanorpidae). Because the anten-
nae are anteromedially displaced, Issiki (1933) designated the area above
the antennae as the postfrons and that below, the antefrons. There is no
justification for the retention of these terms since there is no suture to
demark the divisions. For the same reasons it is difficult to retain Issiki's
subdivision of the clypeus into an anteclypeus and a postclypeus (terms
borrowed from Crampton, 1921), since there is no suture separating the
two. In Mecoptera there is no distinct limit of the frons except the epistomal
suture.

The clypeus of Mecoptera is an elongate sclerite which contributes to the
formation of the anterior side of the rostrum. As might be expected, its
relative length varies with the length of the rostrum. The clypeus of all
Mecoptera is constricted at the level of the hypocondylar mandibular articu-
lation. That portion above the mandibles is subquadrate; that below the
mandibles is gradually tapered to a point in those families having long and
intermediate rostra, while in those families having a short rostrum the
clypeus is approximately quadrate (Figs. 21-2^).

There is no clypeolabral suture in Mecoptera but there is an apodeme
which separates these parts, as was initially shown by Miyake (1913), but
with some errors. Miyake thought that the clypeal constriction at the man-
dibular bases represented the junction oi labrum and clypeus in Panorpidae,
but his interpretation of other families was correct. Issiki (1933) questioned
this interpretation and Heddergott (1938) correctly described the situation
in Panorpidae. Heddergott made the same mistake as did Issiki in his
interpretation of an anteclypeus and a postclypeus and in thinking that the
clypeal constriction separated the two. Both Issiki and Heddergott did,
however, point out that the labrum has become laterally displaced by the
ventral extension of the clypeus. The apex of the clypeus is heart-shaped
and supports the membranous labrum. The labrum then lies laterad of
the clypeus and ventrad of the mandibular bases. The clypeolabral apodeme
varies from a well developed ridge to a barely perceptible line. A proximal
or basal limit of the labrum is indicated by the insertion of the labral com-
pressor muscles (discussed later).

Parietal Region

The parietals are lateral parts of the cranium delimited dorsally by the
coronal suture, anteriorly by the frontal suture, and posteriorly by the
occipital suture when these sutures are present. Each of the parietals usually
bears a lateral ocellus, an antenna, and one of the compound eyes. In
generalized insects the dorsal parts of the parietals form the vertex and the
portions posterolateral to the eyes form the genae (Snodgrass, 1935).

734 The University Science Bulletin

The limits of the parietals as described for generalized insects are not
present in Mecoptera. The only indication of a dorsal limit of the parietals
occurs in those Mecoptera which retain temporal sutures. A definitive
vertex and genae do not exist in Mecoptera. There is no occipital suture in
Mecoptera. The coronal portion of the ecdysial cleavage line is a reduced
trace confined to the ocellar region in Choristidae, Panorpidae, and Panor-
podidae.

These areas are arbitrary because they are not defined by sutures, a
condition also found in Hymenoptera and Diptera (Heddergott, 193S). The
vertex might arbitrarily be defined in the absence of sutural limits as that
area mesal and dorsal to the lateral ocelli. The genae can be considered as
approximately ventrolateral to the compound eyes and unseparated from
the vertex (Figs. 39-47). Heddergott (193S) realized the true relationship
between the gena and subgena and that the lateral margins of the rostrum
are formed by the subgenae; however, he misinterpreted the clypeogenal
suture as a continuation of the subgenal suture. He also pointed out that
the postgenae are fused with the occiput and the vertex through the sup-
pression of both the occipital and postgenal sutures.

Posterior Region

In Mecoptera, the head approximates the orcephalic type in which the
orientation is hypognathous and the foramen magnum (or occipital fora-
men) is subdivided by the corporotentorium into a dorsal alaforamen and
a ventral neuraforamen. The foramen is dorsolaterally confined by the
postoccipital suture and ventrally by the fusion of the postgenal processes.
The postocciput bears a laterally projecting occipital condyle (or odon-
toidea) above the level of the corporotentorium which extends a short
distance into the foramen. These condyles articulate with the cervical
sclerites. As usual, the posterior tentorial pits are situated at the ventral
extremity of the postoccipital suture and are closed but visible (Figs. 30-38).

There is no corporotentorium in Apteropanorpidae, Choristidae, Mero-
peidae, and Notiothaumidae, but it is present in the other Mecoptera; hence
the Mecoptera fall into two groups, those with double foramina and those
with a single occipital foramen. I can think of no explanation based on
function why this should be so. The usual explanation of modification for
rostrum formation does not apply because some families with short rostra
and others with long rostra have double foramina while those of inter-
mediate length do not.

Occasionally the postoccipital suture extends below the posterior tentorial
pits enclosing a sclerotized area known as the gula (Snodgrass, 1935). Each
portion of the postoccipital suture ventrad of the foramen is designated as

The Skeleto-Muscular System of Mecoptera: The Head 735

the gular suture and accompanies the prognathous condition of some
Coleoptera and Neuroptera. The presence of a gula in Mecoptera has been
claimed by some mecopterists (i.e., Heddergott, 1938, for the Panorpidae;
Imms, 1944, for the Nannochoristidae) who thought that a gula was
squeezed in between the mesally projecting postgenal processes. The nature
of the insect gula was not satisfactorily known until DuPorte (1962) studied
the problem in detail. He showed that the gula can be either of cervical
origin or in combination with postoccipital and postgenal contributions.

It should be noted that in Panorpodidae, Choristidae, and Notiothaumi-
dae the incomplete mesal fusion of the postgenae leaves a small area, but
not a gula. Evans (1942) recognized this in Nannochoristidae and pointed
out that the area misinterpreted by others (Heddergott in this instance) is
in reality the hypostomal bridge. This is not, however, the case in the
families just mentioned. The sclerotized area enclosed by the incompletely
fused mesal processes of the postgenae appears to be formed by contributions
from the ventral edge of the postoccipital ridge and by the mesal extension
of the posterior tentorial arms. This interpretation is only tentative and
would require developmental studies for its justification.

PREORAL CAVITY— Introduction

There is an area at the apex of the rostrum ventrad of the pharynx which
serves for the uptake of food into the pharyngeal aperture, or mouth. This
space has been termed the buccal cavity or mouth cavity. As Snodgrass
(1960) pointed out, there never has been a satisfactory treatment of this
area by morphologists, and the terminology applied to it is very inconsistent.
Actually, this space is the functional preoral cavity, formed by the enclosure
of the gnathal appendages around it. Snodgrass considered that its upper or
inner wall is the true ventral wall of the head and that the hypopharynx
arises from it. The mouth (aperture of the pharynx) is anterior to the base
of the hypopharynx. The depressed hypopharynx forms the floor of a pocket
which has been erroneously termed the pharynx, but is now known as the
cibarium of chewing insects. It is variously modified and in liquid-feeding
insects is the sucking pump (Mundpumpe of the Germans) as in some
Mecoptera. Because of incorrect terminology through the years (calling
the cibarium the pharynx), the term "epipharynx" has been applied to the
anterior wall of the preoral cavity and the implications of this term are
erroneous. Unfortunately, no alternative is at hand, so the terminology is
retained and the features discussed accordingly.

Otanes (1922) was the first to discuss these relationships in Mecoptera
and used a scheme no worse than any others. He designated the pharyngeal
complex as a dorsal "postpharynx" and a ventral "prepharynx"; in the

736 The University Science Bulletin

latter he included the epipharynx and hypopharynx. Although this is a
misuse of the term "pharynx," Otanes did adequately describe the relation-
ships of the parts.

Epipharynx

As Otanes (1922) pointed out, the size of the epipharynx varies with the
size of the clypeolabrum which in turn varies with the size of the rostrum
(Figs. 1-2, eph). The epipharynx of Mecoptera is a thin, membranous
continuation of the posterior part of the clypeolabrum and bears a pair of
epipharyngeal rods or sclerites on either side (Steiner, 1930; Heddergott,
1938). The larger pair is anterior and supports the apex of the epipharynx
while the other pair is situated in the middle of the plate. At the base of
the labrum there are two sclerotized rods on the posterior side which
various authors have interpreted somewhat differently. Steiner (1930) sug-
gested that these were epipharyngeal tendons, which is curious because there
are no muscles attached to them. Heddergott's (1938) interpretation is that
these rods represent the remnants of reduced epipharyngeal tormae, which
appears more reasonable. The epipharynx has been adequately described
for Meropeidae, Panorpodidae, and Boreidae by Otanes (1922) and for
Panorpidae by Heddergott (1938).

The musculature of the generalized labrum as found in the lower orders
(Matsuda, 1965) consists of a set of extrinsic effectors and some intrinsic
compressors. Only the intrinsic labroepipharyngeal compressors (or di-
lators) are retained by Mecoptera (Fig. 1, lem). These muscles consist of
several discrete bundles oriented somewhat obliquely and which arise on
the highly sclerotized anterior labral plate and insert on the membranous,
nonsclerotized opposing epipharyngeal wall. The secondary displacement
of these muscles might be accounted for on functional grounds. Heddergott
(1938) indicated that this arrangement facilitates sensory contact of the
epipharynx with the food source to which it is appressed during feeding.
The epipharynx is richly invested with a variety of sensoria which pre-
sumably function as organs of taste, an idea first advanced by Packard
(1889) and further supported by Grell (1938) and Heddergott (1938).
There has been no experimental work to test the nature of these receptors.

In the construction of the epipharyngeal complex, Mecoptera are refer-
able to two distinct groups. The Panorpidae and Boreidae follow the pat-
tern discussed above. The other Mecoptera retain only the lateral longi-
tudinal sclerotic rods. These rods gradually become indistinct as they
approach the apex of the rostrum. They are continuous with the pharyngeal
trough at the level of the Apodemalwalze and the mandibular bases. The
musculature of this group is identical with that of the first.

The Skeleto-Muscular System of Mecoptera: The Head 737

mbm

tcm

tsm

1 2

Fig. 1. Notiothauma reedi. Longitudinal view showing major skeletal muscles oi the
head except those of the mandibles.

Fig. 2. Notiothauma reedi. Longitudinal view with all muscles except those of the
mandibles removed.

Hypopharynx

The hypopharynx of generalized hypognathous insects forms the base
of the cibarium or preoral cavity and the opening of the salivary duct is
caudal to its posterior wall (Snodgrass, 1935, I960). Matsuda (1965) indi-
cated that embryologically the hypopharynx is a composite organ and may
be formed by contributions from more than one head segment. The devel-
opment of the hypopharynx in Pterygota is extremely variable and Matsuda
described it as having evolved in two directions, toward reduction and
toward enlargement. This organ is occasionally musculated in some of the

738 The University Science Bulletin

lower orders but there is no hypopharyngeal musculature in Mecoptera.
The ontogenesis of the hypopharynx in Mecoptera is not known.

Issiki (1933) adequately discussed the hypopharynx of Panorpidae and,
if taken somewhat grossly, the account will serve for all Mecoptera. The
hypopharynx is a simple tongue-shaped process, setose, and usually well
developed in Mecoptera. It is situated between the mandibles and maxillae
and hangs downward in the preoral cavity (Fig. 1, hyp). The posterior
surface of its base is continuous with the lateral edges of the labium at the
distal end of the mental plate. The anterior surface is laterally continuous
with the flexible transverse process formed by the medial extension of the
subgenae. The bridge formed by these processes is completely fused with a
ventral apodemal process of the pharyngeal trough (designated the Apo-
demalwalze by Heddergott, etc.) and apparently unique to Mecoptera. On
either side of the hypopharynx there is a thin sclerotized bar or strip that
is confluent with the mesal subgenal bridge. These sclerotized processes
have been subject to various interpretations. Heddergott (193S) suggested
that they represent traces of the superlinguae of other insects and Matsuda
(1965) used the same interpretation. Since these processes are intimately
fused with the hypopharynx laterally and lack any trace of a lobed structure
(as is typical of superlinguae), this interpretation should be accepted as
tenuous at best, pending developmental studies.

The hypopharynx in Mecoptera is somewhat variable in length and
width, but is present in all families. It is least developed in Boreidae and
Nannochoristidae, in which it is reduced to a simple rounded lobe-like
process. In all other Mecoptera it has the general shape already described.
Representatives of several families have been illustrated by Otanes (1922),
Setty (1940), and Issiki (1933). Matsuda (1965) mistakenly reported that
there is no hypopharynx in Boreidae, and that the suspensorial superlinguae
are present in Mecoptera.

TENTORIUM

The insect tentorium has had a long history of study, often yielding
contradictory evidence. The ontogenetic contributions to the tentorium vary
greatly (Snodgrass, I960), and many theories have been proposed to account
for its modifications. The history of these theories has been recently dis-
cussed by Matsuda (1965). Whatever its phylogenetic significance, the
tentorium is an important structure formed by invaginations of the head
capsule. Heddergott (193S) pointed out that there is a separation of the
pterygote tentorium and hypopharynx in which the tentorial apophyses are
volumetrically reduced and the hypopharynx is increased in volume. This
does not occur in Mecoptera, where the tentorium is well developed, and
the hypopharynx is ventrally displaced and greatly reduced.

The Skeleto-Muscular System of Mecoptera: The Head 739

The mecopteran tentorium (Figs. 30-34) is composed of the following
parts: 1) a corporotentorium (ct); 2) two posterior tentorial arms (w) ;
3) two anterior tentorial arms (q); 4) two dorsal tentorial arms (v);
5) two lateral processes of the anterior arms (j). The corporotentorium
(metatentorium of some authors) is situated in the posterior part of the
head. It is variously developed in Mecoptera and may form a complete
bridge across the occipital foramen, dividing it into a dorsal alaforamen
(Fig. 30, af) and a ventral neuraforamen (nf), or it may remain as protrud-
ing but unfused processes extending into the undivided occipital foramen
(fo). When the occipital foramen is divided, the alaforamen accommodates
the alimentary canal and the neuraforamen the nerve cord and salivary
ducts. The ventral margin of the occipital foramen, or of the neuraforamen,
is reinforced by contributions from the genae, across the back of the head,
termed the genaponta. The lateral margins of the foramen are reinforced
by ridges from the posterior tentorial arms, which are buttressed caudally
and dorsally at their bases, an unusual condition in insects. The fusion of
the posterior arms with the postocciput is such that there is no external
opening marking their termination.

The anterior tentorial arms are formed by invaginations in the epistomal
area. Whether they always arise in this way has been subject to dispute
(DuPorte, 1946). Whatever the case, in Mecoptera the anterior tentorial
arms are always associated with the epistomal suture and are externally indi-
cated by the anterior tentorial pits on either side of the epistomal suture.
The anterior tentorial arms extend caudally at an oblique jngle to the
longitudinal axis of the body and terminate in the sides of the postocciput
at the occipital foramen (or neuraforamen). The broad, flattened, fused
cranial plate of some lower insects is not formed in Mecoptera. Instead, the
posterior tentorial arms recess into the postoccipital area and do not project
into the foramen. The anterior tentorial arms remain as paired, but not
fused, thickened hollow rods that cross the head and form a suspensorium
for various organs.

From the middle of the dorsal side of each anterior arm, there is an
antero-dorsal process, the dorsal tentorial arm, which extends to the base of
the antenna. In those families having a rotator of the scape, the dorsal arm
extends up to the base of the antenna, where a portion of the rotator inserts
on its apex. Those families lacking a rotator retain the dorsal arm as a
shortened projection extending toward the antennal base.

Originating on each tentorial arm near the bases of the dorsal arms are a
pair of medial processes which come together but do not fuse. In all
Mecoptera (except Panorpidae and Boreidae) these medial processes bear
the origins of the tentoriocardinal muscles. That this is primitively so and
that the migration of these muscles to a clypeocardinal position (in Panorpi-

740 The University Science Bulletin

dae and Boreidae) is of secondary origin is further supported (aside from
being widespread in other Mecoptera) by the fact that in the lower ptery-
gotes these muscles are always tentoriocardinal (Matsuda, 1965).

The tentorium of Mecoptera plays a major role as the internal skeleton
of the cranium and serves for the origins of the antennal muscles (Figs.
18-20), the tentoriocardinal muscles (Figs. 11-12), and the tentoriostipital
muscles (Figs. 11-12). The medial processes of the anterior tentorial arms
are ventrally curved at their apices and form a trough in which the pharynx
is situated (the "postpharynx" of Otanes). Although the genaponta is not
ontogenetically related to the tentorium, it does serve for the origin of the
premental retractors. The tentorium proper is remarkably constant, and
its variations are limited to thickness, the extent of arching in the center of
the head, the degree of ventral slope in the anterior arms, and the amount
of splay from the occipital foramen.

The tentorium of Mecoptera has been previously studied by others:
Setty (1940) discussed and illustrated it in Bittacidae; Otanes (1922) in
Panorpodidae; Issiki (1933), Heddergott (1938), and Ferris and Rees (1939)
in Panorpidae. In general terms my findings confirm theirs. Ferris and
Rees made a curious error in their study of Panorpidae when they stated
that the only external evidence of the anterior tentorial arms were darkly
pigmented spots. In actual fact, the anterior arms of all Mecoptera retain
their external openings (anterior pits). The anterior pits are generally sub-
circular but in Bittacidae are often oblique slits. In Pazius (Bittacidae) the
pits at first seem absent but they are tucked into recesses beneath the ex-
tremely bulging compound eyes.

MANDIBLES

Although the mandibles (mn) are not involved in the elongation of the
head into a rostrum, they have undergone certain modifications concomi-
tant with this change. They have become displaced toward the apex of
the rostrum and articulate with the head capsule below the level of the
other gnathal appendages (Figs. 52, 39, 47). The mandibles attach to
the head by a dicondylic articulation consisting of an anterior epicondyle
(eci) and a posterior hypocondyle (Fig. 3, he). The point of articulation
varies depending on the ventral extension of the subgena and the length
of the rostrum.

Unlike other insect epicondylar articulations, the epicondyle of Mecop-
tera is retained only as a slight, relatively unsclerotized knob that articulates
with a ventrolateral inflection of the clypeus. The more strongly developed
hypocondyle articulates in a cavity of the subgena (Fig. 3). The mandibles
are decussate in repose, each projecting between the opposite galea and

The Skeleto-Muscular System of Mecoptera: The Head 741

lacinia (Fig. 32). In cross section the mandible is triangular at the base
(its relative thickness can be seen in lateral view) but flattens out below
the articulation into a tapered blade (Fig. 5).

This displacement of the mandible is presumably related to a change in
feeding habits, and a partial loss of the normal masticatory function of a
chewing mandible has resulted. The mandibles of Mecoptera vary in their
width/length ratio and in dental armature on the inner surface of the
mandible which is variously toothed. The number of mandibular teeth is
constant in each family. Both the mesal and lateral edges of the mandible
are more heavily sclerotized than is the flat central plane (Figs. 4. 6). The
only Mecoptera with an unusual mandible are the Nannochoristidae, which
have very reduced mandibles and lack mandibular tendons.

The apical displacement of the mandibles combined with retention of a
generalized musculature has been compensated for by the development of
two large apodemes, an abductor (abt) and adductor (adt) tendon for each
mandible (Figs. 3-4, 6-7). The muscles that move the mandible insert on
the distal ends of these apodemes. The more strongly developed adductor
apodeme attaches to a small protuberance on the base of the inner edge of
the mandible. It splays above the level of the corporotentorium, providing
an extensive area for muscle attachment. The abductor a|xjdeme is well
developed but not as extensive as the adductor. The abductor apodeme
attaches to the outer edge of the mandibular base in an inflection of the
subgenal wall (Fig. 3). It is ventrolateral to the adductor apodeme.

The number of mandibular muscles is inconstant in Mecoptera. The
only detailed study of these muscles was of Panorpidae by Heddergott
(1938) and I have found that most of the families correspond with his
descriptions.

Tergal adductors (Fig. 2, mdm). The adductors are subdivided into
three parts by the adductor apodeme, which is T-shaped in cross section;
part 1 arises on the vertex; part 2 on the postgena; part 3 is formed by two
convergent bands that arise on the gena, postgena, and vertex. Since these
bands are divided by the mandibular apodeme, Heddergott (193N) named
each band accordingly; but since they function together as a single adductor
it is not necessary to formally name each band.

Clypeo-apodemal adductor (Fig. 2, cdm). This muscle is the second
mandibular adductor of Heddergott; it arises on the clypeus and inserts
directly on the shaft of the mandibular apodeme below the insertion of the
tergal adductors. Portions of the muscle arise dorsally on the antennarium
and the circumocular sclerite. Heddergott (193S) considered this arrange-
ment to be secondarily derived and that the clypeoapodemal muscles are
derived by fragmentation of the tergal adductors. He predicted that in those
Mecoptera with a shortened rostrum (Panorpodidae), the clypeoapodemal

742

The University Science Bulletin

,pht>

Fig. 3. Bittacus chlorostigma. Caudal view of mandibular complex after labium and maxil-
lae have been removed. Note complexity of skeletal elements and the Apodemalwalze (aw)
and status of the articulations of the mandibles.

Fit;. 4. Chorista australis. Mandible with tendons and condyles.

Fig. 5. Notiothauma reedi. Lateral view of articulation of mandible with rostrum.

I"k.. 6. Notiothauma reedi. Lateral view of mandible with tendons and condyles.

Fig. 7. Bittacus chlorostigma. Lateral view of mandible with tendons and condyles.

The Skeleto-Muscular System of Mecoptera: The Head 743

muscles would be greatly reduced and perhaps reunited with the tergal
adductors. I have found this the case in Panorpodidae.

Tergal abductor (Fig. 2, mbm). This muscle is the antagonist of the
adductors and is ventrolateral to them. It arises on the gena and vertex and
inserts on the abductor apodeme. The abductor is not subdivided as are
the adductors.

It is difficult to describe a basic plan of mandibular musculature for
Mecoptera and to designate portions as secondary. In terms of other Ptery-
gota, the division in Mecoptera of the adductors first into a tergal group
which is further subdivided and second into a group of clypeoapodemal
muscles should be regarded as a secondary development. The situation is
complicated by the fact that there are three possible combinations in Mecop-
tera. One group (Panorpidae, Panorpodidae, Meropeidae, Apteropanorpi-
dae, Boreidae and probably Choristidae although none could be checked)
possesses the full set of tergal and clypeoapodemal adductors and abductors.
This group also contains rostra of all three lengths. A second group contains
Bittacidae and Notiothaumidae in which there are no clypeoapodemal ad-
ductors. The other muscles are present and well developed, as in the first
group. Heddergott (193S) indicated a similarity between Bittacidae and
the nematocerous Diptera. A third group contains Nannochoristidae, which
completely lacks mandibular apodemes. I have been unable to find any trace
of a mandibular musculature in this family, and the conclusion that the
mandibles are functionless was previously advanced by Evans (1942).

Because of the nature of the mandibular articulation with the head
capsule, the mandibles are almost or wholly limited in their movement to
abduction and adduction. Some slight protraction and retraction might pos-
sibly occur as well.

MAXILLAE

Although the maxillae participate in the formation of the rostrum by
the elongation of their stipites, their structure is typical of that of many
pterygote chewing insects. The maxilla (mx) of Mecoptera consists of a
cardo (cr), stipes (st), lacinia (lc), usually a galea (gl), and a maxillary
palp (mxp) (Figs. 30-32, 34-3S, 41). The cardines are approximately tri-
angular sclerites by which the maxillae articulate with the cranium below
the postgenal bridge. The stipites are elongate, their length varying with
the length of the rostrum. The maxillary palps are five segmented in all
Mecoptera and arise at the apex of the stipites by membranous palpifers.
The stipes bears two apical lobes, an inner lacinia and an outer galea. These
lobes vary in their dimensions and degree of basal fusion, and in Nanno-
choristidae one of the lobes (the galea) is absent. When both lobes are
present (Fig. 9), there is an apodemal framework consisting of two heavily

744

The University Science Bulletin

ism

ccm

8

Fig. 8. Bonus sp. Longitudinal section of head with extrinsic effectors of the maxilla.
Note clypeocardinal muscle.

Fir,. 9. Chorista australis. Maxilla with its internal apodemal framework.

Fit.. Id. Panorpa iiuptialis. Longitudinal section of head with extrinsic effectors of the
maxilla. Note that both muscles are clypeal rather than tentorial in origin.

The SkeletoMuscular System of Mecoptera: The Head 745

sclerotized rod-like processes (y) in the lacinia and one in the galea (ua),
all of which are basally fused. The outer process of the lacinia continues
dorsally as an elongate lacinial tendon (let) on which the tergolacinial
muscles (tlm) insert at the level of the posterior tentorial arms. This tendon
or apodeme is characteristic of the maxillae of all Mecoptera regardless of
the degree of development of the galeal and lacinial processes or whether
the galea is present or not. The tendon varies with the length of the stipes.

The major skeletal variation in the maxilla involves the lacinia and
galea. Because of the extreme degree of fusion of these parts in Panorpidae,
and the absence of the galea in Nannochoristidae, there has been general
disagreement over the correct interpretation of these parts. By superficial
comparison with Diptera, Otanes (1922) regarded the lacinia as suppressed,
with the result that the two lobes represent a subdivided galea. Issiki (1933)
agreed with this interpretation. Heddergott (193S) rejected the former view
in favor of the one described above. The most convincing arguments for
the interpretation of these parts as presented here are in the comparative
study by Imms (1944). On the basis of musculature, Imms concluded that
the inner lobe is, in fact, the lacinia and the outer lobe, when present, the
galea. I concur with this view.

There is a graded series in the development of the galea and lacinia from
well developed and separated lobes in Bittacidae to the absence of the galea
in Nannochoristidae. Between these endpoints, the variations are of three
kinds. In Notiothaumidae, Meropeidac, and Apteropanorpidae, the galea
and lacinia are quite distinct and are fused only at the a[>ex of the stipes. In
Choristidae, Panorpodidae, and Boreidae, both lobes are shorter than those
in the preceding group and there is more fusion of the lobes. Lastly, the
Panorpidae are unique in that both lobes are present but reduced in size and
are completely continuous with the stipes, forming a single piece. The galea
and lacinia of Mecoptera are characteristically covered with a variety of
spines and setae, some of which form a conspicuous brush on the apex of
the lacinia as in Meropeidae (Figs. 35, 42). The anatomical details of these
setae are well illustrated for Panorpidae by Heddergott (193S) and tor
other families by Otanes (1922).

The maxillary musculature of Mecoptera is more variable than that of
either the mandible or labium, but there is a basic plan which consists of
the following elements:

Tentoriocardinal muscles (Figs. 11-12, tcm). These muscles arise on the
medial tentorial process and insert on the cardo. They are perpendicular to
the long axis of the head (parallel to the longitudinal axis of the body) and
are adductors or depressors of the maxillae.

Tentoriostipital muscles (Figs. 1, S, 11-12, tsm). These muscles arise on
the medial tentorial process (anterior to the tentoriocardinals) and on a

746 The University Science Bulletin

portion of the anterior tentorial arms and extend obliquely to the stipes,
where they insert on both sides of the stipital process. There are two distinct
tentoriostipital muscles. The first and more dorsal muscle is the abductor
and is antagonistic to the tentoriocardinal muscle. The second (often
termed the stipital promotor) is like the first except that its insertion is below
that of the first and is consequently more obliquely oriented. This muscle
is the promotor of the stipes and those parts below the stipes. In some
families these two muscles often appear continuous; in others, they are
distinctly separated.

Tergolacinial muscles (Figs. 11-13, tlm). These muscles arise as two
distinct groups, an anterior occipital and a posterior postgenal group, and
insert on the maxillary apodeme (an extension of the base of the lacinia).
The two muscles function as one flexor of the lacinia (which is the cranial
flexor of most other insects).

Stipitopalpal muscles (Fig. 11, spm). These muscles are the extensors
of the palp. They arise near the base of the stipes and insert on the first
segment of the palp.

Palpal muscles. The first of these is the second extensor of the palp and
arises in the first segment and inserts in the second segment. This muscle
is followed by flexors in the third and fourth segments of the palp.

The maxillary musculature as outlined above occurs in Panorpodidae,
Bittacidae, Meropeidae, Notiothaumidae, Apteropanorpidae, Nannochoris-
tidae, and probably Choristidae, though none were checked. The two excep-
tions to the above scheme are Panorpidae (Fig. 10) and Boreidae (Fig. 8).
This is unfortunate because most of the generalizations that have been made
about Mecoptera and passed on in the textbooks are based on studies of
Panorpidae by Issiki (1933), Hetrick (1935), and Heddergott (193S) and
do not reflect the condition of Mecoptera as an order. Panorpidae deviates
from the general scheme in two respects. The tentoriocardinal muscles of
other Mecoptera are clypeocardinal in Panorpidae and both groups of
tentoriostipitals of other Mecoptera are clypeostipital muscles (csm) in
Panorpidae. Although these three muscles have shifted their origins in
Panorpidae, they function as do their counterparts in other Mecoptera. The
interpretation of the condition in Panorpidae as secondary is further sup-
ported by the situation in Boreidae. The Boreidae are intermediate between
the majority of Mecoptera and Panorpidae. In Boreidae the "tentorio-
cardinal muscles" are clypeocardinal (as in Panorpidae), but the tentorio-
stipitals are extremely oblique and some of the fibers arise on the tentorium
(as in most Mecoptera) while others arise on the clypeus. Matsuda (1965)
claimed that the secondary development of the tentoriocardinal muscles
migrating to a clypeocardinal position is correlated with the elongation of

The Skeleto-Muscular System of Mecoptera: The Head

tlm

11

12

13

Fig. 11. Bittactts chlorostigma. Anterior view of maxilla and its muscles. Note that most
of the extrinsic effectors arc tentorial in origin.

Fig. 12. Bittactts chlorostigma. Lateral view oi same.

Fig. 13. Bittactts chlorostigma. Anterior view of same with tsm and tcm muscles removed.

the rostrum, as in Panorpidae and Boreidae. This conclusion is tenuous at
best since there is no constant proportional difference in the relative distance
of the tentorium from the apex of the rostrum in tentoriocardinal families
and clypeocardinal families. Some of the tentoriocardinal families (Bittaci-
dae, Notiothaumidae) have a greater distance between tentorium and the
apex of the rostrum than does the clypeocardinal family Boreidae. The
Panorpidae do, however, exhibit the greatest distance between the tentorium
and the apex of the rostrum.

748 The University Science Bulletin

LABIUM

The labium of Mecoptera is composed of two parts, a proximal post-
mentum and a distal prementum (Figs. 30-3S). The postmentum (pk)
varies in extent and degree of sclerotization. In most Mecoptera the post-
mentum is membranous and has a distal mental plate (mp). The excep-
tions are Meropeidae (Fig. 35) and Nannochoristidae (Fig. 31) in which
the postmentum is entirely membranous, the Boreidae in which it is com-
pletely fused into a single sclerite (zygostipes of Crampton, 1942) accom-
panied by the mesal fusion of the maxillary stipites (Fig. 30, mlc), and
lastly the Nannochoristidae in which it is greatly reduced. It lies between
the maxillary cardines and stipites and is separated from the foramen mag-
num by a postgenal or hypostomal bridge (pgb, hyb). The prementum
(rk) is usually sclerotized and may extend over the base of the labial palps.
The labium and maxillae are laterally fused and form the posterior wall of
the rostrum. The prementum bears two membranous palpifers, each bearing
a two segmented labial palp (lp). The mesal line of fusion between the
presumably once separated lobes of the prementum is visible in all Mecop-
tera, although it has become slightly obliterated in Panorpodidae. There is
no remnant of a ligular structure in any Mecoptera.

The base of the postmentum is always fused with the maxillary cardines
and stipites but the extent of the lateral fusion varies. In Notiothaumidae,
Choristidae, Panorpidae, and Apteropanorpidae the line of fusion extends
to the base of the mental plate. In the remaining families the line of fusion
extends below the mental plate to the base of the prementum.

The position and configuration of the mental plate is variable in Mecop-
tera but it is generally quadrate except in Bittacidae, where it is extremely
narrow and elongate (Figs. 32-33). The prementum is likewise variable.
The premental lobes of Choristidae are separated at their apices but fused
at their bases with a discernible fusion line (Fig. 37). According to Bier-
brodt (1942) both the larva and pupa of Panorpidae have a three-segmented
palp, which is reduced to two segments in the adult. Presumably this holds
for the other families as well. The basalmost segment of the preimaginal
palp becomes the palpiger of the adult palp. The palpiger is reduced and
continuous with the apex of the prementum and bears the base of the palp.
It varies in size and is completely absent in Bittacidae.

There are four possible sets of muscles that operate the labium of
Mecoptera:

Retractors of the prementum (rnm). These are paired muscles on either
side of the labium that arise on the ventral edge of the postgenal bridge and
insert either a) directly on the base of the prementum (Fig. 16), or b) on an
apodeme (ra) which extends dorsally from the base of the prementum
(Figs. 15, 17).

The Skeleto-Muscular System of Mecoptera: The Head

749

Extensors of the labial palp (epm). These muscles arise either on a) the
base of the prementum (Figs. 16-17), or b) the retractor apodeme (Fig. 15).
In both instances the muscles insert on the base of the first segment of the
palp (Figs. 15-17).

Flexors of the palp (fpm). These muscles arise on the mesal line of
fusion of the prementum and insert on the base of the first segment of
the palp.

Distal flexors of the palp (dfm). These muscles arise on the base of the
first segment and insert on the base of the second (Figs. 14-17).

mxp

dfm

17
14 15 16

Fig. 14. Nannochorista dipteroidcs. Caudal view of labium and maxilla. Note absence of
rnm and fpm muscles.

Fig. 15. Panorpa nuptialis. Caudal view of labium. Note absence of fpm muscle and
origin of epm muscle on tendon (ra).

Fig. 16. Merope tuber. Caudal view of labium. Note absence of tendon and fpm muscle.

Fig. 17. Bittacus chlorostigma. Caudal view of labium. Note presence of fpm muscle
which is unique to the Bittacidae.

750 The University Science Bulletin

Only the Bittacidae possess all four of these muscle sets, and in addition
they possess well developed retractor apodemes (Fig. 17). Nannochoristidae
(Fig. 14) and Boreidae lack both a premental retractor and a flexor of the
first segment of the palp. In the remaining families there is a premental
retractor, an extensor of the first segment of the palp, and, according to
Imms (1944), a flexor of the second segment of the palp. These families
lack a flexor of the first segment of the palp. The relative lengths of these
muscles and of the retractor apodeme naturally vary with the develop-
ment of the rostrum. The retractor apodeme is notably long in Panorpidae
with a consequently shortened premental retractor. There is no retractor
apodeme or flexor of the second segment of the palp in Meropeidae and
Notiothaumidae.

There has been much confusion about the labium of Mecoptera, most
of which stems from old controversies about the nature of the insect labium
in general. Otanes (1922), for example, interpreted the apex of the pre-
mentum as a fusion product of former ligular components, imagined that
it was unique to Mecoptera, and termed it the mecaglossa. Although there
is no trace of ligular remnants (stipulae, glossae, paraglossae) in Mecoptera,
there is no basis for this interpretation. Otanes also stated that the palpigers
are also part of this fusion product, which is incorrect (as shown by Bier-
brodt, 1942). Subsequently there was a long controversy between Crampton
and MacGillivray over the interpretation of the labium that did not resolve
the problem. In his study of the labium of Holometabola, Crampton (1925)
illustrated and discussed the labium of Bittacidae, Panorpidae, Choristidae,
and Nannochoristidae but confused the postmentum with the mentum and
the prementum with the palpiger. Issiki (1933) and Heddergott (193S)
followed a different interpretation in which they designated the postmen-
tum as the submentum in Panorpidae and Panorpodidae. This interpreta-
tion is unacceptable since the mental plate occurs alone without a separate
sclerotized submental plate. There is no basis for the recognition of a
submentum as such and the single plate of the postmentum should be desig-
nated as the mental plate to be consistent with the generalized insect labium
(Snodgrass, 1935). Setty (1940) interpreted these structures correctly in his
study of Bittacidae.

The greatest difficulties in the interpretation of the labium of Mecoptera
attend the Nannochoristidae. In a study of this family, Evans (1942) dis-
pelled the notion of a gula in Nannochoristidae (an interpretation retained
by Matsuda, 1965) but misinterpreted the components of the labium. Evans
stated that the prementum is reduced to a narrow membranous band con-
fluent with the palpigers and that the single sclerite of the postmentum
represents the mentum (which is actually the prementum). In an otherwise
fine comparative study of the labium of Mecoptera and Diptera, Imms

The Skeleto-Muscular System of Mecoptera: The Head 751

(1944), apparently unaware of Evans' paper, offered yet another interpreta-
tion of the labium in Nannochoristidae. He stated that: 1) there is a gula
which separates the occipital foramen from the postmentum; 2) that the
retractors of the prementum arise on the apex of the gula; 3) that the pre-
mental lobes are completely separated; 4) that the flexor muscles of the
second segment of the palp arise on the base of the prementum; and 5) that
the labial palps of all Mecoptera are two segmented. He incorrectly criti-
cized Tillyard (1917) and Crampton (1921) for having designated the
prementum as the first segment of the labial palp and attempted to support
this claim with evidence from musculature (since the muscles of the labial
palps arise from the prementum and labial retractors insert on its base).

Imms' interpretation is rejected on the basis of the following general
criteria, based mostly on Snodgrass (1935) and Matsuda (1965). The pre-
mentum is denned and identified by the presence of the retractors on its
base. The postmentum should be termed the submentum if it includes a
distinct mental plate. A correct interpretation of Nannochoristidae is diffi-
cult because 1) the postoccipital sutures partially extend into the postgenal
bridge (creating the illusion of a gula where there is none), 2) there is no
retractor of the prementum, 3) the postmentum is greatly reduced. The
similarities which Imms sought between Mecoptera and Diptera also in-
volved a misinterpretation of the labium in certain nematocerous Diptera.
Wenk (1962) showed that the postmentum of some Nematocera is often
greatly reduced and that the postgenal processes are simultaneously well
developed, a situation identical to that in Nannochoristidae. What Imms
illustrated and termed the postmentum still retains the mesal line of fusion
(as in all other Mecoptera) between once separated lobes; however, the post-
mentum of an insect labium never exhibits a trace of a fusion line anywhere.
It is the prementum which has the median fusion line. The actual post-
mentum of Nannochoristidae is greatly reduced. Likewise, it is impossible
to claim a two-segmented labial palp (as Imms tried) if his designations
are accepted, because this interpretation would leave only one segment for
the palp. This same observation was made by Hoyt (1952) in his study of
the evolution of the head and mouthparts of Diptera. The Nannochoristidae
(Figs. 14, 31) are unusual in that they lack the normal labial retractors and
a well defined postmentum. In other respects they conform to usual mecop-
teran labial structure and have a prementum, a small palpiger, two-seg-
mented labial palps, an extensor of the first segment of the palp, and a
flexor of the second segment of the palp.

ANTENNAE

The antennae of Mecoptera (Figs. 1S-29) are of the flagellar type charac-
teristic of the Pterygota (Snodgrass, 192S; Imms, 1939; Schneider, 1964;

752 The University Science Bulletin

Matsuda, 1965). Each antenna consists of three parts, two basal musculated
segments — a proximal scape (sc) and a distal pedicel (pe) — and a flagellum
consisting of a varying number of flagellomeres, none of which is muscu-
lated. The antennae are connected to the head capsule in the frontal area
between the compound eyes (co) and below the ocelli (oi) by an extensive
membranous socket, the antacorium (act). The antennal base and the
antacorium are circumscribed by the antennal sclerite or antennarium (aj),
which is reinforced by an internal submarginal ridge, the antennal suture
(ans). The antennarium bears a ventrolateral antennifer (anf) which
functions as the single pivotal point of articulation for the base of the scape
with the head capsule.

In Mecoptera, as in other Pterygota, movement of the antenna is effected
by extrinsic muscles which arise on the dorsal side of the anterior tentorial
arms and insert on the base of the scape. The general pterygote antennal
musculature consists of 1 or 2 pairs of levators and 1 or 2 pairs of depressors
of the scape (Snodgrass, 192S; Imms, 1939; Schneider, 1964; Matsuda, 1965).
There is a single pair of levators and depressors for each antenna of Mecop-
tera. Heddergott (193<S) worked out the basic structure in Panorpidae;
Setty (1940), the Bittacidae; and I have corroborated their findings and also
examined the other families. The basic pattern is as follows:

Depressor muscles of the scape (dsm). The depressors of the scape arise
on the dorsal surface on the anterior tentorial arms (q) and insert on the
ventral basal rim of the scape (Figs. 18-20). The attachment is off-center
and simultaneously depresses the antenna and turns it slightly to the outside.

