see pemen te Car hig Ce Meyer: HAUT nT eM wen my we sourequdar ye
we cwke ae fe wipes inte ewebe ge Hamre cheteme Fe tae Pe

ereew er igue

wee ore ep ae t “OP ete es tae per tet 2 mw Fone oy FH ee re eee eres
hee ree ee ee a eet ee i ee ee en oo rie ot ot tole
Ce i hee ee ee er
ee ee eT a a i a aero! Weim
ee Te a? ee ee So er oes
pee ewe ef eR ONE ® CaN e RAE ed Km He

nen a a
er ee ed

FON mgs Oct gt Weeemr ip shee oe

wees

any © eae yin cowerde

Pe eee Se

oe Teepe

re ee ee ee : woe ‘ wo oar : 7 aus ri peas a

er GY os hee eK we pene See aren ate en a ee

YS 8 tee 5 POP er ne Oe ee ee
eR! tang ae lonerge ce TANS He . hd ie . . mtet we - : : : ee oa oe ep 8 FoR SPO ce wR Ie OR Gere g Fas er pRNe WE rE Ty F
: rere erie ray ta at 2 ae One Tes

+33 oe ee ee Pe cr
Co ee ee ee eS

NOW TET wR Re ene
Oe Oe a

oe we for Oe OR Mem pe Re aiheieee en he we

ae eet

nae osioke sy
ope aap

EM Meets Peery FOES Fe ATE Noe! pret OniS
re er ae ee ae ee ee ae ee

a ae ee ee oe es eT Pee

ha
a

eat ad }aaed

t
ft
.
i
‘
fi
ta
e
:
a
:
wt ‘
-
‘
"a:
1 ’ - :
7 ca ‘ .
* 7 . .
\ t 7
I , ’ : i
¥ 7 1 rn : % ‘
i - i
4 J
| , “es F io A
] : r : :
. a ez : . ‘ n 7
‘ +
{ ‘
f 7 *
~
a ‘
’ + a s
.
. >.
‘ ‘
2 {
~ -
a - é . : .
- . 7 L
Ls Aa i
Be ' ee * i
¢ : 4
OTP
Oo te : i :
is poe Me : ' ' ;
a Rs - :
, , ‘ ae
con - =, j : 1"!
ct - Gs “
. 4 :
fo ad ‘ -
7 ; ~ * j
’ 7 '
a a & - . *
a
het ~~ f i -
a y i
I , - ae
. He
cic i ae
ie ‘
- ' . ~ ~ 1
. 7 ' .
foe I -
- i * * -
7 \ ss ih
: : : x : 7 4
7
se) 5
- ty . . : , :
‘
>
} y 0 :
i, : oa 7 ‘
a
’
' y :
D + 7
am; =
- ee * fa {
s : . >
3. EA
A f
mo, -
r .
: :
: .
5 ~~ 7
<x - :
a H ue
7 i _— :
is) *
is 3
4D ig ”
On 7 a a
, nea |} ™ 2
' E Ae :
Sess 7 ime is
fea ’ ‘
1
> :
a , a :
I ‘ =n
sual! *, = 1
“i ie -
- : 7
7 “
a
oa
1
: :
a
%
‘
” ~
Us 4
2
:
‘
*v, . - a -
: 3
‘ =
* *
‘
‘
1 -_
a : 7
‘ S
' , nis - - = '
ar rn

“~*

ae

a

ae

i
ae. 8b
:
3 Pal
“soe
s
4
'
4
1
c]
:
7
i
;
7
:
“i
inal ft
7 “i

oi my ae yt

ara

‘
.
a :
s
‘
# ‘
7 S
1
- ‘
P ‘
f ;
+ "
.
od
4
f
'
.
me a
1%
.
v
one
.
oi
‘ «
j
1
‘
,
'
‘
‘
\
“e
‘
1
5
1
;
i
a fom
{
‘ ;
‘
a
} qn
1
1
7 “4
‘ =
: : y
ri

re
i
mm de = 2
; A
4
‘
fl
*
ba
re a
:
5°
,
i
A
,
i
‘
;
z
i ]
: \
H
7
‘

DanikL ‘TREMBLY MacDouGaL

Frontispiece to Volume 9, Maprono

MADRONO.