Levator muscles of the scape (Ism). The levators arise just behind the
depressors at the point of origin of the dorsal tentorial arms (v) and insert
on the dorsal edge of the base of the scape (Figs. 18-20). Being thus oblique-
ly oriented, they simultaneously raise the antennae and rotate them medially.

Rotator muscles of the scape (rsm). In some Mecoptera there is an
unusual muscle, the rotator of the scape, which arises either on the Irons or
the dorsal tentorial arm and inserts on the dorsolateral side of the anta-
corium. In those families in which the muscle arises on the frons, the dorsal
tentorial arm continues up to and terminates on this muscle. Because of its
oblique course, it rotates the scape mesally and might contribute slightly to
the elevation of the whole antenna.

The intrinsic muscles of the antennae are those that arise on the base
of the scape and insert on the base of the pedicel.

Levator muscles of the pedicel (1pm). The levators of the pedicel arise
on a reinforced ridge that circumscribes the base of the scape and insert on
the base of the pedicel (Figs. 19-20).

Depressor muscles of the pedicel (dpm). The depressors of the pedicel
are situated opposite the levators and function antagonistically to them

The Skeleto-Muscular System of Mecoptera: The Head

753

(Figs. 19-20). The relative size of the levators and depressors varies but the
depressor is usually the larger of the two.

The flagellomeres are only passively movable with respect to one another
and the basal segments. With the exception of the rotators of the scape, all

Fig. 18. Panorpa nupticdis. Three-quarters profile of longitudinal section of right antennal
connection with the head. Note the disposition of the antenniferal pivot (anf) as well as
the peculiar rotator of the scape (rsm).

Fig. 19. Bittacus chlorostigma. Longitudinal section of antennal connection with head.
Note the absence of rsm.

Fig. 20. Panorpa nuptialis. Longitudinal section of antennal connection with head. Note
rsm and the oblique placement of dm and 1pm.

754 The University Science Bulletin

families fit the above description. A rotator of the scape ocurs in Panorpidae,
Panorpodidae, Apteropanorpidae, Meropeidae, and Notiothaumidae but
not in the other families.

There are slight variations in the external morphology of the antennae
that involve the antennifer, antacorium, antennarium, antennal suture, and
the number and shape of the flagellomeres. An antennifer is present in all
families of Mecoptera except the Bittacidae. Although the extent of the
antacorium varies, it is generally more extensive in Mecoptera than in other
insects. An antennarium is present in all Mecoptera and is most extensive
in Nannochoristidae; it is unusual in Boreidae inasmuch as it is ventro-
laterally incised. In general, the circumantennal suture is well developed
and conspicuous in Mecoptera, excepting Panorpidae, in which it is evanes-
cent. The scape is usually obliquely compressed and the pedicel subcylin-
drical. The flagellum is filiform with the exception of Meropeidae in which
it is moniliform. The number of flagellomeres is inconstant both in
individuals and in species (Miyake, 1913).

The number of muscles that effect the scape is usually two or four. Since
there are only two in Mecoptera, plus the rotator, there are two possible
explanations for the origin of the rotator of the scape. It might be regarded
as derived from one of the dorsal effectors, with the loss of its ventral
antagonist, or as having resulted from enlargement and subsequent splitting
of the levator of the scape. The second alternative seems more feasible and
would involve fewer developmental modifications to effect this change, as
evidenced from various experimental work on insect muscle (Matsuda,
1965).

DISCUSSION AND CONCLUSIONS

Most of the generalizations made about Mecoptera are compared with
what is known about other insect orders and about departures from what
might have been expected on the basis of a general plan of insect anatomy.
Most of the comparative information is taken from Matsuda (1965), who
has covered the literature on the insect head thoroughly.

1. In the process of rostrum formation, the relationship between the gena
and the clypeus has changed so that the subgena is ventrally displaced along
the lateral margin of the clypeus and is posteriorly involuted, forming an
articulation for the hypocondylar process of the mandible. The line of
fusion between the now extended subgena and the clypeus results in the
formation of the clypeogenal suture. A similar formation has occurred in
Hymenoptera but not to the same extent as in Mecoptera.

2. The suppression of cranial sutures is not unique to Mecoptera, but
progresses from the exopterygotes through the holometabolous orders. A
consequence of this trend is the indefinability of cranial regions. The post-

The Skeleto-Muscular System of Mecoptera: The Head 755

occipital region has moved dorsally and cephalad, as has the occipital region
which is fused with the vertex. There is no separation of vertex and genae,
hence no distinct parietals. The occurrence of a temporal suture is basic
to the Panorpoid Complex but occurs only sporadically among lower
pterygotes.

3. The maxillae are generalized and typical of chewing insects. The
cardo and stipes are unremarkable except for the formation of a zygostipes
(Crampton, 1942) in Boreidae. There has been a trend toward reduction
of the galea and lacinia in Mecoptera and in the other Panorpoid orders.
The tentoriocardinal and tentoriostipital muscles of Mecoptera (though often
absent in higher orders) are modified only in Boreidae and Panorpidae
through a secondary shifting of their insertions. The tergolacinial muscles
are unusual in that they arise as two bundles and jointly insert on a common
lacinial apodeme. The retention of stipitopalpal muscles must be regarded
as unusual since they do not occur in other Panorpoid orders. In terms of
the theoretically maximum number of maxillary muscles, the tendency in
Mecoptera has been reduction in the total number of muscles and the
retention of primitive sites of origin and insertion of those retained. Other
common maxillary muscles such as tergocardinal, tergostipital, stipitolacinial,
and stipitogaleal, while present in lower orders, generally do not occur in
the Panorpoid Complex and are completely absent from Mecoptera.

4. The mandibles of Mecoptera vary from extremely reduced to elongate
well developed appendages. Their articulations are modified by a close
association of the hypocondyle with the subgena] wall and the reduction of
the epicondyle and its close association with the Apodemalwalze, which is
unique to Mecoptera. It is generally held that the shape of the mandible is
directly related to feeding. While this may be so, not enough is known
about the feeding habits of Mecoptera to test the assumption. The mandibu-
lar muscles show the same tendency toward reduction as do those of other
pterygotes and there is nothing unique about these muscles in Mecoptera.

5. The labium of Mecoptera exhibits a combination of primitive charac-
ters plus certain losses not usually associated with a primitive state. There is
no gula in Mecoptera but the postgenal bridge is well developed and results
in a reduction of the base of the labium. The retention of an entirely
membranous postmentum and a sclerotized mental plate is regarded as
secondarily derived. The complete absence of ligular elements is basic to
the Panorpoid Complex. The labial musculature is variable in Mecoptera
as it is in other Panorpoid orders. The overall tendency is toward reduction
in the number of muscles.

6. A hypopharynx is present in all Mecoptera and is intimately associated
with the labium and the pharyngeal trough. Of a half dozen possible
hypopharyngeal muscles in insects, none occur in Mecoptera. It is difficult

756 The University Science Bulletin

to generalize about this organ since it is so variable and there is no apparent
overall trend in insects.

7. The antennae of Mecoptera are unique only in the possession by some
families of a rotator muscle of the scape. Compared to other insects, the
antacorium of Mecoptera is extensive.

8. Of the five tentorial elements, only the corporotentorium is variable
in Mecoptera, occurring in about half of the families. In lower Pterygota
there is a tendency toward fusion of the anterior and posterior arms into a
single plate, while the reverse occurs in the higher Pterygota. Mecoptera are
somewhat intermediate in this regard since they have a complete fusion of
the anterior and posterior arms on either side but no mesal fusion into a
single plate. The development of the tentorium in other Panorpoid orders
is variable and does not show any particular trend.

9. The modifications of the mecopteran head compared to a generalized
head have been twofold: 1) reduction of elements through specialization
into a new kind of sucking organ; 2) the retention of many primitive
features. All of the structural elements present can be homologized with
similar structures in other groups. The modifications discussed have gen-
erally been simply the elongation of previously existing structures without
change of the basic form of the element concerned, or in some cases reduc-
tion and suppression of these elements.

LITERATURE CITED

Applegarth, A. G. 1939. The larva of Apterobittacus apterus MacLachlan (Mecoptera:
Panorpidae). Microentomology 4:100-120.

Bierbrodt, E. 1942. Der larvenkopf von Panorpa com mums und seine Vcrwandlung m it
besondercr Beriicksichtigung des Gchirns und der Augen. Zool. fahrb., Anat. 68:
49-136.

Bigelow, R. S. 195-1. Morphology ol the face in Hymenoptera. Canad. Jour. Zool. 32:378-
392.

Byers, G. W. 1965 (1964). Families and genera of Mecoptera. Proc. 12th Int'l. Congr. Ent.
p. 123.

Crampton, G. C. 1921. The sclerites ol the head and mouthparts of certain immature and
adult insects. Ann. Ent. Soc. Amer. 14:65-103.

. 1925. A phylogenetic study of the labium of insects, with particular reference to

the Diptera. Proc. Ent. Soc. Wash. 27:68-91.

.1942. The external morphology of the Diptera. Conn. Geol. and Nat. Hist. Survey

Bull. #64, pp. 10-165.

CRICHTON, M. I. 1957. The structure and function of the mouthparts ol adult caddis (lies
(Trichoptera). Philos. Trans. R. Soc, London B 241:45-91.

DuPorte, E. M. 1946. Observations on the morphology ol the face in insects. Jour. Morph.
79:371-417.
— . I('62. Origin ol the gula in insects. Canad. lour. Zool. 40:381-384.

DuPorte, E. M., and R. S. Bigelow. 1953. The clypeus and the epistomal suture in Hymenop-
tera. Canad. Jour. Zool. 31:20-29.

The Skeleto-Muscular System of Mecoptera: The Head 757

Ehrlich, P. R. 1958. The comparative morphology, phylogeny and higher classification of
the butterflies (Lepidoptera: Papilionoidea) . Univ. Kansas Sci. Bull. 39:305-370.

Evans, J. W. 1942. The morphology of Nannochorista maculipennis Tillyard (Mecoptera).
Trans. Roy. Soc. S. Australia 66(2) :2 18-225.

Ferris, G. F., and B. E. Rees. 1939. The morphology of Panorpa nuptialis Gerstaecker (Me-
coptera: Panorpidae). Microentomology 4:79-108.

Grell, K. G. 1938. Der Darmtraktus von Panorpa communis L. und seine Anhange bei
Larve und Imago. Zool. Jahrb., Anat. 64:1-86.

Heddergott, H. 1938. Kopf und Vorderdarm von Panorpa communis L. Zool. Jahrb., Anat.
64:229-294.

Hetrick, L. A. 1935. The morphology of the head of the scorpionfly (Panorpa nuptialis
Gerst. Mecoptera). Proc. La. Acad. Sci. 2( 1 ) :1 13-120.

Hovt, C. P. 1952. The evolution of the mouthparts ot adult Diptera. Microentomology 17:
61-125.

Imms, A. D. 1939. On the antenna! musculature in insects and other arthropods. Quart. Jour.
Micros. Sci. 81:273-320.

.1944. On the constitution of the maxillae and labium in Mecoptera and Diptera.

Quart. Micros. Sci., N.S. 85:73-96.
Issiki, S. 1933. Morphological studies (>t the Panorpidae of Japan and adjoining countries and

comparison with American and European tonus. Jap. Jour. Zool. 4:315-416.
Levreault, P. 1936. The morphology of the Carolina mantis. Univ. Kansas Sci. Bull. 25:

577-633.

Matsuda, R. 1965. The morpholog) and evolution oi the insect head. Mem. Amer. Ent.
Inst. no. 4, 334 p.

Miyake, T. 1913. Studies on the Mecoptera ot Japan. Jour. Coll. Agr. Tokyo 4:265-401.

Otanes, F. Q. l'»22. I lead and mouth-parts of Mecoptera. Ann. Ent. Soc. Amer. 25:310-323.

Packard, A. S. 1889. On the occurrence ol taste organs in the epipharynx ol the Mecoptera.
Psyche 5:15'M64.

Schneider, I). 1964. Insect antennae. Ann. Rev. Ent. * ' : 1 <> -5 - 1 22.

Settv, L. R. 1940. Biology and morphology ot some North American Bittacidae (Order
Mecoptera). Amer. Midland Naturalist 23(2) :257-353.

Snodgrass, R. E. 1935. Principles <>l Insect Morphology. McGraw-Hill, New York. 667 pp.
.1947. Tin- insect cranium and the "epicranial suture." Smiths. Misc. Coll. 107(7):

1-52.

.I960. Facts and theories concerning the insect head. Smiths. Misc. Coll. 142(4):1-61.

Steiner, P. 1930. Studien an Panorpa communis. Zeits. Morph. Okol. Tiere 17:1-67.

Tillyard. R. J. 1917. Studies in Australian Mecoptera. Proc. Linn. Soc. N.SAY. 42(2):

284-301.

Wenk, P. 1962. Anatomic des Kopfes von Wilhelmia equina L. (Simuliidae syn. Melusinidac,
Diptera). Zool. Jahrb., Anat. 80:81-134.

Yie, S. T. 1951. The biology of Formosan Panorpidae and morphology of eleven species of
their immature stages. Mem. Coll. Agr. Nat. Taiwan Univ. 2:1-111.

758

The University Science Bulletin

mxp-

Fig. 21. Neobittacus blancheti. Anterior view of head. Fig. 22. Nannochorista dipteroidei
Same, Fig. 23. Apterobittacus apterus. Same. Fig. 24. Boreus unicolor. Same.

The Skeleto-Muscular System of Mecoptera: The Head 759

a

o

>

C
<

!

s

•3

K)

fe

760

The University Science Bulletin

£

W5

O

as

6

£

~<3

oo
tN

O

-a
u

o
u

c

<

«

©

«

ft,

o

The Skeleto-Muscular System of Mecoptera: The Head 761

mxp

31

mn

Fig. 30. Boreus unicolor. Caudal view. Fig. 31. Nannochorista dipteroides. Same.
Fig. 32. Apterobittacus apterus. Same. Fig. 35. Harpobittacus tillyardi. Same.

762

The University Science Bulletin

c

es
CO

IT)
00

p

6

-a

3
ca

o

<3

ft.

ft.

tt)

0

The Skeleto-Muscular System of Mecoptera: The Head 763

-<3

od
en

0

V

E
c/5

o

a

3

u

.§

S
«

s

5

B
o

Ci.

rn

764

The University Science Bulletin

•a
a

The Skeleto-Muscular System of Mecoptera: The Head 765

Fig. 43. Brachypanorpa carolinensis. Lateral view. Fig. 44. Nannockorista dipteroides.
Same. Fig. 45. Bittacus chlorostigma. Same. Fig. 46. Boreas unicolor. Same. Fig. 47.
Apterobittacus apterus. Same.

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLVIII Pages 767783 November 14, 1969 No. 18

The Morphology and Anatomy of
Alethopteris lesquereuxi Wagner1

Robert W. Baxter and M. R. Willhite
Department of Botany, Universit) oi Kansas

ABSTRACT

The morphology and anatomy of numerous specimens of Alethopteris leaves
found in Middle Pennsylvanian Kansas coal halls are descrihed from observations
hased on fractured surfaces showing entire pinnae with pinnules, and serial peel
thin sections showing details oi internal anatomy. The specimens are identified
as heing equivalent to the compression species, A. lesquereuxi Wagner. Ohser-
vations are made on the comparative anatomy of leaves of Alethopteris sulhvanti
and A. lesquereuxi and the extant Cycas revoluta.

INTRODUCTION

The form genus Alethopteris was established by Sternberg (1N25) lor
certain Upper Paleozoic fern-like leaf compressions which have since proven
to be an abundant and widespread element of the world's Pennsylvanian
flora. A wide range of forms have been recognized, with the recent excel-
lent monographic work of Wagner (1968) listing 29 species and 4 varieties.

The affinity of the leaves to the pteridosperms has long been recognized,
both because of the constantly sterile pinnules (lacking any of the usual
fern sporangia) and also because of the occasional discovery of specimens
with large seeds attached (Halle, 1927). The frequent association of the
Alethopteris leaf compressions with compressions of the medullosan pollen
bearing organ, Dolerotheca, and seeds, Pachytesta, also suggest, even more
specifically, that Alethopteris constituted at least one of the leaf types borne
by the stem organ genus Medullosa. This suggestion is strengthened by the
anatomical evidence found in coal balls, where it has been shown (Leisman,

1 This is part of a general investigation of the Pennsylvanian Coal ball flora supported by
the senior author's National Science Foundation grant GB 4933.

768

The University Science Bulletin

— 3 ~
•J c -3

x '3

— (j c

It UJ £

K t3 «i

3 S 12

t 2 w

C -O .

u en C

Sji'S

fa bC >

2 c ~

3 ns 2

ho <N

f x

3 'c3

^0 2 u in

c a V! J2

'S 'So M .3
•Si"""*

u

u

3 ^
3 3

3 - 5 £
C C_

— i~i o ^

>

3 'C b

Hug

.5 73 C

~ *-" 4-1 U

? u " >
fa o >>

3 nl *

■« X

- b~

O y

*"* >

u 3
ns

4i Ji v

53 Cl, U Or,

o

u

u

CO

^ 3i

■3 C U

•~ — ca
?«•-«.£
<-l <■» n

.^ J= c"

Morphology and Anatomy of Alethopteris lesquereuxi Wagner 769

I960) that similar mucilage canals, sclerotic strands and multicellular hairs,
as well as numerous scattered collateral vascular bundles, are found in both
the medullosan petiole Myeloxvlon and the rachis and mid-ribs of Alethop-
teris sullivanti pinnules.

A comprehensive review of the synonymy and classification of the genus
Alethopteris is given by Wagner (1968) and need not be repeated here.
However, the revised generic diagnosis, based on his exceptionally complete
monographic study, is worth quoting as a background for the following
descriptions.

Generic Diagnosis (Wagner, 196S) : "Repeatedly pinnate fronds which
may attain large dimensions. Rachides usually striate. Pinnules strongly
asymmetric, fused at the base, decurrent at the basiscopic side and decur-
rent, straight or slightly constricted at the acroscopic side. Pinnule lamina
generally rather thick (giving a vaulted aspect to the pinnules). Lobing
may be rather abrupt or more gradual, but lobing parts of the frond never
seem to predominate. Nervation characterised by a well-marked midvein
and numerous, non-anastomosing laterals attaining the pinnule borders at
about right angles or somewhat obliquely (depending on the species). The
lateral veins fork at irregular intervals. Subsidiary veins occur at the basi-
scopic side of the pinnule and generally at the acroscopic side as well. They
either spring immediately from the supporting rachis or, more likely, derive
from reclining basal lateral veins. Female Iructifications ('seeds') of the
Trigonocarpus and Pachytesta types. Male fructifications (as far as known)
are bell-shaped synangia of the Whittleseya and Dolerotheca types."

Except for the paper by Leisman (I960) referred to above, all of the
species of Alethopteris are known solely from compressions and impressions,
and the much to be desired correlation of the details ot the internal anatomy
with those of the external morphology are generally lacking.

The purpose of our present study is to present such a correlation for
specimens we believe to be equivalent to Alethopteris lesquereuxi Wagner
which show both external morphology characters and internal anatomy
comparable to that presented by Leisman for Alethopteris sullivanti. While
Leisman originally described his material as Callipteridium sullivanti, both
Wagner (1968, p. 151) and Cridland, Morns and Baxter (1963, p. 72) have
pointed out that the proper assignment should be to Alethopteris.

MATERIALS AND METHODS

The present study is based on numerous specimens of Alethopteris found
in coal balls from the Fleming coal of southeast Kansas. This is in the
Cabaniss Formation. Cherokee Group, Desmoinesian Stage of the Middle
Pennsylvanian which we believed to be equivalent to the upper part of
the Westphalean D.

770

The University Science Bulletin

Fig. 2. Alethopteris lesquereuxi. Fractured surface of coal hall showing portion of frond.

Note lack of visible lateral venation anil deeply revolute lateral margins and apex of pinnules.

Cross sections of pinnules exposed in fracture are shown at (t) with verj large hypodermal
cells clearly visible as white layer. X 6.

Morphology and Anatomy of Alethopteris lesquereuxi Wagxer 771

The specimens are contained in several dozen coal balls selected from
some of the thousands which have been collected and cut in our laboratories
over the past twenty years. From these, five different specimens of pinnae
with pinnules showing the surface features (usually only seen in compres-
sions) were obtained by splitting some of the coal balls along their natural
fracture lines (Figs. 2-4). Numerous peel thin sections were also made in
the three planes shown in Figure 1, the sagittal sections particularly being
informative of details of venation which could not be observed in surface
views.

GENERAL MORPHOLOGY

The pinnules average 3-4 mm wide by 8-10 mm in length, or propor-
tionally around 2-3 times as long as they are broad. Both the pinnule lateral
margins and apices are strongly revolute. The midrib is deeply sunken on
the upper side, while the lateral venation pattern is very faint to lacking as
viewed from the upper surface. The pinnule shape varies from nearly
parallel sided, pecopterid forms on lower portions of the pinnae (Fig. 2) to
pinnules with strongly decurrent lower sides (Figs. 3, 7) near the more
apical portions of the pinnae. The bases of the pinnules are frequently
swollen, forming a characteristic double bulge where they attach to the rachis
(Fig. 4). The later. il veins, which generally can be seen only in internal,
sagittal sections (Figs. 7-8), depart from the midrib at a slightly oblique
angle at an average density oi 32 per cm. The terminal pinnule is elongated
and somewhat lanceolate.

INTERNAL ANATOMY

Transverse sections of the pinnules (Figs. 1. 6) present a somewhat M-
shaped pattern due to the strongly revolute margins and the sunken,
abaxially ridged midrib. The pinnules measure 720-750 ju, in thickness near
the middle of the lateral mesophyll, tapering gradually to a margin of
scarcely 150 fi. This is much less than the equivalent measurements for
Alethopteris sullivanti, where Leisman (1960) gives a 1445 fi thickness
through the midvein and 635 /' in the middle mesophyll with little tapering
at the non-revolute margins.

The upper epidermis consists of small, rectangular cells, which we have
been able to see satisfactorily only in side views of cross and longitudinal
sections of the pinnules (Figs. 1, 10, 13). In this view the cells measure 50 ^
along their periclinal surface and only 20 ^ in their anticlinal thickness.
This layer is most frequently very poorly preserved and, along with the
small size of the cells, quite difficult to see (Figs. 6, 11). Below this true but
inconspicuous epidermis is a hypodermal layer of exceptionally large cells
measuring 140 x 100 x 70 \>.. In surface view their outer periclinal wall has

o

Morphology and Anatomy of Alethoptens lesquereuxi Wagner 773

an irregular, pentagonal shape with the greatest width, which is parallel to
the lateral veins, reaching 140 p. In this view, where the upper epidermis
has either naturally sloughed off or been removed by etching, they may-
simulate a fine reticulate venation reminiscent of the genus Lonchopteris
(Fig. 5). Longitudinal sections of the pinnules, at right angle to the lateral
veins, show the hypodermal cells are large rectangles measuring 70 /j. across
the top (periclinal) wall and 100 fi down the side (anticlinal) wall (Figs.
10-11). In cross sections of the pinnules, parallel to the lateral veins, the
hypodermal cells measure 140 ^ across the top (periclinal) wall and 100 ^
down the side (anticlinal) wall (Figs. 1, 13).

This hypodermis seems identical to what Leisman (I960) called the
inner layer of a double epidermis in Alethoptens sullivanti. However, since
the determination of a double epidermis is dependent on knowledge of its
ntogeny, it is obviously impossible to use the term here with certainty.
Also, while a true double epidermis is known to occur in only a very few
living plants, such as in the leaves of Fiats and Peperomia and roots of the
orchids (Esau, 1965), hypodermal layers are quite common in many leaves,
particularly in those of somewhat xerophytic character. Accordingly, the
term hypodermis would seem most appropriate for this most distinctive
tissue apparently common to both A. sullivanti and A. lesquereuxi. Indeed,
while the relative development of the cuticle, palisade and spongy tissues <>t
an extant species may vary with ecological changes, the presence or absence
of a hypodermis will normally be a constant taxonomic character. Accord-
ingly, the proof that two such dissimilar species of Alethoptens have this
concurrence of a minute upper epidermis subtended by a large hypodermis
suggests that it may be a character common to the genus. It is not, to our
knowledge, a feature found in any other coal ball leaves.

Immediately below the hypodermal layer is the palisade tissue, con-
sisting of one to two layers of vertically elongated cells which are character-
ized in nearly all of our material by their opaque, black contents (Fig. 11).
This tissue constitutes approximately one half of the mesophyll thickness
below the hypodermis, the balance consisting of a loosely arranged spongy
tissue with numerous air spaces (Figs. 9, 11).

The midrib vascular bundle enters directly from the lower side of the
flattened rachis (Figs. 7-8). It consists of 16-20 primary tracheids, the small-

Figs. 3-4. Alethoptens lesquereuxi.

Fh.. 3. Fractured surface ol coal ball showing apical portion of a frond. Original cut of
coal hall resulted in loss of tip of apical pinnule, hut visible portion indicates the elongate-
lanceolate shape characteristic of the species. A fracture through mesophyll of pinnule shows
a sagittal section at S. X 6.

Fig. 4. Fractured surface of coal ball of still another specimen. Note deeply sunken
midrib, revolute margins, double swellings at pinnule bases, and absence of visible lateral
veins. X 6.

774

The University Science Bulletin

w » y i » J . * *

0

•■ * J V

% . . ♦ i

• ' » •*

»

.\

• >

.

Figs, 5-6. Alethopteris lesquereitxi.

Morphology and Anatomy of Alethopteris lesquereuxi Wagner 775

est of which are on the abaxial side, making the structure apparently exarch.
The xylem is directly enclosed in a bundle sheath 200 ^ in diameter with
collenchyma type sheath extension cells extending to the upper hypodermis
and lower epidermis. No phloem can be recognized.

The lateral veins depart from the midrib at a slightly oblique angle to-
wards the pinnule apex and usually dichotomize once before reaching the
margin (Figs. 7-9). They pass through the spongy tissue slightly below the
palisade layer (Fig. 11) and are constantly of such small size that the term
"veinlet" seems appropriate for them. The xylem of each veinlet consists of
only 3-5 primary tracheids, ranging in diameter from 10 to 20 ik showing
an apparent exarch development (Fig. 10).

The conspicuous bundle sheath tightly encloses the xylem strand, and
no tissue recognizable as phloem could be identified. The bundle sheath
consists of parenchymatous tubular cells, 40 x 500 \i long, elongated parallel
to the veinlet (Figs. 9-11). Similar bundle sheath extension cells extend
upwards to or through the palisade layer but never beyond the hypodermis.
On the lower (abaxial) side, the sheath extensions widen out and become
more sclerotic, forming the buttresses for the conspicuous ridges which
characterize the lower surface of the pinnules (Figs. 10-11). The entire
diameter of the veinlets, including the bundle sheath, is 120-140 fi, with their
density (measured near the pinnule margin) averaging 32 per cm.

Subsidiary veins can be seen entering the base of the pinnules directly
from the upper side of the flattened rachis (Fig. 7). Accordingly, in this
species at least, there can be no doubt that separate subsidiary veins do exist
in Alethopteris and that the situation is not as described by Bochehsky
(1960) but rather as diagramed by Wagner (1968, p. 25, text fig. 1). Con-
sequently, Wagner's generic diagnosis, quoted earlier, should probably be
revised to emphasize this direct rachis origin of the subsidiary veins.

There is certainly no evidence, in all of the many sagittal sections of A.
lesquereuxi examined, to support the concept of a single, strong, decurrent
midvein. Quite to the contrary, as many as 5-6 subsidiary veins appear to
enter the bases of decurrent pinnules directly from the upper side of the
flattened rachis, with the independent midrib vein entering the pinnule
from a lower level of the flattened rachis. This feature of independent
vascular strands is undoubtedly related to the numerous separate vascular

Fig. 5. Pair of pinnules exposed on fractured surface of coal ball. Note reticulate pattern
of large hypodermal cells exposed here by etching away the upper epidermis. Fracture of tip
of lower pinnule has exposed a cross section showing the large hypodermal cells in .side view
and deeply incurved revolute pinnule margins. /;, hypodermis in face and side views. X 10.

Fig. 6. A peel cross section of a pinnule. Adaxial midrib furrow lacking due to section
being near pinnule apex (see pinnules upperleft part of Fig. 2). Compare to Fig. 1. which
represents cross section near middle of pinnule, e, epidermis; /;, hypodermis; /, multicellular
epidermal hairs. X 60.

The University Science Bulletin

©

-

&E

IS

;

.•

n

•• krlj >.

*

>

+: >>

-

' " h '''

:

1

«!

t

*
1

%

1

■ . * '

.'* > ;

i

*

•

-

^t-

•• ' •

V .

^} •

* *

t

' ' 1

**«

~,

•

-
•

■

$*£

'■■ft .,.,.

r

•■ i - : •;

I

-£ **> " • • • • . -

.'■ t

i

-

VJ

Figs. 7-8. Alethopteris lesquereuxi. Pee] thin sections in sagittal pl.inc through pinnules
showing pattern of lateral veinlets.

Morphology and Anatomy of Alethopteris lesquereuxi Wagner 777

bundles already known to occur in the Alethopteris rachis (Leisman, I960),
which is apparently an extension of the multi-bundled anatomy of the
Medullosa petioles.

The lower epidermis appears to be a continuation around the pinnule
margins of the same uniseriate layer of very small cells found on the upper
side. As on the adaxial side, they are poorly preserved and usually only
evident as a thin amorphous black layer subtending the veinlet ridges and
the alternate furrows. Efforts to obtain free epidermal tissues by maceration
were not successful, nor could stomata be pinpointed with any assurance.
It would seem very likely, however, that stomata were present on the lower
surface in the furrows between the veinlet ridges. There seems no reason
to doubt, considering other basic similarities between A. sullivanti and
A. lesquereuxi, that the stomata are probably also of the haplocheilic type.

Numerous uniseriate, multicellular hairs arise from the lower epidermis.
They are particularly abundant on the midrib and veinlet ridges and also
near the inner side of the revoltile pinnule margins (Figs. 1, 6). Pappilate
cells such as described for A. sullivanti (Leisman, 1%0) appear to be
lacking.

DISCUSSION

The assignment of our coal ball specimens to A. lesquereuxi Wagner is
based on strong similarities in the size, shape and revolute margins of the
pinnules. The angle and number of lateral veinlets per cm also agrees with
Wagner's specific diagnosis as does the form ol the terminal pinnule. Photo-
graphs of the specimens exposed on the fractured surfaces of the coal balls
(Figs. 2-4) were also sent to Dr. Wagner, who independently arrived at an
identification with A. lesquereuxi.

Since Alethopteris sullivanti and A. lesquereuxi now constitute the only
two species in which both external morphological and internal anatomical
characters are fully known and correlated, it may be worthwhile to sum-
marize the anatomical features in which they agree versus the ones in which
they differ. In this way we may be able to emphasize what seem to be
generic versus specific characters. On this basis, the characters common to
the genus appear to be: (1) the presence of a thin, small celled upper
epidermis subtended by a hypodermis of conspicuously larger cells; (2) dif-

Fk.. 7. A portion of the rachis (R) with two attached decurrent pinnules. Section passes
above sunken midrib, the position of which is shown at M. Susidiary veins (V) can be
seen entering base of pinnule directly from rachis. X 10.

Fig. 8. Nearly complete pinnule at upper left shows midrib and lateral veins near apex,
while section in basal part has been cut above sunken midrib through large cells of hypo-
dermal layer. The bases of the two pinnules on the right margin appear as double-
compartmented boxes due to the sagittal section passing below the main pinnule and showing
only the sunken midrib and the revolute margins. /;. hypodermis; T, tissue of sunken
midvein; R, revolute pinnule margin. X 10.

778

The University Science Bulletin

« •

0

e*V&kU -*\

% *

^ ~

® *

if

*&r

>t^ *

A

^

r

*

^ H Jr'if* H <■■ —

s

if 'Si

*•* •<. -\

'0

fci:

. j

Figs. 9-10. Alethopteris lesquereuxi.

Morphology and Anatomy of Alethopteris lesquereuxi Wagner 779

ferentiation of the mesophyll into an upper palisade layer and a lower
spongy layer; (3) parenchymatous bundle sheathes and bundle sheath exten-
sions; (4) numerous bundles in rachis; (5) uniseriate multicellular hairs
on lower epidermis.

The anatomical characters in which the species differ are as follows:
(1) Size of epidermal and hypodermal cells. In A. sullivanti the epidermis
consists of cells 19 x 31 /* with the cells of the hypodermis averaging
43 x 43 /J-. In A. lesquereuxi the comparable measurements are 20 x 50 ,u
for the epidermis and 100 x 70 ^ for the hypodermis as seen in longitudinal
section of the pinnules (our Fig. 10 and Leisman, I960, fig. 7); (2) the pin-
nules are thicker in A. sullivanti, measuring 635 /j. in the blade to 540 /x for
A. lesquereuxi, with a corresponding variation in thickness of the palisade
and spongy layers of the mesophyll; (3) the lower surface of the pinnule
of A. lesquereuxi is ridged below the veins while the lower surface is smooth
in A. sullivanti; (4) A. sullivanti pinnules are flat with little tapering at
the margins contrasted to A. lesquereuxi pinnules which have strongly
revolute, tapered margins.

The primary xylem of the veins in A. sullivanti is described as endarch
while that of the veinlets and midvein in A. lesquereuxi seems to be exarch.
It is doubtful that this is a real difference since observations regarding the
relative position of protoxylem and metaxylem, which are interred solely
from tracheid size as seen in cross sections, are dubious to say the least.
The size range of the tracheids and the shape of the xylem strand in our
specimens (Figs. 10, 12) seems suggestive of exarch development but
certainly cannot be held to be conclusive.

While the leaves of both species seem to show some xerophytic charac-
ters, e.g. hypodermis, thick mesophyll, and abundant multicellular epi-
dermal hairs, the trend appears more strongly developed in A. lesquereuxi.
This inference is based on the much thicker hypodermal layer, the probable
restriction of stomata to furrows sunken between the veinlet ridges and the
deeply revolute margins which may have also functioned to reduce air flow
over the stomatal area. The hypodermal layer of A. lesquereuxi is particu-
larly striking since, while uniseriate, it still may contribute as much to the
pinnule thickness as the entire palisade tissue.

Wagner's (196S) monograph of Alethopteris indicates a wide distribu-
tion for A. lesquereuxi in strata of late Westphalean D to strata of lower

Fig. 9. Peel section through sagittal plane of pinnule tip showing dichotomous lateral
veinlets (V) passing through spongy tissue (S). Elongate cells shown are those of the veinlets'
buntllc sheaths. /;, hypodermal layer exposed at revolute margin. X 33.

Fig. 10. Longitudinal section of pinnule near margin, e, epidermis; /;, hypodermis; .v.
xylem; s, bundle sheath; se, bundle sheath extension. Palisade and spongy tissues are
diminished due to section being very close to pinnule margin. Note abaxial ridges below
veinlets. X 120.

780

The University Science Bulletin

:»i

l

1

J

Morphology and Anatomy of Alethopteris lesquereuxi Wagner 781

and middle Stephanian age in Spain, Great Britain and North America.
It is certainly the dominant species in the Middle Pennsylvanian coal ball
flora of Kansas since, other than the occasional A. sullivanti specimen,
almost all of the identifiable material in our collections now seems assign-
able to A. lesquereuxi.

Since a possible relationship between the genus Medullosa (which almost
certainly bore these leaves) and the living cycads has frequently been postu-
lated on various features of stem anatomy, seed position, etc., a comparison
of the leaf structure in A. lesquereuxi and the extant Cycas revoluta is
shown in Figures 13 and 14. It may be noted that both have a thin epidermis
subtended by a much larger hypodermis. Also, while Cycas revoluta is
described (Coulter and Chamberlain, 1917) as lacking lateral veins, it does
show the presence of long tubular parenchyma cells with large conspicuous
pits which extend outwards from the midrib to the pinnule margins and
possibly function as a kind of transfusion tissue. As can be seen in Figures
13 and 14, these cells are strikingly similar to the bundle sheath cells in
A. lesquereuxi which, because of the minute amount ot vascular tissue,
probably were also functional in water conduction. A reduction theory
might be developed leading from the rather large veins of Alethopteris
sullivanti to the small veinlets of A. lesquereuxi to Cycas revoluta where
the vascular portion of the lateral veinlets has been completely lost.

Certainly the lateral veinlets in Alethopteris lesquereuxi arc so small and
separated from the upper blade surface by the large hypodermis, that it is
doubtful that the living, uncompressed pinnules would have shown any
pattern of lateral venation on the upper surface (Figs. 2, 5), bur instead
would seem to have only the single midvein as in Cycas revoluta.

The extremely primitive megasporophyll of Cycas revoluta makes it a
particularly attractive subject for comparison to the seed ferns, but one
could as equally well postulate an evolvement of the leaves ot the other
genera of extant cycads from the seeming vascular reduction in Alethopteris
sullivanti and A. lesquereuxi. For example, the numerous basal dichotomies
leading to the apparent parallel venation of Zamia, Dioon, etc. could be
derived from the individual "subsidiary veins" of Alethopteris with a reduc-
tion in size of the midvein, while Stangeria. with its midrib and dichoto-
mous lateral veins, could represent a form with relatively unmodified
venation.

Fk,. 11. Pee] longitudinal section through thicker portion of pinnule than shown in
preceding figure, e. epidermis; /;, hypodermis; x, bundle sheath; se, bundle sheath extension.
Some palisade tissue visible below hypodermis. Note abaxial ridges below veinlets. X 120.

Fig. 12. Transverse section of pinnule mid-rib. x, primary xylem of bundle, apparently
exarch. X 50.

7S2

The University Science Bulletin

na&jks/' . /h ®

Fig. 13. Alethopteris lesquereuxi. Cross section of a portion of a pinnule cut parallel to
a lateral veinlet. c, epidermis; //, hypodermis; s, elongated cells of bundle sheath. X 1 2U.

Fig. 14. Cycas n valuta. Cross section of a portion of a pinnule showing elongate "trans-
fusion" cells (t). c, epidermis; A, hypodermis, X 135.

Morphology and Anatomy of Alethopteris lesqaereuxi Wagner 783

If one continues in this purely theoretical argument, it can be postulated
that the obvious xerophytic modifications in the extant cycads led to a
reduction in frond size from the very large multi-pinnate leaves of Medul-
losa to the once pinnate (bipinnate in Bowenia) leaves of the present cycads,
along with a reduction in lateral venation as a mechanism to diminish
water loss.

LITERATURE CITED

Bochexski, T. 1960. Evolution of pinnule venation in the carboniferous seed ferns Alethop-

terides (Alethopteris and Lonchoptcris ) and the meaning of pinnule venation analysis

for diagnosis of species. Prace Inst. Geol. Warszawa. 20:1-42.
Coulter, J. M., and C. J. Chamberlain. 1917. Morphology of Gymnosperms. University

of Chicago Press.
Cridlaxd, A. A., J. E. Morris and R. \Y. Baxter. 1963. The Pennsylvanian plants of Kansas

and their stratigraphic significance. Palaeontographica (B) 112:59-92.
Esau, K. 1965. Plant Anatomy. John Wiley S< Sons, New York. 2nd Edition.
Halle, T. G. 1927. Palaeozoic plants from central Shansi. Palaeontologia Sinica (A) 2:1-316.
Leisman, G. A. I960. The morphology and anatomy of Callipteridium sullivanti. Amer. Jour.

Bot. 47:281-287.
Sternberg, K. von. 1 825 . Versuch einer Geognostisch-botanischer Darstellung der Flora

Vorwelt. Teil 4:1-48.
Wagner, R. H. 1968. Upper Westphalian and Stephanean Species of Alethopteris from

Europe, Asia minor, and North America. 2 Volumes. Uitgevers-Maatschappij.,

Ernest Van Aelst, Maastricht. Netherlands.

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLVIII Pages 785-790 November 14, 1969 No- 19

On the Status of Caecilia occidentalis Taylor

Edward H. Taylor*

While re-examining caecilians in the Academy of Natural Sciences of
Philadelphia (ANSP), I found four specimens from Popayan and Mos-
copan, Cauca, Colombia, which had been overlooked on my previous
examination of the collection. These had been labeled Caecilia pachynema
by a previous caretaker. Careful examination of these specimens causes
me to associate them with Caecilia occidentalis Taylor described from a
specimen obtained from Dr. Max Hensley and likewise from Popayan.
Despite the fact that the specimen serving as the type had been injured
badly, practically all significant characters were discerned.

Of these four specimens, ANSP Nos. 25566-25569, the first three are
topotypes of C. occidentalis, the fourth is from Moscopan, Cauca, Colombia.
They provide data to substantiate the validity of the species. A skull has
been prepared from ANSP No. 25568, which is described and figured.

Comparative data on measurements, tooth counts, etc., are presented
in the following table [data on the type, (EHT-HMS 4665), included].

From the table, it will be noted that the process of growth entails a
great lengthening of the body, the width of head and body growing but
little proportionally.

The absence of secondaries in certain specimens of a species has been
reported in other species of the genus Caecilia (i.e., guntheri, pachynema)
while in certain other species secondaries have not as yet been found to
occur (i.e., caribea and elongata).

The collars in these four specimens are not as well marked as in the
type. The transverse dorsal grooves are either absent or only dimly indi-
cated. The two collars are not clearly separated dorsally, and the second is
fused dorsally and ventrally with the first primary fold.

The lateral cream or yellowish stripe is similar to that of the type, better
defined anteriorly and posteriorly, while in the area between, it may be
reduced to a dim broken line which may be absent for a distance. The

* Research Associate, Kansas University, Museum of Natural History.

786

The University Science Bulletin

Table 1. Comparative data on Caealia occidentalis. All measurements are

in mm.

Museum EHT-HMS

No. 4665

Locality Popayan

Total length 1035

Head width 10

Body width 9.3

Eye to tentacle 5.4

Tentacle to nostril 1.7

Primary folds 218

Complete folds 4 or 5

Secondary folds 0

Premax-max. teeth 12-13

Prevom-pal. teeth 10-11

Dentary teeth 11-11

Splenial teeth 5-(5)

Scales (rows) posteriorly 1

Width in length (times) Ill

Vertebrae 225

ANSP

25566

Popayan

ANSP

25567

Popayan

ANSP

25568

Popayan

ANSP

25569

Moscopan

425

564

868

995

7

8

8

9

8.2

8.5

8

9

3.2

3.5

4

4.1

1.1

1.25

1.5

1.6

191

209

205

221

5

8

17

20

12

0

4

8

8-8

10-9

11-11

11-10

10-9

11-12

12-12

12-11

12(12)

12-12

12-12

11-11

3-3

5-3

3-3

3-3

1

1

1

1

52

663

1085

110

197

207

212

229

subterminal markings are practically the same in all. The occipital region
may have three dim cream spots and there is a more or less distinct light
spot at or anterior to the tentacle and the nostril; one also on the tip of
the snout.

Scales in the grooves are present in more than two thirds of the body.
Subdermal scales, in evidence throughout much of the body, are very small,
rarely more than 0.2 mm in diameter. The head is generally tapering
somewhat, the snout extending about 2.0 to 2.3 mm beyond the mouth.

Within the mouth the choanae are very close to the prevomerine teeth.
The narial plugs of the tongue are elevated and teatlike.

The skull of this species shows strong similarity to the skulls of the
Caeciliinae, especially to species of the Caecilia sensu strictit, as the data
here recorded indicate. The data presented are from the skull of ANSP
22568. The skull agrees with skulls of the family in having the reduced
number of bones. Thus the prefrontals, septomaxillae and oculars are not
present as separate bones. The nasals and the premaxillary elements are
fused to form the nasopremaxillae. The maxillae and palatines are fused
to form the maxillopalatines. The stapes are present. The mesethmoid
does not appear on the dorsal surface of the skull.

The anterior dorsal surface of the skull is covered by the paired naso-
premaxillae, the lateral edges of which are roughly notched along the

Ox the Status of Caecilia occidentalis Taylor

787

788

The University Science Bulletin

sutures. The frontals are narrowed anteriorly, flaring out posteriorly, in
contact only for about one half of their length. The parietals are a little
longer than the frontals, bending down posteriorly with very uneven
sutures. The two lateral posterior parts of the compound basisphenoid meet
narrowly preceding the foramen magnum.

The side of the skull anteriorly is formed by the maxillary part of the
maxillopalatine, followed by the squamosal, a bone somewhat shorter than
the preceding, and in contact posteriorly with the quadrate and also to a
process of the stapes.

Fig. 2. X-ray of Caecilia occidentalis Taylor. ANSP No. 22569, Moscopan, Cauca, Colom-
bia. Vertebrae, 229. Actual length, 995 mm.

On the Status of Caecilia occidentalis Taylor 789

The stapes fits into a notch in the lateral part of the basisphenoid. The
posterior part of the suture between the parietal and squamosal is joined
by cartilage.

Anteriorly on the ventral surface of the skull there is a shelflike forward
projection between the nasal openings. The premaxillary parts of the
nasopremaxillae bear seven teeth (3-4). On the ventral side of these parts
there are three posterior projections which interdigitate with two blunt
processes from the anterior part of the prevomers preceding the prevomerine
dental series. These series consist of seven teeth (3-4). The prevomerine
bones extend posterior to the teeth, separated for much of their length by
an anterior spine of the basisphenoid and reaching one fourth of their length
behind the internal nares, forming half of the border of the nares on their
inner sides. The outer sides of the nares are bordered by the palatine portion
of the maxillopalatine.

The basisphenoid sends forward a spinelike process that separates the
prevomers for more than half their length, and also two processes between
the prevomers and the relatively small diastema between the basisphenoid
and the pterygoid process of the quadrate. The "wings" of the basisphenoid
are large, laterally curving down slightly. Below the region of the otic
capsules there are two short, transversely flattened, blunt, processes.

The skull measurements in mm and tooth counts are as follows: total
length of skull, 11.2; greatest width, 6; width at orbits, 5; length of jaw, 7.2;
length of basisphenoid (ventral), 8; width of same at "wings," 4; width at
otic capsules, 4; length of prevomers, 5.2; width of prevomers, greatest, 3;
length from anterior edge of internal nares to occipital condyle, 6.4; pre-
maxillary teeth, 3-4; maxillary teeth, 7-7; prevomerine teeth, 4-4; palatine
teeth, 8-7; dentary teeth, 13-13; splenial teeth, 4-4. Total length of the
preserved specimen, 868.

I desire to extend my gratitude to The Director of the Philadelphia
Academy of Sciences, Dr. RadclyfTe Roberts, and to Dr. James E. Bohlke,
Curator, for the loan and tor assistance while at the Academy.

790

The University Science Bulletin

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLVIII Pages 791-799 January 9, 1970 No. 20

Dental Abnormalities in North American Bats.

II. Possible Endogenous factors in Dental Caries in

Pliyllostomus hastatus

Carleton }. Phillips* and J. Knox [ones, Jr.

Mu cum ot Natural History, The University of Kansas

Dental decay is a complex disease that lias been studied intensively in
man, strains of the Norway rat (Rattus norvegicus), and in a few other
laboratory mammals. Many different endogenous and exogenous factors
have been analyzed (see Keyes and Jordan, 1963). It is thought, however,
that the basic biochemical, physiochemical, and physical factors that interact
to cause dental caries are comparable in all mammals (Konig, 1965). Al-
though laboratory and clinical studies are numerous, there have been only a
few reports of the incidence and causes of dental caries in wild mammals
(Colyer, 1936; Hall, 1940, 1945), and no previous study of this disease in
bats has been undertaken. In the course of studies of the dentitions of North
American Chiroptera, we discovered a high incidence of dental caries in
the spear-nosed bat, Pliyllostomus hastatus, but no evidence of the disease in
related species. When disease of any kind is found commonly in one species,
but not in closely related kinds, the causes and evolutionary implications of
that disease become important areas for investigation.

Among 52 specimens (all adults) of Pliyllostomus hastatus examined
from Nicaragua, Panama, Trinidad, and Venezuela, 21 (40.4%) of the
individuals had readily detectable dental caries, resulting, in many cases, in
the loss of one or more teeth. The incidence of carious lesions in males
(75% of 20 specimens) was found to be significantly greater (p— .99) than
in females (18.7% of 32 specimens). Occurrence of the disease evidently is
not geographically variable, because specimens from each of the regions
listed had carious lesions. No evidence of dental caries was found, however,
in 103 specimens of Pliyllostomus discolor from Chiapas, Guatemala, Nica-
ragua, and Trinidad, nor in 12 specimens of P. elongata from Peru, nor were

* Present address: Systems Analysis and Design, Grumman Aerospace Corporation, Beth-
page, New York 11714.

792 The University Science Bulletin

carious lesions found in individuals of other phyllostomatine genera studied
(Micronycteris, Macrotus, Chrotopterus ; Mimon, Tonatia, Trachops, Lon-
chorhina, and Macrophyllum) . Furthermore, among 1500 specimens exam-
ined of species in the families Emballonuridae, Noctilionidae, and Chilonyc-
teridae, but one bat was found with dental caries (Phillips and Jones, 1969).

Specimens of Phyllostowus were examined grossly for carious lesions,
which, when discovered, were recorded as to site and severity. External
morphological features of the teeth visible with a binocular dissecting scope
were correlated with the occurrence of lesions. For comparative histological
studies and further analysis of possible endogenous factors, teeth of several
P. hastatus and P. discolor were demineralized in 5% nitric acid (48 hours),
embedded in tissuemat and piccolyte at 56.5° C, sectioned at 9 ^-, and stained
with Mallory's triple. The material used for histological examination had
been preserved in 10% formalin and stored in 70% alcohol. The teeth of
P. hastatus and P. discolor were compared with regard to all visible surface
characteristics as well as to histological features.

Dental caries in P. hastatus often begin in the third upper and lower
molars and progress anteriorly. We base this statement on the fact that
carious lesions were present and generally more advanced in these teeth in
the 13 diseased individuals in which there were extensive caries in the molars
(Fig. 1). Dental caries are of the acute type in P. hastatus, which results in
the complete destruction of the crown and cervical region of diseased teeth.
Following initial penetration of the enamel, the disease destroys the dentin
leaving an enamel shell that is easily fractured and exfoliated (Fig. 1).
Diseased teeth respond to infection in essentially the same ways as do
human teeth: in some instances, there is extensive resorption of the ce-
mentum layer and dentin; in other instances, new layers of cement are
added to the apical portion of the root of the traumatized tooth (Fig. 2).

The third molars of P. hastatus differ from the other cheekteeth in being
the last to develop and attain the functional eruptive stage and in being the
smallest. Furthermore, the labial surface (parastyle) of the upper third
molar is covered by a thick ridge of tissue. This tissue covering varies in-
dividually but always is well developed in specimens having dental caries.
The ridge of tissue begins on the labial surface of the mandibular epithelium
and ends on the surface of the maxilla at the corner of the mouth. Conse-
quently, a flap covers the labial coronal and cervical surfaces of the third
upper molars. Debris accumulate (as can be seen in specimens preserved in
alcohol) in the pocket between the surface of the tooth and the flap of
tissue. Presumably, such an environment would provide an excellent sub-
strate for bacterial infection. In Phyllostomus discolor the ridge and flap of
tissue are either absent or are so reduced that the surface of the third molar
is free and therefore subject to self-cleaning.

Dental Abnormalities in North American Bats

793

.- c
o w

> -t:

c .„ o

zr w u
3 "n 3

2, C C

o C —

7" 3 c

u .2Pm-(

^3 5 —

C _

S a O
■a « E

3 C O
U u x

■B 8 &

U

c
73 e C

— 0

u

8- °

E 5 «

= 1 -

W C o

"Co

W 3

I*

*-J y:

3 U

794

The University Science Bulleti

N

wm i

Fig. 2. Longitudinal section of carious (and now nonfunctional) upper premolar in P.
hastatus. New layers of cement (c) have been added to the root of the traumatized tooth and
principal fiber bundles of the periodontal ligaments (p) have been destroyed. Other lower case
letters denote the alveolar hone (a) and gingiva (g). Stained with Mallory's triple.

Carious lesions in teeth of P. hastatus are found at certain sites, which
are listed here in approximate order of importance based on incidence and
severity of lesions: (1) the labial and posterior surface of the third upper
molar; (2) the internal margin (entocristid to entoconid) of the talonid of
the third lower molar; (3) the buccal pit anterior to the hypocone on the
first and second upper molars; (4) the depression in the cingular shelf
posterior to the metacone on the first and second upper molars, and (5) the
site of intersection of fissures at the base of the metacone in these teeth;
(6) the slight depression on the posterior medial surface of the upper
incisors; (7) the depression on the surface of the cingular shelf of the upper
canines; and (8) the crevice between the entoconid of one lower molar and
the paraconid of the next. Generally, the sites of the most extensive carious
lesions are characterized by a depression in the enamel and therefore are
places where the surface is not subject to self-cleaning.

The enamel of teeth of P. hastatus typically is scored with stained fissures
(Figs. 1, 3). The edges of some fissures can be detected with a sharp probe.

Dental Abnormalities in North American Bats 795

but we did not consider them carious unless we could detect a definite
lesion in the enamel. Frequently, several fissures intersect in the same pit
or depression. Such combinations of fissures and depressions provide an
ideal site for penetration of the enamel (Fig. 3). This is especially true
because the fissures extend through the enamel and into the dentin, appear-
ing under the microscope as V-shaped wedges that can be traced along the
surface of a tooth, at the dentino-enamel junction, in serial sections (Fig. 4).
No relationship was found in P. hastatus between amount of wear on teeth
and numbers of stained fissures. Some individuals with numerous stained
fissures had only slight to moderate wear and others, having almost no
fissures, also had moderately worn teeth.

The teeth of Phyllostomus discolor and P. elongate! are strikingly different
morphologically from those of /'. hastatus at the very sites where most cari-
ous lesions occur in the latter. The most extreme modification is seen in
P. discolor, which lacks the depressions that commonly are sites of infection
in P. hastatus. It is of additional interest that fissure lines seldom are found
on teeth of P. discolor and P. elongata and, when present, only rarely exceed
one fissue per tooth. The teeth ol P. hastatus arc, overall, considerably more
primitive than those of either /\ discolor or /'. elongata. Furthermore, the
later two species differ considerably in size from hastatus in that they are
much smaller.

In summary, carious lesions on the teeth of /'. hastatus can be correlated
with three endogenous factors of apparent cariogenic importance: (1) de-
pressions on the surface of certain teeth, (2) numerous fissures in the enamel
of teeth, often extending into the dentin, and (3) a flap of tissue that covers
the labial coronal and cervical surfaces oi the third upper molars. We are,
of course, unable to judge the effect oi the exogenous factors and the oral
environment. The endogenous factors found in /'. hastatus either are absent
or are present to a much lesser degree in /'. discolor and P. elongata. This
and the apparent lack of dental caries in the latter two species does not
prove, however, that P. discolor and P. elongata are somehow resistant to
caries, because apparent resistance can result from the partial or total lack
of exogenous cariogenic factors rather than the absence of endogenous factors
(Dirks, 1%5; Darling, 1965). In this regard, it is especially interesting that
males of P. hastatus have a higher incidence of the disease than do females.
We presently have no explanation of this phenomenon because we were
unable to find any morphological difference in the teeth between the sexes.
This discrepancy may be due to our small sample sizes.

The relationship between the occurrence of carious lesions and fissures
and depressions in the enamel has been studied both in man and in labora-
tory rodents. Hunt and Hoppert (1950), for example, correlated broad fis-
sures in teeth of rats with known susceptibility to caries; the broad fissures

796

The University Science Bulletin

1mm

ta *>%

V \

JBKEM

to

Fig. 3. Second upper molar of P. hastatus, showing exposed dentin in relatively snull
carious lemons (a anil c), stained fissure lines extending through enamel (b), and stained fissure
lines intersecting on the surface of the tooth (d). The metacone (not in focus) is labeled (e)
for reference.

were thought to allow a greater chance for frequent impaction of food. In
man, according to Dirks (1%5), resistance to caries in individual teeth is
lowest in pits and fissures and highest on free, smooth surfaces. This, of
course, agrees with our conclusions about P. hastatus. Some workers (for
example, Brucker, 1944) have debated the importance of fissures, but their
arguments probably do not apply here because most were concerned with
the problem of clinical diagnosis in man. The cause of the fissures that can
be seen on teeth of P. hastatus is as yet unknown. It is possible that struc-
tural deficiencies in the dentin or enamel, or incomplete mineralization,
allow for fissures to form along lines of stress. Because number of fissures
does not appear related to wear, or to sites of surface depressions, it is
doubtful that occlusion alone is responsible for the production of these
barely macroscopic cracks (Fig. 3).

Feeding habits sometimes can be correlated with the incidence of dental
caries. Hall (1940), for example, thought that diet played an important
role in the production of caries in the teeth of bears. The dietary factor in

Dental Abnormalities in North American Bats

797

o

c ^

C cd

^^

« y.

*H

O 3

fci-S

H ~

*->

<— _

w cfl

'M

c

"3 l>

u — :

_^

>- ^-v

5 ~

u

c ^

3,

'-C

V.

c i

:/;

i-H

rt

~

-. r™

e fe

s

c

J3

m_

—

~ 5

S

•— > l—

c

X 3

V s

-

y.

w '4=

y.

5 g1

S.

a

!'!

u

u

" -£

;/:

3 "S

c ,

a b

<

■a £

c_ o

s -

a

C u.

c

u u

OJ

_ J=

~

g bt

»■* • —

V

-5

«u

c

. 1)

E/3

3>

w

« .

i-

^s

—

■<;

rs

^

~- c

~

u-l "S

U

Id

798 The University Science Bulletin

dental caries in Phyllostomus hastatus is unknown, but it might prove note-
worthy that this species is known to feed on certain small vertebrates as well
as on insects and fruit, whereas P. discolor apparently feeds on fruit and
possibly also insects (Goodwin, 1946:307; Goodwin and Greenhall, 1961:238;
Arata et al., 1967). It is possible that mastication of certain kinds of food is
one factor related to the formation of fissures in the enamel of teeth of
P. hastatus. Dunn (1933) reported that captive spear-nosed bats easily
crushed the bones of mice and small bats, which were readily eaten and
possibly preferred to fruits.

Although all of the endogenous and exogenous cariogenic factors cer-
tainly cannot be determined at this time, it is noteworthy that acute dental
caries occur commonly in P. hastatus. The occurrence of the disease and
the phenotypic dental characteristics with which it is associated in this species
suggest speculation in terms of their evolutionary significance. Two possible
interpretations of the occurrence of caries in P. hastatus are: (1) one or
more of the genetic factors that result in a high degree of acute dental caries
are associated with some characteristic (s), for example large size, that is
of significant survival value; (2) some exogenous factors, such as a recent (in
an evolutionary sense) shift in food habits, are of great cariogenic import,
and selection has not yet produced a phenotype capable of ameliorating
these environment-produced conditions, or that the adaptive value of exploi-
tation of a new food source is of far greater benefit to the species than
prevention of dental caries. It might be argued, of course, that dental caries
are not particularly important to the survival of an individual bat, but this
seems unlikely when one considers that in severe cases of the disease entire
teeth are lost and the gingivae undoubtedly are infected and inflamed.
Future ecological and laboratory studies of all species of the genus Phyllo-
stomus could shed additional light on the reasons for a high incidence of
dental caries in P. hastatus.

Financial support for this study was from a contract (DA-49-193-MD-2215)
with the United States Army Medical Research and Development Command and
a grant (3453-5038) from The University of Kansas General Research Fund.
Richard G. Van Gelder and Karl F. Koopman, The American Museum of
Natural History, kindly loaned us specimens of Phyllostomus elongata and
fames W. Bee and G. Lawrence Forman, The University of Kansas, helped us
with certain aspects of histological studies.

Dental Abnormalities in North American Bats 799

LITERATURE CITED

Arata, A. A., J. B. Vaughn and M. E. Thomas. 1967. Food habits of certain Colombian bats.
J. Mamm. 48:653-655.

Brucker, M. 1944. Studies on the incidence and cause of dental defects in children. VII.
Fissures and canes. J. Dent. Res. 23:101-105.

Colyer, F. 1936. Variations and Diseases of the Teeth of Animals. John Hale, Sons and
Daniels' on Ltd., London, viii -4- 750 pp.

Darling, A. I. 1965. The physical features of caries-resistant teeth. In Caries-Resistant Teeth
(G. Wolstenholm and M. O'Connor, eds.), pp. 149-161. Little, Brown and Company,
Boston.

Dirks, O. B. 1965. The distribution of caries resistance in relation to tooth surfaces. In Caries-
Resistant Teeth (G. Wolstenholme and M. O'Connor, eds.), pp. 66-83. Little, Brown
and Company, Boston.

Dunn, L. H. 1933. Observations on the carnivorous habits of the spear-nosed bat, Phyllostomus
hastatus panamensis Allen, in Panama. J. Mamm. 14:188-199.

Goodwin, G. G. 1946. Mammals of Costa Rica. Bull. Amer. Mus. Nat. Hist. 87:271-474.

— , anil A. M. Greenhall. 1961. A review of the bats of Trinidad and Tobago: Descrip-
tions, rabies infection, and ecology. Bull. Amer. Mus. Nat. Hist. 122:191-301.

Hall, E. R. 1940. Supernumerary and missing teeth in wild mammals of the orders In-
sectivora and Carnivora, with some notes on disease. J. Dent. Res. 19:103-119.

. 1945. Dental canes in wild bears. Trans. Kansas Acad. Sci. 48:79-84.

Hunt, H. R., and C. A. Hoppert. 1950. The distribution of cavities in the lower molars ol
caries-susceptible and caries-resistant rats. |. Dent. Res. 2l':157-164.

Keyes, P. H., and H. V. Jordan. 1963. Factors influencing the initiation, transmission, and
inhibition of dental caries. In Mechanisms of Hard Tissue Destruction (R. F. Sognnaes,
ed.), pp. 261-283. Publ. 75, Amer. Assoc. Advancement Sci., Washington, D.C.

Konig, K. G. 1965. Caries resistance in experimental animals. In Caries-Resistant Teeth (C.
Wolstenholme and M. O'Connor, eds.), pp. 87-106. Little, Brown and Company,
Boston.

Phillips, C. }., and J. K. foNES, Jr. 1969. Dental abnormalities in North American bats.
I. Emballonuridae, Nbctilionidae, and Chilonycteridae. Trans. Kansas Acad. Sci.
71:509-520.

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLVIII Pages 801-844 May 1, 1970 No. 21

The Skeleto-Muscular System of Mecoptera: The Thorax1,

H. Randolph Hepburn3

ABSTRACT

This is the second of a series of morphological studies on the order Mecoptera,
and treats the thoracic skeleton and musculature. The cervix is well developed
and is unique among all orders in the possession of musculated dorsocervicalia.
The prothorax generally tends toward reduction and fusion of sclerites, especially
in the lateroventral field; however, the prothoracic musculature is fairly constant
despite skeletal variations. The pterothoracic segments are extremely generalized
in hoth sclerite arrangement and in musculature but certain modifications do
accompany hrachyptery and aptery. The tendency toward involution of the
sternum resulting in cryptosterny reaches an extreme condition in Mecoptera,
where there is no longer any external evidence of a sternum. The normal articu-
lations of the legs with the thorax are more restricted in Mecoptera than in other
orders; i.e., the trochantinocoxal articulation is absent and there is no involvement
of the meron in the coxopleural articulation of the mesothoracic leg. The
pterothoracic musculature is constant in most cases but some variation does occur
in the pleural series. With the exception of raptorial modifications in one family,
the legs of Mecoptera are extremely generalized.

INTRODUCTION

This investigation of the thorax is one of a series of morphological studies
on the order Mecoptera (Part 1, Hepburn, 1969). Two priorities underlie
the approach used in this paper. The first is that a basic knowledge of the
thoracic skeleton and musculature was needed for the order as a whole. The
second is that once such information was obtained, a more critical study
could then be made of the Panorpoid Complex. Only the first of these prob-
lems is treated here, except for a few comments on the musculature with
reference to other orders.

1 Contribution number 1436 from the Department of Entomology, University of Kansas,
Lawrence, Kansas.

2 A portion of this work was supported b\ NSF Grant GB-7045X to Dr. George W. Byers,
Dept. of Entomology, The University of Kansas.

3 Present address: Dept. of Zoology, The University of Bristol, Bristol, England BS-8.

802 The University Science Bulletin

The thoracic morphology of Mecoptera has received little notice in the
past; however, the few papers which do treat the order are outstanding. The
principal pertinent work was that of Hasken (1939) who discussed the tho-
racic skeleton and musculature of Panorpidae in detail. Issiki (1933) studied
the exoskeleton of Panorpidae and Panorpodidae; Maki (1938), the mus-
culature of Panorpidae; Ferris and Rees (1939), the skeleton of Panorpidae.
More recent work includes excellent papers on Boreidae (Fuller, 1954, 1955),
the exoskeleton of Meropeidae (Mickoleit, 1967), and the pleurosternal
musculature of the Panorpoid Complex (Mickoleit, 1969).

Historically, the paper by Hasken and that of Ferris and Rees assume
importance in morphology beyond the scope of Mecoptera anatomy. These
authors, especially the latter, were among the first to interpret the thoracic
sternum of Mecoptera in terms of the cryptosterny thesis of Weber. This
view, which has been adopted in this paper, has been further advanced by
Matsuda (I960, 1963a, b) in his comparative study of the insect pterothorax.

Other problems related to the thoracic skeleton include structural dis-
crepencies between the prothoracic and pterothoracic pleural regions, the
modifications of this region through reduction to brachyptery and aptery in
Boreidae and Apteropanorpidae, and the nature of the coxal articulations.
The role of the mesothoracic meron in the coxopleural articulation is unique
in Mecoptera. The thoracic tergum is straightforward and offered no special
problems.

The muscular system ordinarily would have included the problem of
homologizing the muscles of the three segments but this is worked out by
Hasken (1939). Hasken's work is used as a standard of comparison for all
of the other families. The loss of muscles in the flightless forms presents no
special problems of interpretation. The thoracic musculature of Mecoptera
is in some instances compared with that of other orders of the Panorpoid
Complex by reference to the tables given by Maki (1938).

Terminology remains a problem central to insect morphology and one is
caught between following an existing nomenclature with faults as opposed
to modifying it or coining a new one. The former course is taken in this
paper. Two different systems of terminology are used in this paper. The
terminology applied to the skeletal region is based on Snodgrass (1927, 1935)
with refinements by Matsuda (I960). The terminology applied to the muscu-
lature is that of the "European system" used by Hasken (1939) and Fuller
(1955) in which more emphasis is placed on a consistent classical name than
implying functional significance. This system was followed here both be-
cause it facilitates an easy comparison of all of the literature on the thoracic
musculature of Mecoptera and because this study is not a functional analysis
of the musculature.

The Skeleto-Muscular System of Mecoptera: The Thorax 803

ACKNOWLEDGMENTS

I would like to credit Dr. George W. Byers for his interest and support of this
project, Dr. C. D. Michener and R. W. Lichtwardt, all of the University of
Kansas for a review of the manuscript; Dr. C. O. O'Brien and Dr. L. B. O'Brien
for collecting much of the material used in this study.

METHODS AND MATERIALS

The specimens employed in this study are representative of all nine families of Mecoptera
currently recognized in the classification of Byers (1965). The justification for the selection is
expressed in the first part of this study (Hepburn, 1969) on the skeleto-muscular system of the
head. The species used and their classification are as follows:

Panorpidae Apteropanorpidae

Panorpa nuptialis Apteropanorpa tasmanica

Panorpa communis (from the literature) Choristidae

V a nor pa lugubris Chorista australis

Vanorpa sibirica Tacniochorisia pallida (from the literature)

Ncopanorpa harmandi Meropeidae
Bittacidae Metope tuber

Bit ta< us chlorostigma Borcidae

Bittaats italic ns (from the literature) Bonus unicolor

Bittacus pilocoinis (from the literature) Boreas calijornicus

Bittacus strigosus (from the literature) Nannochoristidae

Aptcrobittacus apterus Kannochorista diptcroides

Harpobittacus australis Nannochorista maculipennis

Harpobittaais tillyardi Notiothaumidae

Pazius obtusus Notiothauma reedi

Anabittacus iridipennis Panorpodidae

Kalobittacus microcerus Panorpodes paradoxa (from the literature)

Austrobittacus anomaltis Brae hypanorpa carolintnsis

Neobittactis blancheti

Fluid-preserved specimens (70% ethanol, Bouin's soln., Kahle's soln., or Dietrich's soln.)
were dissected in 70% ethanol, with razor blade fragments and jewelers' forceps, under a
standard dissecting microscope. Occasionally the tissues required histological staining to accen-
tuate certain features of skeleton and musculature. After initial dissection, specimens were
placed in a 0.5% solution of Congo red and 70% ethanol for 30 minutes, washed in 70%
ethanol and then examined. The residual effect is that the stain partially washes out of the
muscles but is retained by the endoskeleton. This greatly facilitated locating origins and in-
sertions of muscles. A solution of 0.5% methylene blue in 70% ethanol was also used lor
high-lighting muscles, but this stain has no residual effect.

The external skeletal morphology was examined in several ways. In some instances pinned
specimens were sufficient (and in some cases were the only ones available) to yield gross infor-
mation. Fine details of external structure and the details of internal skeletal structure were
usually elucidated by boiling the specimen in a 10% KOH solution and washing in 70%,
ethanol. After such treatment the specimens are often too hyaline to discern fine elements, and
the specimen must then be stained. Again, Congo red proved useful. If a specimen is first
stained in Congo red, washed, and then boiled in a 10% KOH solution, the soft tissues are
rendered while the cuticle retains the stain perfectly. In those specimens which were not boiled
but simply stained for muscle work, the residual stain was particularly concentrated along
apodemes and apophyses, presumably owing to chemical differences in these areas. No attempt
was made to identify the cuticular layers, but this reaction is useful in following apodemes.

Where possible, drav/ings were made after the examination of several individuals to avoid
being misled by teratological specimens. The illustrations were made by use of an ocular grid
and grid paper.

804

The University Science Bulletin

LIST OF ABBREVIATIONS

a

apodeme

lpn

lateropostnotum

aa

anterior apodeme

m

meron

ab

abdomen

pa

pleural apophysis

ac

anapleural cleft

pea

pleuro-coxal articulation

acs

antecostal suture

pes

precoxal suture

aes

anepisternum

pes

preepisternum

ar

arolium

ph

phragma

as

anapleural suture

pla

pleural apodeme

awp

anterior notal wing process

pin

planta

ax

auxillia

poa

posterior apodeme

b

basalare

pon

postnotum

bas

basisternum

ps

pleural suture

be

basicosta

Pt

pretarsus

bs

bsasalar suture

pwp

posterior notal wing process

dc

dorsocervicale

s

spiracle

ec

eucoxa

sb

subalare

em

epimeron

sea

sterno-coxal articulation

ep

episternum

scl

scutellum

fa

furcal arm

set

scutum

fs

furcasternum

si

scutal lobe

fu

furca

sr

sternal ridge

ic

intrasegmental conjunctiva

tn

trochantin

ks

katepisternum

trt

trochanteral tendon

ke

laterocervicale

ts4 . .

s tarsomere

lec

lumen of eucoxa

u

unguis

up

unguitractor plate

CERVIX

There have been only two comparative treatments of the insect cervix,
those by Verhoefr (1903) and by Crampton (1926). Since the ontogenetic
development of the cervix has yet to be studied in detail, it is not known to
what extent cephalic and thoracic elements contribute to its formation in most
instances. The cervix is included with the thorax in this paper for two
reasons, tradition and the fact that several cervico-thoracic muscles point to a
more intimate relationship with the thorax than with the head. Nevertheless,
the inclusion of the cervix with a discussion of the thorax should be taken
as arbitrary and not in any way as an espousal of the old "microthorax" theory
(Snodgrass, 1909).

The cervix or neck of Mecoptera is, as in other insects, a mostly mem-
branous area between and joining the head and prothorax. In most insects it
is a small and inconspicuous region, but in Mecoptera it is quite extensive.
It includes a variable number of sclerites: a pair of usually lateral laterocer-
vicalia (lc, Figs. 5, 7-9, 19, 22) and occasionally a pair of dorsal dorso-
cervicalia (dc, Fig. 24). A pleural element, the trochantin, lies within the
ventral cervical field, but it is not functionally related to the other sclerites of
the neck. Not only is the structure of each of these elements variable, but the

The Skeleto-Muscular System of Mecoptera: The Thorax 805

entire complex bears a close relationship to the prothorax and varies with
prothoracic development. Each of these elements is considered separately.

The degree to which the dorsum of the cervix is membranous varies with
the anterior extent of the pronotum. In Meropeidae, Boreidae, and Bittacidae
the anterior edge of the pronotum extends to and covers the base of the head.
The Panorpidae and Apterpanorpidae exhibit the most extensive cervicoria,
hence the smallest relative pronotal size. The remaining families are inter-
mediate. Considering the first case (the first three families mentioned), a
consequence of pronotal enlargement has been a lateroventral displacement
of the laterocervicalia: a) in Meropeidae the laterocervicalia have become en-
tirely ventral and extend parallel to the longitudinal axis of the body
(Mickoleit, 1967) ; b) Boreidae are the same but, in addition, the pronotum
has extended laterally and anteriorly such that the posterior bases of the
laterocervicalia have been shifted forward, with the result that they lie in a
dorsoventral plane perpendicular to the long axis of the body (Fuller, 1954) ;
c) in Bittacidae the laterocervicalia are in the generalized position but are not
externally visible because they are obscured by the lateral swelling of the
pronotum.

In the intermediate forms (Notiothaumidae, Fig. 24; Panorpodidae, Fig.
19 and Issiki, 1933; Choristidae, Fig. 23; and Nannochoristidae, Crampton,
1926) the relationship of pronotum and laterocervicalia is the usual one, the
laterocervicalia extending obliquely from prothorax to postocciput in a
longitudinal direction (with the exception of Nannochoristidae in which
they extend dorsoventrally). In Panorpidae (Hasken, 1939) the Aptero-
panorpidae (Fig. 22), the pronotum is reduced and the laterocervicalia occupy
a generalized position.

Laterocervicalia

The most persistent cervical sclerties of insects generally, and more spe-
cifically of Mecoptera, are the laterocervicalia. In general, the laterocervicale
is a long, narrow sclerotic process which articulates anteriorly with the head
at the occipital condyle and posteriorly with the propleuron. The latero-
cervicale is incised by a suture for most of its length. Internally a well de-
veloped apodeme is formed along the suture and serves as a site for muscle
attachment. Although the anterior articulation with the occipital condyle is
constant, the relationship of the laterocervicale with the propleuron is variable.
In the majority of families the laterocervicale is produced into a process
which articulates with the propleuron (usually the proepisternum, but in
some instances the propleurites are fused and so reduced that it it not possible
to distinguish which propleurite is involved). The exceptions are the
Notiothaumidae and Panorpidae, in which the laterocervicale is basally fused
with the propleuron (Fig.2) and the Bittacidae in which the single propleural
sclerite is longitudinally fused with the long axis of the laterocervicale.

806 The University Science Bulletin

DoRSOCERVICALIA

In addition to the laterocervicalia, some families of Mecoptera possess a
pair of transverse, oblong dorsal sclerites of the cervix situated immediately
behind the postocciput, the dorsocervicalia (dc). These sclerites are present
in Panorpidae, Boreidae, Notiothaumidae, and Panorpodidae (only in the
Asian Panorpodes). Their occurrence in Mecoptera is unusual in that they
serve as sites for muscle attachment, which is not typical of these sclerites in
other insects (Snodgrass, 1935).

In those Mecoptera with dorsocervicalia, the dorsal longitudinal muscles
extend from the first phragma and terminate on the posterior edges of the
dorsocervicalia, while those familes lacking dorsocervicalia have the anterior
termination of these muscles on the postoccipital ridge. There are several
explanations for these differences, the most plausible of which is that the
dorsocervicalia arose by fragmentation from the original postocciput. Al-
though there is no available evidence from development studies, this view
appears more acceptable than a tie novo origin of dorsocervicalia.

Crampton (1926) adequately discusses the skeletal anatomy of the
mecopteran cervix as it relates to the other orders of the Panorpoid Complex;
he illustrates several examples of the mecopteran cervix. Issiki (1933) and
Hasken (1939) discuss Panorpidae. Ferris and Rees (1939) in their study of
Panorpidae incorrectly suggest that the dorsocervicalia are unique to
Mecoptera.

Cervical Musculature

The musculature of the mecopteran cervix was first elucidated by Hasken
(1939) for Panorpidae; this work is used as a standard of comparison for the
musculature of the cervix and thorax of other Mecoptera. The presence or
absence of muscles is tabulated by families (Tab. 1), although variations are
discussed separately.

Intersegmental Muscles

longitudinal muscles (Fig. 2)

Odlmi. Anterior side of phragma 1 at the midline to the dorsocervicale
or the postoccipital ridge. In those families with dorsocervicalia, the insertion
is on the posterior side of the sclerite; in those lacking dorsocervicalia, the
insertion is on the postoccipital ridge. This muscle elevates the head.

Odlmi. Middle of the pronotum to the dorsocervicale or occipital ridge;
retracts the cervicorium, and the resulting creasing facilities the antero-
posterior movement of the head.

Ovlm\. Anterior side of the prof urea near its middle to a small tendon on
the lateral arm of the tentorium; depressoi of the head in conjunction with
the next muscle.

The Skeleto-Muscular System of Mecoptera: The Thorax 807

Table 1. Cervical Musculature.

<

Q

P

UJ

<

Q

<
Q

ft.

PJ
<

C

H

5
o

o

X

o
o
z
z

<
Q

O
<

<

Q

Q

O
ft.
oi

O

z

Id
<

Q

5
O

z

<

D

H

U3
<

Q

s

ft,

o

<
DC
h
O

O

z

<

O
tA
W

X

<

H

<

<

g

UJ

0

ft,

'J

z

m

ft.

—

2

^

<

Odlmi

+

+

+

+

+

+

+

+

+

Odlm2

—

3

—

—

—

+

—

+

—

Ovlm-i

+

+

+

+

+

+

+

+

+

Ovlm2

+

? +

+

+

+

+

+

+

+

Oism i

+

! —

—

+

+

+

—

+

+

Oism2

+

' +

+

+

+

+

—

+

+

Oism.,

+

J

+

+

+

+

—

+

+

Oism i

+

+

+

+

+

—

—

+

+

Oisnh,

+

5 ?

+

+

+

+

—

+

+

Odvm-i

+

+

+

+

+

+

+

+

+

Odvm2

+

+

+

+

+

+

+

+

+

Ovlmi. Immediately ventrad of Ovlmi and has the same insertion and
function as Ovlmi.

OBLIQUE MUSCLES (Fig. 2)

Oismi. Junction of the profurca and the propleuron to a short tendon
either on the lateral edge of the dorsocervicale or on the postoccipital ridge
near the insertion of Odlnn ; elevates the head.

Oismi. Laterally on the anterior pronotal furrow to the dorsal side of the
laterocervicale; elevates the laterocervicale. Indirectly effects movement of
the head.

Oisni:-,. Laterally on the base of phragma 1 to the laterocervicale between
the insertions of Oismi' and Oisiru; with Oism., it elevates the anterior end of
the laterocervicale.

Oism\. Anteroventrally of Oismi to a short tendon on the laterocervicale;
elevates the laterocervicale.

Oisrrio. Basicosta of the procoxa to the anteroventral end of the latero-
cervicale; indirectly elevates the head.

Segmental Muscles

DORSOVENTRAL MUSCLES (Fig. 2)

Odvmx. Dorsal side of the laterocervicale to either the dorsocervicale or
the postoccipital ridge; elevates the head.

808 The University Science Bulletin

Odvm-i. Immediately behind Odvmi on the dorsal side of the latero-
cervicale to either the posterior end of the dorsocervicale or the postoccipital
ridge; elevates the head.

PROTHORAX: Pronotum

The pronotum of Mecoptera is in general a heavily built sclerite that is
extensive both dorsally and caudoventrally. Anteriorly it may overlap the
vertex (see cervix) and laterally the propleurites. It is produced caudoven-
trally into a tapered process which is connected by a hinge to the mesepister-
num at its apex. Very rarely the process is continuous with the mesepisternal
spur; i.e., in Panorpa stbirica as reported by Issiki (1933) and confirmed by
me. Thus Hasken's criticism (1939) of Issiki on this point is unjustifiable.

The pronotum is creased by longitudinal and transverse furrows, some of
which internally produce apodemes for muscle attachment. Since these fur-
rows are not sutures in the classical sense (no apodemes), Miyake's attempt
(1913) to homologize the resulting subdivided areas with pterothoracic
sclerites related to flight is unacceptable, as pointed out by Hasken (1939) and
Snodgrass (1935). The first furrow simply traverses the pronotum; the
second sets off a protuberance on either side of the midline; the third begins
on the protuberances and extends into the caudoventral process at the
mesepisternum, serving internally as a site for muscle attachment {Odlm^
and Oism-z). Likewise, the first furrow bears an apodeme for muscle attach-
ment (ldlm ). This constitutes the generalized case in Mecoptera to which the
majority of families correspond.