A WEST AMERICAN JOURNAL OF
BOTANY

VOLUME IX
1947 - - 1948

Published by the

California Botanical Society, Inc.,
4004 Life Sciences Building, University of California, Berkeley
North Queen Street and McGovern Avenue, Lancaster, Pennsylvania

Board of Editors

Hersert L. Mason, University of California, Berkeley, Chairman.
LeRoy Asrams, Stanford University, California.

Epcar ANvDERSON, Missouri Botanical Garden, St. Louis.

Lyman Benson, Pomona College, Claremont, California.

Hersert F. Copetann, Sacramento College, Sacramento, California.
Ivan M. Jonnston, Arnold Arboretum, Jamaica Plain, Massachusetts.
Mitprep EK. Marnuitas, University of California, Los Angeles.

Bassett Macuire, New York Botanical Garden, New York.

Marion Ownsey, State College of Washington, Pullman.

Secretary, Editorial Board

ANNETTA M. Carter, Department of Botany, University of California, Berkeley.

Business Manager

Rimo Bacicauturr, Natural History Museum, Stanford University, California.

To Dr. Daniel T. MacDougal, botanist, explorer, philosopher
and savant, the California Botanical Society dedicates this ninth
volume of Madrofio. Now in his eighty-third year, Dr. Mac-
Dougal has achieved eminence in the field of plant physiology, as
well as in the direction and pursuance of productive research. For
fifty years he has been a dominant force in the organization of
American science. He has been a life-long student of the botany
of Western America and is an internationally known authority on
the plants of arid regions and their environmental relations. We
wish him continued good health and ali the joy that comes from a
life rich in work well done.

ili

CONTENTS

Frontispiece: Daniel Trembly MacDougal
Certain Plant Species of the Canyon of Hurricane Creek, Wallowa County,
OREO Cag Wve x Ng ge re gles Deacon a Morton EF. Peck 1
Generic Names of Algae Proposed for Conservation. I.
George F. Papenfuss 8
Germination of Phacelia Seeds ....................... Clarence R. Quick 17
The Genus Mimulus in or Adjacent to Sinaloa, Mexico.
Howard Scott Gentry 21

Location of Extraneous Materials in Redwood ....... Irving H. Isenberg 25
ERE WHE MUS Ot ee i SA ee AEE Ph ONE cee iat aes oe AS Se ec 30, 135, 165, 193, 228
Vegetation and Climate of Coahuila, Mexico ........ Cornelius H. Muller 33
Viadimir L. Komarov, 1869-1946 ............ 0.0000. e eee N. T. Mirov’ 57
The Present Status of the Genus Polemoniella Heller .. John F. Davidson 58
Wilhissimni EPSOM hess on sree be ieee Gees oe eb tS Herbert L. Mason 61
INOUE SHANG ING WS. chee! cat ee ath ee Sb ae oR Ce 64, 168, 199, 232

Observations on the Structure and Classification of the Pyroleae.
Herbert F. Copeland 65
A New Species of Blennosperma from California .. Charles B. Heiser, Jr. 103
The Polygonal Graph for Simultaneous Portrayal of Several Variables in
Population Amalysis! 2 2.3 24h es ad tos ees « John F.. Davidson 105
Maturation of the Gametes and Fertilization in Nicotiana.
T. H. Goodspeed 110
Nomenclatorial Changes in Elymus with a Key to the California Species.
Frank W. Gould 120
Two New Varieties of Condalia from Texas ................. V.L. Cory 128
A. New Violet:trom:Mexico 2.2.2... 000i ete i bet we! Milo S. Baker 131
A New Species of Oxytropis from the Central Rocky Mountains.
C. L. Porter 133
Environmental Extremes and Endemism .............. LeRoy E. Detling 137
Noteworthy South American Plants, I and II.
Paul C. Standley and Fred A. Barkley 149
Notes on Pacific Coast Marine Algae .................. G.J. Hollenberg 155
On Two Perennial Caespitose Lepidiums of Western North America.
Reed .C. Rollins 162
Concentration of Environmental Extremes as the Basis for Vegetation

PASE AS eR HEE i MeN Pe! oot ART acdsm LAVA RI Schull Stith a LeRoy EH. Detling 169
The Genus Helianthella in Oregon .................. William A. Weber 186
A New Polemonium from Mexico .................... John F. Davidson 187
Notes on the Taxonomy of Some Eastern Asiatic Ferns of the Genera