The development of the lateral process of the pronotum (aw, Figs. 19-21,
23, 24) varies from a well produced arm in Notiothaumidae to a very
reduced one in Apteropanorpidae, Boreidae, and Bittacidae, the other fami-
lies being intermediate. However, the internal apophysis of this arm is ths
site of attachment of the Idvm muscle in all families of Mecoptera. The in-
ternal development of the furrows is also variable. In some families (Aptero-
panorpidae, Boreidae, and Bittacidae) there are only slight indications of
apodemes; in others (Notiothaumidae) a well developed apodemal ridge
extends from one side of the pronotum to the other. The number of pronotal
apodemes varies from 0 to 2. Apteropanorpidae are unique in having a single
external furrow and no apodemes; in addition, the first thoracic spiracle is
completely imbedded in the posterior part of the pronotum and not in the
membranous region behind the lateral process (aw) of the pronotum (also in
Boreidae). Besides transverse furrows, the Meropeidae possess a longitudinal
apodeme and a short process in the lateral wall of the pronotum, both parallel
to the long axis of the body (Mickoleit, 1967). Since there is no consistent
plan of development in the prothorax of insects (Snodgrass, 1935) it is diffi-

The Skeleto-Muscular System of Mecoptera: The Thorax 809

cult to make phylogenetic inferences here, and by extension difficult to assess
variation found in the pronota of Mecoptera.

PROTHORAX: Propleuron

The propleuron of Mecoptera contains both of the sclerites usually associ-
ated with this region — the proepisternum and proepimeron. Externally the
propleuron is situated below the pronotum and dorsal to the coxal bases.
Because of the extensive cervicoria, the propleuron is posteriorly displaced
such that a portion of it is concealed by the cervical membrane and the lateral
extension of the pronotum. At the anterior end of the propleuron there is a
transverse apodemal process which extends to the furca. Internally the
proepisternum extends caudally (though externally obscured by cervicorium
and pronotum) to the furca.

The proepimeron is situated immediately behind the proepisternum and
is extremely reduced. However, the two propleurites are separated from each
other by a suture (probably corresponding to the pleural suture) at the ventral
end of which is an apophysis. The junction of propleuron and furca is repre-
sented by a fusion of proepimeron, profurca, and pleural apodeme, and only
a vestige of the proepimeron is recognizable in most cases. That the proepi-
meron is present as part of the fusion product is indicated by the attachment
of the dorsoventral intersegmental muscle (Iism).

The proepisternum articulates anteroventrally with the laterocervicale and
posteroventrally with the base of the coxa which comes to a point.

The epimeron continues ventrally into the depression of the furca, pos-
terior to the episternum. Anteriorly there is a process of the furca which
serves for muscle attachment. The monocondylic articulation of the procoxa
with the proepisternum is unique and in contrast to the dicondylar state of
the pterothoracic coxae, in that the procoxal connection with the episternum
is immovable while those of the latter are not (Hasken, 1939).

As indicated by Snodgrass (1935), the propleurites of insects are always
separated into episternum and epimeron, though the latter tends toward re-
duction. This is true of Mecoptera and has been verified (various authors)
and confirmed by me. Crampton (1926) claimed that there is a tendency for
fusion of laterocervicalia and propleuron in Mecoptera based on an approxi-
mate condition in Choristidae and on an actual fusion in Panorpa higiibns
(Panorpidae). A close examination of Panorpidae (including P. higiibns)
does not support this claim. In fact, there is no fusion between the latero-
cervicalia and propleurites in any of the Mecoptera examined in this study.

The propleural elements anterior to the pleural suture are termed the
precoxale or precoxal bridge, an extremely reduced area in the prothorax of
Mecoptera. The only remnant of a coxolpleurite in Mecoptera is a very
reduced, isolated, ventral trochantin. In the lower Pterygota this sclerite is

810 The University Science Bulletin

derived from the primitive coxopleurite (Snodgrass, 1935) and usually articu-
lates with the coxa as well as serving as a site for muscle attachment of tergal
promotor muscles. In the higher Pterygota the trochantin becomes reduced,
as in all Mecoptera. The trochantin of the prothorax does not receive any
muscles nor does it articulate with the procoxa. A trochantin is present in
the prothorax of all Mecoptera, though Evans (1942) claimed the contrary.

Issiki (1933) adequately discussed the trochantin and interpreted it cor-
rectly in Panorpidae and Panorpodidae, as did Ferris and Rees (1939) in
Panorpidae. Other cited papers on Mecoptera make no mention of this
sclerite.

PROTHORAX: Prosternum

Discounting minor variations in shape and size, the prosternal field is
uniform in its constituents. The basic structure closely approximates that of
the generalized lower Pterygota (Snodgrass, 1935). The only sternal inter-
segmental element, the spinasternum, is intimately fused with the prosternum
and is externally visible as a reduced spike-like caudal process of the furca-
sternum (or sternellum). Internally it is produced into an apophysis to which
are attached the ventral longitudinal intersegmental muscles (lvltn) that ex-
tend from the mesothorax into the prothorax. Snodgrass (1935) reports that
there are occasionally two such spinasternal elements in insects; there is one
on either side in Mecoptera.

The sternal elements in the prothorax of Mecoptera include, in an antero-
posterior direction: a eusternum, apophyseal pits (or furcacavae), and a
spinasternum. There are no pleurosternites in the prothorax. The eusternum
is generally triangular to trapezoidal and is wedged between the bases of the
closely appressed coxal bases. Its anterior basisternum (bs) and posterior
furcasternum (fs) are externally divided by a Y-shaped sternal suture (in
most other insects this is a simple transverse suture). The furcal pits are
posterolateral on the eusternum and give rise to an extensive sternal apophysis
which forms the major endoskeletal element of the prothorax. The degree of
sternal involution has obliterated the sternacostal suture and sternacosta seen
in other insects. The position of the furcal pits partially delimits the basister-
num and furcasternum. The sternal apophyses are mesally fused and at the
level of the base of the procoxae they branch laterally into an anchor-shaped
process continuing to the pleuron, where they fuse with the pleural apophyses,
as well as serve as sites for muscle attachment (see propleuron).

Compared to generalized Pterygota (i.e., of the Orthopteroid level) the
Mecoptera have an extremely involuted prosternum that exhibits a trend
toward fusion of the elements into a single undifferentiated sclerite, as evi-
denced by the fusion of the spinasternum with the furcasternum, the appres-
sion of basisternites and furcasternites of either side into a single unit, lack of

The Skeleto-Muscular System of Mecoptera: The Thorax 811

a strenacostal suture and sternacosta, and the extreme development of the
endoskeletal furca within the prothorax. Inasmuch as the sternum of the
prothorax is homodynamous with that of the pterothorax, the trend is further
borne out in the latter in which sternal involution is complete and the
development of the furca reaches an extreme among insects.

Gross descriptions of the prosternum are available for Boreidae (Fuller,
1954, 1955), Meropeidae (Mickoleit, 1967), and Panorpidae (Hasken, 1939).
Although Crampton (1926) illustrated three families, his discussion is nebu-
lous. Issiki (1933) discussed the prosterna of Panorpidae and Panorpodidae.
My own findings and interpretations parallel his. The prothoracic leg is
discussed with the pterothoracic legs.

Prothoracic Musculature

The musculature of the prothorax was first worked out in Panorpidae by
Hasken (1939). As with the cervical muscles, those of the prothorax are
compared, with Hasken as a standard; the presence or absence of muscles is
listed in Table 2.

Intersegmental Muscles

LONGITUDINAL MUSCLES (FigS. 2, 25)

Ulhn. Mesonotum at the base ot phragma 1 to the apodeme of furrow 2 or
3 of the pronotum; effects a close appression of pronotum and mesonotum.

Ivlm. Posterior side of the profurca anteriorly and somewhat obliquely to
a long tendon on the anterior side of the mesofurca; effects appression of
pro- and mesosterna. This muscle is subdivided into lvlmi and lvlm-j in
Notiothaumidae but in no other families.

OBLIQUE MUSCLES

lism. A projection of the caudal aspect of the proepimeron to the meso-
notum behind Oism;.; maintains the position of the propleuron.

Segmental Muscles

DORSOYENTRAL muscles

It! cm. Posterolateral edge of the pronotum to the inner edge of the
merocosta. It is laterad of lism and posterior to the pleurocoxal articulation;
a remotor of the coxa.

INTRAFURCAL MUSCLES

lifitm. Between the arms of the furca; antagonistic to the strenal leg
muscles by drawing the furcal arms toward each other.

812

The University Science Bulletin

Table 2. Prothoracic musculature.

<

Q
P

w

w
<

Q

<
Q

<
Q

P

O
X
U

C/3

Pi

o
a:

u
o
z
z
<
z

<
Q

a
<

H
H

3

<
Q

3
o

Pi
O

z

<

<
g

5
o
z

<

<
Q

3

&
O

EG

<

0

w

&
0

w

D

<

O
P

o
z

5
o
z

<

0
b5

uq
H

<

kllm

+

+

+

+

+

+

+

+

+

lvlm

+

+

+

+

+

+

+

+

+

lism

+ •

+

—

+

+

+

+-?

+

_?

Idvm

+

+

+

+

+

+

+

+

+

1 ifum

+

1 —

+

+

+

+

+

p

+

1pm i

+

+

+

+

+

+

+

+

+

lpm2

+

+

+

+

+

+

+

+

+

1 pm:)

+

+ ?

+

+

+

+

+

+

+

1pm 4

+

+?

+

+

+

+

+

+

+

1 bmi

+

+

+

+

+

+

+

+

+

lbm2

+

+

+

+

+

+

+

+

+

1 bill:]

+

+

+

+

+

+

+

4-

+

lcxmi

+

+

+

+

+

+

1

+

+

lcxm^

+

+

+

+

+

+

+

+

+

lcxm:i

+

+

+

+

+

+

+

+

+

Icxm4

+

+

+

+

+

+

+

+

+

PLEURAL MUSCLES

lpnn. Dorsal side of the episternum and to the hasicosta between the
pleuro-coxal hingle and Oismr,; a coxal promotor.

lpm.2. Dorsal side of the episternum behind lpmi ventrally to the tro-
chanteral tendon where it inserts; a femoral abductor.

Ip/n-.u Anterolateral part of the pronotum to the dorsal side of the pro-
pleuron where it joins a process of the pleural apodeme; maintains the posi-
tion of the propleuron.

lpnn. Dorsal side of the episternum to the apodeme of furrow 2 or 3 of
the pronotum; in conjunction with lprri3 it maintains the position of the
pleuron with respect to the notum since they connect the two sclerites. They
are also antagonistic to lpm, and lpmi>.

Leg Muscles

STERNAL LEG MUSCLES

lbm\. Inside of the sternum to the posterior side of the hasicosta; a coxal
promotor.

Ibm-i. A projection of the anterior or ventral side of the profurca to the
trochanteral tendon; a femoral abductor.

The Skeleto-Muscular System of Mecoptera: The Thorax 813

Ibmz. Ventral side of the furca to the basicosta, being disposed between
the pleuro-coxal articulation and Ilcxmi; a coxal remotor.

COXAL MUSCLES

lcxm\. Anterior surface of the coxa to the proximal end of the trochanter
at the articulation between coxa and trochanter; a femoral adductor.

Icxmi. Basicosta to the lateral proximal edge of the trochanter posterior to
the termination of lcxiru; a femoral adductor.

lcxmz. Posterior side of the coxa to the lateral proximal edge of the
trochanter immediately posterior to lcxmj; a femoral adductor.

Icxnu. Medial edge of the coxa to the medial side of the trochanteral
tendon; a femoral abductor.

MESOTHORAX: Mesonotum

The mesothoracic tergum is of the generalized type described by Snod-
grass (1927, 1935, etc.). Ferris and Rees (1939) pointed out that the thoracic
terga of Panorpidae are of the textbook variety and offer no difficulty of
interpretation. As a generalization, this is true of all Mecoptera; however,
certain modifications do occur in those forms which approach aptery, in
which there is a tendency towards consolidation of sclerites into a single
indistinguishable plate (e.g., Apteropanorpidae, Fig. 22).

The mesonotum of Mecoptera is initially divisible into two plates, an
anterior alinotum and a posterior postnotum. The alinotum is further sub-
divided into two plates, a large anterior scutum (set) and a small posterior
scutellum (scl) by the scuto-scutellar suture (scs), which is internally pro-
duced into a reenforcing apodeme (the "V-formige Leiste" of German
authors, Figs. 1, 4, and 12). In most Mecoptera the alinotum and postnotum
(pon) are separated by a membranous area of secondary intrasegmental con-
junctiva (ic). In the higher Pterygota the tendency is for reduction of this
conjunctival region in connection with more efficient and complex flight
(Snodgrass, 1935), a modification possessed only by the Bittacidae among
Mecoptera (Fig. 12). There is no comparative information available on the
relative flight ability of various Mecoptera to assess the value of this modi-
fication, but Bittacidae appear to be the most accomplished flyers. This would
at least circumstantially support Snodgrass' contention. This principle is well
founded, however, with respect to insect flight in general (Pringle, 1957).

The anterior limit of the scutum is the antecostal suture (acs). This suture
is also an indication of an invaginated phragma of two lobes, one on either
side of the midline, to which the dorsal longitudinal muscles attach. Anterior
to the suture is a very short sclerite, the acrotergite (precosta of Issiki, 1933),
which is extremely reduced in Mecoptera and in other higher Pterygota
(Snodgrass, 1935). Separate prescutum and parapsidal furrows do not occur
in Mecoptera.

814 The University Science Bulletin

The convex scutum bears an anterior notal wing process or tergal lever
(awp) as a flattened flanged process on either side (Figs. 1, 4, 12). This
process is bilobed, with the anterior lobe slightly below the level of the pos-
terior lobe. The first axillary sclerite articulate between the lobes. Hasken
(1939) discussed these processes as distinct from each other but they collec-
tively function as a single notal process. At the posterior end of the meso-
scutum in most Mecoptera there is a slender, forked posterior notal wing
process (pwp). The wing is connected to and articulates with the notum
between these processes. The scutellum is variable in shape (oval to pen-
tagonal), but in all cases it is laterally drawn out into a thin process on either
side and is continuous with the axillary cord of the posterior edge of the wing.

The transverse postnotum is subdivided into a median portion (pon) and
a quadrangular lateropostnotum (lpn) on either side. Ventrally the latero-
postnotum ends on the dorsal edge of the epimeron. Posteriorly the post-
notum is delimited by a transverse suture (acs), the antecostal suture of the
next segment, that indicates the position of the next phragma and the end of
the mesonotum. Anteriorly the scutum is separated from the episternum by
a conjunctival area (Figs. 19-21, 23, 24).

This is a general account of the mesonotum of fully winged Mecoptera
(Panorpidae, Panorpodidae, Choristidae, Nannochoristidae, Notiothaumidae,
Meropeidae, and Bittacidae) and does not apply to brachypterous or apterous
forms. Minor variations as width/length and degree of convexity can be seen
by comparison of dorsal and lateral views illustrated (Figs. 4, 12, 19, 21, 23,
24). The Bittacidae are noteworthy in this group because 1) there are two
large anterior scutal lobes which receive the tergosternal muscles; 2) the
posterior notal wing process is simple, unforked; and 3) in some genera there
is no secondary conjunctiva between alinotum and postnotum (Fig. 12).

The modifications of the mesonotum in Boreidae are discussed in detail by
Fuller (1954). In addition to being brachypterous, males of Boreidae utilize
their wings as accessories in mating, the details of which were described by
Cooper (1940). In summary, the situation in Boreidae is that 1) the mem-
branous areas of the tergal region have become enlarged, 2) the scutum, post-
notum and lateropostnota are all reduced, as are the phragmata, 3) the scutum
is medially divided by the anterior extension of the scutellum which is rela-
tively large (i.e., the scuto-scutellar suture almost reaches the antecostal
suture), and 4) the postnotum is anteroventrally produced below the scutel-
lum. The net effect is that the notal sclerites are generally reduced and that
all of the mesonotal sclerties have become anteroventrally displaced. The
anterior and posterior notal wing processes are present but widely separated
and apparently functionless. There is a strong sexual dimorphism in the
mesonotum (Fuller, 1954).

The Skeleto-Muscular System of Mecoptera: The Thorax 815

The most extreme example of thoracic amalgamation and reduction is
represented by the wingless Apteropanorpidae (Fig. 22). In this family all
three of the thoracic nota and pleura are fused into a single saddle-shaped
unit that is traversed by furrows. There is no trace of either dorsal longi-
tudinal or tergal muscles so that it is difficult to interpret exactly what these
furrows (sutures?) really represent. In the mesothorax there is a single
transverse furrow which terminates at the coxal bases on either side. It is not
possible to assign the demarcated areas to either alinotal or postnotal regions.
There is no trace of a wing structure, associated sclerites, internal apophyses,
or phragmata. Internally the furrows are represented by fine trace apodemes
but there are no muscles associated with them.

In the two flightless bittacid genera, Apterobittacus (which is completely
apterous) and Anomalobittacus (which is brachypterous), the thoracic seg-
ments are clearly distinguishable from one another; distinct nota! and pleural
regions are visible as well (Fig. 21). While there is some reduction, there is
no fusion of the kind seen in Apteropanorpidae. There is no rearrangement
nor superposition of sclerites as in Boreus, only a loss of those elements directly
related to flight such as anterior and posterior wing processes and reduction
of phragmata. Likewise, there are no tergosternal or dorsal longitudinal
muscles in Apterobittacns (no specimens of Anomalobittacus were available
for dissection).

METATHORAX: Metanotum

The tergum of the metathorax is extremely similar in structure to the

mesothorax (Figs. 1, 4, 12). The metanotum is rigidly fused with the

mesopostnotum. The account given for the mesonotum can be directly
applied to the metathorax; differences are discussed below.

Panorpidae, Panorpodidae, and Choristidae possess a pair of lateral fur-
rows which arise on the anterior end of the metascutum. These have been
termed parapsidal furrows by Issiki (1933), with some reservation. Whether
they are would ordinarily have been a minor point, but a condition in Bit-
tacidae complicates the interpretation. There is the additional basic problem
as to what exactly are the parapsidal furrows, lines or sutures. There is no
general agreement on the use of this term and, when it is encountered,
extreme reservation should be exercised in attempting homologies. Those
familes which have "parapsidal furrows" have a corresponding internal, an-
teriorly directed apodeme which arches down to and is fused with the pre-
ceding phragma. The structural support provided here is obvious, especially
for a relatively weak and highly flexible cuticle as is found in winged
Mecoptera. There is no evidence of parapsidal furrows in the other families.

The Bittacidae possess a quadrangular anteromedial plate (Fig. 12, set)
on the metascutum which Setty (1940) termed the prescutum. Whereas this

816 The University Science Bulletin

designation may be correct, unequivocable evidence is not available to support
it. There are at least two possible explanations for this plate. First, it may
indeed be the prescutum; however, by definition (Snodgrass, 1935), the
prescutum is that area between antecostal and prescutal sutures. In Bittacidae
there is a single anterior transverse suture that by its connection with the
phragma is the definitive antecostal suture. The only other suture-like lines
are those longitudinal ones on either side of the plate. These lines could have
been regarded as the parapsidal furrows, such as are found in some other
families; however, they continue mesad posteriorly so that they become a
single U-shaped line. This condition could represent fused parapsidal fur-
rows (a condition otherwise unknown in insects) or the prescutal suture
(which would be unique to Bittacidae among the Mecoptera), a condition
approximated by the adephagous Coleoptera (Snodgrass, 1908). Neither
conclusion is entirely satisfactory.

The metapostnotum is fused with a part of tergum 1 of the abdomen, but
the separation of the two is indicated by the antecostal suture.

In Boreidae, the metanotum is essentially like the mesonotum except that
the metapostnotum is almost twice the size of the mesopostnotum. The
condition of the metanotum in Apteropanorpidae is the same as that described
for the mesonotum. The metanotum in the apterous and brachypterous
Bittacidae conforms with the mesonotal pattern described for these genera.
The relative sizes of mesonotum and metanotum vary from one family to
another.

PTEROTHORAX: Pleuron

The sternopleural region of insects has long been a subject of controversy
and remains so today. The details of this controversy and the major problems
which remain to be solved were recently documented by Matsuda ( 1963b)
and DuPorte (1965). The conclusions of the Ferris-Matsuda position stand
in contrast to the earlier interpretations of the traditional views advanced by
Weber (1928) and Snodgrass (1935). This controversy is particularly ger-
mane to a discussion of Mecoptera because some of the arguments of Matsuda
(1963b) refer to interpretations of this order and other lower Holometabola.
I have accepted the terminology of Matsuda (1960, 1963b) for the ptero-
thoracic pleural region but with reservations on the morphological impli-
cations of this stand. This terminology is similar to that of Snodgrass (1927,
1935).

Each pterothoracic pleural field is divided by the vertical pleural suture
(ps) into an anterior episternum (eps) and a posterior epimeron (em). The
suture extends from the base of the coxa, at the pleural coxal process, dorsally
to the base of the wing, where the pleurites taper to form the pleural wing
process. Internally the pleural suture is represented by a well developed

The Skeleto-Muscular System of Mecoptera: The Thorax 817

pleural apodemal process. From approximately the middle of the process the
pleural apophysis (pa) extends ventromesally into the thoracic cavity (Figs.
5-6).

The episternum is delimited dorsally from the basalare by the basalar
cleft, which in most Mecoptera has become a suture (basalar suture, bs), and
ventrally by the coxa. The episternum proper is subdivided into three regions.
The dorsalmost anepisternum (aes) is delimited ventrally by the anapleural
cleft (ac) or anapleural suture (as, episternal suture of authors). The extent
of this suture varies, but it does not reach the pleural suture (except in
Nannochoristidae). Immediately below the anepisternum is the preepister-
num (pes), which is delimited ventrally by the precoxal suture (pes, Figs. 19-
24). This region has been termed the precoxal bridge (Issiki, 1933). Ventral
to the preepisternum is the katepisternum (ks) which is delimited dorsally by
the precoxal suture and ventrally by the coxa. The precoxal suture is in-
complete and does not reach the pleural suture in most Mecoptera. Generally
the extent of the anapleural cleft of the metathorax exceeds that of the meso-
thorax. In Notiothaumidae (Fig. 24), the precoxal suture of the metathorax
is incomplete but attains the pleural suture, while it does not in the meso-
thorax. The epimeron is usually a single elongate sclerite of simple con-
struction. Situated above the epimeron and surrounded by membrane is the
subalare (sb).

The above description is of the mesopleural field but applies also to the
metathorax generally. This description holds for the majority of Mecoptera:
Panorpidae, Panorpodidae, Choristidae, Nannochoristidae, and Notiothau-
midae (Figs. 19-24). The major trend toward reduction of the supracoxal
sutures is considered as derived from and representing modifications of the
more generalized state seen in Meropeidae, which was elucidated by Matsuda
(1960, 1963b) and Mickoleit (1967). The Meropeidae very closely approxi-
mate Matsuda's hypothetical pterothoracic segment in which the anapleural
ring is completely separated from the ventral katapleural ring by the pre-
coxal suture. The Meropeidae are the only Mecoptera retaining a complete
precoxal suture, such as is found in Paleoptera and Neuroptera (Matsuda,
1963b). As a consequence the epimeron of Meropeidae is subdivided into
anepimeron and katepimeron.

The Bittacidae are intermediate between most of the families and
Meropeidae with respect to pleural development. In Bittacidae the anepi-
sternum is deeply incised by the anapleural suture and the basalare is com-
pletely fused with the anepisternum (Fig. 20).

Peculiarities associated with wing reduction are evident in Boreidae and
Apteropanorpidae. As Fuller (1954) indicated, the pterothorax of Boreidae
is peculiar in that the mesepimeron is fused with the metepisternum so that
the pterothoracic pleuron is essentially a single sclerite. Boreidae also lack an

818 The University Science Bulletin

anapleural cleft (suture) and a precoxal suture; consequently there is only
an episternum separated from an epimeron by the pleural suture (the meron
is also greatly reduced). In Apteropanorpidae there is a total fusion of all
three thoracic segments with essentially no differentiation in the pterothorax
(discussed under the section on the pterothoracic terga).

PTEROTHORAX: Sternum

Interpretation of the sternal region of Mecoptera encounters the same
difficulties as those of the pleural region inasmuch as it has not been resolved
for insects in general. However, Weber's study (1928) of the pleurosternal
region of Lepidoptera and subsequent studies of other Holometabola by
Ferris and Matsuda corroborate the conclusion that in the majority of the
Holometabola the sternal field has become invaginated within the thoracic
cavity and that in some orders, including Mecoptera, there is no external
sternum. The concept of cryptosterny as advanced by Weber (1928, 1933)
holds that all of the primary sternal elements and the ventropleurite are in-
vaginated and reduced to a single median longitudinal ridge. Thus, the
tendency towards sternal reduction, amalgamation, and invagination seen in
the prothoracic sternum reaches its extreme in the pterothorax. The exter-
nally visible sclerotized area of the ventral field in the pterothorax of Mecop-
tera is composed solely of pleural elements.

In the ventral field the preepisternum and katepisternum have shifted
anteromesally and terminate at the bases of the coxae (Fig. 8). This is true
of the fully winged as well as for the flightless and wingless forms of
Mecoptera. The trochantin is retained as a very slender sclerite in the
katepisternal area but does not articulate with the coxa or pleuron and has no
muscles. The preepisternal and katepisternal elements (Figs. 7-9, 19-24)
which occupy the anteroventral field of the pterothoracic segments have col-
lectively been termed by mecopterists the precoxal bridge, a term which has
no classical meaning or usage (DuPorte, 1965) but which does have useful
descriptive connotations. In his excellent study of the Panorpidae and
Panorpodidae, Issiki (1933) regarded the ventral field anterior to the coxae
as a combination of the precoxal bridge and the basisternite, that posterior to
the coxae as furcasternite, the ventralmost part of which forms an articulation
with the mesal side of the coxa. This interpretation is based on external
studies, and as a consequence Issiki was not able to observe the well developed
internal structure of the cryptosternite (Figs. 3a-g, 10-11, 13). There are also
some inconsistencies in the terminology of Ferris and Rees (1939) but the
synonomies have been worked out by Matsuda (I960, 1963b).

Internally the pleurosternal region is a complex of apodemes and apophy-
ses which vary depending upon the flight function. In some instances the

The Skeleto-Muscular System of Mecoptera: The Thorax 819

axial orientation depends upon whether the organism is compressed (the
majority of Mecoptera) or depressed (Meropeidae and, to a lesser extent,
Notiothaumidae) .

In the majority of Mecoptera (Panorpidae, Panorpodidae, Choristidae,
Nannochoristidae, Notiothaumidae, Bittacidae and Meropeidae) the pleuro-
sternal region consists of the following elements. The episternum has an
anterior apodeme which begins just below the anapleural cleft and extends
ventrally to the level of the coxa where it turns mesad and becomes con-
tinuous with the sternal apodeme (Fig. 5). There is also a ventral apophysis
of the pleural apodeme at the base of the coxa which extends slightly into the
coxal cavity. At the anterior edge of the sternum there is a dorsally projecting
apophysis. On the posterior edge of the epimeron there is a dorsoventral
apodeme delimited dorsally by the lateropostnotum and terminating ventrally
at the base of the epimeron. The furca is highly developed in Mecoptera and
bears an anterior tendon above its base. Extending ventrally from the median
longitudinal ridge on either side is a V-shaped sclerotized area. Anteriorly
this area is continuous with the precoxal bridge. Ventrally it terminates in
an articulation with the coxa. It is dorsoventrally divided by a suture coming
down from the furca (Figs. 3a-g, 5-6, 10-11, 13).

This sclerite has been regarded as the furcasternite (Issiki, 1933), the
pleurum (Ferris and Rees, 1939), the sternal ridge (Hasken, 1939), and the
sternum (Fuller, 1954; Mickoleit, 1967). The authors after Issiki who have
termed this area the sternum or process thereof are in accord with the inter-
pretation of Matsuda (1960, 1963b). To consider the median longitudinal
ridge, the furca, and these ventral extensions as the sternum of Mecoptera is
the view taken here — and one that is consistent with the Weber-Ferris-
Matsuda line of thinking on the sternum of Holometabola and in particular
of Mecoptera.

The sternal modifications in the brachypterous forms (Boreidae and
Anomalobittacus) and apterous forms (Apteropanorpidae and Apterobtt-
tacus) are minor; that is, they present no special problems of interpretation
but are characterized by a reduction in the development of apodemes and
apophyses in the sternopleural region. Details in the Boreidae are discussed
by Fuller (1954). The sternal structure in Apteropanorpidae is simplified
through suppression of the various apodemes and apophyses.

LEGS

Concomitant with the involution of the sterna, the coxae of all thoracic
segments have become mesally displaced such that they are somewhat ap-
pressed against one another. As Issiki (1933) pointed out, the coxae of
Mecoptera are not as freely movable as are those of other orders in which

820 The University Science Bulletin

they are widely separated by a well developed sternum. In addition there are
some basic differences in the structure of the procoxae and in the pterothoracic
coxae of Mecoptera.

The coxa of the prothorax in all families of Mecoptera is a hollow cylinder
attached to the thorax by a coxal membrane or conjunctiva (the coxal articular
corium). The coxa is laterally tapered at its connection with the trochanter.
At its proximal end it forms a typical monocondylic articulation with the
pleuro-coxal hinge process of the propleuron. Just below its base, the coxa is
circumscribed by a submarginal basicostal suture represented internally by an
apodemal ridge, the basicosta. Proximad of the basicostal suture is a thin,
flattened flange on either side of the pleuro-coxal articulation, the basicoxite,
which is confluent with the coxal shaft toward the mesal plane.

There is no trace of a meron in the prothoracic leg of any Mecoptera,
although Hasken (1939) designates the post-articular basal apodeme as the
merocosta and the rim basal to it as the merocoxite. These designations are
unacceptable because there is no indication of a coxal (or meral) suture in the
prothoracic coxa.

The pterothoracic coxae are more complex than those of the prothorax
and possess a ventral articulation. They are basically cylindrical or conical
and articulate with the respective trochanters by two well developed condyles
(Fig. 15). Like the procoxa, the pterocoxa has an exterior pleuro-coxal articu-
lation formed by a process at the ventral end of the pleural suture and a
process at the proximal end of the coxa, at the base of the coxal (or meral)
suture. On the mesal side the coxa has a condylar articulation with the ventral
process (sternal ridge, sr, Figs. 6, 10-11) of the sternum, which is the best
developed of the articulations. A third articulation, commonly encountered
in other insects (Snodgrass, 1935), is a trochantinal articulation, but in
Mecoptera the trochantin is greatly reduced and the articulation non-existent.
Issiki (1933) analyzed all three of these articular points in Panorpidae and
Panorpodidae and found that all of them are poorly developed in comparison
with other insect groups; this is true of the other families of Mecoptera as well.

The pterothoracic coxa is dorsoventrally divided by the coxal suture into
an anterior eucoxa and a posterior meron (one of the classical characteristics
that define the Panorpoid Complex). In addition, there is the basicostal
suture which circumscribes the proximal end of the coxa as in the procoxa.
That portion anterior to the coxal suture (which appears as a ventral con-
tinuation of the pleural suture) is the basicoxite; below it is the basicostal
flange-like apodeme; that posterior to the coxal suture is the corresponding
merocoxite and merocosta. Ventrally the meron terminates just above the
coxo-trochanteral articulation by the ventrolateral curvature of the coxal
suture in the metathorax. At this point (the apex or distal end of the meron)
the area above the posterior coxo-trochanteral articulation is greatly re-

The Skeleto-Muscular System of Mecoptera: The Thorax 821

enforced. Snodgrass (1935) states that the meron never includes the region
of the posterior articulation and that it is always cut off by the coxal suture.
This is not the case in the mesocoxa of most Mecoptera. In all mecopteran
families except Boreidae and Apteropanorpidae the tip of the meron of the
mesothorax is continuous with the distal end of the coxa and does in fact
bear the posterior coxo-trochanteral condyle (Fig. 15). Hasken (1939) recog-
nized this relationship in Panorpidae.

In Boreidae the only evidence of meral development in the pterothorax is
a very short coxal suture that barely extends beyond the basicosta, with the
result that the meral area is vestigial at the proximal part of the coxa. The
Boreidae are unique in that they are the only jumping Mecoptera. The modi-
fications to meet this mode of locomotion are discussed in fine detail by
Fuller (1954, 1955). The structure of the coxa in Apteropanorpidae is remi-
niscent of the generalized prothoracic coxa of the order. There is no trace of
a meron, coxal suture, merocosta, or merocoxite. The meron plays no part in
the coxo-trochanteral articulation in either pterothoracic leg of Boreidae or
Apteropanorpidae.

Internally the basicosta extends mesoventrally to the sternocoxal articu-
lation in the dorsoventral plane; the other coxal process is parallel to the
longitudinal axis of the body and terminates at the posteroventral end of the
trochantin (though the articulation is functionless). Likewise the coxal (or
meral) suture internally gives rise to a meral apodeme that relates to the
coxo-trochanteral posterior joint (see above). In all coxae there is a trochan-
teral tendon to which coxal muscles attach (discussed under musculature). In
general, the relative size of the coxae increases from prothorax to metathorax
with the notable exception of Nannochoristidae in which the mesothorax and
its legs comprise the largest of the thoracic segments.

The legs of Mecoptera consist of the typical number of segments; coxa,
trochanter, femur, tibia, tarsus (of five tarsomeres) and pretarsus. The coxa
was discussed with the pleurosternal region because of its close affinity with
the latter.

The Mecoptera are divisible, on the basis of leg structure, into walking
(most Mecoptera), jumping (Boreidae) and raptorial (Bittacidae) insects,
each with modifications peculiar to the activity. In a detailed study of Bit-
tacid legs, Rober (1942) analyzed the fine points of their structure and func-
tion (Figs. 18a-c). He likewise based some observations on his comparative
study of other families. Fuller (1955) analyzed the jumping function in
Boreidae. The remaining families have generalized legs with the normal
functions (Snodgrass, 1935).

The trochanter (tr) is a short, slightly compressed, curved segment in
Mecoptera. It has a dicondylic articulation with the coxa. At its distal end it
has an outer and an inner syndetic articulation at the trochantero-femoral

822 The University Science Bulletin

joint, which allows only a very slight horizontal oscillation of the femur on
the trochanter. The trochanter has a proximal tendon on which the coxai
muscles insert. At the distal end there is a tendon for the attachment of
trochanteral muscles.

The femur (Fig. 14) is the largest leg segment. The femoro-tibial joint is
a dicondylic articulation of the anterior-posterior type.

The tibia (Fig. 14) is slightly curved at its proximal end and internally
bears a dorsal extensor tendon and a ventral flexor tendon for the tibial ex-
tensor and flexor muscles, respectively. In the Bittacidae there are inner and
an outer tibial extensor tendons, and the muscles are also divided (Fig. 16).
On the distal inner side of the tibia is a pair of articulated tibial spurs.

The tarsus is composed of five tarsomeres, the basal of which bears both
the tarsal flexor and extensor tendons at its proximal end.

The pretarsus (pt) is essentially the same in all Mecoptera except the
Bittacidae (Figs. 17, 18a-c). The general account of the pretarsus which
Issiki (1933) gives of Panorpidae suffices for the majority of families. The
pretarsus arises from the apex of the last tarsomere to which it is connected
by a membrane. Dorsolaterally there is a pair of tarsal claws (ungues), with
an arolium (ar) between them. The arolium is a pad-like continuation of the
median pretarsal base. It is mostly membranous but has sclerotized rods of
variable dimensions. Basally it is supported by an auxilliary plate (Hilfsplatte
of German authors). Dorsally the unguifer articulates with the claws, is in-
ternally produced and not externally visible. The unguitractor plate is par-
tially invaginated within the distal tarsomere on the ventral side of the
pretarsus. Distal to the unguitractor is a small sclerite, the planta (pi). At
the base of each claw and extending ventrolaterally on each side is a slender
sclerite, the auxillia (ax). From the unguitractor plate the long pretarsal
tendon, to which the tibial and femoral pretarsal effector muscles attach,
extends through the tarsomeres and tibia and into the femur.

This adequately describes the structure of the pretarsus in the majority of
families. There are a few exceptions, such as the absence of an arolium in
Notiothaumidae and Boreidae and the fact that the unguitractor plate is
deeply invaginated within the apical tarsomere in Notiothaumidae and
Panorpodidae. The Bittacidae constitute the single major exception in which
all of the pretarsal elements have become amalgamated into a single heavily
sclerotized claw associated with the raptorial function of the leg in this family
(Figs. 18a-c;R6ber, 1942).

SPIRACLES

It is characteristic of the Pterygota that all of the postembryonic stages
have but two pairs of thoracic spiracles. Primitively, the first belongs to the
mesothorax and the second to the metathorax (Snodgrass, 1935). However,
there is a tendency for the spiracles to become secondarily displaced antero-

The Skeleto-Muscular System of Mecoptera: The Thorax 823

dorsally so that they occupy an intersegmental position — the first between the
prothorax and mesothorax and the second between the mesothorax and meta-
thorax — as is the case in most Mecoptera.

It is characteristic of Mecoptera that the thoracic spiracles are considerably
larger than those of the abdomen and that the muscular effectors in the two
tagmata differ (Hassan, 1944). With the exception of Boreidae and Aptero-
panorpidae, the thoracic spiracles of Mecoptera are obliquely oriented in the
intersegmental area above the anapleural cleft (or suture). Each spiracle
consists of two elongate lips that appress one another in the closed position
and which open into an atrium from which tracheae ramify.

Exceptions to this arrangement occur in the Bittacidae, in which the
metathoracic spiracle has a single external lid, and in the Boreidae and Ap-
teropanorpidae, in which the mesothoracic spiracle has become secondarily
displaced into the posterolateral part of the pronotum and the metathoracic
spiracle lies in the fused mesepimeron-metepisternal area. The thoracic spir-
acles of these last two families are devoid of external lids or covering of any
kind. A comparative account of the spiracles of Mecoptera and their allies is
given by Hassan (1944).

In those families which have a two-lipped closing apparatus, there is an
occlusor muscle which extends from the ventral end of the spiracle on an
apodeme to the anterior apodeme below and behind it. The first spiracular
muscle terminates on the anterior apodeme of the mesepisternum (Figs. 25-
26), the second on the anterior apodeme of the metepisternum. The closing
mechanism and structure are described by Hassan (1944). The thoracic spir-
acles are conspicuously larger than those of the abdomen and differ from the
latter in having: a single occlusor muscle, while there are two occlusor muscles
for each abdominal spiracle. Those families which lack the normal external
lids and in which the spiracles are firmly imbedded in the nota (Boreidae and
Apteropanorpidae) also lack spiracular occlusor muscles.

PTEROTHOR AX : Musculature

The pterothoracic musculature presents several interesting problems. The
first concerns the differences between the prothoracic musculature and that of
the pterothorax. Although all three of the thoracic segments have a homo-
dynamous relationship, their skeletal and muscular elements are quite dif-
ferent. These muscles were initially recorded by Hasken (1939). The prob-
lem of homologizing the muscles of mesothorax and metathorax is lesser and
was also resolved by Hasken. Perhaps the most difficult problem in the study
of the pterothoracic muscles of Mecoptera is resolving modifications (includ-
ing those of the skeleton) as they relate to the flight function, since the order
includes apterous and brachypterous species as well as the majority of normal
fliers.

824 The University Science Bulletin

As Maki (1938) pointed out, the degree of difference between meso-
thoracic and metathoracic musculature varies in different orders; but these
differences are slight in Mecoptera. There is also a close correspondence of
these muscles within the Panorpoid Complex as compared to other Holo-
metabola (Maki, 1938).

The following account of the pterothoracic musculature is based on the
mesothorax. The close agreement of the metathoracic muscles with those of
the mesothorax obviates the need for a separate description. For purposes of
comparison, only the fully winged Bittacidae are used along with the other
families; that is, there is no special consideration of reduction or loss in such
genera as Apterobittacus and Anomalobittacus. The Apteropanorpidae are
treated as the apterous group and the Boreidae as the brachypterous group.
The musculature of the nine families is tabulated in Tables 3 and 4.