Protowoodsia and Pteretis ......................4.. Clyde F. Reed 189
Some Problems in the Genus Gilia .. Herbert L. Mason and Alva D. Grant 201
Potamogeton latifolius in Texas ...................... W.C. Muenscher 220
The Place of Willis Linn Jepson in California Botany .... David D. Keck 223

The Identity and Delimitation of Allium Tolmiei Baker .. Marion Ownbey 233
Notes on the Genus Townsendia in Western North America.
Charles B. Heiser, Jr. 238
Some Parallels between Desert and Alpine Flora in California.
F.W. Went 241

Some Additional Notes on Polemoniaceae ............ Herbert L. Mason 249
A New Species of Phacelia from Sonora, Mexico .... Lincoln Constance 255
Chromosome Number Publication '.................... J. A. Rattenbury 257
IndexatorV.OMImMe INC l yy So asa an eg uaenS baw oh Sak vsu elastin Aha eaes 259

ERRATA

Page 5, line 47: for dnemome read Anemone.

Page 14, line 5: for sedoidis read sedotdes.

Page 58, line 33: for 1941 read 1944,

Page 59, line 40: for Lemmoni read Lemmonii.

Page 61, line 2: for November 6 read November 7.

Page 142, line 13: for columbiananum read columbianum.
Page 162, line 22: for lead read led.

Page 182, table 1: for absoma read absona.

Page 180, line 3: for glauca read glaucus.

Page 202, line 9: delete Coast.

Page 209, lines 8, 9: for miilifoliata read miilefoliata.
Page 209, lines 19, 20: for millifolia read millefolia.
Page 210, line 10: for Abramsi read A bramsii.

Page 214, line 25: for 271 read 279.

Page 214, line 30: end sentence with Mountains. Delete the type.
Page 214: delete line 31.

vi

VOLUME IX NUMBER 1

MADRONO

A WEST AMERICAN JOURNAL OF
BOTANY

vf

Contents
CrerTAIN PLANT SPECIES OF THE CANYON OF HuRRBICANE CREEK, WALLOWA
County, OrnEGon, Morton EB. (Peek ae oko ko ec ee ee ne 1
GENERIC Names oF ALGAE Proposep For CoNsERVATION. I, George F.
PEGA CE RO SLES: UR RG AE A EE ROSIE AGL ROA ONAL I EH ST 8
GERMINATION OF PHACELIA SEEDS, Clarence R. Quick .................... 17
Tue Genus Mimvutus In on ADJACENT TO Stnatoa, Mexico, Howard Scott
CECE Ca TT CCA NORA PEG Misi AN ESN EA 06H SOE, a RU a ee peek ae nA 21

Published at North Queen Street and McGovern Avenue,
Lancaster, Pennsylvania

January, 1947

MADRONO

A WEST AMERICAN JOURNAL OF BOTANY

Board of Editors

Hersert L. Mason, University of California, Berkeley, Chairman.
LeRoy Asrams, Stanford University, California.

Epaar Anperson, Missouri Botanical Garden, St. Louis.

Lyman Benson, Pomona College, Claremont, California.
Hersert F’. Copetanp, Sacramento Junior College, Sacramento, California.
Ivan M. Jonnston, Arnold Arboretum, Jamaica Plain, Massachusetts.
Mivprep EE. Matutas, 2851 North Lake Avenue, Altadena, California.
Bassett Macuire, New York Botanical Garden, N. Y. C.

Marion Ownsey, State College of Washington, Pullman.

4

Secretary, Editorial Board—ANNeEtTTA CaRrTER
Department of Botany, University of California, Berkeley

Business Manager—Reep C. Ro.iins
North Queen Street and McGovern Avenue, Lancaster, Pennsylvania
or
Natural History Museum,
Stanford University, California

Entered as second-class matter October 1, 1935, at the post office at
Lancaster, Pa., under the act of March 3, 1879.

Established 1916. Published quarterly. Subscription Price $2.50 per year.
Completed volumes I to VII inclusive, $35.00; each volume $5.00; single numbers
$0.75.