Intersegmental Muscles

longitudinal muscles (Figs. 25-26)

(//mi. Between phragmata 1 and 2. It completely covers phragma 1 and
part of the scutum on the anterior end and on the posterior end covers
phragma and part of the postnotum. It is cuneiform in cross-section and is an
indirect depressor of the wings.

dlm-i. Lateral edge of phragma 2 on the lateropostnotum to the side of the
scutum. It is an indirect flight muscle which in conjunction with dvm mus-
cles functions as a levator of the wing. As such, it is antagonistic to dlmi and
also depresses the scutum.

dhn-.>,. This is a very small muscle from the scutoscutellar apodeme to the
anterior side of the postnotum.

vim. Base of the posterior side of the mesofurca to the middle of the
anterior side of the metafurca where it ends in a short tendon.

Segmental Muscles

dorsoventral muscles (Figs. 25-26)

dvm\. Lateral part of the scutum ventrally to the anterior side of the
sternal apophysis on the preepisternum.

dvmz. Immediately behind dvim on the scutum ventrally to the tro-
chantin. This is the only muscle of the trochantin.

Jvm-.\. Lateral part of the scutum anterior of the scutellum and behind
dvm j to the long trochanteral tendon; a tergal depressor and femoral ab-
ductor. It is especially well developed in Boreidae (Fuller, 1955).

th'nii. Scutum behind dvm:: and ventrally at an oblique angle to the
posterior side of the meron. The dorsoventral muscles collectively depress the
tergum through their action on the scutum and are indirect flight muscles in

The Skeleto-Muscular System of Mecoptera: The Thorax 825

conjunction with the dorsal longitudinal muscles (dim:.-). In addition, dvm?
and dviru impart a slight amount of flexion to the coxa.

transverse muscles (Figs. 25-26)

zm. Apex of the furcal arm to the pleural apophysis.
ijum. Between the arms of the furca. It is antagonistic to the sternal leg
muscles and draws the furcal arms towards the midline.

PLEURAL MUSCLES (FigS. 27-2S)

pmi. Precoxal bridge to the inner dorsal side of the episternal tendon; an
alary abductor.

pm-2. Outer edge of the basicosta dorsally to the anepisternal tendon and
to the tendon above the anapleural cleft; elevates the wing.

pm:\. Inner edge of the trochanteral tendon to the dorsal anepisternal
tendon; an alary abductor and femoral adductor.

pnu. Merocosta just behind the pleuro-coxal articulation to the subalare;
an alary adductor.

pm:>. Pleural apodeme to the first axillary sclerite; an alary rotator.

pmn. Immediately behind pm.-, and to the side of the scutum; an alary
rotator.

pmi. Pleural apodeme immediately behind pme to the posterior hinge
process. All three of these rotators, pm.v?, arise on short tendons one behind
the other and effect rotation of the wing. The muscle pm<; is also a flexor of
the tergum (Hasken, 1939). Occasionally, as in Boreidae (Fuller, 1(>55), there
is a single alary rotator instead of three.

pnis. Anepisternal hinge process to the tegula; a protractor of the wing.

pmo. Anapleural cleft to a tendon of the third axillary sclerite; a retractor
of the wing.

pmio. Below the end of the episternum to the lateral edge of the basicosta
in a tendon near pm?; a promotor of the coxa.

pmu. Dorsal side of a tendon process on the base of phragma 1 anteriorly
to the side of the scutum; an accessory of the large dorsal longitudinal muscle
dlmi.

pmn. Dorsal end of the posterior epimeral apodeme to a tendon cap of
the subalare.

pm\z. Anterior episternal apodeme dorsally across the anapleural cleft to
the anepisternal side of the cleft; when the cleft is membranous as opposed to
sutured, the muscle closes the cleft.

pmu. Below pmi3 on the anterior apodeme dorsally to the tendon process
on the side of pmi::; effects closure of the anapleural cleft.

pnixr,. Just above pmi3 and pmu to the lateral edge of the scutum; it
functions antagonistically to pmi3+i4.

pniuu Pleural apodeme to a dorsal process of the subalare.

826 The University Science Bulletin

pm\-. Ventral end of the pleural apodeme to the suhalare in a tendon
next to pnu to which it is an accessory; an alary adductor.

During the course of this study it was not possible to dissect any speci-
mens of Choristidae; consequently, only conjectures were available as to the
probable musculature of the members of this family. Shortly after the com-
pletion of this work, Mickoleit (1969) published a report on the pterothoracic
pleurosternal musculature of Panorpidea and Neuropteroidea. His findings
for Taeniochonsta (Choristidae) confirm the supposition that the pleuro-
sternal muscles ("pm" series) are indeed present in the same number and
arrangement as suggested in this paper. The findings of Storch and Chnd-
wick (1968) for Bittacidae agree with my own findings in this family.

Leg Muscles

STERNAL MUSCLES (FigS. 25-26)

bm\. Anterior side of the mesofurca to a terminal apophysis of the
pleural apodeme at its ventral end; antagonistic to cxmi.

bm-2. Membranous area on the inner side of the sternal apodeme near the
base of the furca to the medial side of the trochanteral tendon on the middle
of its base, laterad of dvma; a femoral abductor.

bm?,. Sternal apodeme to the posterior side of the basicosta; a coxal pro-
motor. This muscle is regarded as subdivided (Hasken, 1939). The second
band is parallel to the first and has the same origins and insertions as the
former.

bw^. Base of the posterior side of the sternal apodeme to a tendon on the
distal end of the meron.

bms. The first band from the posterior side of the sternal apophysis to the
posterior side of the basicosta; the second from the posterior side of the sternal
apophysis to the medial side of the meron; a coxal remotor. This muscle is
also considered as divided (Hasken, 1939).

coxal muscles (Figs. 25-27)

cxm\. Anterior surface of the eucoxa to the trochanteral tendon in front

of CXIT12.

cxtn-2. Basicosta (sometimes part of the origin is on the merocosta) to the
proximal end of the trochanteral tendon.

cxm?,. Ventral process of the pleural apodeme to the lateral proximal end
of the trochanteral tendon; cxmi-z are femoral adductors.

cxm±. Anteromedial side of the coxa to the base of the trochanteral tendon
on the ventral side.

cxm-,. Anterior side of the coxa between cxmi and cxiru to the trochan-
teral tendon.

Table 3. Mesothoracic Musculature.

w

UJ

<
Q

<

<

Q

P

W

Q

E

<
Q

P

CO

5

CO

5
O
X
u
O
z

<

Q

G

<

<
Q

a
0

OS

0

<

Q

OS
O

<

Q

w

<
Q
3

Oh

O

1—1

<
X

H
O

O

z
<

0-
O
pi
w

0

z

H

z

z

ei

Cf

P

H

33
u

<
z

H
PQ

<

Oh

<

a,

0

w

O
Z

<

tllmi

+ '■

+

+

+

+

—

+

+

—

dlma

+ -:

+

+

+

+

—

+

+

—

dlm3

+ •'

—

+

+

—

—

—

—

vim

+

+

+

+

+

—

—

+

+

dvmi

+

+

+

+

+

—

+

+

—

dvni2

+

+

+

+

+

—

+

+

—

dvrri3

+

+

+

+

+

+

+

+

—

dvrru

+ •

+

+

+

+

—

+

+

—

zm

+

+

+

+

+

—

+

+

+

pmi

+

+

+

+

+

—

+

+

+

pma

+

+

+

+

+

+

+

+

+

pm:!

+

+

+

+

+

—

+

+

+

pmi

+

+

+

+

+

+

+

+

—

pills

+

+

+

+

+

+

+

—

pma

+

+

—

+

+

+

+

+

—

pmT

+

+

+

+

+

+

+

—

pma

+

+

+

+

+

—

+

+

—

pm„

+

+

+

+

+

—

+

+

—

pmw

+

+

—

+

+

+

+

+

—

pnin

+

+

+

+

+

—

+

+

—

pmi2

+

+

+

+

+

—

+

+

—

pmia

+

+

+

+

+

—

+

+

—

pmH

+

+

+

+

+

—

+

+

—

pmio

+

+

—

+

+

—

+

+

—

pmie

+

+=

+

+

+

—

+

+

—

pnin

—

3 —

—

—

—

+

—

—

—

bmi

+

5 +

+

+

+

+

+

+

+

bm«

+

3 +

+

+

+

+

+

+

+

bm.T

+

3 +

+

+

+

+

+

+

+

brru

+

3 +

+

+ ?

+

+

+

+

+

bnir,

+

? +

+

+

+

+

+

+

+

CXITli

+

? +

+

+

+

+

+

+

+

CXITI2

+

3 +

+

+

+

+

+

+

+

cxiri3

+

3 +

+

+

+

+

+

+

+

cxm4

+

? +

+

+

+

+

+

+

+

cxm5

+

? +

+

+

+

+

+

+

+

cxmo

+

? +

+

+

+

+

+

+

+

cxm?

+

? +

+

+

+

+

+

+

+

mrf

+

? +

+

+

+

+

+

+

+

mft

+

? +

+

+

+

+

+

+

+

met

+

? +

+

+

+

+

+

+

+

mfpi

+

? +

+

+

+

+

+

+

+

m£p2

+

? +

+

+

+

+

+

+

+

mfa

+

? +

+

+

+

+

+

+

+

mea

+

? +

+

+

+

+

+

+

+

ifum

+

? +

—

+

+

—

—

—

5

om

+

? +?

+

+

+

+

+

+

—

Table 4. Metathoracic Musculature.

<
Q

1— 1

H
c/3

5

O

5

<

G
P

5
O
X
u
0

Z
Z

<

z

<

<r
h

3

<
Q

5
0

OS
O

z

<

U-1

<

Q

OS

O

Z

<
a,

<

Q

OS

O
m

<
Q

W

a*
O
oS
W

<

Q

<

X
H
O

P
O
Z

w

<;

Q
K

OS

O
Z

<
a*
O
oi
pj
H
a.
<

dim i

+

? +

+

+

+

—

+

+

—

dim,
dlm3

vim

+

? +

?

+

+

+

+

+

+

—

+

? +

+

—

+

+

+

—

—

dvmi

+

? +

+

+

+

—

+

+

—

dvm.

+

? +

+

+

+

—

+

+

—

dvms

+

? +

+

+

+

+

+

+

—

dvm4

+

+

+

+

+

—

+

+

—

zm

+

? +

+

+

+

—

+

+

+

pnii

+

? +

+

+

+

—

+

+

1
~r

pm.

+

? +

+

+

+

+

+

+

+

pin-:

+

> +

+

+

+

—

+

+

—

pm4

+

? +

+

+

+

+

+

+

—

pm-,

+

?

+

+

+

+

+

—

pm«

+

+

+

+

+

+

+

+

—

pm7

+

+

—

—

+

+

—

pmg

+

+

+

+

+

—

+

+

—

pma

+

+

+

+

+

—

+

+

—

pmio

+

+

3

+

+

+

+

—

—

pmn

+

+

+

+

+

—

+

+

—

prrii2

+

+

+

+

+

—

—

—

—

pmi.1

+

> +

+

+ *

+

—

+

+

—

pmn

+

+

+

+ *

+

—

+

+

—

pm ).-,

+

+

5 -+-

+

+

—

+

+

—

pm„,

+

? +

5

+

+

+

—

+

+

—

pniiT
hnii

+

+

+

+

+

+

+

+

+

bm»

+

+

+

+

+

+

+

+

+

bma

+

+

+

+

+

+

+

+

+

lim4

+

+

+

+ ?

+

+

+

+

+

bms

+

+

+

+

+

+

+

+

+

CXIDi

+

+

+

+

+

+

+

+

+

CXIll;

+

+

+

+

+

+

+

+

+

CXI11:

+

+

+

+

+

+

+

+

+

CXITU

+

+

+

+

+

+

+

+

+

cxmr,

+

+

+

+

+

+

+

+

+

cxm«

+

+

+

+

+

+

+

+

+

cxnri7

+

+

+

+

+

+

+

+

+

mrf

+

+

+

+

+

+

+

+

+

ml't

+

+

+

+

+

+

+

+

+

met

+

+

+

+

+

+

+

+

+

mtp,

+

+

+

+

+

+

+

+

+

mfpL.

+

+

+

+

+

+

+

+

+

mfa

+

+

+

+

+

+

+

+

+

mea

1 1 11 m

+

+

+

+

+

+

+

+

+

om

+

+

+

+

+

+

+

+

—

pmi3 ;ind prriu fused

The Skeleto-Muscular System of Mecoptera: The Thorax 829

cxme. Posterior side of the coxa to the posterior side of the trochanteral
tendon; antagonitic to cxm.-,; cxm^-e are femoral rotators.

cxm-. Posterior side of the basicosta to the proximal edge of the trochanter
behind cxm^; an accessory femoral adductor to cxmi-3.

TROCHANTERAL MUSCLES (Fig. 14)

mrf. Proximal part of the trochanter to a tendon that originates on the
anterolateral side of the femur; a femoral rotator. This is a large muscle
which completely occupies the trochanter.

FEMORAL MUSCLES (FigS. 14, 16)

;?;//. Trochanter to the tibial tendon; works in conjunction with the next
muscle; tibial flexor.

/>;//. Ventral wall of the femur to a corresponding position on the base of
the large tibial tendon which occupy most of the rest of the femur; tibial
flexor.

met. Lies opposite the inferior flexor dorsally from its origin to the cor-
responding position on the dorsal proximal end of the tibia; a tibial extensor.

mfpi. Proximal end of the femur to the long pretarsal tendon which
passes through the femur, tibia and tarsus and terminates on the pretarsal
unguitractor plate; a pretarsal flexor.

TIBIAL MUSCLES (Fig. 14)

mjp-z. Proximal end of the tibia to the pretarsal tendon; an accessory of
the superior pretarsal flexor.

mfa. Ventral distal end of the tibia to the proximal end of the first tarso-
mere; a tarsal flexor.

mea. Lies opposite the tarsal flexor and has the corresponding origin and
insertion points; dorsal to the flexor and antagonistic to it; a tarsal extensor.

DISCUSSION AND CONCLUSIONS

The cervix of Mecoptera is well developed and has a pair of latero-
cervicalia and in some families a pair of dorsocervicalia. It is suggested that
the dorsocervicalia arose by fragmentation from the postocciput and that their
function as points of muscle insertion is derived.

The dorsal longitudinal muscles of the cervix (Odlmi-j) have no direct
relationship to the prothorax in terms of origins and insertions, since in gen-
eralized insects they extend from phragma 1 to the head capsule (Snodgrass,
1935). The occurrence of Odlm,, in Boreidae and Notiothaumidae is prob-
ably secondarily derived. The ventral longitudinal muscles (Ovlmi-2) are
also considered to be derived by fragmentation from a single band (Hasken,
1939).

830 The University Science Bulletin

The oblique intersegmental muscles (Oismi-5) are variable; reduction in
this series occurs in those families in which there is a general trend toward
reduction of the entire cervical field (Boreidae, Bittacidae, and Meropeidae).
Only the Notiothaumidae possess the full complement of cervical muscles
with secondary subdivisions (i.e., with Oistru subdivided into two bands).

The pronotum of Mecoptera is variously creased with transverse furrows
whose inner surfaces form apodemes for muscle attachments. The pleural
region tends toward reduction and fusion into a single complex involving the
proepimeron, proepisternum, profurca, and the pleural apodeme. A tro-
chantin is present but does not articulate with the coxa nor have any muscles
attached to it, an arrangement seen also in the higher Pterygota. The
prosternum is reduced, compared with that of generalized insects. The sternal
elements are fused into a plate containing the eusternum, furcacavae, and
spinasternum. The entire sternal complex is involuted and there is no trace
of a sternacostal suture.

Although there is considerable variation in the prothoracic pleural region
and in the extent of the pronotum, the number of prothoracic muscles is
rather constant for the order. Nothing of the trends in the evolution of these
muscles is evident within Mecoptera, but when Mecoptera are compared with
other orders of the Panorpoid Complex the following patterns emerge.

The intersegmental prothoracic dorsal longitudinal muscles (ldlm) of
Mecoptera are disconcerting because they occur only as median dorsal mus-
cles, while in other Panorpoid orders there are lateral and anterior dorsals as
well. Three possible sets of these muscles are present in the lower orders
(Maki, 1938), and it is presumed that the described reduction is restricted to
Mecoptera. The prothoracic ventrals offer nothing noteworthy.

Of the prothoracic segmental muscles, the dorsoventral and pleural mus-
cles are constant within Mecoptera and in the other orders of the Panorpoid
Complex. The intrafurcal muscle (lifum) occurs only in Mecoptera and
tipulid Diptera (Maki, 1938). The dorsoventral muscles are also constant
within Mecoptera and the Panorpoid orders, but the overall trend is toward
reduction in the number of these muscles in the Panorpoid Complex as com-
pared with Neuroptera. There is also a reduction and loss of anterior external
dorsoventral muscles from the entire Panorpoid Complex, while these mus-
cles are retained in Neuroptera (Maki, 1938; Kelsey, 1957).

The pleural prothoracic muscles are uniformly present in Mecoptera but
vary within the Panorpoid Complex and in the lower orders. There is no
clear pattern to their presence or absence within the Complex. Sternopleural
muscles are absent from the prothorax of Mecoptera and from most of the
Panorpoid orders. There is a reduction in the total number of procoxal
muscles (cxm series) within Mecoptera and allies as compared with the
lower orders.

The SkeletoMuscular System of Mecoptera: The Thorax 831

The pterothoracic notum of Mecoptera is of the generalized type. Modi-
fications which do occur include a fusion of the notal sclerites into a single
plate in brachypterous and apterous species. It is remarkable that the alino-
tum and postnotum of Mecoptera are separated by a secondary intrasegmental
conjunctiva since this conjunctiva is lost in higher Pterygota. An extreme is
reached in Apteropanorpidae in the complete fusion of the notal and pleural
sclerites into a single plate.

The pterothoracic pleuron of Mecoptera is well developed and contains
most of the sclerites of the hypothetical type (Matsuda, 1963b). This is espe-
cially true of Meropeidae and Notiothaumidae (sometimes placed together
in a separate suborder, Protomecoptera, from the other families of Mecop-
tera). In other Mecoptera the trend is toward a loss of the coxopleural ele-
ments posterior to the pleural suture. This is considered to be a derived state,
resembling the situation in higher Pterygota. Modifications associated with
wing reduction, in Boreidae and Apteropanorpidae, include fusion of the
pleurites.

The tendency toward involution ot the sternum, seen in the prothorax,
reaches an extreme in the pterothorax. There are no sternal elements ex-
ternally visible in the pterothorax. The view taken here follows the Weber-
Ferris-Matsuda line of thought, namely, that the pterothoracic sternum of
Mecoptera and of other Panorpoid orders consists only of the internal median
longitudinal ridge, the furca, and the sternal ridge of the sterno-coxa]
articulation.

The articulations of the coxae with the thorax are restricted and reduced
compared with those of other higher Pterygota. The normal trochantino-coxal
articulation found in other insects (Snodgrass, 1935) is absent from Mecop-
tera. The coxa is dorsoventrally subdivided into a eucoxa and a large meron.
In addition, it is laterally circumscribed by the basicostal suture which sets off
a small basicoxite and merocoxite. The mesocoxa of Mecoptera is unusual in
that the posterior articulation of the coxa with the trochanter directly involves
the meron.

The peculiarities and limitations of leg movement in Mecoptera are largely
due to alterations in articulations and not to loss of muscles, since there are
sufficient promotor and remotor muscles to allow versatile movement. The
prothoracic leg muscles are essentially the same as those of the pterothorax.
In the other Panorpoid orders there is a tendency toward modification and
loss of some of these muscles (Maki, 1938).

The intersegmental pterothoracic longitudinal muscles (dlmi-i>) of Me-
coptera are present in all of the normally winged species, with loss occurring
in the apterous species. This pattern varies slightly in other orders of the
Panorpoid Complex. An unusual, very small muscle (dim.;) is found in the
mesothorax of Panorpidae and Panorpodidae. The ventral longitudinals

832 The University Science Bulletin

(vim) are more variable within the order but are generally present, although
they do not exhibit any clear pattern.

The dorsoventral segmental muscles are essentially the same for both
pterothoracic segments but are absent from apterous Mecoptera. These
muscles correspond with those of other orders of the Panorpoid Complex and
collectively show a reduction in number from those of lower Pterygota. There
is nothing unique about the furcal muscle (zm) in Mecoptera or related
orders, but it is curious that it is absent from the brachypterous Boreidae but
present in the apterous Apteropanorpidae, in both pterothoracic segments.

The complex set of pleural muscles is the most variable in the pterothorax
of Mecoptera and other Panorpoid orders, a fact easily anticipated in view of
the modifications of the pterothorax related to flight. Maki (1938) pointed
out that there is a greater uniformity in the meso- and metathorax of neu-
roptera than in any of the other higher Pterygota. In Mecoptera, as well,
there is a very close correspondence in these muscles between the segments
within a family but more variation between families. Those muscles related
to flight are naturally present in the flying forms but are totally lost as aptery
is approached. Variation in the occurrence of these muscles among fully
winged species must relate in some manner to peculiarities of flight within
the order, but to establish this would require a detailed comparative study of
flight dynamics.

The series of pleural muscles (pmi-4) present in the pterothoraces of all
fully winged Mecoptera tend to reduction with aptery. The series pm.v?
varies within the order and comprises very small alary rotators. There is a
tendency to reduction in the number of these muscles within the metathorax,
but it is difficult to determine exactly which ones of the series remained and
which were lost. The muscles pm.s-;> lie in the same plane as pm.-, - and are
present in all of the fully winged species but are absent from Boreidae and
Apteropanorpidae. The muscle pmio is peculiar in that it is constant among
all winged families (except Bittacidae) but is absent from the brachypterous
and apterous species and from the metathorax of Notiothaumidae. There is
nothing noteworthy about the series pmn-12, pmi.vi.-,, and pmie. The muscle
pmi7, unique to the mesothorax of Boreidae, does not represent a special case
or a new muscle but an unidentified one of the pleural series not homologized
by Fuller (1955).

The major trend in the evolution of the pterothoracic musculature in-
volves a reduction in the total number of muscles, as compared to the thoracic
musculature of the lower Pterygota. In the existing muscles, there is a trend
toward secondary splitting.

This study has shown that there is a definite, basic anatomical pattern in
the order Mecoptera, that is, one to which all of the families in general
correspond. Naturally, some families fit the generalized condition more pre-

The Skeleto-Muscular System of Mecoptera: The Thorax 833

cisely than others do. Skeletal and muscular structures of the head and thorax
suggest certain conclusions to the relative closeness of one family to another
and the position of the order with respect to other insect orders.

Meropeidae is judged the most primitive family within the order since
it shows the greatest concentration of primitive or generalized features, par-
ticularly in the thoracic skeleton and musculature. Notiothaumidae, while
not as generalized as Meropeidae, retains a greater number of these primitive
characters than do the other families and hence appears to have the greatest
taxonomic affinity to Meropeidae. I think that these two families are aptly
placed in the suborder Protomecoptera.

The families Panorpidae, Panorpodidae, Choristidae, and Nannochoristi-
dae share a more advanced state of organization than is seen in the Proto-
mecoptra. These four families represent an intermediate stage in the gradu-
ation of anatomical features within Mecoptera. Boreidae, in my estimation,
also belongs with these intermediate families, in spite of its many thoracic
modifications. In this connection, I think that attempts to remove the
Boreidae (Hinton, 1958) to a separate order Neomecoptera are unjustifiable.

Bittacidae are structurally the most advanced family of Mecoptera. This
family is a good example of the transition from a morphologically primitive
state, as in Meropeidae, to that ot the higher holometabolous orders. The
family Apteropanorpidae remains enigmatic because of the extreme reduc-
tion in its thoracic development. I am inclined to think that this family
represents an extreme modification of the intermediate level of organization.

In terms of classification, these findings (which are based on an entirely
different set of anatomical structures) in general support the recently pro-
posed classification of Byers (19(d), although he regarded the Choristidae as
the most generalized family. I think the soundness of that classification is
demonstrated by the fact that, in the course of this study, an anatomical
"family pattern" was found which corresponded in each case to the familial
taxa recognized by Byers.

The order Mecoptera as a whole has emerged at the base of the Panorpoid
Complex and is structurally more primitive or generalized than any of the
sister orders within the Complex. Accordingly, Mecoptera retain a level of
organization at the base of the Holometabola, alongside the Neuroptera, just
as Tillyard (1918a & b, 1919, 1935) suggested in his study of half a century ago.

LITERATURE CITED

Ryers, G. W. 1965 (1964). Families ami genera of Mecoptera. Proc. 12th Int'l. Congr. Ent.,

p. 123.
Cooper, K. W. 1940. The genital anatomy and mating behavior of Bonus brumalis Fitch (order

Mecoptera). Amer. Mid. Nat. 23:354-367.
Crampton, G. C. 1926. A comparison of the neck anil prothoracic sclerites throughout the

orders of insects from the standpoint of phytogeny. Amer. Ent. Soc. Trans. 52:199-248.
DuPorte, E. M. 1965. The lateral and ventral sclerites of the insect thorax. Canad. J. Zool.

43:141-154.

834 The University Science Bulletin

Evans, }. W. 1942. The morphology of Nannochorista maculipennis Tillyard (Mecoptera).

Trans. Roy. Soc. S. Australia 66(2) :2 18-225.
Ferris, G. F., and B. E. Rees. 1939. The morphology of Panorpa nuptialis Gerstaecker

(Mecoptera: Panorpidae). Microentomology 4:79-1 (IK.
Fuller, H. 1954. Das Thorakalskelet von Boreas westwoodi Hag. Zool. Jahrb., Anat.

73:425-449.
. 1955. Die Muskulatur des Thorax von Boreus westwoodi Hag. Zool. Jahrb., Anat.

74:189-210.
Hasken, W. 1939. Der Thorax von Panorpa communis L. Zool. Jahrb., Anat. 65(3) :295-338.
Hassan, A. A. G. 1944. The structure and mechanism of the spiracular regulatory apparatus in

adult Diptera and certain other groups of insects. Trans. Rov. Ent. Soc. Lond.

94(1):103-153.
Hepburn, H. R. 1969. The skeleto-muscular system of Mecoptera: The head. Univ. Kansas

Sci. Bull. 48:721-765.
Hinton, H. E. 1958. The phylogeny of the Panorpoid orders. Ann. Rev. Ent. 3:181-206.
Issiki, S. 1933. Morphological studies of the Panorpidae of Japan and adjoining countries and

comparison with American and European forms. Jap. J. Zool. 4:315-416.
Kelsey, L. P. 1957. The skeleto-motor mechanism of the dobsonfly, Corydalus cornutus. II.

Pterothorax. Cornell Agr. Exp. Sta., Mem. 346:1-31.
Maki, T. 1938. Thoracic musculature of insects. Mem. Fac. Sci. and Agr., Taihoku Imp.

Univ. 24(3):l-343.
Matsuda, R. 1960. Morphology of the pleurosternal region of the pterothorax in insects. Ann.

Ent. Soc. Amer. 53:712-731.
. 1963a. Evolution of the thoracic musculature in insects. Univ. Kansas Sci. Bull.

44(ll):509-534.

. 1963b. Some evolutionary aspects of the insect thorax. Ann. Rev. Ent. 8:59-76.

Mickoleit, G. 1967. Das Thorakskelet von Merope tuber Newman (Protomecoptera). Zool.

Jahrb., Anat. 84:313-342.
. 1969. Vergleichend-anatomische Untersuchungcn an der pterothorakalen Pleurotergal-

muskulatur der Neuropteria und Mecopteria (Insecta, Holometabola). Z. Morph. Tiere

64:151-178.
Mivake. T. 1913. Studies on the Mecoptera of Japan. Jour. Coll. Agr. Tokyo 4:265-401.
Pringle, J. W. S. 1957. Insect Flight. Cambridge Univ. Press, Cambridge, 133 pp.
Rober, H. 1942. Die Raubbeine der Bittaciden. Zool. Jahrb., Anat. 68:399-414.
Setty, L. R. 1940. Biology and morphology of some North American Bittacidae (Order

Mecoptera). Amer. Mid. Nat. 23(2) :257-353.
Snodgrass, R. E. 1908. A comparative study of the thorax in Orthoptera, Euplexoptera and

Coleoptera. Proc. Ent. Soc. Wash. 9:95-108.
. 1909. The thorax of insects and the articulation of the wings. Proc. U.S.N.M.

36:511-595.

. 1927. Morphology and mechanism of the insect thorax. Smiths. Misc. Col. 80 ( 1 ) : 1 - 1 08.

. 1935. Principles of Insect Morphology. McGraw-Hill, New York, 667 pp.

Storch, R. H., and L. E. Chadwick. 1968. Thoracic structure of the adult mecopteran,

Bittaats strigosas Hagen (Mecoptera: Bittacidae). J. Morph. 126:199-21(1.
Tillyard, R. J. 1918a. The Panorpoid Complex. Proc. Linn. Soc. N. S. Wales 43:265-319.

. 1918b. The Panorpoid Complex. Proc Linn. Soc. N. S. Wales 43:626-657.

. 1919. The Panorpoid Complex. Proc. Linn. Soc. N. S. Wales 44:533-718.

. 1935. The evolution of the scorpionflies and their derivatives. Ann. Ent. Soc. Amer.

28:1-45.
Verhoeff, K. W. 1903. Thorax tier Insekten. Nova Acta Leop. Carol. 82: (cited from

Crampton, not seen by me)
Weber, H. 1928. Die Gliederung der Sternopleuralregion des Lepidopteren-Thorax. Eine

vergleichende morphologische Studie zur Subcoxaltheorie. Zeitschr. wiss. Zool. 131:181-

254.
. 1933. Lehrbuch der Entomologie. Jena, 726 pp.

The Skeleto-Muscular System of Mecoptera: The Thorax 835

E

>

O

3
E

C

c

bo

c

o

c

836

The University Science Bulletin

awp

pwp

Fig. 3. Bittacus chlorostima. a) Dorsal view of sternal region with tergum removed.
b-d) Lateral view of the respective furcae. e-g) Dorsal view of tureae alone.
Fig. -1. Nannochorista maculipennis. Dorsal view of the thorax.

The Skeleto-Muscular System of Mecoptera: The Thorax 837

Poa/SCLpor

Fig. 5. Brachypanorpa carolinensis. Lateral internal view of the thorax showing the skeletal
elements.

Fig. 6. Bittaats chlorostigma. Lateral internal view of the thorax showing the skeletal
elements.

838

The University Science Bulletin

CT>

p

C
>

00

£

X

«

c

u

5

u

« °

£- >

2 S

■£._

3 g
^ >

0

The Skeleto-Muscular System of Mecoptera: The Thorax 839

10

11

12

13

Fig. 10. Bittacus chlorostigma. Anterior view of the skeleton of the mesothorax.

Fig. 11. Bittacus chlorostigma. Posterior view of the skeleton of the mesothorax.

Fig. 12. Bittacus chlorostigma. Dorsal view of the thorax. Note the absence of any inter-
segmental conjunctiva. Broken lines indicate humps.

Fig. 13. Brachypanorpa carolinensis. Dorsal view of sternal region with tergum removed.
Broken lines indicate the continuation of the coxae on the ventral side.

met

mea

Fig. 14. Pan or pa communis. View of the leg and its musculature. (Generalized from
Hasken, 1939).

Fig. 15. Notiothaitnia rcedi. Generalized view of the mesothoracic coxo-trochanteral articu-
lation. Note the role of the meron in the posterior articulation.

Fig. 16. Bittacas italicus. Lateral view of the femoral musculature. (Generalized from
Rober, 1942).

Fig. 17. Panorpa nuptialis. Ventral view of the fifth tarsomere and pretarsus, typical of
Mecoptera. (Generalized from Ferris and Rees, 1939).

Fig. 18. Bittactis chlorostigma. a) Lateral view of the last two tarsomeres and the pretarsus.
b-c) Details of the teeth that occur on the inner surface of the tarsomeres.

The Skeleto-Muscular System of Mecoptera: The Thorax 841

21

Fig. 19. Brachypanorpa carolinensis. Lateral view of the thorax.
Fig. 20. Bittacus chlorostigma. Lateral view of the thorax.
Fig. 21. Apterobittacus aptcrus. Lateral view of the thorax.

842

The University Science Bulletin

~<3

o

-^:

CO

<o

o

rn

(^1

6

Uh

ti

M

8.!

5 2

\-0-

^ 0

i *

S.rb

^t

8 u

C\J

^; o

CM--

-O

o -^

-a

The Skeleto-Muscular System of Mecoptera: The Thorax 843

CM

B
u
-^

DC

ID

~*

>

n

E

CO

jd

c

s.

SJ

it

C

r-

u

ro

O

CA

u^

^o

CN

r^i

u

<i

u.

-

844

The University Science Bulletin

00 78

fco

c
o

ro

c
u

-^

E
o

y=

Ih

3

ts

■ „;

n

'J

*tf-

3

3'

E

3

Q.

en

0

in

_S

o

-3

Ih

-J

O

n

Ih

11

Q.

>

0

n

N

«

• —'

t/>

P8

cs

Lh

r^

O
3

0

eg

a

c/3

r^

on

CM

c-i

d

q

Uh

—

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLVIII Pages 845-848 May 1, 1970 No. 22

An Aquatic Caecilian from the Magdalena River,

Colombia, S.A.

Edward H. Taylor*

Dr. Edwin Cooper, of the University of California Medical School,
recently sent me a caecilian which he had used in certain experimental work.
It appears to belong to the recently described genus Nectocaecilia Taylor, a
genus comprising four previously known species that are widely distributed
in South America. As yet only an exceedingly small number of individuals
of this genus have been taken. It is presumed that this species is aquatic, as
are the species of the related genera ot the family Typhlonectidae.

Nectocaecilia cooperi, sp. nov.
(Figs. 1-2)

Holotype. American Museum of Natural History, No. A82255. From
the Rio Magdalena at Barranquilla, Colombia, South America.

Diagnosis. Having the generic characters. An elongate slender typhlo-
nectid caecilian lacking a dorsal "fin." Approximately 86 primary folds, some
rather dim, seemingly none complete; 96 vertebrae; eyes visible, in socket.
Width of body in length approximately 45 times. Dentition in four series.
Internal nares very large. Deep black throughout except for light coloration
about eyes and at vent, and slightly lighter coloration on the head and jaws.

Description of the Holotype. Head a little wider than body; total
length 356 mm. Eye visible, in a socket; tentacular opening very small, close
behind the much enlarged nostril, the distance from eye, 4 mm, from nostril
0.45 (the tip of the snout has been injured the skin missing, so that the
deeply sculptured bones are exposed). Snout tip to first nuchal groove, 12 mm
(lateral measurement), to third nuchal groove, 22 mm. The two nuchal col-
lars are indistinct. Following the collars there are approximately 86 primary
folds (some dim and difficult to count), the body with a large unsegmented
"shield" at termination. The last centimeter of body strongly triangular in
cross section, flattened on the ventral side. The disc of the anal region whitish,

* Research Associate, Museum of Natural History, Kansas University.

846

The University Science Bulletin

Fig. 1. Nectocaecilia cooperi sp. nov. Holotype. AMNH No. AK2255. From Rio Magdalena,
Barranquilla, Colombia, S.A. Dorsal (upper) and ventral (lower) views. Actual length,
556 mm.

An Aquatic Caeciliax from Colombia, S.A. 847

somewhat oval, the grooves between the surrounding denticulations termi-
nating at vent, do not reach to edge of disc. There are 5 denticulations pos-
terior to the vent, two lateral to it, each of which bears a small black anal
gland, and two larger ones preceding the vent. A vague suggestion of a
dorsal ridge but no dorsal "fin." Choanae very large relatively, the length 3
mm, the width approximately 1.5 mm. The distance between choanae, 1.1
mm. No scales or secondary folds. The general surface relatively free from
the wrinkles usually present in the typhlonectids.

Dentition. Premaxillary-maxillary series, 18-1-18; prevomeropalatine, 18-
1-18; dentary 16-17; splenial series, 3-(l?). The counts, if not exact are very
close approximations. The palatine and splenial teeth are the smallest, the
anterior dentaries largest.

Measurements in mm. Total length, 356; head width, 9; body width about
8; width preceding vent, 4.4; body height, 8; snout projects 3.

Remarks. A figure of the type specimen is presented. The spine in the
posterior part of the body has been broken and healed before capture. An
X-ray shows the presence of 96 vertebrae.

The species is presumably related to Nectocaeciiia haydee (Roze), a species
which differs in having a "fin" from head to terminus. The number of
splenial teeth is different.

The type is most probably a young specimen. It has been named for Dr.
Edwin Cooper who has made the specimen available to me for study.

848

The University Science Bulletin

Fig. 2. Nectocaecilia cooperi Holotype. From Rio Magdalena at Barranquilla, Colombia,
S.A. X-ray showing 96 vertebrae.

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLVIII Pages 849-854 May 1, 1970 No. 23

A New Caecilian From Ethiopia

Edward H. Taylor*

Two caecilian specimens were sent to the British Museum of Natural
History from Ethiopia by Mr. M. J. Largen of the Haile Sellassie I University,
Addis Ababa. These in turn have been made available to me for study by
Miss Alice G. C. Grandison, Curator of Herpetology at the British Museum.

The first specimen, BMNH No. 1969- 15S9, was found at Aleku village
12 km N of Dembidollo, Wallega, Ethiopia at an elevation of approximately
1846 m (34° 37'E; 8° 39'N). The second, BMNH No. 1969-976, was ob-
tained by the Great Abbai Expedition, at Ghimbi, Wallega, Ethiopia at an
elevation of approximately 2180 m (35° 50'E; 9° 10'N). The latter locality
is "no more than 125 km (in direct line) NE of Aleku."

These localities are several hundred kilometers north of previously known
points that have yielded caecilians on the eastern side of Africa; however, on
the western side they have been taken in southern Senegal some 5° farther
north than the Ethiopian localities.

The caecilians seemingly belong to a new species of the genus Gcotrypetes,
and it is herewith described:

Geotrypetes grandisonae sp. nov.
(Figs. L-4)

Type. British Museum (Natural History) No. 1969-1589, Aleku, 12 km
N Dembidollo, Ethiopia, 1846 m elev. Paratype. BMNH, No. 1969-976,
Ghimbi Wallega, Ethiopia, 2180 m elev.

Diagnosis. A rather short plump species, reaching a known length of
259 mm. The body width in length approximately 24 times. Eyes visible in
a socket not continuous with the tentacular groove. Tentacle distinctly closer
to the eye than to the nostril. Primary folds 84-95, complete dorsally but
narrowly incomplete on the anterior ventral part of the body. Secondary folds
69-72 with 32-33 complete. Splenial teeth, 3-3 to 5-5. Four to five scale rows
in each fold on the last two cm of body. Tongue with two small narial plugs.
A diastema between the squamosal bone and the parietal bone.

* Research Associate, Museum of Natural History, Kansas University.

850

The University Science Bulletin

Fig. 1. Gcotrypetes grandisonae sp. nov. Holotype, (9 :) British Museum (Natural History)
No. 1969.1589; Aleku, 12 km N Dembidollo, Ethiopia, 1846 m clcv. Dorsal (upper) and
ventral (lower) views. Actual length, 259 mm.

A New Caecilian from Ethiopia 851

Descriptions of Type. A small, rather thick-bodied species tapering
slightly posteriorly, the length 259 mm, the body width 11 mm; body width
in length, approximately 23.5 times. Head narrowing somewhat anteriorly,
with rounded snout projecting 1.2 mm beyond mouth; length of lower jaw
from tip to rictus oris, 8 mm. Eye in socket which is not continuous with the
tentacular groove. Tentacle distinctly closer to eye (1.9 mm) than to nostril
(2.7 mm). Snout tip to first nuchal groove, 12.4 mm; to 2nd groove, 15.4 mm;
to 3rd nuchal groove, 19 mm (lateral measurements). Tentacle about equi-
distant between edge of lip and a line from eye to nostril, the external opening
minute, somewhat horseshoe-shaped, very slightly elevated. The two collars
following occiput not very distinct, seemingly somewhat swollen (perhaps
due to a small tumor in mouth and throat). The first nuchal groove distinct
laterally, vague dorsally and ventrally; a transverse groove vaguely evident on
collar; second groove limiting first collar distinct below and on side but very
dim or absent dorsally; the third groove rather distinct dorsally and ven-
trally, except it fuses with the first primary fold for a short distance. Second
collar wider than first, with no dorsal transverse groove evident. Primary
folds following second collar, 84, complete dorsally throughout but narrowly
incomplete ventrally on most of anterior half of body. Secondary folds, 69, of
which about 33 are complete.

Scales beginning on primary folds at a point near first secondaries. At
midbody, 2 to 3 scale rows, which may not be complete ventrally, in each
fold; posteriorly, 4-5 scale rows in each fold, the scales variable in size, the
largest 1.2-1.4 mm in greatest width. No subdermal scales found.