Papers up to 15 or 20 pages are acceptable. Longer contributions may
be accepted if the excess costs of printing and illustration are borne by the
contributor. Range extensions and similar notes will be published in con-
densed form with a suitable title under the general heading “Notes and News.”
Articles may be submitted to any member of the editorial board. Manuscripts
may be included in the forthcoming issue provided that the contributor pay the
cost of the pages added to the issue to accommodate his article. Reprints of
any article are furnished at a cost of 4 pages, 50 copies $4.10; 100 copies $4.50;
additional 100’s $0.85; 8 pages, 50 copies $5.95; 100 copies $6.60; additional 100’s
$1.30; 16 pages, 50 copies $8.35; 100 copies $9.35; additional 100’s $2.00. Covers,
50 for $2.75; additional covers at $1.65 per hundred. Reprints should be
ordered when proofs are returned.

Published at North Queen Street and McGovern Avenue, Lancaster,
Pennsylvania, for the

CALIFORNIA BOTANICAL SOCIETY, INC.

President: Palmer Stockwell, Institute of Forest Genetics, California Forest
and Range Experiment Station. First Vice-President: Adriance S. Foster,
University of California, Berkeley. Second Vice-President: Jens Clausen, Car-
negie Institute of Washington, Stanford University, California. Secretary:
George F. Papenfuss, Department of Botany, University of California, Berke-
ley. Treasurer: Reed C. Rollins, Natural History Museum, Stanford University,
California.

Annual membership dues of the California Botanical Society are $2.50,
which includes a year’s subscription to Madrofio. For two members of the same
family the dues are $3.50, which includes one copy of Madrofio and all other
privileges for both. Dues should be remitted to the Treasurer. General
correspondence and applications for membership should be addressed to the
Secretary.

CERTAIN PLANT SPECIES OF THE CANYON OF
HURRICANE CREEK, WALLOWA
COUNTY, OREGON

Morron E. PecKx

From the geologist’s standpoint the Wallowa Mountains pre-
sent the most interesting and diversified section of Oregon. Flo-
ristically they are no less noteworthy. The boundaries of the
Wallowa mountain area are rather vague. In the broader sense
the area embraces all of Wallowa County except the northwest
corner, a considerable portion of eastern Union County and a
strip of northern Baker County. Between six and seven hundred
square miles of this territory, including what are commonly
known as the high Wallowas, and the territory eastward to the
Snake River, are of exceptional botanical interest. For the most
part this is an extremely rugged region with deep, narrow can-
yons, precipitous slopes, numerous small lakes, and high peaks
reaching an altitude of 10,000 feet.

The geological history of the Wallowa Mountains is extremely
complex, but scarcely concerns us in the present connection.
There are large exposures of Permian metamorphosed lavas,
Upper Triassic crystalline limestone, great areas of quartz diorite
(Cretaceous) and equally large areas of Tertiary Columbia River
basalt mantling the slopes or even covering the summits of many
of the high peaks. The floors and lower slopes of many deep
canyons and small valleys are covered with Quaternary morainic
material, mostly water-worn gravel and small boulders mixed
with alluvial deposits.

As to their origin, the Wallowas form an outlying island of the
Rocky Mountain system, and they present a larger exposure of
Paleozoic and early Mesozoic formation at higher altitudes than
occurs elsewhere in the state. These two circumstances deter-
mine to a great extent the character of the flora. The number of
recorded species of Oregon plants found only in the Wallowas
is very large. While not a few of these are endemics, most of
them show a close relationship with Rocky Mountain species.
In many cases the difference is subspecific, the distinction often
being very slight.

While the actual distance of the high Wallowas from the
Rocky Mountains is not great, their separation from the latter,
at least as far as the distribution of alpine species is concerned,
is very effectual. The extremely deep gorge of the Snake River,
with no high peaks on the Oregon side for nearly 20 miles, must
constitute a complete barrier to all high altitude forms except
under unusual conditions. Added to the relatively rare accident

Corrected date line: Maprono, Vol. 8, pp. 241-272. November 6, 1946.
Date line for Maprono, Vol. 9, pp. 1-32, will appear in the April, 1947, issue.

2 MADRONO | [Vol. 9

of seed transportation across the intervening depression, the un-
certainty of deposition on favorable soil with other suitable
environmental conditions in so extensively barren a terraine make
the chances of successful establishment of a species in the new
territory very slight. The degree to which the migrant forms
differ from the probable parent stock may thus be taken, in
general, as an indication of the relative remoteness of the date
of their first arrival. Of these differences, as stated above, we
find varying degrees. At least some of these conclusions are well
illustrated by a rather detailed examination of the vegetation of
a very small but exceptionally favorable section of Wallowa
Mountain territory recently made by the writer.