Glands in skin visible, but not especially conspicuous. In the grooves the
elongate glands are directed forward and downward but are less conspicuous
than in many caecilians. No anal glands visible ( 9 ?).

Scales not found in the first secondaries. Terminal "shield" very small (3
mm wide). Vent subcircular, the surrounding denticulations elevated (may
not be typical).

Dentition. Premaxillary-maxillary series, 19-1-21; prevomeropalatine, 20-
1-21; dentary, 17-17; splenial, 5-5. Dentary teeth for the most part larger than
maxillaries or premaxillaries and these in turn larger than prevomeropala-
tines. The splenials equal to or a little smaller than prevomeropalatines.

Measurements in nun. Total length, 259; head width (greatest), 9; body
width (middle third), 11; width near vent, 8.5; vent to terminus, 2; width
in length, approximately 23.5 times.

Color. The general color is a dull bluish to violet slate nearly uniform
above, perhaps somewhat more violet anteriorly. Top and sides of head and
chin more grayish, lighter than dorsum. A distinct light spot over and sur-
rounding eye, one at nostril, one at tentacle, and one covering the denticles in

852

The University Science Bulletin

3

rrj

~+j

<N

■si

3

r.

2

4— i

SO

_,

B

/

u

u

rt

PQ

a

o

*o

<

,

(/J

a

£

u

7^

>

*"^

_c

Si

9 c

"T3 >

3« B

O 5

CM i-

J Q

A New Caecilian from Ethiopia 853

vent area. Folds on sides have very narrow inconspicuous light edges on
anterior part of body.

Remarks. Most of the characters of the skull are unknown. However,
there is a diastema between the squamosal and parietal bones and the orbit of
the eye is not continuous with the tentacular groove. A tumorous growth is
present in the posterior palatal area extending back into the throat. The
tongue is swollen as if it too might be affected by the disease. There seems to
be only a relatively small passage left for food.

A second specimen (paratype) referred to this species is BMNH No.
1969-976. It offers the following data: primaries, 85; secondaries, 72; com-
plete secondaries 52; premaxillary-maxillary teeth, 20- 1-18; prevomeropalatine
teeth, 17-1-18; dentary, 15-1-15 (seemingly a median tooth); splenial, 3-3.
Measurements in mm are: length, 231 9 ; head width, 7.9; body width, 11.2;
tentacle to nostril, 2.1; tentacle to eye, 1.7.

The specimen has much the same general color as the type save that the
anterior section of the body is darker in places probably due to air exposure
and slight dehydration. The penis is partially extended. The collars are very
clearly delineated, the second seemingly not fusing below with the first
primary fold.

While the two specimens appear to agree in most easily observable char-
acters, there seems to be variation in the secondaries and in scales. I cannot
find any scales in the first few secondaries; elongate skin glands are present
bordering the folds above, filled with a white cheeselike material. These
glands also appear in the primary folds in the same region but here there may
be a few scales. Posteriorly, scales appear in both primary and secondary
folds, and are similar to those in the type.

On the external surface one may discern certain larger skin glands that
may be diseased or possibly the result of parasitism. These contain much
cheeselike material. Occasionally some are seen in which the cheeselike
material appears to have shrunk leaving a round shallow depression on the
surface above it.

The placing of this species in Geotrypctes is tentative. It would appear to
agree more closely to this than to other African genera. It differs specifically,
however, from other known species of this genus in the number of scale rows
in the posterior folds, and in that it has fewer splenial teeth, fewer primary
folds and a larger number of secondary folds, many of which are complete,
and the tentacular aperture is distinctly closer to the eye than to the nostril.
Usually the aperture maintains a fixed position within a genus with relation
to the eye and nostril.

The species is named for Miss Alice G. C. Grandison who has been help-
ful in providing these specimens for study.

854

The University Science Bulletin

X, ..

Fig. 3. Geotrypetes grandisonae sp. nov. Terminal view of the partially extruded penis of
No. 1969.976, paratype.

Fig. 4. Geotrypetes grandisonae ^>. nov. Holotype. British Museum (Natural History) No.
1969.1589. X-ray showing 87 vertebrae and relatively long ribs. Total length 259 mm.

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLVIII Pages 855-860 May 1, 1970 No. 24

Notes on Brazilian Caecilians
Edward H. Taylor*

A small collection of caecilians from Duke University, sent to me by Dr.
Joseph R. Bailey, contains four specimens representing three species. Two
species are examples of rare, recently described forms. The third is a well-
known species of wide distribution. All seem worthy of comment. I have
also examined a specimen of the recently described Nectocaecilni ladigesi
Taylor, and include it here also.

Siphonops paulensis Boettger

(Fig. I)

Siphonops paulenis Boettger, 1S('2. Katalog der Batrachier-Sammlung in Museum der Sencken-
bergischen Naturforschenden Gesellschaft, pp. 62-63 (type-locality Sao Paulo, Brasil).

The specimen, Duke Univ. No. 9628 from Barra Bonita, Sao Paulo mea-
sures (in mm) 291 in length, body width about 25, head width 16, terminal
"shield" width 20, and a circumference of 82. The length divided by width is
approximately 23.6. The primary folds are 104, the vertebrae 108 or 109. The
color is a typical dark bluish slate, each fold marked laterally and ventrally
with white lines which on the back become dulled and greatly narrowed so
that they are not or scarcely discernible. Anal glands are present ( 6 ).

This, I believe, is the largest known specimen of this species.

A smaller specimen, Duke Univ. No. A 9629, is from Jarimu, near Jundiai,
Sao Paulo. This has 109 primary folds, and a length of 251 mm.

The range of this species is chiefly in eastern and southeastern Brasil.
Specimens are known from the States of Rio Grande do Norte, Bahia, Minas,
Gerais, Espirito Santo, Guanabara, Sao Paulo, Rio Grande do Sul, Southern
Goias and southern Mato Grosso; also from northern Argentina, Paraguay
and Bolivia.

Specimens from Goias seemingly have the lowest number of primary
folds, the four specimens examined having counts of only 101-104. The
known range of counts elsewhere is 104-119 folds.

* Research Associate, Museum of Natural History, Kansas University.

856

The University Science Bulleti

v

h

3C

<

c
D

o

o

Notes on Brazilian Caecilians 857

This species, like S. annulatus, is adapted to relatively dry situations, and
like that species has a wide distribution. However, it appears to be confined
to the drainage areas of the Rio Parana and the short coastal rivers emptying
into the Atlantic, while annulatus has a far wider range, extending into
western and northern South America.

Oscaecilia hypereumeces Taylor

Oscaecilia hypereumeces Taylor, 1968, The Caecilians of the World, pp. 607-611, figs. 331-334
(type-locality, Joinville, Santa Caterina, Hrasil).

This specimen, Duke Univ. No. A 9627 (locality unknown), the second
known of this species, has the following characters:

The eye is present under bone, the tentacle almost directly below the
nostril, very much closer to nostril than to eye. The scales begin near the
middle of the body, only a few in a fold laterally. More posteriorly there is a
single row of scales around the body with occasionally one or more extra
scales appearing dorsolateral^. The elongate glandules lying above the scales
seemingly are more numerous than the scales.

Its counts and measurements are as follows; numbers in parentheses refer
to the type: The primary folds, 208 (226) are incomplete above and below
except in the posterior part of the body. There are 21 (4) secondaries of which
6 are complete. There are four dental series (the tooth counts, if not correct,
are close approximations): premaxillary-maxillary, 10-1-10 (8-1-7); prevo-
meropalatine, 11-1-11 (10-1-9); dentary, 10-10 (9-9); splenial, 2-2 (3-3).

Measurements in mm. Total length (body broken), approximately 400
(640); body width, 5 (7); snout projects l.S (1.9); tentacle to eye, 2.6;
tentacle to nostril 1.1 (1.0); tip of snout to angle of jaw, 5.2; tip of snout to
first nuchal groove, 7.2, to second groove, 9.1, to third, 11.5 (all lateral mea-
surements); width in length, approximately 80 (91) times.

Remarks. The head is whitish to light brown from the eyes forward to
near the tip of the snout, with light areas about nostrils and tentacles. A
cream spot is present posterior to the brown area. A pair of brownish spots
is behind and above jaw angle, in front of which is an indefinite light lateral
streak. The edge of the upper lip is light with a vaguely darker streak above.
The entire lower jaw is whitish or cream but with a fine scattering of darker
pigment. The anterior 10 cm of the body is an ill-defined violet brown, the
venter a dull cream with a thin scattering of pigment. Posterior to this, the
body is variegated vaguely (suggested in the figure of type, no. 331, loc. at.)
and brown in color. In practically all other characters this specimen agrees
with the type.

The differences in the count of primaries and secondaries from those of
the type are not greater than has been reported in several species of Caeciha.

An x-ray picture of this specimen prepared by Dr. Joseph Bailey shows
214 vertebrae. The subdermal scales are scattered and difficult to find in the

858

The University Science Bulletin

Fig. 2. Typhlonectes anguillajormis Taylor. Duke Univ. No. A %j(). "Probably vicinity of
Manaos," Brasil. Total length, 372 mm; body width, approximately 19 mm.

Notes on Brazilian Caecilians 859

thin connective tissue under the skin. They are small, circular and nearly
transparent, usually less than .5 mm in diameter. Posteriorly some of the
scales in the grooves are vaguely visible externally along sides.

Typhlonectes anguillaformis Taylor
(Fig. 2)

Typhlonectes anguillaformis Taylor, L968, The Caecilians of the World, pp. 235-238, figs. 117-
120 (type-locality unknown).

This specimen, Duke Univ. No. A 9630, is in good condition except that
a portion of the side of the snout has been dissected. The color has become
almost uniform light brown; presumably it originally was some shade of slate
or violet. The type specimens are somewhat grayish but I believe that this is
not the color in life, since when the loose epidermis was removed, the color
below appeared violet-brown. Since this is the third known specimen, I am
recording the following data:

The head is short and somewhat flattened. The bodv length is 373 mm,
its width approximately 19 mm, the head width, 13.6 mm. The width in
length is approximately 19.6 times. The minute tentacle is close (0.7 mm)
behind the large nostril. There are 82 primary folds and no true secondaries.
Most folds have at least one transverse crease or wrinkle. There is an tin-
segmented terminal "shield," triangular in cross section. The subterminal
area is bounded laterally by two slightly sinuous ridges extending forward for
about 3 cm. These are slightly more compressed than is depicted in the figure
of the type (Taylor, loc. cit., fig. 119).

The dorsal skinfold (fin) begins at the second collar and continues to the
terminus. At first it is about 2 mm high, becoming gradually higher pos-
teriorly, and reaching a height ot 6-7 mm on the last 2 cm of the body. The
tentacular opening is minute and is not evident in the type figure (loc. cit..
fig. 119).

In practically all other characters the specimen agrees with the type.

Nectocaecilia ladigesi Taylor
(Fig. 3)

Nectocaecilia ladigesi Taylor, 1968, Caecilians of the World, pp. 275-27'', figs. 139-142 (type-
locality, Rio Moju near its mouth, | junction with the Tocatins| near Belem, Brasil).

A recent acquisition of the U.S. National Museum (No. 154035) is a speci-
men of the above species, collected by Dr. Philip Humphrey in 1964 at Utinga,
Belem, Brasil. Data on the specimen are as follows; numbers in parentheses
are those of the type*; measurements are in mm: length, 3S9 (416); head
width, 12 (13); neck width near the head, 8.8; greatest body width, 14 (9.5);
height of body, 18.5 (17); width at vent, 5.3. There are 92 (97) vertebrae.

* In the type description, p. 277, line -I, lor "6" read "9.5"; p. 279, line 12, for "eye" read
'tentacle."

860

The University Science Bulletin

TJfejft&t&X?^'

Fk.. 3. N ectocaecilia ladigesi Taylor. USNM No. 154085, Rio Moju near its mouth, near
Belem, Brasil. X-ray .showing 92 vertebrae. Total length 38() mm.

The counts of the dental series are close approximations of the actual tooth
numbers: premaxillary-maxillary, 20-1-24 (19-1-20); prevomeropalatine; 20-
1-20 (20-1-20); dentary, 19-18 (17-16); splenial, 6-7 (5-5). The skin is very
smooth, the color nearly uniform brownish slate. The denticles surrounding
the vent (11 in number) are almost identical in arrangement and proportion-
ate size to those of the type.

I am under obligation to Mr. R. A. Tuck, Jr. of the U.S. National Mu-
seum for an x-ray (Fig. 3) mu\ 6 photographs of this, the second known
specimen of this aquatic species.

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLVIII Pages 861-868 May 1, 1970 No. 25

The Lateral-line Sensory System in the Caecilian Family,

Ichthyophiidae (Amphibia: Gymnophiona.)

Edward H. Taylor1

The sense organs commonly known to anatomists as the lateral-line or
neuromast system developed in early aquatic Chordata (Ostracoderms) and
continued to appear in subsequently developed aquatic groups. It was a
sensory system believed to serve the organism in adjusting to an aquatic
environment. When an aquatic animal having such a system changes to a
terrestrial environment, the system seemingly no longer functions and tends
to disappear. This is true of certain species of fish that assume a semiterres-
trial habitat (Symbranchidae). It is also true of many Amphibia that have a
free-swimming larval period after hatching and which, at transformation
from larval to adult form, usually assume a terrestrial habitat. The system is
present in numerous Caudata during the larval period. One species of the
genus Taricha is said to lose the system on transformation to a land habitat.
Later in life when it again resumes the aquatic habitat the system reappears
and presumably again functions! Most Salientia that have a free-swimming
larval period seek a terrestrial habitat on transformation and the system is
lost. However, in certain species that are, as adults, semiaquatic, the system
may be partially retained (Rami hexadactyla and R. cyanophlyctis). Others
may remain permanently aquatic and retain more of the system (Xenopits).

The neuromast organs of the lateral-line system are usually described as
groups of sensory cells, each having a hairlike process at its free end and
surrounded by supporting cells. These organs are arranged in series that fol-
low the nerves. They are in pi ts or depressions in the skin which are usually
connected by canals filled with a mucous secretion. The series are symmet-
rically arranged on the two sides of head and body. In most fishes there is a
series crossing the head that connects the two sides. If this is present in the
caecilians, it is not obvious externally.

The purpose of this paper is to inquire into the character of this sense
system in the third order of the Amphibia, the Gymnophiona, especially in

1 Research Associate, University of Kansas, Museum of Natural History.

862 The University Science Bulletin

the family Ichthyophiidae, and to demonstrate that the general pattern occur-
ring in the Asiatic genera is duplicated in a South American genus of the
family.

In the order Gymnophiona, Taylor (1968, 1969) recognized four families:
the Ichthyophiidae, Scolecomorphidae, Typhlonectidae and Caeciliidae. The
fourth family was divided into two subfamilies, the Caeciliinae and the
Dermophiinae.

This sense organ was first reported in Gymnophiona by the Sarasins, who
found it in Ichthyophis glutinosus (1887a,b). They discuss the system
(Seitenorgane) and present figures representing the general distribution of
the organs (1887a, pi. 5, figs. 42, 51-54); their structural details are discussed
and figures are given (1887b, pi. 6, figs. 11-18). Taylor (I960, 1968) has given
figures illustrating the general pattern of the cranial part of the system in
Ichthyophis mindanaoensis, I. youngorum and / supachaii. The first two
species have a larval life lasting one or two years, during much of which time
the lateral-line system is retained. The third species has a larval life of only
a few weeks. At the end of this time the animal transforms and the system
is lost as it becomes terrestrial in habitat.

Four genera are recognized. Besides the Asiatic Ichthyophis, there is
another Asiatic genus, Caudacaecilia, and two South American genera,
Epicrionops and Rhmatrema. Larvae of Caudacaecilia weberi, C. larutensis
and C. asplenia have been examined, all showing traces or parts of the system
but none showing it satisfactorily, probably because none was of proper age.
It is most probable that the details when found will differ but little from
those of Ichthyophis.

The characteristics of the system in Ichthyophis both in the diagram (Fig.
1) and in the retouched photograph (Fig. 2) may be described as follows:
On the dorsal surface of the head two lines of neuromasts, seemingly a con-
tinuation of the premaxillary series, pass up over the tip of the snout and
form two short groups that terminate back of the level of the nostrils. These
are designated internasals. Just back of the nostrils the canthals begin, curve
outward and turn backward. They may be somewhat continuous with the
supraorbitals that curve above the eyes and terminate slightly above and
behind the eyes. Below the nostrils, beginning on the upper lip, is the
infraorbital series that passes up toward the eye, then curves under it and
terminates behind the eye but at a higher level. Beginning somewhat back of
the eye and running diagonally to near the level of the line of the mouth is
the genal series. Farther back and about on a level with the mouth are the
supraspir oculars terminating above the anterior edge of the open gill slits.
Beginning on the lower lip are the mentals that curve back on the chin to
near the level of the corners of the mouth, their terminal points near to each
other.

The Lateral-line Sensory System in the Ichthyophiidae 863

2 34 5

6 7

A 8

B C

Fig. 1. Neuromast scries based on the larval head of Ichthyophis. Lateral (A), ventral (B)
and dorsal (C) views: 1. internasal series; 2. canthal; 3. infraorbital; 4. supraorbital; 5. genal;
6. supraspiracular; 7. gill slits; 8. mental; 9. infralabial; 10. premaxillary; 11. gular. The three
body series are not indicated in the diagrams.

Bordering the lower lip the infralabials extend directly backward to near
the posterior terminus of the gcnal group. From this series short branches
extend backward and inward for a short distance, there being one to three
neuromasts visible on each branch. On the beginning of the neck short

864

The University Science Bulletin

Fig. 2. Head of Ichthyophis peninsularis Taylor. Larva. E.H.T.-H.M.S. No. 1836,
Bangalore, India. Head width, 5.1 mm; total body length, 140 mm. Retouched. Usually the
neuromasts appear whitish. This color may disappear before the neuromast is lost.

groups of neuromasts extend back to near the front level of the gill slits.
These are designated the gulars.

While not shown in the cranial diagram, there are three lines of neuro-
masts on each side of the body that reach from the gill region to, or near to,
the end of the body. They are designated the upper, middle and lower lateral
series. Only rarely are all three visible at the same time on a specimen. It is
quite possible that there are other parts of this system which are not obvious
to the eye and which might be discovered by histological methods.

The characteristics of the lateral-line system, as seen from the lateral and
ventral aspects of Epicrionops p. petersi, is best shown in Figure 3. The in-
fraorbital series is nearly complete, as is the genal. The mentals, beginning
near the middle of the lower lip, extend back behind the angle of the mouth,
more or less contacting the genals posteriorly. Ventrally the infralabials show
the numerous short branches to a greater or lesser degree, behind which the
mentals may be distinguished. In this species the neuromasts appear distinctly
larger and perhaps fewer than in Ichthyophis (Fig. 2) and differ somewhat in
their general arrangement from that in the diagrammatic Figure 1. The
neuromasts on the dorsal surface of the head (Fig. 3) are relatively indistinct,

The Lateral-line Sensory System in the Ichthyophiidae 865

Fig. 3. Epioionops petersi petersi Taylor. Larva. Kl'MNH No. 119402 (three views of
head). Head width, 4.6 mm; total length of specimen, 117 mm. Cordillera del Due, above Rio
Coco, 1150 m elev., Napo, Ecuador.

which suggests that they are already tending to disappear. The three linear
lateral series are more or less evident, continuing most of the length of the
body. Four other specimens of Epicrionops (Figs. 4, 5) are shown. What
might be regarded a variation is, most probably, only the degree of loss of the
neuromasts in older specimens that are preparing to assume the adult char-
acteristics. Measurements recorded suggest that the specimens differ in age
and in the degree of development.

In conclusion, insofar as the genera Ichthyophis and Epicrionops are
concerned, one may state that the lateral-line sense organs follow the same
general design and the function may be presumed to be the same. Such
slender evidence as is available shows that the system in Caudacaecilia follows
the characteristics of those of Ichthyophis. Rhinatrema, the most aberrant
genus of this family, may or may not be similar. Only two specimens of that
genus are known, both adults.

866

The University Science Bulletin

Table 1. Data on Epicrionops p. petersi Taylor.

Number

119399

199397

119400

119398

119401

119402

Total length

150

135

135

125

125

117

Head width

7

6.1

6

5.1

6

4.6

Tail length

S.l

8.2

9

8

5.2

7.5

Body folds

Total

Primary

Total

Primary

Primary

Primary

circa 234

circa 64

circa 231

circa 66

circa 65

circa 66

Tail folds

13

11

11

11

?

?

Premax-max teeth

12-1-13

12-1-12

11-1-11

10-1-11

10-1-9

11-1-11

Prevom -palatine

17-1-12 +

18-1-18

17-1-18

16-1-17

17-1-17

15-1-15

Dentary

13-13

13-14

13-14

12-12

15-1-15

14-1-15

Splenial

5-5

6-7

5-5

5-5

6-6

6-5

Gill-slit

1

1

1

1

1

1

Neuromasts

Intcrnasal series

""

5-4

5-5

4-5

Premaxillary

0

0

0

0

0

b

Supraorbital

4-4

5-5

5-5

4-4

5-5

5-4

Infraorbital

12-12

13-14

13-13

14-14

13-13

13-13

Infralabial

18-18

16-18

17-18

16-16

20-19

18-17

Supraspiracular

4-3

6-4

6-6

3-3

4-6

3-3

Genal

2-3

3-4

3-5

3-3

3-5

3-5

Mental

4-4

4-4

4-4

3-3

3-5

3-5

Lateral (middle)

22

21-2(1

25

20

20-25

18-?

Literature Cited

Liebermann, Jose. 1939. Distribucion geografica de los caecilidos Argentinos y observacion
acerca de la biologi'a. Physis (Buenos Aires), tomo 16, no. 48, pp. 83-88, figs.

Sarasin, Paul, and Fritz Sarasin. 1887-1890. Ergebnisse naturwissenschaftlicher Forschungen
auf Ceylon in den Jahren 1884-1886. Zweiter Band.

.1887a. Heft 1. Zur Entwicklungsgeschichte und Anatomie der ceylonesischen Blind-

wiihle lchthyophis glutinosus. Erster Theil: Einlcitung, das Ei, Befruchtung und
Brutpflege, Entwicklung der Korperform, Historisches, Systematisches und Ver-
gleichendes. pp. 1-40, pis. I-V. Kreidel's Verlag, Wiesbaden.

. 1887b. Heft 2. Zur Entwicklungsgeschichte und Anatomie der ceylonesischen Blind-

wiihle lchthyophis glutinosus. Zweiter Theil: Die Seitcnorgane tier Larve, die letzten
Endigungen der Blutcapillaren in den Intercellularraumen der Epidermis Becherzellen
und Cuticulaborsten, Korperringel und Schuppen, Bau und Entwicklung der Cutis-
driisen. pp. 41-94, pis. VI-XI. Kreidel's Verlag, Wiesbaden.

. 1889. Heft 3. Zur Entwicklungsgechichte und Anatomie der ceylonesischen Blind-

wiihle lchthyophis glutinosus. Dritter Theil: Das Schicksal des Dotters; Uebcr die
Homologie der Keimblattcr im Thierreiche auf Grund des Satzes, dass die beiden
Keimschichten der Gastrula nicht clem Ektoderm und Entoderm, sondern dem Blasto-
derm und Dotter tier Vertebraten entsprechen; der Enddarm der Embryonen. pp. 95-
150, pis. XII-XIV. Kreidel's Verlag, Wiesbaden.

— .1890. Heft 4. Zur Entwicklungsgeschichte und Anatomie der ceylonesischen Blind-
wiihle lchthyophis glutinosus. Vierter Theil: Der Schadel; Nasc; Jacobson'sches Organ
und Thrancnnasengang; Der Tentacle; Das Gchororgan; Einc Notiz liber das Gehirn
von lchthyophis; Driisen der Mundhohle; Bemerkungen iiber das Gcfasssystem ;
Kiemenkorb, Zunge, aussere Kiemen, Kiemenspalten; Die Spermatozoen; Schluss-
bemcrkungen; Nachtrag; Verzeichniss der Orginallitcratur iiber die Cacciliiden. pp.
151-263, pis. XV-XVI. Kreidel's Verlag, Wiesbaden.

Taylor, E. H. 1960. On the caecilian species lchthyophis monochrous and lchthyophis glu-
tinosus. Univ. Kansas Sci. Bull., vol. 40, no. 4, Apr. 20, I960, pp. 37-120, '38 figs.

. 1968. The Caecilians of the World: A Taxonomic Review. Univ. Kansas Press, 848 pp.,

425 figs.

— . 1969. A new family of African Gymnophiona. Univ Kansas Sci. Bull., vol. 48, no. 10,
May 16, 1969, pp. 297-305, 5 figs.

The Lateral-line Sensory System in the Ichthyophiidae

867

Fig. 4. Epicrionops petersi petersi Taylor. Larvae Upper figures (three views of head),
KUMNH No. 119398. Head width, 5.1 mm; total length of specimen, 125 mm. Lower fig-
ures (three views of head), KUMNH No. 119401. Head width, 6 mm: total length of specimen,
125 mm. Roth specimens from Cordillera del Due, above Rio Coco. 1150 m clev., Napo,
Ecuador.

868

The University Science Bulletin

Fig. 5. Epicrionops petersi petersi Taylor. Larvae. Upper figures (three views of head),
KUMNH No. 119400. Head width, 6 mm; total length of specimen, 135 mm. Lower figures
(three views of head), KUMNH No. 119397. Head width, 6.1 mm; total length of specimen,
134 nun. Both specimens from Cordillera del Due, above Rio Coco, 1150 m elev., Napo,
Ecuador.

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLVIII Pages 869-949 July 2, 1970 No. 26

An Introduction to the Phytogeography of Kansas

Janet E. Bare and Ronald L. McGregor

Botany Department
University of Kansas

ABSTRACT

The vascular flora of Kansas is composed primarily of species which are
not confined to the Great Plains. Most of these are believed to have migrated
into the area during Tertiary or later times, though it is probable that some
evolved within the Plains. Paleobotanical evidence, recorded accounts of
the native flora at the time of the white man's arrival and present distribution
patterns suggest that the flora of Kansas is related to eight major centers of
frequency — the east, northeast, southeast, south, southwest, north, interior
and Great Plains — and also contains a number of widespread species. Major
patterns of distribution exhibited by species entering the state are described
in this paper, and tentative conclusions regarding the ecological factors which
seem to be most influential in determining these patterns are discussed. It
is suggested that much more work on the biosystematics of plains species
will yield information of real value in interpreting phytogeographical prob-
lems.

INTRODUCTION

Kansas has only a single species of Embryophyta, the moss Aschisma
kansanum, which, as far as is known, is entirely confined to the state. An-
other moss, Grimmia teretinervis , collected in Scott and Norton counties, is
otherwise known only from the Hudson Bay area and from the Alps. Among
the angiosperms, Oenothera jremontii is limited to north-central Kansas
and extreme south-central Nebraska. Clematis jremontii has a similar dis-
tribution in Kansas and Nebraska but also has a disjunct population in
eastern Missouri. Phlox ol^lahomensis is known only from Woodward
County, Oklahoma, and from Butler, Cowley, Elk and Chautauqua coun-
ties, Kansas. With the exception of these five species, Kansas is populated
by plants which are relatively widespread and which are believed to have
migrated into the area during later Tertiary and more recent times, al-

(S70 The University Science Bulletin

though it is probable that some have evolved within the plains. As one
studies the distributions of these migrant taxa, certain recurring patterns
become obvious. For example, plants which are related to the flora of
the southeastern United States tend to share a common area within the
state, as do species which are characteristically northeastern. Some mem-
bers of these two floras overlap in distribution, but even where they do
not overlap, there are indications that some of the same ecological factors
limit both elements in their westward distribution. It is the purpose of
this investigation to provide a more thorough understanding of the rela-
tionships of the Kansas flora to those of other regions and, when possible, to
consider those major ecological and evolutionary factors which define areas
of distribution within the state. Nomenclature for the vascular plants is
according to Gleason and Cronquist (1963) except for certain recent
monographs and revisions.

PALEOBOTANICAL RECORDS
FROM THE CRETACEOUS TO THE RECENT

The distribution of any species in a modern flora may depend in part
upon past ecological conditions as well as upon present ones; thus it is
pertinent to consider the paleobotanical history of central North America
since the close of the Cretaceous Period. During the Cretaceous, much of
the interior of the continent was covered by inland seas and the climate was
warm and humid. With the uplift of portions of the Rocky Mountain sys-
tem at the end of the Cretaceous, this interior region was exposed, and the
land became available to terrestrial plants. By the Eocene, the Rocky
Mountains were much reduced; by the Oligocene, peneplanation was com-
plete. The Mississippi Embayment, together with moist winds from the
Pacific, allowed for a rather uniform oceanic climate with summer rains
and temperatures that rarely fell below freezing. During the Early Eocene,
most of the eastern United States, Canada and Alaska were occupied by the
Arcto-tertiary flora. Subtropical conditions extended as far north as present
North Dakota, and temperate conditions reached nearly to the North Pole.
Samuel Aughey (1880) records Populus, Quercus, Salix, Ficus, Liquidambar,
Sassafras and palms from deposits in Nebraska which Chaney and Elias
(1936) assign to this early Tertiary Period. MacGinitie (1962) includes
Ailan thus, Cercidiphyllum, Dipteronia, Engelhardtia (Asiatic type), Ex-
buchlandia, Glyptostrobus, Idesia, Ketelleria, Koelrettteria, Metasequoia,
Paliurus, Paulownia, Pterocarya, Trapa and ZelJ^pva as being characteristic
of the Arcto-tertiary flora during the Oligocene and into the Miocene. He
lists Acer, Betida, Carya, Castanea, Fagus, Fraxinus, Hydrangea, Lirioden-
dron, Nyssa, Quercus, Ulmus and others as being characteristic of the later
Miocene Arcto-tertiary flora.

An Introduction to the Phytogeography of Kansas 871

The western part of the United States as well as the Gulf Coast and
Florida were occupied during the early Eocene by the Neotropical-tertiary
flora which was characterized by tropical and subtropical plants, especially
by trees with broad, thick evergreen leaves as in the Lauraceae. According
to MacGinitie, this group increased in importance until late Eocene or
Oligocene and included genera such as Cedrela, Engelhardtia, Thoitinia,
Meliosma (Mexican type), Cissampelos, Rhamnidium, Davilla, Chrysophyl-
lum, Petrea and Per sea.

The Mexican Plateau was the center of origin of the Madro-tertiary flora
which began perhaps as early as the Eocene "on scattered dry sites and on
the lee sides of high ridges and mountains" and which consisted primarily
of small trees, shrubs and grasses (Kendeigh, 1961).

Toward the end of the Oligocene, the North American climate was
gradually becoming cooler, although large trees still prevailed well to the
north. Beginning during the Miocene and increasing into the Pliocene
(decreasing thereafter), there was extensive uplifting in the Rocky Moun-
tains, Appalachians, Ozarks and Ouachitas. With the continuing cooling
process, there was a general southward shift of the northern boundaries of
temperate and tropical species. The erection of high mountains between
the Pacific Ocean and the interior of the continent, together with the uplift
of the Ozark region and the subsequent drainage of much of the Mississippi
Embayment, produced a much less humid climate in the Great Plains region.
Apparently as a result of reduced summer rainfall and decreased reliability
of rainfall, many of the East Asian and East American elements had disap-
peared from this area by the end of the Miocene, and hardy drought-resistant
plants from northern Mexico (elements of the Madro-tertiary flora) were
migrating northward into the Great Plains and the western slopes of the
Sierra Nevada (Chaney, 1947).

The fossil record in the Great Plains is scanty, and interpretations of the
material available do not agree as to the type of vegetation which occupied
this region during the late Tertiary. Chaney and Elias (1936), primarily on
the basis of fossil floras and faunas from Brown County, Nebraska (Upper
Miocene to Lower Pliocene); Beaver County, Oklahoma (Lower Pliocene);
and Logan County, Kansas (Middle Pliocene), postulate the existence of
extensive grasslands with trees limited to stream borders and lake margins.
MacGinitie (1962), working with a larger but older Upper Miocene flora
(Kilgore Flora) from Cherry County, Nebraska, believes the area was
occupied by a "savanna type of vegetation with well-forested river bottoms
and an open pine-oak forest of subhumid aspect on the interfluves," and
that "the present prairie type of treeless vegetation is a product of the
Pleistocene climates." Wells (1965) also states that "the extensive treeless
grassland in the plains region may be a relatively recent development." It

872 The University Science Bulletin

is interesting to note, however, that in contrast to the megafossil flora de-
scribed by MacGinitie, which contained only trees and shrubs, the Pliocene
strata of the Ogallala formation as described by Segal (1966) contain 32
grasses and 10 forbs but only one tree, Celtis. Modern counterparts of the
fossil grasses and forbs are typical prairie species, and Celtis is commonly
found along streams in semi-arid grassland regions. Other indications of a
drier climate during the Pliocene, such as modifications in leaf structure
and stratigraphical evidence of a declining water table, are discussed by
Chaney and Elias (1936) and by Frye and Leonard (1957). The relative
paucity of Pliocene fossils is also suggestive of a more xeric environment
since fossilization is thought to occur less readily in a dry situation than in
a wet one.

It is further interesting to note that of the modern species most similar
to MacGinitie's 28 megafossil species, nine (according to MacGinitie) are
presently associated with that flora which extends from southeastern Nebraska
to the Ozarks and eastward to the Appalachians; six are now restricted to
the southwestern United States and adjacent Mexico; and three occur in
both of these regions. Five of the modern related species are growing now
at the fossil site. Three of these, Celtis occidentals, Fraxtnus pennsylvanica
and Ulmus americana, are related to eastern floras; and two, Acer negundo
and Kibes aureum (including K. odoratum), occur in both areas.

Bessey (1887) describes an interesting modern locality approximately
35 miles east and a few miles north of the Kilirore fossil site, where a small
tributary of the Niobrara River, Long Pine Creek, runs through a deep and
winding canyon. Bessey refers to the country surrounding the canyon as
"absolutely treeless . . . the surface in many places thrown up into rounded
hills of what must have once been drifting sand." In this canyon, he found
great numbers of Finns ponderosa var. scop ulor um, Primus demissa, Kibes
aureum, Rhus trilobata, and an oak approaching Quercus undulata, all
western species, growing together with Juglans nigra and Ostrya virginica,
both at the western-most limits of their ranges. Other plants growing in
the canyon also indicate that the area is, as Bessey calls it, "a meeting place
for two floras." Might not the Kilgore flora have been growing in just such
a canyon as the one at Long Pine? The protected nature of the canyon
would have allowed for the survival of more mesic species than could have
existed on the surrounding upland and at the same time would have made
fossilization of plant parts more likely. Although MacGinitie feels there is
no definite paleobotanical evidence indicating the existence of widespread
treeless prairies or steppes in the Tertiary, we have been unable to discover
any data which would unquestionably exclude the existence of a type of
vegetation such as that postulated by Chaney and Elias.

Ax Introduction to the Phytogeography of Kansas 873

During the Pleistocene, the North American climate became more humid.
The cooling trend which had begun in the Oligocene culminated in four
major periods of glaciation which alternated with inter-glacial stages thought
to have been as warm, or nearly so, as the present (Polunin, 1960). Only the
first two major advances— the Nebraskan and the Kansan— reached into
Kansas, and these entered only the northeastern corner of the state. The
Wisconsin glaciation, being the latest and covering much of the same area
as the previous three, is the most important when one is considering modern
distribution patterns. At its maximum, approximately 18,000 years BP or
even later, the Wisconsin glacier reached into southeastern South Dakota,
extended northwestward toward North Dakota, roughly paralleling the
present course of the Missouri, and east-southeastward across northern Iowa
(Jones, 1964).

During the glacial advances, some species of plants and animals were
forced to exist southward, while others were eliminated completely. Workers
do not agree upon the distance south of the glacial boundary to which the
effects of the ice masses were felt. According to Braun (1928), the arctic
and coniferous belts intervening between the ice and the more southern
deciduous forest probably were not wide, "for the effects of glacial refrigera-
tion do not seem to have extended far beyond the limits of the ice cap."
Deevey (1949) feels, however, that "glacial chilling in the southeastern states
must have been fairly extensive and that the warmth-loving species, including
many or perhaps most of the 'Miocene relicts,' survived in peninsular Florida
and in Mexico, and have subsequently migrated to their present localities."
The effects of Pleistocene glaciation upon speciation, particularly in regard
to isolation of populations, creation of new habitats, hybridization and
polyploidy, have been discussed in some detail by Anderson (1949), Stebbins
(1950), Braun (1955) and others.

Pleistocene fossils from the Great Plains are primarily limited to pollen
and other materials from bogs and marshes. Horr (1955) discusses a pollen
profile from Muscotah Marsh in Atchison County, Kansas. The material,
according to Horr's interpretation, gives evidence of a trend from a cool,
moist climate (Abies dominant, with Tsuga, Picea and Larix) to a warmer,
drier climate (decrease in coniferous pollen accompanied by an increase in
Corylus, Popuhts and Quercus) followed by a reversal to a cool, moist climate
with coniferous vegetation. The remaining layers of the marsh indicate the
development of a grassland association, dominated by Cyperaceae, Cheno-
podiaceae and Gramineae, probably brought about by a continuing decrease
in moisture and an increase in temperature. McGregor (1968) gives a C-14
date of 15,500 years BP ± 1,500 for Abies needles obtained from the bottom
layer of a core from the same marsh. This would seem to coincide with the
Cary Substage of the Wisconsin glaciation at which time a lobe of the glacier

874 The University Science Bulletin

reached into north-central Iowa (Wright and Ruhe, 1965). Evidently the
climate approximately 200 miles southwest of the glacial margin was cool
enough to permit the growth of a boreal type of forest. According to Mc-
Gregor, "it would appear that the transition from boreal forest to prairie
began about 11,000 BP and was completed about 7,000 BP. Between 7,000
and 4,000 BP the present vegetation of the area of northeast Kansas was
reached." These dates agree with those given by Wright (1968) for Illinois
and the general area of the Prairie Peninsula.

During the Late Pleistocene, the Great Plains is thought to have ex-
perienced a warm, wet period usually referred to as the early Hypsithermal
Interval. Jones (1964) imagines a "westward extension far onto the plains
of elements of the eastern deciduous forest and probably also tallgrass
prairie — forest species along the river systems and on favored drier sites,
and tall grasses on the uplands" during this time. Disjunct populations of
two species of woodland mammals, Peromyscus leucopus, the white-footed
mouse, and Neotoma floridana, the wood rat, support this theory. Distribu-
tion patterns of various plants, including the species mentioned in Bessey's
paper of 1887, also uphold this theory. Rydberg (18%) years ago noted the
similarity between the flora of the Black Hills and that of the Great Lakes
region. Watts and Wright (1966) also discuss western and eastern species
which occur in the Black Hills and state that conifers from Canada and
from the eastern Rockies probably were in contact in the western Great
Plains during the Wisconsin and that they probably provided a forest "bridge"
by which exchange of species between the two types of forest was possible.

As the Pleistocene drew to a close, the climate became progressively drier
and warmer. A period (or perhaps several periods) of frequent, widespread
and prolonged drought conditions, the later Hypsithermal Interval, allowed
for an eastward extension of the central grasslands through the lower Great
Lakes region perhaps as far east as the Mohawk Valley, New York. Borchert
(1950) presents a brief and lucid account of events as they are thought to have
occurred during this time. Briefly they are as follows. As the continental
glaciers retreated northward, the boreal spruce forests followed. These were
not replaced immediately by the birch-beech-maple-hemlock association which
is presently situated between the boreal forests and the more southern
deciduous forests. In the northeastern United States, the area vacated by the
boreal forests was occupied first by a wedge of oak forests, indicating relatively
drier conditions, and then, for a longer period, by grasslands. Schmidt (1938)
and Smith (1957) present zoogeographical evidence which substantiates this
theory of an eastward extension of the prairie peninsula. Davis (1965)
maintains that New England may not have felt the effects of the late
Hypsithermal Interval but that enough information to reach a conclusion is
not yet available. Wright (1968) synthesizes much of the most recent data

An Introduction to the Phytogeography of Kansas 875

regarding the history of the Prairie Peninsula. In the interior plains region,
the spruce forests were replaced immediately by a grassland type of vegetation.
After the drier late Hypsithermal Interval, as conditions became cooler,
the grassland peninsula was invaded by the deciduous forests, leaving
remnants of the prairie stranded in Iowa, Illinois, Indiana, Ohio, Pennsyl-
vania and New York. During the driest parts of the Hypsithermal Interval,
the flora of eastern Kansas would have included fewer mesic species than
it does now, but the present flora certainly was established long before white
men arrived in the New World.