The period from July 21 to 25, 1944, was spent in collecting
and studying the plants growing on a two-mile segment of the
canyon of Hurricane Creek, Wallowa County, beginning at a
point about 4.5 miles south of Enterprise, just above the place
where the stream emerges from the narrow, steep and rugged
portion of the canyon. The altitudinal range of the two-mile
stretch is from a little over 6000 feet to a little under 7000.

The total length of Hurricane Creek is scarcely over 25 miles.
Its general direction of flow is northward. The lower third of its
course is through rather open, moderately sloping country. Above
this the descent is very steep and the canyon walls are high,
rugged, and often precipitous. It is fed by melting snow on some
of the highest peaks of the Wallowas, especially Sackajawea and
Matterhorn, both of which reach an altitude of about 10,000 feet.
It bears a remarkably large volume of water for so short a stream,
and the temperature is very low, even in midsummer. So steep
is its bed that nearly half its course in the section studied consists
of a series of noisy cataracts. Within these two miles it receives
two small tributaries from the west and one, a little larger, from
the east. The canyon floor in its broadest place is not much over
100 yards wide, from which it varies to only the width of the
stream, so that in some places it is impossible to find footing on
the margin.

The eastern slope of the canyon is very steep and consists in
part of metamorphosed lava. On the west side the slope is
broken, some distance up, by a more or less continuous bench,
in places nearly a half-mile wide. The formation here is partly
shale and conglomerate. Much the greater part, perhaps three-
fourths, of the whole drainage area of the stream consists of
metamorphosed limestone.

The canyon floor is made up of glacial deposits—the usual
coarse sand, water-worn gravel and small boulders, with a mix-
ture of silt in varying proportions, but often scant. At several
points along the margins of the stream are sand and gravel bars,
usually only a few yards in width, but in one or two cases up to
a hundred yards or more in length. In one place where the can-

1947] PECK: HURRICANE CREEK CANYON 3

yon floor is unusually broad there is a peat bog (not sphagnum)
covering perhaps an acre. Here there is abundant silt and over
much of it has accumulated a rather deep layer of spongy peat.
The bog seems to have been built up to its present state largely
through the work of beaver.

With the set of conditions thus briefly indicated it is not difh-
cult to understand why the flora of the canyon of Hurricane
Creek presents features of uncommon interest. There is first, the
great depth and narrowness of the gorge; second, the remarkable
development of sand bars and peat bog, unusual in a stream bed
of this character; third, the very low temperature of the water,
varying little through the growing season; finally, and most sig-
nificant of all, is the fact that the mountain masses from which
the stream descends are largely calcareous. Such a combination
of conditions can scarcely be duplicated elsewhere in the state.
Add to this the comparative isolation with respect to the Rocky
Mountains, where similar situations ‘are probably not unusual,
and we might well expect to find an interesting assemblage of
species.

Many of the forms named in the following list are ordinarily
found at much higher elevations than our records show. As so
often happens in deep narrow canyons carrying streams from
melting snow, alpine species have descended to a much lower
level than that to which they are accustomed, the icy water proba-
bly favoring the downward extension of their range. To what
extent the calcareous content of the water may have influenced
the plant population as a whole may be judged by the list follow-
ing. Included in this are only those species that for one reason
or another appear to be of particular interest. Most of these are
from the canyon floor or moist banks just above it, though some
are from higher slopes. Many others might, perhaps, have been
appropriately included.

Botrycuium virGinianum (L.) Sw. Several specimens were
found on moist banks above the stream. The species is rare in
Oregon.

EQuiseTUM vARIEGATUM Schleich. Some of the bars along the
creek are covered with a dense mat of this species. Otherwise it
seems rare and local in the state.

Pinus FLexitis James. One of the rarest of Oregon conifers.
There is a small colony on the high eastern slope of the canyon
of Lostine River, Wallowa County, about twenty miles above the
mouth. This second colony was found on the western slope of
Hurricane Creek canyon. None of the trees in either case are very
large. In both localities they grow in dry calcareous soil.

Poa aupina L. Noted in several places on wet banks and bars
far below its usual altitudinal range.