RECORDED HISTORY OF THE INTERIOR GRASSLANDS

Kansas is situated in a region variously referred to as the Great Plains,
the Prairies and Plains, the Central Grasslands or the Interior Plains. Eastern
botanists, ecologists and geographers writing about the region which extends
from North Dakota to central Texas and from the eastern base of the Rocky
Mountains to the edge of the eastern deciduous forests have tended to treat
it as a rather homogeneous area consisting of vast expanses of monotonous
grassland fringed on the east by remnants or invaders from the eastern
forests. The names applied to these central states reflect this trend in thought.
In vegetational and climatic descriptions, the gradual increase in elevation
and gradual decrease in rainfall from east to west are usually emphasized,
while physiographic regions and north-south changes in length of growing
season and average temperature (as well as the floristic changes that ac-
company them) often go unmentioned. A reader unfamiliar with the central
states could easily fall (and many do) into the habit of regarding the region
as a flat to rolling grassland differing from place to place only in the height
or sparseness of its plant cover and in the amounts of its meager rainfall.

This traditional notion of homogeneity and barrenness was continuous
from the 16th century until the end of the 19th century and still has not
been completely dispelled. This seems to be based upon the history of the
region as much as upon its physiography or flora. It is furthered by the fact
that major highway systems in the plains states run east and west — parallel
to the drainage patterns — thus affording much less variation in relief and
vegetation than one would observe if traveling from north to south. The
role of the history of the region in painting an image of barrenness and
monotony in the central prairie and plains states is evident in the following
summary. Those interested in a more detailed account are referred to
Baughman (1%1), Miller, Langsdorf and Richmond (1961) and Wedel
(1959).

Spanish conquistadores who in 1541 came as far as Rice County, Kansas,
in their search for Quivera were among the first white men to leave a written
description of that part of the state. "The country is like a bowl," they de-

876 The University Science Bulletin

clared, "so that when a man sits down, the horizon surrounds him at the
distance of a musket shot" (Farb, 1964). Had they accidently found their
way to the Chautauqua Hills in southeastern Kansas, their impression might
have been much different.

After Louis Jolliet and Jacques Marquette traveled down the Mississippi
River in 1673, the French began to explore the Missouri River as a possible
route to California. French explorers produced some maps — the first in fact
which made reference to the Kansa Indians for which the state of Kansas was
named — but left very little in the way of written descriptions of the land or
its plants.

Few Anglo-Americans other than traders and trappers reached the central
states until the early 1800's. The acquisition of the Louisiana Territory in
1803 prompted the government to sponsor exploratory expeditions such as
those led by Lewis and Clark, 1803-1806, and Pike, 1806. Lewis and Clark
spoke disparagingly of the northern grasslands, and Pike, who explored the
grasslands of Nebraska and Kansas in the fall of 1806, was also left with a
poor opinion. He called the Great Plains "a desert — a barrier — placed by
Providence to keep the American people from a thin diffusion and ruin"
(Jackson, 1966). In 1819, Major Stephen Long led another group into the
area. It was on his map, according to Miller, et al. (1961), that the High
Plains or short grass prairie was first labeled the "Great American Desert."
Forty years later, describing for the readers of the New York Tribune his
overland journey to Colorado and California, Horace Greeley was still
proliferating the myth of the "American Desert . . . the acme of barrenness
and desolation" (Caldwell, 1940). As James Malin (1967) points out, "this
desert tradition was so firmly embedded in the public mind that it became
an obstacle to thinking about the area, and has today been eradicated only
in part."

Academic activities during the first decade of the 19th century were
primarily confined to the eastern seaboard — to cities such as Boston and
Philadelphia. There were, however, a handful of naturalists already con-
tributing to the inventory of the plants of North America. Among these
was a young English naturalist, Thomas Nuttall, who was the first person
to bring back significant observations and materials relating to the flora of
the central United States. During 1810-11, he traveled into what is now
Nebraska and the Dakotas (Graustein, 1967) and in 1818-20 he traveled
westward across Arkansas and into eastern Oklahoma. His longest trip
(1S34-36) began at Independence, Missouri, and continued up the Missouri
and Kansas rivers, north to the Platte River, up the North Fork of the
Platte and across the Rocky Mountains. From these expeditions Nuttall
brought back many previously undescribed plants. A study of Gates' Flora
of Kansas (1940) shows that of the 1,626 native taxa listed, 164 of them or

An Introduction to the Phytogeography of Kansas 877

about 10% were first described by Nuttall. Another 11 species were named
after him (e.g., Zygadenus nuttattii, Viola nuttallii, Sedum nuttallianum) .
A remarkable record for one person!

As more people began to move west, two major highways were responsible
for the founding of trading posts and primitive settlements to serve travelers.
The Santa Fe Trail, first used by wagon traffic in 1822, was used primarily
for trade with the Spanish Southwest. The Oregon Trail, especially im-
portant during the late 1830's to 1850's, was, however, primarily a road of
emigration (Miller, et ah, 1%1). Little botanizing took place during the years
of active settlement, but journals of travelers and settlers provide some
descriptions of the native flora at that time. Many of the settlers moving
west looked at the plains somewhat as Pike did; they felt it was a barrier to
be crossed before their destination farther west could be reached. It is
interesting to note, however, the difference in comments made by travelers
who crossed through the grasslands in the spring and early summer and
those who came through in late summer and fall as Zebulon Pike did.
Heinrich Mollhausen, who traveled the Santa Fe Trail Erom Santa Fe to
Leavenworth in 1858, passed through Morton County, in the southwestern
corner of Kansas, in late June. He speaks of water shortages and hot weather,
but just as frequently describes green, grassy valleys and violent thunder-
storms.

In 1842, Fremont led a mapping expedition across the plains to the
Rocky Mountains, following the Oregon Trail to the forks of the Platte and
continuing as far west as the Wind River Mountains of Wyoming. They
returned along the Platte and Missouri rivers. The expedition of 1843-44,
according to Fremont's journal, left from "the little town of Kansas, on the
Missouri frontier, near the junction of the Kansas River with the Missouri
River." The route followed on this expedition was more southern than that
of the previous year, following the valley of the Kansas River. They then
continued west to the head of the Arkansas River and then north and west
to Oregon and California. On the return journey, they crossed the Rockies
near the headwaters of the Arkansas River and followed the Arkansas and
Smoky Hill river valleys back to Independence. The original report of the
expeditions of 1842 and 1843-44 (published by Gales and Seaton, Printers,
Washington, 1845) includes a catalogue of plants collected by Fremont, with
a preface by John Torrey and an appendix containing the hitter's descriptions
of four new genera and 13 new species collected during the second expedition.

Fremont's journal contains frequent references to species in particular
and to the aspect of the vegetation in general. His daily records of latitude
and longitude make it possible to determine with reasonable accuracy his
location at any given time and to compare his observations of flora and soils
with present knowledge of them. For example, on the morning of June 16,

878 The University Science Bulletin

1843, Fremont's party was approximately 265 miles west of the mouth of the
Kansas River (in the vicinity of present Lucas). An entry in his journal
pertaining to that day's travel reads as follows: "After a few hours' travel
over somewhat broken ground, we entered upon an extensive and high level
prairie, on which we camped toward evening at a little stream, where a single
dry cottonwood afforded the necessary fuel for starting supper. Among a
variety of grasses which today made their first appearance, I noticed bunch
grass (festitca), and buffalo grass (seslaria dactyloides). Amorpha canescens
(lead plant) continued to be the characteristic plant of the country, and a
narrow-leaved lathyrus occurred frequently, with a psoralia near psoralia
floribunda, and a number of plants not hitherto met, just verging into bloom."
Fremont and his group had that day emerged from the Dakota sandstone
area of north-central Kansas and entered the High Plains of short-grass
prairie region. All of the plants he describes above are presently known to
occur in the same area.

The journals of these expeditions, diaries and letters of early settlers, and
other historical documents are an aid to determining the native flora of the
central states at the time of the white man's arrival. Since the original land-
scape has by now been largely destroyed by cultivation, over-grazing and
other activities connected with human habitation, these records can be of
value to a person studying natural vegetation areas in general or distribution
of species in particular. Such documents must be used with a bit of caution,
however. Most of the early settlers had no formal training in natural history
and their identification of plants may not always have been correct. Common
names also pose a problem, as they may not always apply to the same plant
in different parts of the country. One must also bear in mind certain
prejudices that the settlers brought with them. To lumbermen or to people
who had just traveled many miles through forests, a region which was
relatively treeless seemed barren and unprofitable. This "forest-mindedness"
of the easterners was one of the most powerful factors in establishing the
"Great American Desert" myth mentioned previously. Another point is
made by Braun (in Steyermark, 1959) who warns that "the unusual was
often emphasized," in these old accounts. At the same time, the usual was
often neglected. The following quotation from Mollhausen's 1858 journal
(see Mollhausen, 1948) illustrates this point well: "On July 5, we continued
our trip along the river valley; the road was good, the weather extremely
pleasant, but there was so little change in the scenery that it could be con-
sidered as non-existent. When we therefore observed indistinct forms of
three or four cottonwood trees on the far horizon or went past islands on
which willows grew, we turned our full attention to them, and found objects
beautiful and pleasant which would have gone unnoticed in other regions."

In 1856 the southern boundary of Kansas was surveyed. Several accounts

An Introduction to the Phytogeography of Kansas 879

of the surveying expedition are available and all include frequent references
to the flora, soils, topography and wildlife of the areas through which the
party traveled. Sufficient landmarks are given in these reports to permit
comparison of the vegetation with that of the present. It is easy to recognize
the Chautauqua Hills, for example, which are described as country "with
a broken and irregular appearance . . . many ridges very rocky and covered
with a dense growth of black jack/' West of the Chautauqua Hills, the
surveying party left the "timbered country" and moved into prairie. "The
dividing ridge ... a plateau about 300 feet above the valley" is the eastern
escarpment of the Flint Hills. The plateau is described as having "sides very
abrupt and rocky" with "limestone near the summit" (Miller, 1932).

In addition to organizing the territories of Kansas and Nebraska, the
bill of 1S54 also repealed the Missouri Compromise of 1820 and during the
era of violence which resulted, little botanizing was done in Kansas. In spite
of this political unrest, some academic progress was made during the 1860's.
In 1863, the Kansas State Agricultural College (now Kansas State Univer-
sity) was established at Manhattan, and Lawrence was chosen as the site
of the University of Kansas. F. H. Snow, one of the first three faculty
members at the University, and his students assembled the first nucleus of
an herbarium. In 1867, "moved by the impulses of the age," 17 Kansas
naturalists founded the Kansas Natural History Society which, in 1871,
became the Kansas Academy of Science (Parker, 1872). Many of the papers
dealing with the Kansas flora have been published in the Transactions of
this academy.

In the 1870's, serious work on the flora of Kansas began and check lists
were prepared. From 1872 to 1884, Reverend James H. Carruth, a staff
member at the University of Kansas, devoted full time to the study of the
state flora. Carruth (1877) published his Centennial Catalogue of the Plants
of Kansas in which he included 1,082 plants and indicated that 40 were
introduced and 142 not known east of the Mississippi. For the first time,
specimens were kept in Kansas rather than sent east or to Europe. Ministers,
physicians, teachers and others sent from a few to over 500 specimens to
Carruth for his study. A series of papers reporting additions to the state's
flora resulted. In 1883, Carruth reported that "the additions now made are
but few. The field has been reaped and henceforth I can only give the
gleanings." Carruth's last paper (1885) brought the check list of Kansas
plants to 1,515, and he stated that "the work of making a catalogue of the
plants of Kansas is so nearly completed, and my salary as a state officer is
so very meager, that I have done but little myself." Just prior to this,
Carruth's disagreements with Snow over Darwinism had precluded further
work by Carruth. Unfortunately, none of Carruth's collections are present
at the University of Kansas and their disposition is unknown.

8<S0 The University Science Bulletin

Carruth's anti-Darwinism did not deter others from floristic studies and
field work. Individuals from over the state began contributing articles on
the flora. Plank (1883) published A Preliminary Notice of the Flora of
Montgomery County, Kansas and Scribner (1885) produced A Contribution
to the Flora of Kansas: Gramineae in which he stated "probably only half
of the grasses in Kansas are known."

During the 1880's, M. A. Carleton and W. A. Kellerman established the
Kansas State Herbarium at Kansas State University. During the 1890's,
A. S. Hitchcock developed a remarkable field program aimed at obtaining
a complete collection of plants from each county of the state. By 1900 a
minimum of 100 species had been obtained from each of the 105 counties.
These collections were the basis of Hitchcock's (1899) Flora of Kansas Maps
which consisted of a series of dot maps giving the distribution of flowering
plants by counties.

Upon the transfer of Hitchcock to the National Herbarium in Washing-
ton, D.C., B. B. Smyth, curator of the State Museum of Natural History in
Topeka, continued further work on the state's flora. Many individuals sent
material to him and he published a series of lists of Kansas plants. In 1900,
he published his Plants and Flowers of Kansas as his fourth contribution.
He planned a fifth and complete list but lived to complete only a third of
the task which was published in 1912. Smyth's herbarium contained over
4,000 specimens and bundles of unlabeled material. In 1926, it was given
to Kansas State University. Unfortunately, specimen evidence for many of
his Kansas reports was not present in his collections.

In the period of 1908-1912, Professor F. U. G. Agrelius was employed
during the summers by the University of Kansas to collect plants in the
northern and western areas of Kansas. His collections, containing many
new state records, formed the basis of the University Herbarium. Agrelius
also salvaged the collections of F. H. Snow, his students and other donors
and put the herbarium in order.

In 1919, F. C. Gates joined the staff at Kansas State University at Man-
hattan and, though trained as an ecologist, became the most productive
worker on the state's flora until his death in 1955. During his tenure, nearly
25,000 specimens of Kansas plants were added to the herbarium. Using the
material gathered earlier through the efforts of Hitchcock, he was able to
produce a series of handbooks on plants of Kansas. His most important
contribution was his Flora of Kansas published in 1940. This work for the
first time included a carefully prepared annotated list of all Kansas vascular
plants, as well as maps giving distribution of each species by counties. He
continued publishing notes, additions and corrections until his death.

In the period from 1933 to 1954, Professor W. H. Horr at the University
of Kansas conducted fieldwork over much of Kansas and by 1954 had ac-

An Introduction to the Phytogeography of Kansas 881

cumulated additions to the University Herbarium totaling several thousand
sheets. He was assisted in the latter phases of the work by R. L. McGregor
who took over the operation in 1954. By 1968, the Kansas specimens in the
University Herbarium numbered over 75,000.

Throughout the history of study of the state's flora, the emphasis has
been on completing the inventory and much work still remains to be done.
Basic and applied ecological studies have been made on prairies and range
lands. Very little has been written, however, on the relationship of the
Kansas flora to that of other areas. Hitchcock (1892) published a short paper
on The Relations of the Composite Flora of Kansas. Of 47 genera studied,
Hitchcock related 31 to the Mexican flora; 28 to the southeast; 23 to the
northeast; and eight to the Rocky Mountain flora. He includes 40 of these
in two to three different floras. Fearing (1952), briefly alludes to regional
affinities of the Kansas flora but does not consider the matter in depth. Mc-
Gregor (1955) states that in the Hepaticae "a relationship to the southern,
southeastern and southwestern hepatic floras is evident/' He lists only one
species which would indicate a relationship to northern floras. According
to Smith (1966), only one of the moss species characteristic of the western
two-thirds of the state is eastern in distribution, and this single species is
abundant only in the Blue Hills physiographic area. Smith states that "from
east to west in the Great Plains there is a decrease in the percentage of mosses
of widespread distribution and an increase in the percentage of mosses of
western affinities" until, in the High Plains area, 60% of the characteristic
species are related to western floras and 40% are of widespread distribution.
Lathrop (1958) presents a carefully prepared study of the flora and ecology
of the Chautauqua Hills in southeastern Kansas. His data indicate that
the flora of that region is influenced primarily by elements of the Texas
Biotic Province to the south but also contains elements from the eastern
and southeastern United States.

METHODS USED IN THIS STUDY

Gates (1940) included 1,872 species in his check list of Kansas plants.
Of these, 779 have not been considered for various reasons: 326 are intro-
duced, 177 are known from a single collection only, and another 175 are too
poorly known to be considered in a study of distribution patterns. An addi-
tional 101 species have been deleted because of nomenclatural changes,
monographs, revisions and other studies. On the other hand, 108 new species
have been added to the list since 1940, bringing the number of species which
would be useful in a survey of distribution patterns to a total of 1,201. For
this study, 479 well-collected vascular species native to Kansas were selected
as having sufficient data available concerning them, and their distributions
in the central states area were mapped by counties. Distribution data were

882 The University Science Bulletin

obtained from the herbaria at the University of Kansas, Kansas State Uni-
versity, the University of Nebraska, the University of Oklahoma and Okla-
homa State University. For distributions in Missouri, the detailed maps in
Steyermark's (1963) Flora of Missouri were used. This information was
supplemented by our own field studies and by general statements of distri-
bution as given in manuals, monographs and other literature. The species
included in this investigation represent 15 pteridophyte genera and 250
angiosperm genera from 85 families.

Once the species were mapped, they were sorted into groups according
to their present centers of frequency within the United States. All plants
considered fell into one of nine groups: eastern, northeastern, southeastern,
southern, southwestern, northern, interior, Great Plains or widespread.
These groups are defined in the section on geographical affinities of the
Kansas flora. Once sorted into these categories, the species were further
subdivided according to the minor patterns or "sub-patterns" which they
displayed within the major patterns. In the final stages of the investigation,
an effort was made to correlate these patterns and sub-patterns with past
and present ecological conditions.

PROBLEMS ENCOUNTERED

In a study of this type which depends heavily upon the research and
collections of previous workers, there are a number of inevitable problems,
particularly in regard to determining the center of frequency of any given
species. When transferring dots from maps in manuals and monographs, it
is impossible to tell whether one dot represents hundreds of native plants or
a single introduced specimen collected in a freight yard. Statements of dis-
tribution as given in manuals are usually very general and do not accurately
indicate a plant's distribution or center of frequency. Remarks on herbarium
labels are quite helpful, but all too frequently information concerning
abundance or general distribution is found wanting. Dot maps also tend
to reflect areas of collecting (or lack of it) and number of specimens ex-
amined. We have relied upon our experience in the field to at least partially
solve this problem.

Species which involve taxonomic problems also present difficulties. Should
Monarda fistitlosa, for example, be mapped as one species ranging from
southwestern Quebec and western New England, south to Maryland, upland
Georgia, and west to British Columbia and Arizona? Or should it be mapped
as two species, Monarda fistidosa with an eastern distribution and Monarda
menthaefolia with a western distribution? In cases such as this one, we fol-
lowed the most recent and most comprehensive treatment of the genus
available. In many groups, of course, the taxonomic problems have not been
solved; in fact, it appears that many are as yet unrecognized. As a matter of

An Introduction to the Phytogeography of Kansas 883

convenience, these species or species complexes were eliminated from this
study.

A third problem arises when one begins to sort maps into groups of like
distribution patterns. All of the plants whose center of frequency is in the
northeastern United States do not have exactly the same area of distribution,
nor do the plants from the southwest, or from any other area. Within any
one group, there is a species with a "minimum area of distribution" and
another with a "maximum area of distribution," so to speak; the remaining
species show a continuous gradation from one extreme to the other. This
can be seen in Figure 1 which shows both the general range of southwestern
species as denned in this paper and the individual areas of distribution of
several species within the group. There is also a continuous gradation be-
tween any two adjacent centers of frequency. Some species are clearly con-
fined to the northeastern United States, while others clearly occur through-
out the eastern half of the country. However, there are other species which,
while occurring south of the boundary between the northeast (Gray's Manual
range) and the southeast (Small's Manual range), do not extend southward
to the Gulf States. Into which series should these species be placed? The
same situation arises when one attempts to determine whether a plant is
primarily southeastern or southern in distribution or whether another is
southern or southwestern. In the case of each individual species, a decision
was made based upon frequency of occurrence at the margin of its range.

GEOGRAPHICAL AFFINITIES OF THE KANSAS FLORA

Eastern Species

Of the 479 species studied, 169 or one-third occur throughout the eastern
half of the United States and typically range from Maine to Florida, west
to Minnesota and Texas (Fig. 2). Fifty-nine of these are woody plants.
According to their habitats, the eastern species can be broadly categorized
as elements of the forest, forest border and open areas, stream banks and
marshes, tallgrass prairie, or sandy open areas and gravelly hillsides (see
Table 1). Species which are most common in forested areas include

Table 1. Habitat preferences of eastern species.

Habitat No. species

1. Forest 64

2. Forest border and open areas 37

3. Stream banks and marshes 27

4. Tallgrass prairie 26

5. Disturbed areas, sandy open places and gravelly hillsides 6

6. Occur in both (2) and (3) or in (3) and (4) 9

169

884 The University Science Bulletin

Asplenium platy neuron, Cormts florida, Hystrix patula, Osmorhiza
longistylis, Ostrya virginiana, Podophyllum peltatum, Quercus stellata,
Sanguinaria canadensis and Vaccinium stamineum. Species usually associated
with forest borders or open areas include Clematis virginiana, Corylus
americana, Anonymous atropiirpureus, Panicum sphaerocarpon, Phytolacca
americana, Quercus prinoides, Sassafras albidum and Silene stellata, while
Anemone virginiana, Betula nigra, Carex fran\ii, Onoclea sensibilis and
Salix caroliniana occur primarily along stream banks and in marshy areas.
Tallgrass prairie species include Camassia scilloides, Castilleja coccinea,
Eryngium yuccifolium, Plypoxis hirsuta and Penstemon digitalis. Species
such as Aristida oligantha, Helianthus mollis and Rhus copallina are usually
found in disturbed areas, sandy open places and on gravelly hillsides, while
Carex cephalophora, C. granulans and Juncus biflorus may occur in moist
areas along forest borders, stream banks and marshes or moist grasslands.

When one considers the distributions of these eastern species within
Kansas and within the Great Plains as a whole, five types of patterns become
immediately obvious. The first is displayed by Carex rosea and Orchis
spectabilis which are representative of a small group of plants that reach
their westernmost limits of distribution in northeastern Kansas. Both are
forest species and are rarely collected.

A second small group of eight species extends a short distance into the
northeastern and southeastern corners of the state but not into east-central
Kansas. These eight — Amelanchier arborea (Fig. 3), Anemonella thalic-
troides, Cubelium concolor, Dicentra cucullaria (Fig. 4), Helianthemum
bic\nellii, Quercus imbricaria, Sanguinaria canadensis, and Uvularia grandi-
flora — occur primarily in forests or forest border situations and tend to skirt
the arm of tallgrass prairie which extends eastward from the Flint Hills
through Coffey, Anderson and Linn counties, Kansas, into Bates County,
Missouri.

Forty-one species, approximately one-fourth of the eastern species studied,
are limited to the southeastern corner of the state (Fig. 5). Seven of these
— Alnus serrulata, Ascyrum hypericoides, Botrychium dissectum var. dis-
sectum, Corn us florida, Gillenia stipulata, Pontederia cordata and Vaccinium
stamineum — occur in Cherokee County only, where they are accommodated
by a small area (approximately 20 square miles) of Ozarkian topography
with oak-wooded hills and cherty soils. The remaining 33 are more wide-
spread, with some distributed westward into the Chautauqua Hills and into
the southern tip of the Flint Hills in Cowley County.

A fourth type of pattern is displayed by a group of 74 species whose
western-most boundaries are roughly north-south lines within Kansas. In
Nebraska and the northern Great Plains some of these tend to extend as
much as 150 miles farther west than in Kansas. Some species such as

An Introduction to the Phytogeography of Kansas 885

Carex retroflexa, Rosa Carolina, Cystopteris protrusa and Crataegus crus-galli
reach only into eastern Kansas, while others such as Populus deltoides,
Gkditsia triacanthos and Celtis temtijolia cross the entire state. Most fall
somewhere in between these two extremes. Westward across Kansas, the
number of eastern species present gradually decreases, as does the number
of individuals within any one species. Within this particular pattern, there
are two minor sub-patterns. Some plants, particularly those of forest, forest
border or stream bank, tend to move westward along river systems, producing
an effect of feathering or dove-tailing. Others have their westward dispersal
abruptly curtailed. In the case of Podophyllum peltatum, Quercus marilan-
dica, Q. velutina, Rosa setigera and others, this limitation seems to correspond
with the eastern edge of the Flint Hills, while with species such as Aristida
oligantha and Parthenocissis quinquejolia it seems to parallel the eastern
edge of the High Plains. It should be emphasized, however, that it is impos-
sible to find any single factor which limits all or even a large number of
these eastern species in their westward distribution. They simply do not
march as a group to some one barrier and then stop. Instead, there is a
gradual reduction in the number of species and of individuals from east to
west as those less able to withstand the habitats of central and western
Kansas drop out.

The fifth pattern displayed by the eastern species is one of a great arc
northward and westward around the southern half of the Great Plains,
sometimes extending southward again in the Rocky Mountains (Fig. 6).
In Nebraska, the southern boundary of this arc approximates the southern
extent of glaciation. Of the 44 species which exhibit this pattern, only 29
have a continuous distribution across the plains. These include Mimulus
nn gens, Impatiens ca pen sis, Carex bland a, Hypoxis hirsuta, Bra m us pur guns
and Smilax herbacea. Within Kansas, they extend westward various dis-
tances. The remaining 15 extend westward into eastern Kansas and Nebraska
and sometimes into eastern North and South Dakota and then have disjunct
populations in the canyons of the Niobrara River in north-central Nebraska,
in the Black Hills of South Dakota and/or in the Rocky Mountains, suggest-
ing that at one time they also had a continuous distribution across the plains.
The more mesic climate of the early Hypsithermal Interval could have al-
lowed these species to migrate westward along the river systems or perhaps
across the uplands as well. The drier climate of the later Hypsithermal Inter-
val and of Recent times apparently eradicated many suitable habitats in the
plains area and left relict populations stranded in the regions already men-
tioned. Species in this group include Anemone virginiana (Fig. 7), Aquilegia
canadensis (Fig. 8), Arisaema triphyllum, Corylus americana, Hystrix
patula, Osmorhiza longistylis (Fig. 9) and Quercus muhlenbergii.

886 The University Science Bulletin

Northeastern Species

A second major pattern of distribution is exhibited by a group of 32
species whose present center of frequency is in the northeastern United
States. Plants in this group typically range from Maine and southeastern
Ontario west to eastern North Dakota and south to Virginia, Kentucky,
Missouri and eastern Kansas (Fig. 10). The habitats most frequently oc-
cupied by these northeastern species are upland forest; forest border and
open or disturbed situations; lowland forest; tallgrass prairies, hillsides and
thickets; and aquatic habitats (see Table 2). Upland forest species include
Acer saccharum , Celtis occidentaUs,HydrophyUum appendiculatum, Quercus
borealis var. maxima, and Viola sororia. Ritbus allegheniensis and R.
flagellars are found in forest borders and open or disturbed areas, while
Carex hitchcoc\iana and Quercus palustris occur in lowland forests. Species
which occur primarily in tallgrass prairies, on hillsides and in thickets in-
clude Rosa blanda, Ceanothus ovatus and Sporobolus heterolepis. A rep-
resentative aquatic species is Potamogeton obtusifolius.

Table 2. Habitat preferences of northeastern species.

Habitat No. species

1 . Upland forest 14

2. Forest border and open or disturbed areas 2

3. Lowland forest 7

4. Tallgrass prairie, hillsides and thickets 8

5. Aquatic 1

32

Within Kansas, the northeastern species can be divided into three minor
patterns or sub-patterns. Six species — Carex hirtijolia, C. hitchcoc1{iana, C.
cristatella, C. sparganioides, Hydrophyllum appendiculatum (Fig. 11) and
Rosa blanda (Fig. 12) — are restricted to the northeastern corner of the state.
Three species — Carex squarrosa (Fig. 13), Physocarpus opidifolius (Fig. 16),
and Vaccinium vacillans (Fig. 14) — occur only in southeastern Kansas al-
though they are found throughout much of Missouri. Most of the remaining
22 occur within the eastern third of Kansas. Examples include Quercus
palustris, Rub us allegheniensis, Carex jamesii and Viola sororia. Except for
Parthenocissus vitacea, Ceanothus herbaceous and Celtis occidentalis, species
displaying this sub-pattern do not extend westward much beyond the Flint
Hills.

In all of the minor groups just discussed, there are some species which
have the same tendency to arc northward and westward in Nebraska as do
some of the eastern species described in the previous section of this paper.
Examples include Carex crawei, C. normalis, C. bicl{nellii, Sporobolus

An Introduction to the Phytogeography of Kansas 887

heterolepis, Parthenocissas vitacea, Ceanothas herbaceous and Tilia ameri-
cana (Fig. 15). Physocarpus opitlijolius (Fig. 16), restricted within the state
to Cherokee and Neosho counties in the southeast corner, has disjunct pop-
ulations in north-central Nebraska, the Black Hills and the Rocky Mountains
of Wyoming and Colorado. Liparis loeselii, of the north and east has been
collected in Shannon County, Missouri, Pottawatomie County, Kansas, and
north-central Nebraska. It is described by Steyermark (1963) as "another one
of the isolated relict survivors which has been stranded in the swampy
meadows . . . following one of the advances and retreats of an ice sheet
during Pleistocene times."

Southeastern Species

Thirty-five of the 479 species studied occur primarily in the southeastern
United States and typically range from Florida west to Texas, north to
Virginia, Kentucky, Missouri and eastern Kansas (Fig. 17). This area in-
cludes the southern half of the Appalachian Upland and the Ozarkian
Plateau which Fernald (1931) describes as the oldest large section in eastern
North America which has been available for occupation by flowering plants.
Adams (1902) and Deevey (1949) discuss the southeastern United States
as a Pleistocene refugium and a center of geographical distribution of flora
and fauna.

The general types of habitats in which these species are found are as
follows: forest; forest border and open areas; tallgrass prairie and rocky
hillsides; and streambanks and marshes (see Table 3). Examples of forest
species are Carya illinoensis, Celtis laevigata, Hexalectris spicata, Hex decidua,
Phoradendron flavescens and Vacciniitm arboreum. Plants usually found
in forest borders or disturbed and open areas include Ampelopsis cordata,
Calycocarpum lyoni, Crate gits viridis, Passi flora incarnata and Verbesina
virginica. Bumelia lanuginosa and Nothoscordum bivalve usually occur on
hillsides or in tallgrass prairie, while Cyperus densicaespitosus, Dicliptera
brachiata and Hibiscus lasiocarpus are found along streambanks and marshes.

Within Kansas and the Great Plains, the southeastern species display only
a single major distribution pattern. All but a few, such as Sesbania exaltata,

Table 3. Habitat preferences of southeastern species.

Habitat No. species

1. Forest 10

2. Forest border and open or disturbed areas 8

3. Tallgrass prairies and rocky hillsides 5

-\. Stream banks and marshes 12

35

888 The University Science Bulletin

f uncus validus and /. scirpoides which occur sporadically in Kansas and
Missouri, arc across the southeastern corner of Kansas (Fig. 18). Some
species — Ulmits alata, V accinium arboreum, Calycocarpon lyoni and Hibiscus
lasiocarpus, for example — occur only in Cherokee County, while others such
as Nothoscordum bivalve and Ampelopsis cordata occur as far west as Barber
and Kiowa counties, Kansas. The northwestern limits of most of the south-
eastern species lie somewhere between these two extremes.

Southern Species

Of the species studied, 42 have a center of frequency south of Kansas.
The general area of their distribution is from Texas and northeastern Mexico
north to Oklahoma and Kansas (Fig. 19), although some have a broader
distribution extending west to Arizona, east to Florida and north to Virginia,
Ohio and Indiana. Only three woody species — Cephalanthus occidentalis,
Sapindus drummondii and Primus mexicana — are included in this group.
The southern species tend to occur mainly in the following type of habitats:
midgrass prairie hillsides and canyons; tallgrass prairie and rocky banks;
sandy tallgrass prairie; and streambanks, marshes and moist ditches (see
Table 4). Plants which occur on midgrass prairie hillsides and in canyons

Table 4. Habitat preferences of southern species.

Habitat No. species

1. Midgrass prairie hillsides and canyons 5

2. Tallgrass prairie and rocky banks 19

3. Sandy tallgrass prairie 12

4. Streambanks, marshes and moist ditches 6

42

include Androstephium coeruleum, Echinacea an gusti folia var. an gusti folia,
Sapindus drummondii, Nama stevensii and Oenothera o/(lahomensis. The
latter two are particularly common in the red gypsiferous soils of Clark,
Comanche, Barber and Harper counties, Kansas, and of western Oklahoma
and Texas. Aster sericeus, Acacia angustissima, Callirhoe involucrata,
Desmanthus illinoensis, Eryngium leavenworthii, Echinacea atrorubens,
Penstemon cobaea, Nemastylis geminiflora, Sporobolus pyramidatus and
Tradescantia tharpii are species frequently found in tallgrass prairie and on
rocky banks. Species of sandy tallgrass prairie include the following: Ber-
landiera texana, Cleomella angustifolia, Croton glandidosus, Gaillardia
fastigiata, G. suavis, Liatris squarrosa var. glabrata, Monarda citriodora and
Primus mexicana. Plants usually found along streambanks, in marshes or
moist ditches include Cephalanthus occidentalis, Eleocharis montevidensis,
Carex fissa, Marsilea mucronata, and Heteranthera reniformis.

An Introduction' to the Phvtogeography of Kansas 889

Within Kansas and the Great Plains, four sub-patterns of distribution
are apparent. There is, however, a great deal of variation within each of
these sub-patterns and much overlap among them. A group of six species —
Androstephium coeruleum (Fig. 20), Gaillardia jastigata, G. suavis (Fig.
21), Sporobolus pyramidatits, Nama stevensii and Oenothera ofylahomensis
— come into southwestern or south-central Kansas and do not occur north of
Kansas. Cleomella angustifolia (Fig. 22) has a similar distribution but also
has a disjunct population in southwestern Nebraska.

A second sub-pattern is displayed by nine species which occur primarily
in eastern and southeastern Kansas and southwestern Missouri where they
inhabit rocky (especially calcareous) tallgrass prairies, limestone glades and
bald knobs. Examples include Aster sericeus, Camassia angusta (Fig. 23),
Nemastylis geminiflora, Parthenium hispidum, Cissus incissa and Acacia
angustissima (Fig. 24). Aster sericeus (Fig. 25) is the only member of this
series which extends north of Kansas.

Twelve species, including Croton glandidosus, Eryngium leaven worthii,
Mammillaria missouriensis, Liatris squarrosa var. glabrata, Callirhoe in-
volucrata and Monarda citriodora, occupy drier habitats than do the mem-
bers of the previous two groups, and are more widely distributed than the
members of those groups. Most of these species occur throughout central
and eastern Kansas and some are found in Missouri; about half of them ex-
tend north of Kansas in the Great Plains.

The fourth group consists of 10 species which are more mesic than the
three previous groups and which are more widely distributed than the first
two groups. Seven of these are found primarily in the eastern half of Kan-
sas and in much of Missouri, Arkansas, Oklahoma and Texas, but some
extend northward into Nebraska and the Dakotas as well. Species in this
group include Cephahinthus occidentalis, Helianthus salicijolius, Ophioglos-
sum engelmannii. Primus mexicana and Marsilea mucronata.

It is perhaps noteworthy that the southern species seem to be taxonomically
distinct (more so than in any of the other major groups described). The
only two exceptions to this would be Berlandiera texana and Lesqtterella
gracilis. These two have areas of distribution intermediate between the first
and second sub-patterns described in this section, and it appears that there
may be two varieties of each involved.

Southwestern Species

Eighty-two of the 479 species studied have their present center of frequency
in the southwestern United States. These have a typical range of Arizona
and New Mexico, east to Texas, north to Colorado and Kansas and some-
times extend into the northern Great Plains (Fig. 26). Mimosa borealis,
Prosopis juliflora, Primus gracilis and Rhus trilobata are the only woody

890 The University Science Bulletin

Table 5. Habitat preferences of southwestern species.

Habitat No. species

1 . Shortgrass prairie 36

2. Midgrass (often rocky) prairie hillsides 11

3. Sandy tallgrass prairie 31

4. Alkaline prairie 4

82

species which come in with the southwestern flora. The southwestern
species are most comonly found in the following habitats: shortgrass
prairie; midgrass (often rocky) prairie hillsides; sandy prairie; and
alkaline prairie (see Table 5). Those usually found in shortgrass prairie
include Aristida divaricata, A. Ion gi seta, Baccharis wrightii, Buchloe dacty-
loides, Ditaxis humilis, D. mercurialina, Eriochloa contractu, Gaura villosa,
Gilia acerosa, Triodia pilosa and Zinnia grandiflora. Species which are
found primarily on midgrass, often rocky, prairie hillsides include Berlandiera
lyrata, Castilleja citrina, Hybanthus verticillatus, Mimosa borealis, Oenothera
greggii, Teucriitm laciniatiun and Rhus trilobata. Examples of plants which
occur on sandy prairie are Artemisa filifolia, Baccharis salicina ,Bouteloua
eriopoda, Chloris virgata, Eriogonum longifoliutn, Gaillardia palchella,
Haplopappus annuits, H. ciliatus, Heliotropium convolvidaceum, Linittn
pratense, Penstemon ambiguus and B silo strophe villosa. Representatives of
alkaline prairie plants are Distichlis stricta, Eustoma russellianiim, Banicum
obtusum and Vernonia mar gin at a.

Within the southwestern group, there are three sub-patterns of distribu-
tion. Forty-two are found within the southwestern quarter of the state and
do not occur north of Kansas. Some of these arc across the southwestern
corner of the state. Others such as Castilleja citrina (Fig. 27) enter farther
east in the red gypsiferous soil areas of extreme south-central Kansas or
in the sandy soils south of the Cimarron and Arkansas rivers. Species placed
in this group include Aristida divaricata, Berlandiera lyrata, Convolvulus
incanus, Dalea jamesii, D. lanata, D. nana, Eriochloa contractu, Eriogonum
longifoliutn, Hilaria jamesii, Opuntia imbricata, Brunus gracilis and Zinnia
grandiflora.

A second group of 37 species occurs within the western half of the state.
These tend to extend north of Kansas (Fig. 28). Included in this group are
Abronia fragrans, Aristida longiseta, Artemisia filifolia, Buchloe dactyloides,
Engelmannia pinnatifida, Eurotia lanata, Gutierrezia sarothrae, Haplopappus
annuus, Hedeoma drummondii, Iva xanthifolia, Lygodesmia juncea, Opuntia
polycantha and Ratibida tagetes.

A third minor pattern is exhibited by Eustoma russellianiim, Gilia longi-
flora (Fig. 29) and Sporobolus airoides which are found in the sand hill areas

An Introduction to the Phytogeography of Kansas 891

of southwestern and south-central Kansas and then skip to the sand hills of
Nebraska.

Northern Species

Only 13 of the 479 species studied have a center of frequency north of
Kansas. Some span the continent from east to west, others are more centrally
distributed, but all come into Kansas from the north (Fig. 30). Two of
the 13 are woody; 6 are grasses. These species occur in the following general
types of habitats: marshes; tallgrass prairie; shortgrass prairie; canyons,
prairie hillsides, and streambanks; and alkali and salt flats (see Table 6).

Table 6. Habitat preferences of northern species.

Habitat No. species

1 . Marshes 1

2. Tallgrass prairie 2

3. Shortgrass prairie 6

4. Canyons, prairie hillsides and streambanks 3

5. Alkali and salt flats 1

13

Spartina pectinate! occurs in marshes; Stipa spartea and Penstemon grandi-
florits, in tallgrass prairie; and Hymenopappits filifolius, Linum rigid urn var.
compactum , Stipa viridula, S. comata and Poa arid a are species of shortgrass
prairies. Plants usually found in canyons, on prairie hillsides or along stream-
banks include Muhlenbergia racemosa, Rosa woodsii and Symphoricarpos
occidentalis. Heliotropium spathulatum is representative of species common
on alkali and salt flats.

Within Kansas and the Great Plains, a tendency toward three sub-
patterns is apparent. Heliotropium spathulatum, Rosa woodsii, Poa arida,
Stipa comata (Fig. 31) and S. viridula (Fig. 31) are primarily restricted to
the western half of the plains, while Penstemon grandiflorus, Spartina
pectinata and Stipa spartea (Fig. 31) occur in most abundance within the
eastern half. Muhlenbergia racemosa is representative of the third and largest
group of species which are widespread throughout the Great Plains but are
more abundant in the north.