Guyceria Otis Hitche. A few small colonies were found in

4 MADRONO [Vol. 9

the peat bog. Apparently this is a very local species not hitherto
reported from Oregon.

CALAMAGROSTIS NEGLECTA (Ehrh.) Gaertn. Found sparingly
on wet banks.

MUHLENBERGIA ANDINA (Nutt.) Hitche. A single colony was
discovered on a wet bank. The species is rarely met with in
Oregon.

Kosresia simpLicruscuta (Wahl.) Mack. A single large clump
was discovered on a moist bank. The species occurs from Alaska
to Greenland and southward in the Rocky Mountains to Colorado,
also in Eurasia, but it has not been reported previously from
Oregon. It is always found, apparently, in calcareous soil.

CAREX DISPERMA Dew. Plentiful in the peat bog. The species
is widely distributed in the state at moderately high altitudes,
but is nowhere very common.

CarREx GyNocrAates Wormsk. Grows in considerable abun-
dance in the peat bog. Its normal range is far northward and
eastward. It is known from British Columbia and Colorado, but
hitherto not very near to Oregon. Calcareous soil seems to be
one of its requirements.

CaREX LEPTALEA Wahl. Plentiful in the peat bog. A widely
distributed species but rare in Oregon except in bogs along the
coast.

CAREX PSEUDOSCIRPOIDEA Rydb. Infrequent on wet banks.
Mainly a Rocky Mountain species, reaching eastern Washington
and Oregon. It has previously been collected in the Steens Moun-
tains and at least once in the Wallowas.

Carex concinna R. Br. Rather plentiful in several places on
slightly moist banks well above the water. This is another spe-
cies confined to calcareous soil. It occurs in the Rocky Moun-
tains and eastward, the present collection being a great extension
of the known range.

Carex capitutaris L. Plentiful on wet banks above the stream.
Our material is perhaps var. elongata Oln. Both the species and
the variety are known from British Columbia, the Rocky Moun-
tains, and far eastward, but not hitherto from Oregon.

Carex Vauiiu Schk. Only two small stands were found, both
on banks well above the water. It is a widely distributed species
on caleareous soil, occurring from Yukon to Greenland and in the
Rocky Mountains south to Idaho and Utah, also it has an exten-
sive distribution in the Old World. Apparently it has not been
found before in Oregon.

Juncus Reger Buch. A small form occurs sparingly on bars
along the stream.

SISYRINCHIUM IDAHOENSE Bickn. A low slender form with pale
flowers and stems scarcely at all clustered was found in some
abundance on wet gravel bars. It might prove to be a well-
marked variety.

1947 | PECK: HURRICANE CREEK CANYON 5

Hasenaria optusata (Banks.) Rich. A good-sized colony was
discovered on a shaded, mossy bank some distance above the
stream. The distribution, as hitherto known, extends from the
Aleutian Islands to New Brunswick and southward in the Rocky
Mountains to Idaho and Colorado. Ours seems: to be the first
record for Oregon.

CoRALLORRHIZA TRIFIDA Chat. A few small plants were found
on the low western slope of the canyon. Of wide distribution
northward and eastward, in the far west this species seems not
to have been previously recorded south of southern Washington.

Satix Woxtri Bebb var. taHoensis Ball. Found sparingly on
wet banks. The distribution is mainly Rocky. Mountain, but the
species has been previously reported from Wallowa County.

SALIX BRACHYCARPA Nutt. var. Sansonr Ball. Occurs sparingly
on wet banks and margins of the peat bog, growing close to the
water. The species occurs in the Rocky Mountains and far north-
ward. The variety was taken by the writer some years ago in
Walowa County on the margin of Ice Lake at the southern base
of Matterhorn.

SaLix vesTITA Pursh. A single specimen was found on a wet
bank. The species has been previously taken in the Wallowa
Mountains but is rare. It ranges northward to British Columbia
and eastward to Newfoundland.

BEeTULA GLANDULOsA Michx. A few specimens were noted on
low banks close to the stream. They were unusually large,
attaining a height of 2 to 3 meters.

Erioconum Kine T. & G. A few plants were found high on
the western slope of the canyon. This species has been collected
once previously by the writer in the Wallowa Mountains near
Aneroid Lake.

PotyGonum vivirarum L. Plentiful on wet banks and bars.
This is below its usual range.

ARENARIA PROPINQUA Rich. Plentiful on bars along the creek.

ARENARIA Rossit Rich. <A single specimen was collected on a
wet bank. We have no other dependable record of its occurrence
in Oregon. It is an arctic and arctic-alpine species of Alaska and
occurs southward in the Rocky Mountains to Colorado; it is also
reported from Washington.

SILENE ACAULIS L. var. SUBACAULESCENS Fern. & St. John. Plenti-
ful on sand bars, often forming large mats, a few of them flower-
ing profusely but the greater number only sparingly. The species
is rare in the Wallowa Mountains and unknown elsewhere in
Oregon.

THALIcTRUM aLpinum L. Three or four good-sized colonies
were found in damp ground near the stream. So far as we know
this is the first recorded occurrence of the plant in Oregon.

ANEMOME PARVIFLORA Michx. Rather plentiful in moist places
near the creek. A species occurring in calcareous soil, its main

6 MADRONO [Vol. 9

range is in the Rocky Mountains and far northward and eastward,
the Wallowa Mountains being an isolated area, where we know
of but two or three previous records.

ANEMONE GLoBosa Nutt. Very abundant on dry grassy benches
well above the stream; elsewhere in the Wallowa Mountains it
occurs more sparingly and it is known from no other section of
the state.

AQUILEGIA FoRMOoSA Fisch. var. FLAVEScCENS (Wats.) Frye &
Rigg. Very plentiful on the benches along the west slope of the
canyon. Neither typical A. formosa nor even an intergrade with
the yellow-flowered variety were seen. Elsewhere in the Wallowa
Mountains all intermediates between the two may be found.

DraBa PRAEALTA Greene. Quite plentiful in open woods on
dry slopes.

Drasa nivauis Lilj. var. ELoNGaTA Wats. Several specimens
were found on sand bars.

LEsQUERELLA SHERWOoDII Peck. Occurs sparingly on dry stony
banks.

ARABIS HIRSUTA Scop. var. pycnocarPA (Hopk.) Roll. Appar-
ently rare. Two or three specimens were found on damp ground
near the creek. The variety is well marked by the rather crowded
cauline leaves, small flowers, strictly erect fruit, and spreading,
mostly simple trichomes of the stem.

SAXIFRAGA OPPOSITIFOLIA L. A single specimen, not flowering,
was found on a damp bar.

Dryas Drummonpi Rich. Plentiful on damp sand bars, form-
ing large dense mats, often 1.5 m. across, and flowering abun-
dantly. In the Wallowa Mountains this species, as well as D. octo-
petala, are found mainly, if not altogether, on calcareous soil.

ASTRAGALUsS IMPENSUs (Sheld.) Woot. & Standl. Several plants
were noted on dry banks near the creek.

AsTraGALus Pursuit Dougl. On dry benches; not common.

ASTRAGALUS ALPINUS L. Quite plentiful on gravel bars. This
matches very closely material from alpine summits of the Wal-
lowas. Two other related forms occur less frequently. One of
these is very dwarf, the stem only 2 to 3 em. long, the leaves 4 to
6 cm. with fifteen to twenty-five leaflets 6 mm. long or less, pilose
beneath, the flowers very few and small, about 8 mm. long, the
calyx white-pilose and pods much longer than in the usual form,
white-pilose or with some blackish hairs intermixed. The other
plant has stems up to 2 dm. long, leaves 5 to 6 cm., with the thir-
teen to fifteen leaflets 8 to 10 mm. long and sparsely strigose
beneath. The flowers are 10 mm. long, with black-hairy calyx,
and the pods small and strongly black-hairy as in typical A.
alpinus. ‘‘Hurricane Creek”’ is given as the type locality of Atelo-
phragma alpiniformis Rydb. We suspect that the second of these
two forms is very close to Rydberg’s type; it differs, however, in
the larger flowers and somewhat pubescent, instead of glabrous,

1947] PECK: HURRICANE CREEK CANYON 7

leaflets. The dwarf form might easily be regarded as specifically
distinct. Astragalus alpinus, however, is a highly variable, often
puzzling species, and one might hesitate about making further
segregation.