Interior Species

Of the species studied, 33 have an inland distribution which in general
ranges from Michigan and Indiana west to Minnesota, Nebraska, Kansas,
Oklahoma, Texas, Arkansas and Missouri (Fig. 32). Ten of these are woody.
The plants in this group are more difficult to divide into habitat types than
any other group, primarily because they occur in a wider variety of habitats.

892 The University Science Bulletin

Table 7. Habitat preferences of interior species.

Habitat No. species

1. Moist woodlands, creek valleys and canyons 7

2. Open forest, thickets and rocky banks 8

3. Forest border, rocky hillsides and tallgrass prairie 8

4. Tallgrass prairie - - - 1

5. Disturbed areas, fields and rocky hillsides 2

6. Sandy tallgrass prairie, hillsides and limestone glades 1

7. Moist tallgrass prairie, ditches and marshes 6

33

The following list of habitat types reflects some of the difficulties involved:
moist woodlands, creek valleys and canyons; open forest, thickets and rocky
banks; forest border, rocky hillsides and tallgrass prairie; tallgrass prairie;
disturbed areas, fields and rocky hillsides; sandy tallgrass prairie, hillsides
and limestone glades; and moist tallgrass prairie, moist ditches and marshes
(see Table 7). Aesculus glabra var. arguta, Carex shortiana, Gymnocladus
dioica, Viburnum prunifolium and Forestiera acuminata are found in moist
woodlands. Species which occur in open forest, thickets and on rocky banks
include Carex microdonta, Prunus hortulana, P. munsoniana, Pyrus ioensis,
Rhamus lanceolata and Kibes missouriense. Plants usually associated with
forest borders, rocky hillsides and tallgrass prairie include Amorpha canes-
cens, Asclepias hirtella, Com us drummondii, Clematis pitcheri, Echinacea
pallida and Habenaria leucophaea. Phlox pilosa var. fulgida is common in
tallgrass prairie, and Oenothera triloba and Heliotropium tenellum are us-
ually found in disturbed areas, fields and on rocky hillsides. Liatris squarrosa
var. hirsuta occurs in sandy tallgrass prairies, hillsides and limestone glades,
while Carex ar\ansana, C. triangularis, Juncus interior and Bidens polylepis
are frequently found in moist tallgrass prairie, ditches and marshes.

Within Kansas, the interior species exhibit four of the same sub-patterns
found within the group of eastern species. Some, including Habenaria
leucophaea, Pyrus ioensis, Rhamnus lanceolata (Fig. 33) and Ribes mis-
souriense, arc across the northeast corner of the state and tend to occur farther
westward in Nebraska than in Kansas. Another subgroup of ten species
including Isoetes butleri, Forestiera acuminata, Carex triangularis, and Phlox
pilosa var. ozarkana arc across the southeastern corner of the state (Fig. 34).
The remaining species enter Kansas directly from the east. Examples include
Carya texana, Viburnum prunifolium , Aesculus glabra var. arguta (Fig. 35),
Asclepias hirtella, Bidens polylepis and Clematis pitcheri (Fig. 36). Liatris
squarrosa var. hirsuta (Fig. 37) enters from the east, but does not occur in
the tallgrass prairie peninsula (previously described) which extends from
the Flint Hills through east-central Kansas into Bates County, Missouri.
As within other major groups of plants already described, there is a tendency

An Introduction to the Phytogeography of Kansas 893

for the interior species to skirt the High Plains areas of western Kansas and
to cross the plains farther north in Nebraska.

Widespread Species

Of the 479 species studied, 67 are so widespread in distribution that it
seems impossible to associate them with any one of the centers of frequency
already discussed. Some of these such as Typha latijolia, T. angustifoha,
Sorghastrum nutans and Salix interior are abundant throughout much of
their range. Others such as Carex buxbaumii, Potamogeton gramineus, and
Azolla mexicana are collected only at scattered stations. Acer negundo,
Salix amygdaloides and S. interior are the only woody plants placed in this
group. The plants in this group occur in the following types of habitats:
forest; forest border to grassland; streambanks and ravines; open areas of
grassland, disturbed areas, and fields; marshes and other wet places; and in
aquatic habitats (see Table 8). Botrychium virginianutn, Asplenium trichom-

Table 8. Habitat preferences of widespread species.

Habitat No. species

1 . Forest

2. Forest border to grassland 5

3. Streambanks and ravines

4. Open areas of grassland, disturbed areas and fields .... 31

5. Marshes and other wet places °

6. Aquatic 12

67

anes and monotropa uni flora occur most often in forests, while Selaginella
rupestris, Celastrus scandens, and Danthonia spicata are usually found any-
where from forest border situations to grassland. Acer negundo, Prunella
vulgaris, Salix amygdaloides and S. interior are found along streambanks
and ravines. Species of moist open areas of grasslands, disturbed areas, and
fields include Carex buxbaumii, Cyperus acuminatus, Myosurus minimus,
Juncus torreyi, J. dudleyi, Ecjuisetum laevigatum and others. Plants usually
found in drier open areas in grasslands, disturbed areas and fields are
Leptoloma cognatum, Sorghastrum nutans, Leptochloa fascicularis, Glyc-
yrrhiza lepidota and Festuca octoflora. Plants which occur primarily in
marshes or wet places include Heteranthera hmosa, Phragmttes communis,
Scripus americanus and Sparganium eurycarpum. Aquatic species include
Azolla mexicana, Ceratophyllum demersum, Najas guadalupensis, Potamo-
geton diversijolius, P. foliosus var. macellus, P. gramineus, P. illinoensis and
Ruppia maritima var. rostrata.

Within this group 17 species arc northward and westward around the
southern plains in the manner already described for species in other groups.

894 The University Science Bulletin

Species which display this pattern include Acer negundo, Adiantum pe datum,
Antennaria neglecta, Aster novae-angliae, Carex nebras\ensis, Celastrus
scandens, Danthonia spicata, Equisetum arvense, Koeleria cristata and
Prunella vulgaris.

The other 50 species present a miscellanea of individual patterns. All but
one of the aquatic species studied fall into this group. These tend to have
a scattered distribution, and are dependent primarily on the presence of
ponds, lakes and other bodies of water. A number of relatively weedy species
and plants of open habitats also are included in this group. Examples of
these include Draba reptans, Glycyrrhiza lepidota, Chloris verticillata, and
Cuscuta pentagona.

Species Endemic to Portions of the Great Plains

For purposes of this study, the Great Plains is denned as the region
which extends from the eastern base of the Rocky Mountains to the decidu-
ous forests of the east and from the Dakotas south to the northern edge of
the Edwards Plateau in Texas. Of the 479 species studied, only six (from
six different families) are confined to this region. None of the six occurs
throughout but, instead, each is restricted to a smaller area within the
Great Plains.

Three of the species — Aster fendleri (Fig. 38), Oenothera jremontii (Fig.
39) and Scutellaria resinosa (Fig. 40) — occur on chalk cliffs or dry, rocky
calcareous hillsides and escarpments. Clematis jremontii (Fig. 41) which,
as mentioned previously, has a disjunct population in east-central Missouri,
is found primarily in the same type of habitat but sometimes occurs in rocky
or dry midgrass prairie as well. Phlox olijahomensis (Fig. 42) occurs on
gypsiferous soils in Woodward County, Oklahoma, the type locality, but in
Kansas it is found in rocky limestone tallgrass prairie in the southern Flint
Hills. ] uncus brachyphyllus (Fig. 43), the most widely distributed of these
Great Plains species, is a sandy tallgrass prairie plant which also occurs on
upland sandstone glades and sandy prairies in Missouri.

The distributions of taxa closely related to these species suggest that at
least three of them may have been associated at one time with floras to the
south of Kansas. According to Erickson (1945), Clematis jremontii var.
riehlii is restricted in Missouri "to glades, rocky barrens which occur on
south- and west-facing slopes of otherwise wooded ridges." Floristically,
Erickson relates these glades to the shale barrens of the Appalachians, to
the cedar glades of the Nashville basin in Tennessee, to portions of the
grasslands of Kansas and Nebraska, and to glade-like grassy areas in the
Arbuckle Mountains of Oklahoma and the Edwards Plateau of Texas. On
the basis of the distribution of other plants associated with the glades, par-
ticularly Oenothera missouriensis, Erickson theorizes that "the two Clematis

An Introduction to the Phytogeography of Kansas 895

populations may have been connected by way of the Edwards Plateau. The
separation into two populations may have occurred during the semi-arid
period of the late Pleistocene, or, in view of the importance of competition
from grasses, during the warmer, moister period which followed."

Scutellaria resinosa also has affinities to the south. It occurs in Kansas,
where it is morphologically distinct and homogeneous, and also in northern
Texas and western Oklahoma, where it becomes more variable and difficult
to distinguish from S. wrightii. The latter species occurs primarily in the
Edwards Plateau region, where it is morphologically distinct, and extends
into north-central Texas and southwestern Oklahoma (Epling, 1942).

Wherry (1955) regards Phlox okjahomensis as a descendant of an offshoot
from P. speciosa var. occidentalis (now restricted to the western and north-
western United States) which migrated far southeastward. "Destroyed in
most of its range by the Pleistocene ice steets and xerothermic maxima, it
managed," according to Wherry, "to survive the last of these in the western
Ozark region, and has subsequently been spreading northwestward." Another
possibility might be that the ancestral species was centered in northern Mexico
and/or the southwest and migrated northward in two directions — north-
westward along the west coast and northeastward through Texas and Okla-
homa to Kansas.

DISCUSSION AND CONCLUSIONS

With distribution data available for 479 species, one is tempted to speculate
about the factors which define areas of distribution and, certainly, about the
origins of the Kansas flora. There are, however, certain inherent dangers in
such speculation. When attempting to arrive at broad generalizations it is
always possible to misplace emphasis or to arrive at unwarranted conclusions
because of the fragmentary knowledge of the distribution and biology of
many species. Nevertheless it does seem possible and worthwhile to inter-
pret the major recurring patterns and sub-patterns in the light of past geo-
logical events and present environmental conditions.

Although the species studied have been divided into groups according to
their centers of frequency, one finds that some sub-patterns are duplicated
by members of more than one group. It would appear that once species
reached the plains area, they adjusted genetically to the environment without
regard to their origin or to the species group with which they arrived. Each
species has developed some degree of tolerance which has enabled it to com-
pete and survive in the new association of which it has become an integral
component.

One of the most interesting patterns revealed is the northward and west-
ward arc across the Great Plains, a pattern displayed by species from the east,
northeast, southeast, and interior groups as well as by a number of widespread

896 The University Science Bulletin

species. Braun (1955) mentions this distribution pattern briefly but does not
consider it in detail. Of 56 plants which follow this pattern, 45 are elements
of either forest, forest border to open areas, streambanks and marshes or
moist grasslands. The southern edge of this arc, as mentioned previously,
corresponds roughly with the southern limits of glaciation in Nebraska and
seems to reflect present climatic and edaphic conditions as well. For many
species, the southern edge of this arc passes through the sand hills of Nebraska,
a tallgrass prairie region characterized primarily by a dune topography with
little surface drainage. The loose sandy soils deposited over Brule clay during
the Pliocene and Pleistocene (Tolstead, 1942) readily absorb rain water, thus
reducing water loss by runoff and evaporation. The impenetrable layer of
clay beneath the sand keeps the water table relatively close to the surface
and results in numerous springs and shallow lakes. The sandhills are also
located in a region which, according to Borchert (1950), has a somewhat
less variable summer rainfall from year to year than does the surrounding
plains region. If one plots the mean annual water loss due to transpiration
and evaporation in percent of rainfall for the central states (Fig. 44), there
can be seen a westward extension of the 95% isopleth in northern Nebraska
and South Dakota, bridging the plains area between eastern Nebraska and
the sandhills and corresponding roughly with the arching pattern of some
plant species. North of Nebraska, increased relative humidity, reduced
average temperatures and shorter growing season increase effectiveness of
rainfall even though the amount of prepicitation which falls is actually less
than in the southern plains. These factors together create mesic conditions
in the midst of an area which — with an average annual rainfall of 18 inches
and with high temperatures, low humidity, prevailing south winds and
frequent droughts during the summer — actually has a semi-arid climate.
South of the arc, there are few habitats which mitigate the effects of the xeric
southwestern climate sufficiently to allow for the survival of these mesic
species. Collections of some of these species to the south into Kansas indicate
that they are rare in the area and are either disappearing or are occasionally
introduced and should be thought of as waifs.

As a group, the species which have disjunct populations within the arc
are even more mesic (and hence more restricted in occurrence) than those
which are continuous throughout the arc. Plants from the eastern deciduous
forest (and from the Rocky Mountains to the west) find suitable habitats in
the precipitous canyons of the Niobrara River and in the Black Hills. Both
areas are supplied with springs and streams and afford protection from hot,
drying prairie winds. Humidity within the canyons is higher than that of
the surrounding uplands, and summer temperatures are lower. The more
mesic nature of these areas also reduces damage from winter drought. It
seems likely that these mesic species migrated across the prairie during the

An Introduction to the Phytogeography of Kansas 897

early Hypsithermal Interval of the Pleistocene, at which time they were
probably more widespread than at present. During the later and drier
Hypsithermal Interval, they apparently became localized in canyons and
river valleys and perhaps in moist swales among higher sand dunes where
they occur today as disjunct colonies.

The species which enter Kansas from the north seem to be limited
primarily by the same factors which most influence distribution of species
from the east — relative humidity, average temperatures and rainfall effectivity,
among others. It is interesting to note that some introduced species have
been observed to migrate into Kansas from the north. Gates (1940) reported
Card nits nutans (nodding thistle) from Washington and Marshall counties.
Since 1945, the species has increased in abundance. It has become a serious
pest over all of eastern Kansas and by 1968 had moved into Oklahoma. It
would be interesting to know whether this progressive movement has been
due to development and natural selection of races tolerant of more southern
environment or if it is merely a migration.

A third pattern — the tendency to occur through central Missouri and
across the southeast corner of Kansas — is displayed by species from the east,
northeast, southeast and interior groups. These species in general are tax-
onomically distinct. There are a number of factors, some seemingly un-
related to one another, which share parallel arching boundaries through
southeast Kansas, angling eastward or northeastward through Missouri.
Among these are elevation, topography, surface geology and soils (Schoewe,
1949). Climatic maps by Borchert (1950) reveal clines in amount of winter
rainfall, percent of normal precipitation which falls during drought years
and average departure from normal temperature in July of major drought
years which seem to correlate with (or at least to coincide with) this distri-
bution pattern. For some plants, the northern boundary in Missouri is the
Missouri River or the approximate limit of maximum glaciation which, in
that state at least, roughly parallels the climatic clines described by Borchert.
Steyermark (1959), Braun (1955) and others have discussed plant distribu-
tion in relation to the glacial boundary in some detail. For some species,
it is possible to point to glaciation as a causal mechanism for existing distri-
bution patterns. For other species, the correlation seems to be a geographical
one but not necessarily a causal one. Often, however, there is a tendency to
assume that there must be some sort of cause and effect relationship present
which we are unable to see and then, of course, to continue looking for such
a connection. Perhaps this is being too narrow-minded. If we believe Glea-
son's (1922) statement that the "glacial climate differed from the modern
not so much in kind as in degree," then it seems possible that climatic patterns
during the Pleistocene were similar to those illustrated by Borchert for the
present. Might it not be that the same climatic factors which presently in-

898 The University Science Bulletin

fluence distribution patterns also limited the southward extent of the glaciers
and hence the location of the glacial boundary ? In particular, we are think-
ing of the warm dry southwesterly winds which prevail throughout much
of the year in the Great Plains, along with high summer temperatures, low
humidity and other factors which represent the influence of the southwest
upon the climate of the Great Plains. If this were true, the boundaries of
both the glaciers and the species ranges would be due to the same climatic
factors, rather than the species boundary being dependent upon the glacial
boundary.

The general area of distribution of the interior species suggests that these
plants may have survived in the Ozarkian region of Missouri and northern
Arkansas during the Pleistocene glaciations and late Hypsithermal Interval
and have since radiated out from that region. Presently, they occupy a region
where the assigned limits of the eastern, northeastern, southeastern, southern,
northern and Great Plains centers of frequency come together. The interior
species occur in a larger number of habitats than any other group described,
although the individual species tend to be rather conservative in habitat
preference and occur in fewer habitat types.

The southwestern and southern species are thought to be the oldest mem-
bers of the Kansas flora and ones which were the least affected in their dis-
tribution by the Pleistocene glaciations. It seems likely that the ancestors of
the southwestern species were members of the Madro-tertiary flora which
came into the plains region during the late Tertiary. Being better adapted
for a xeric habitat, they were able to migrate into the Prairie Peninsula dur-
ing the late dry Hypsithermal Interval. Relict colonies of prairie which
remain scattered through the midwestern states today include the following
southwestern and southern species: Opuntia jragilis, Sporobolus heterolepis,
Cristatella jamesii, Callirhoe triangulata, Agave virginica, Galactia volubilis,
Salvia lyrata and Ophio gloss urn engelmannii. In the loess mounds along
the Missouri River in northeastern Kansas and northwestern Missouri, some
exposed slopes and tops of hills are occupied by species more characteristic
of areas to the west. Included among these are Bouteloua hirsuta, B. gracilis,
Yucca glauca var. glauca, Cleome serru/ata, Dalea enneandra, Glycyrrhiza
lepidota, Euphorbia marginata, Liatris punctata var. nebras\ana, Aster
sericeus, Lygodesmia juncea and Oxytropis lambertii. According to Steyer-
mark (1963), "the incursion of this plains flora into these areas may have
taken place in relatively recent times during a period sometime subsequent
to the final retreat of the glaciers and perhaps within the past four thousand
years."

Thus far in the discussion very little has been said about species which
occur throughout the Great Plains, yet it is from these plants that we gain

An Introduction to the Phytogeography of Kansas 899

valuable information concerning the biological phenomena of adaptation
and speciation.

One of our first attempts to correlate distribution patterns with present
climate involved comparing the distributions of various species with zones
of plant hardiness as mapped by the Arnold Arboretum (Fig. 45). We could
find but few species whose distribution seemed to coincide in any way with
the hardiness zones. Most of the species cross at least five zones, and many
occur across seven or eight zones. A closer look at morphological and
physiological variation (features which are not evident on dot maps) within
these species aids in interpretation of this problem.

In areas such as California and Arizona where altitudinal and climatic
changes are abrupt, speciation is thought to occur rather rapidly. In the
Great Plains, however, where there are few marked regional differences in
surface or elevation, latitude is the chief control of mean temperature and
associated aspects of the climate such as insolation and length of growing
season (Borchert, 1950). In the plains, there is a relatively even south-to-
north transition in mean temperature and an east-to-west decrease in pre-
cipitation. The responses of plants to these clines are also transitional and,
although physiologically evident, may not always be morphologically (and
thus taxonomically) detectable. Much more work needs to be done on these
phenomena.

Riegel (1940), studying variations of blue grama grass (Boitteloua
gracilis), observed that plants from the south produced the greatest growth
in height for the season, and those from the north, the least. The southern
plants also produced the most and coarsest textured foliage, while those
plants of the northern section produced the least and finest. The roots of
the southern group of plants penetrated to the greatest depth, while those
of the northern plots (with the exception of Montana) were the shallowest.
The southern plants produced the most spikes per flower stalk and the
largest spikes on the tallest stalks, while the northern plants produced fewer
and smaller spikes on shorter flower stalks. The northern plants began to
flower on June 25 to July 1 and the southern ones began July 23 and 24. Those
from the central group showed a wide variation in flowering time. Since
these differences were observed in an experimental garden where all in-
dividuals were exposed to the same environmental conditions, one would
assume that the variations are genetically fixed in response to differences in
the environments from which the seeds were obtained.

McMillan (1959), studying ecotypic variation in 12 species of grasses
native to the great plains, discovered three general flowering patterns which
were detectable in a transplant garden. The first — displayed by Koeleria
cristata, Boutcloua gracilis, B. curtipenditla, Panicum virgatum, Andropogon
scopariits, the Andropogon gerardi-hallii complex, Sorghastram nutans and

900 The University Science Bulletin

Sporobolus heterolepis — showed early flowering for clones from the northern
and western communities and later flowering toward the south and east.
The second, exhibited only by Ely m us canadensis, presented earliest flower-
ing of clones from southern communities. In the third, flowering was
simultaneous for clones from all communities represented. Species which
displayed this pattern were Stipa spartea, S. comata and Oryzopsis hymen-
oides. McMillan describes these three as "opportunistic species, which re-
sume growth and flower as soon as conditions permit." In McMillan's study,
"shorter stature was correlated with the early maturity of northern and
western material and was replaced by taller growth of the later-flowering
forms toward the south and east. Correlations of height and maturity were
not shown by Elytnus canadensis, the only species with earlier maturity in
southern species-populations."

In the transplant garden at the University of Kansas, clones of Andro-
pogon gerardi and A. scoparius from northwestern Kansas consistently
flower before those from the southeastern corner of the state. This corre-
sponds directly with the length of growing season in the native habitat of
each clone. In addition, plants from the eastern part of the state are always
taller than those from the west, and the same differences in robustness and
numbers of flowers as described by McMillan can also be noted.

It would appear that little correlation between the distribution of plains
species and hardiness zones is apparent because the plains species migrated
into the area and then evolved in place as the climate evolved. Hardiness
zones, on the other hand, reflect the ranges of tolerance of cultivated species
which, for the most part, are not native to this area. Perhaps more correlation
between distribution of native plants and these temperature zones would
be evident if more were known about physiological variation within species
of the Great Plains.

In some groups, morphological differences as well as physiological ones
can be detected, and these shed additional light on the problem of post-
Pleistocene plant migrations and relationships. Unfortunately, there have
been but few taxonomic and biosystematic studies of species complexes within
the Great Plains. One of the more enlightening of these is the work of
Harms (1963) on Variation in the Heterotheca (Chrysopsis) villosa Complex
East of the Rocky Mountains. His study was based on field observations,
analysis of mass population samples, growth of transplants and seedlings
under uniform garden conditions, progeny tests, pollen analysis, determina-
tion of chromosome numbers and analysis of meiosis, artificial crossing,
hybrid analysis, and a critical study of many herbarium specimens. His
study of transplants and plants grown from seed indicate that variation in
the group is genetic.

In the southern areas of the Great Plains, the Chrysopsis villosa complex

An Introduction to the Phytogeography of Kansas 901

appears as several distinct or nearly distinct diploid races. In the northern
plains, the C. villosa forms are mostly tetraploid and intergrade so extensively
that they can hardly be separated in any practical taxonomic manner. The
origin of the tetraploid villosa is highly speculative, since this taxon tends
to morphologically bridge the gaps between the presently known diploid
species.

It is probable that some of the variation now present in villosa is due to
a primary divergence in which different adaptive peaks have been reached
in different areas of the range, since the geographical form is approached
clinally without extreme variation being encountered in local populations.
The tendency for correlation of C. angustijolia characters and the great vari-
ability everywhere accompanying the intergradation of angustijolia with
villosa races strongly suggests a secondary intergraduation of taxa once well
under way toward separate speciation.

According to Harms, the present variation patterns of C. villosa east of
the Rocky Mountains allow several tentative speculations relating to its post-
Pleistocene history. Non-introgressed populations of C. hispida in the Drift-
less Area of Wisconsin and the hispida influence which is apparent elsewhere
in populations north of the Arkansas River suggest that this member of the
complex was the first to move into the northern plains after the Wisconsin
glaciation. During the Xerothermic Period, it withdrew northward and
westward and was replaced by villosa which was spreading at that time
from the foothills of the southern Rocky Mountains or perhaps from a refuge
farther south. The two groups may have come in contact with one another
at that time. At the same time that villosa was moving northward, races of
angustijolia and stenophylla were expanding, presumably from a Texas
refuge. Races of foliosa also began to move across the southern plains, perhaps
from two refuges — one in the southwest and one in Texas. They have since
come in contact with one another again and have intermixed. There has
also been extensive hybridization and introgression between villosa and
angustijolia and between angustijolia and joliosa during post-Pleistocene
times. This complex series of factors makes taxonomic recognition most
difficult.

Other taxonomic problems suggest that detailed biosystematic studies of
plants within the Great Plains could yield information useful in the interpre-
tation of post-Pleistocene plant migrations — information which, conversely,
could be applied to further biosystematic studies. In the genus Liatris, for
example, there are several species complexes which, if carefully worked out,
might shed additional light on this problem. Liatris spicata, L. pycnostachya
and L. land folia were recognized by Gaiser (1946) as distinct species. They
are, however, morphologically quite similar. According to Gaiser and most
manuals, L. spicata is eastern in distribution, occurring from New York to

902 The University Science Bulletin

Florida, westward to the Mississippi River. L. pycnostachya is primarily
southeastern, but extends northward into Kansas, Nebraska and South
Dakota. L. lancijolia is described as being southwestern, distributed from
New Mexico northward to Kansas, Nebraska and South Dakota. These
distribution patterns, together with field observations of their morphology,
suggest that the situation is not as simple as Gaiser describes it, however.

In the Liatris squarrosa complex, Gaiser includes six varieties, two of
which occur in the Great Plains. L. squarrosa var. glabrata occurs from
Texas north to South Dakota and through most of its range shows little
variation. L. squarrosa var. hirsuta has an area of distribution centered in
Missouri and eastern Kansas and is much more variable than glabrata.
Gaiser described a number of specimens which she assumed to be inter-
mediates between glabrata and hirsuta, and Steyermark (1963) includes L.
squarrosa var. glabrata in his Flora of Missouri. In our field work, however,
we have not been able to find glabrata east of the Flint Hills nor hirsuta west
of them. Plants in Missouri which tend to be glabrous although otherwise
resembling hirsuta seem to indicate hybridization with another species,
Liatris cylindracea, which occurs from western New York and southern
Ontario west to southern Ohio, northern Indiana, Michigan, Minnesota and
thence south to Arkansas. The few specimens of glabrata from eastern
Oklahoma which have been observed suggest that here is an area where
glabrata is more variable than it is farther north and an area where it might
come into contact with colonies of hirsuta. Further work with this group
promises to reveal information regarding Pleistocene and post-Pleistocene
relationships and migrations.

LITERATURE CITED

Adams, C. C. 1902. Southeastern United States as a center of geographical distribution of

flora and fauna. Biol. Bull. 3:115-132.
Anderson, Edgar. 1949. Introgressive Hybridization. Wiley and Sons, New York. 109 pp.
Aughey, Samuel. 1880. Sketches of the Physical Geography and Geology of Nebraska. Daily

Republican Book and Job Office, Omaha. 346 pp.
Baughman, Robert W. 1961. Kansas in Maps. Kansas State Historical Society, Topeka Kansas.

104 pp.
Bessey, Charles E. 1887. A meeting-place for two floras. Bull. Torr. Bot. Club 14:189-191.
Borchert, John R. 1950. The climate of the central North American grassland. Ann. Assoc.

Am. Geogr. 40:1-39.
Braun, E. Lucy. 1928. Glacial and post-glacial plant migrations indicated by relic colonies of

southern Ohio. Ecol. 9:284-302.
— . 1955. The phvtogeography of unglaciated eastern United States and its interpretation.

Bot. Rev. 21:297-375.
Caldwell, Martha B. 1940. When Horace Greeley visited Kansas in 1859. Kansas Hist.

Quarterly 26:115-140.
Carruth, J. H. 1877. Centennial catalogue of the plants of Kansas. Trans. Kansas Acad.

Sci. 5:40-59.
— . 1883. Botanical addenda for 1881 and 1882. Trans. Kansas Acad. Sci. 8:32-33.
— . 1885. Botanical addenda for 1883 and 1884. Trans. Kansas Acad. Sci. 9:142-44.
Chaney, Ralph W. 1947. Tertiary centers and migration routes. Ecol. Monogr. 17:139-148.
— , and Maxim K. Elias. 1936. Late Tertiary floras from the High Plains. Carnegie Inst.

Wash. Publ. No. 476:1-72.

An Introduction to the Phytogeography of Kansas 903

Davis, M. B. 1965. Problems in the Quaternary phytogeography of the Great Lakes region.

In The Quaternary of the United States (Wright, H. E., Jr., and D. G. Frey, eds.), pp.

403-416. Princeton Univ. Press.
Deevy, Edward S., Jr. 1949. Biogeographv of the Pleistocene. Bull. Geol. Soc. Am.

60:1315-1416.
Epling, Carl. 1942. The American species of Scutellaria. Univ. of California Publ. in Botany

20:1-146.
Erickson, Ralph O. 1945. The Clematis jremontii var. richlii population in the Ozarks. Ann.

Missouri Bot. Gard. 32:413-460.
Farb, Peter. 1964. The Land and Wildlife of North America. Time Inc., New York. 200 pp.
Fearing, Olin S. 1952. Preliminary study of the taxonomy and ecology of Kansas lichens.

Univ. Kansas Sci. Bull. 35:543-575.
Fernald, M. L. 1931. Specific segregations and identities in some floras ol eastern North

America and the Old World. Rhodora 33:25-63.
Flora, S. D. 1948. Climate of Kansas. Report of the Kansas State Board of Agriculture

58:1-320.
Fremont, John C. 1845. Report of the Exploring Expedition to the Rocky Mountains in 1842

and to Oregon and North California in the Years 1.S43-1844. Gales and Seaton,

Printers, Washington, D.C. 693 pp.
Frye, John C, and A. Byron Leonard. 1957. Ecological interpretations of Pliocene and

Pleistocene stratigraphy in the Great Plains region. Am. Journ. Sci. 255:1-11.
Gaiser, L. O. 1946. The genus Liatris. Rhodora 48:165-183, 216-264, 273-326, 331-382,

393-412.
Gates, Frank C. 1940. Flora of Kansas. Kansas State Printing Plant, Topeka, Kansas. 266 pp.
Gleason, Henry Allan. 1922. The vegetational history of the middle west. Ann. Assoc. Am.

Geogr. 12:39-85.
, and Arthur Cronquist. 1963. Manual of Vascular Plants of Northeastern United

States and Adjacent Canada. D. Van Nostrand and Company, Princeton, New Jersey.

810 pp.
Graustein, Jeannette E. 1967. Thomas Nuttall, Naturalist: Explorations in America, 1808-

1841. Harvard University Press, Cambridge, Massachusetts. 481 pp.
Harms, Vernon L. 1963. Variation in the Hetcrothcca (Chrysopsis) villosa complex east of the

Rocky Mountains. Unpublished thesis. University of Kansas. Lawrence.
Hitchcock, A. S. 1892. The relations of the composite flora of Kansas. Trans. Kansas Acad.

Sci. 13:89-91.

. 1899. Flora of Kansas Maps. Published by the author. Topeka. 13 pp.

Horr, W. H. 1955. A pollen profile stuclv of the Muscotah Marsh. Univ. Kansas Sci. Bull.

37:143-149.
Jackson, Donald, ed. 1966. The Journals of Zebulon Montgomery Pike, with Letters and

Related Documents. Vol. 2. Universitv of Oklahoma Press, Norman. 44u pp.
Jones, J. Knox. 1964. Distribution and Taxonomy of Mammals of Nebraska. University of

Kansas Publications, Museum of Natural History 16:1-356.
Kendeigh, S. Charles. 1961. Animal Ecology. Prentice-Hall, Inc., Englewood Cliffs, New

Jersey. 468 pp.
Lathrop, Earl W. 1958. The flora and ecologv of the Chautauqua Hills in Kansas. Univ.

Kansas Sci. Bull. 38:97-20").
MacGinitie, H. D. 1962. The Kilgore flora: a late Miocene flora from northern Nebraska.

Univ. of California Publ. in Geol. Sci. 35:67-158.
Malin, James C. 1967. The Grassland of North America. Peter Smith, Gloucester, Mas-
sachusetts. 490 pp.
McGregor, Ronald L. 1955. Taxonomv and ecologv of Kansas Hepaticae. Univ. Kansas Sci.

Bull. 37:55-141.
— . 1968. A C-14 date for the Muscotah Marsh. Trans. Kansas Acad. Sci. 71:85-86.
McMillan, Calvin. 1959. The role of ecotypic variation in the distribution of the central

grassland of North America. Ecol. Monogr. 29:285-308.
Miller, Nyle H. 1932. Surveying the southern boundary line of Kansas: from the private

journal of Col. Joseph E. Johnston. Kansas Hist. Quarterly 1:104-139.
, Edgar Langsdorf and Robert W. Richmond. 1961. Kansas: a Pictorial History. The

Kansas Centennial Commission and the State Historical Society, Topeka. 320 pp.
Mollhausen, H. B. 1948. Over the Santa Fe Trail through Kansas in 1858. Transl. by John

A. Burzle, ed. and annotated by Robert Taft. Kansas Hist. Quarterly 16:337-380.
Parker, John D. 1872. Organization and history, Sept. 1, 1868. Trans. Kansas Acad. Sci. 1:3-7.
Plank, E. N. 1883. A preliminary notice of the flora of Montgomery County, Kansas. Trans.

Kansas Acad. Sci. 8:33-34.

904 The University Science Bulletin

Polunin, Nicholas. 1960. Introduction to Plant Geography. McGraw-Hill Book Company,

Inc., New York, Toronto, London. 640 pp.
Riegel, Andrew. 19-10. A study of the variations in the growth of blue grama grass from seed

produced in various sections of the Great Plains region. Trans. Kansas Acad. Sci.

43:155-167.
Rydberg, P. A. 1896. Flora of the Black Hills of South Dakota. Contr. U.S. Natl. Herb.

3:463-536.
Schmidt, Karl P. 1938. Herpetological evidence for the post-glacial eastward extension of

the steppe in North America. Ecol. 19:396-407.
Scribner, F. Lamson. 1885. A contribution to the flora of Kansas — Gramineae. Trans.

Kansas Acad. Sci. 9:115-123.
Segal, Ronald H. 1966. Biorbia (Boraginaceae) in the central United States Pliocene. Univ.

Kansas Sci. Bull. 46:495-508.
Schoewe, Walter H. 1949. The geography of Kansas. Trans. Kansas Acad. Sci. 52:261-331.
Smith, Harold L. 1966. Mosses of the Great Plains and Arkansas River lowlands of Kansas.

Univ. Kansas Sci. Bull. 46:433-474.
Smith, Philip W. 1957. An analysis of post-Wisconsin biogeography of the Prairie Peninsula

region based on distributional phenomena among terrestral vertebrate populations.

Ecol. 38:205-218.
Smyth, B. B. 1900. Plants and Flowers of Kansas. Crane and Company, Topcka, 118 pp.
Stebbins, G. Ledyard, Jr. 1950. Variation and Evolution in Plants. Columbia University

Press, New York. 643 pp.
Steyermark, Julian A. 1959. Vegetational History of the Ozark Forest. University of

Missouri Studies 31. 138 pp.

. 1963. Flora of Missouri. Iowa State University Press, Ames, Iowa. 1725 pp.

Tolstead, William L. 1942. Vegetation of the northern part of Cherry County, Nebraska.

Ecol. Monogr. 12:255-292.
Watts, W. A., and H. E. Wright, Jr. 1966. Late-Wisconsin pollen and seed analysis from

the Nebraska Sandhills. Ecology 47:202-210.
Wedel, Waldo R. 1959. An introduction to Kansas archeology. Smithsonian Institution

Bureau of American Ethnology, Bull. 174. 723 pp.
Wells, Philip V. 1965. Scarp woodlands, transported grassland soils, and concept of grass-
land climate in the Great Plains region. Science 148:246-249.
Wherry, Edgar T. 1955. The Genus Phlox. Morris Arboretum Monograph No. 3. 174 pp.
Wright, H. E., Jr. 1968. History of the Prairie Peninsula. In The Quaternarv of Illinois (R.

E. Berstrom, ed.), pp. 78-88. University of Illinois College of Agriculture Special

Publication No. 14, University of Illinois, Urbana.
— , and R. V. Ruhe. 1965. Glaciation of Minnesota and Iowa, hi Quaternary of the

United States (H. E. Wright and David Frey, eds.), pp. 29-41. Princeton University

Press, Princeton, New Jersey.

An Introduction to the Phytogeography of Kansas

905

Fig. 1. Composite map showing typical distribution of southwestern species (heavy black
line) and individual distributions of four southwestern species which extend into Kansas
(thin black lines).

906

The University Science Bulletin

Fig. 2. General distribution of eastern species.

An Introduction to the Phytogeography of Kansas

907

Fig. 3. Amelanchier arborea

90S

The University Science Bulletin

M ,.- ,^. % -

Fig. 4. Dicentra cucullan

An Introduction to the Phytogeography of Kansas

909

Fig. 5. Distribution of eastern species which are restricted to the southeast corner of Kansas.

910

The University Science Bulletin

Fig. 6. Distribution of eastern species which arc northward and westward around the
southern Great Plains.

An Introduction to the Phytogeography of Kansas

911

Fig. 7. Anemone virginiana

912

The University Science Bulletin

Fig. 8. Aquilegia canadensis

An Introduction to the Pi-iytogeography of Kansas

913

Fig. 9. Osmorhiza longistylis

914

The University Science Bulletin

Fig. 10. General distribution of northeastern species.

An Introduction to the Phytogeography of Kansas

915

Fig. 11. Hydrophyllum appendiculatum

916

The University Science Bulletin

Fig. 12. Rosa bland a

An Introduction to the Phytogeography of Kansas

917

Fig. 13. Car ex squarrosa

918

The University Science Bulletin

Fig. 14. Vaccinium vactlluns

An Introduction to the Phytogeography of Kansas

919

Fig. 15. Tilia americana

920

The University Science Bulletin

Fig. 16. Physocarpus opulijolius

An Introduction to the Phytogeography of Kansas

921

Fig. 17. General distribution <>t southeastern species.

922

The University Science Bulletin

Fig. 18. Distribution of Passiflora lutca in the central states, showing typical arc across
southeast Kansas.

An Introduction to the Phytogeography of Kansas

923

Fig. 19. General distribution of .southern species.

924

The University Science Bulletin

Fig. 20. Androstephium coeruleum

An Introduction to the Phytogeographv of Kansas

925

Fig. 21. Gaillardia sua lis

926

The University Science Bulletin

Fig. 22. Cleomclla angustifolia

An Introduction to the Phytogeography of Kansas

927

Fig. 23. Camassia angusta

928

The University Science Bulletin

Fig. 24. Acacia angttstissima

An Introduction to the Phytogeography of Kansas

929

Fig. 25. Aster serucus

930

The University Science Bulletin

Fig. 26. General distribution of southwestern species.

An Introduction to the Phytogeography of Kansas 931

Fig. 27. Castilleja citrina

932

The University Science Bulletin

Fig. 28. General distribution of southwestern species which occur within the western half
of Kansas and extend north of Kansas in the Great Plains.

An Introduction to the Phytogeography of Kansas

933

Fig. 29. Gilia longi flora

934

The University Science Bulletin

Fig. 30. General distribution of northern species.

An Introduction to the Phytogeography of Kansas

935

Fig. 31. Stipa com at a (solid line), S. spartea (dotted line) and S. viridula (broken line).

936

The University Science Bulletin

Fig. 32. General distribution of interior species.

An Introduction to the Phytogeography of Kansas

937

Fig. 55. Rkamnus lanceolata

938

The University Science Bulletin

Fh;. 34. Phlox pilosa var. ozar\ana

An Introduction to the Phytogeography of Kansas

939

Fig. 35. Aescitlits glabra var. arguta

940

The University Science Bulletin

Fig. 36. Clematis pit (.hoi

An Introduction to the Phytogeography of Kansas

941

■Pimp

Fig. 37. Liatris squamosa var. hirsuta

942

The University Science Bulletin

Fig. 38. Aster fendleri

An Introduction to the Phytogeography of Kansas

943

Fig. 39. Oenothera fremontii

944

The University Science Bulletin

Fig. 40. Scutellaria resinosa

An Introduction to the Phytogeography of Kansas 945

Fig. 41. Clematis jremontn

946

The University Science Bulletin

Fig. 42. Phlox okjahomensis

An Introduction to the Phytogeography of Kansas

947

Fig. 43. ] uncus brachyphyllus

948

The University Science Bulletin

Fig. 44. Mean annual water loss in per cent of rainfall (After Flora, 1948).

An Introduction to the Pmytogeography of Kansas

949

Fig. 45. Plant hardiness zones within the eentral states (after map by the Arnold Arboretum,
1967). Temperatures are given in degrees Farenheit.

3 2044 093 362 275

Date Due

HI