Oxyrropis viscipa Nutt. Plentiful, mostly on dry gravelly
banks. Our material is placed here with some uncertainty. The
characters relied on to separate Aragallus viscidula Rydb. from
Oxytropis viscida are not very satisfactory. Our specimens have
the smaller number of leaflets, mostly twenty-one to thirty-one,
of the former, as well as the rather gradually acuminate pods and
the more or less black-hairy calyx. The hairiness at the base of
the plant is variable. The flowers, on the other hand, in all speci-
mens seen, are whitish as is sometimes the case in O. viscida;
according to the description of Aragallus viscidula, its flowers are
not whitish. .

Oxytropis Cusicku Greenm. A few specimens were found on
gravel bars.

HEpDysaruM BoREALE Nutt. On dry, rocky banks; not common.

Pyrota minor L. On wooded slopes. Infrequent.

PyroLa CHLORANTHA Sw. Found sparingly on shaded slopes.

DopECATHEON ALPINUM (Gray) Greene. A single plant was
found on a wet bank near the stream. This is apparently out of
the usual range of the species.

GENTIANA INTERRUPTA Greene. Plentiful in one locality on an
old gravel bar. This species has been previously reported from
Oregon, but we have seen no definite locality assigned. It is
mainly a Rocky Mountain species, but is known from Washington.

SWERTIA PERENNIs L. Abundant specimens were found on a
moist slope above the stream in one restricted area which is
probably the exact locality where Cusick collected the type of
S. occidentalis Greene. An examination of our material verifies
St. John’s statement (Revision of the Genus Swertia, Am. Midl.
Nat. 26: 9-10. 1941) that the type and isotypes of Greene’s spe-
cies have some plants with 4-merous and some with 5-merous
flowers. In our material we find both forms, in some cases even
on the same plant.

CRYPTANTHA CELOSIOIDES (Eastw.) Pays. A few specimens were
found on the high dry western slope of the canyon.

PENsTEMON Wixtcoxn Rydb. The most abundant species of the
genus. It occurs in open woods, thickets and meadows.

Veronica Cusickit Gray. This occurs sparingly on grassy
slopes well above the stream.

CaASTILLEJA WALLOWENSIS Penn. An apparently rare and dis-
tinct species. Two examples were collected on bars along the
creek,

Pineuicuta vutearis L. Found sparingly on wet shady banks
Hear the creek.

SoLipaco ciLiosa Greene. Plentiful in places on moist banks.

8 MADRONO [Vol. 9

Many of the plants are large and the heads have unusually long
rays.

ANTENNARIA PULVINATA Greene subsp. arpescens E. Nels.
Only a few specimens were found on some of the higher bars
along the creek. This is a Rocky Mountain plant, the variety
being known previously from Idaho but hitherto not from Oregon.
It is easily recognizable by the strongly spreading, light brown
involucral bracts as well as by the densely pulvinate habit, and
the very small, rhombic-spatulate leaves.

Willamette University,
Salem, Oregon.

GENERIC NAMES OF ALGAE PROPOSED FOR
CONSERVATION. I

GeorGe F. PAPeENFUSS

In the course of work on South African and other marine
algae, it has been found that several old and well-established
generic names are invalid. They are hereby proposed for con-
servation.

CHLOROPHYCOPHYTA

CiapopHora Kiitzing (Cladophoraceae), Phyc. general. 262. 1843.
versus
Annulina Link, in Nees von Esenbeck, Horae phys. berol. 4. 1820.

Type species: Cladophora glomerata (L.) Kitz.

Annulina Link (1820) is based upon seven species of which
four, Conferva glomerata, C. albida, C. nigricans, and C. rupestris,
and possibly a fifth, C. rivularis, are representative of Cladophora
Kiitzing (1843). Of the two remaining entities, Conferva sordida
is a species of Conferoa Linnaeus (1753) and C. compacta is a
species of Urospora Areschoug (1866), which name has been
conserved.

Since Cladophora has been an accepted genus for more than a
century, its conservation would probably be welcomed by most
phycologists.

A name usually regarded as, in part, synonymous with Cla-
dophora is Chantransia DeCandolle. This is owing to the belief
that Chantransia was described for the first time by DeCandolle
in the third edition of DeLamarck and DeCandolle’s Flore Fran-
gaise (1805, vol. 2, p. 49); and as there conceived the genus
included representatives of the green algal genera Cladophora
and Oedogonium and the red algal genera Batrachospermum and
Lemanea. In a recent paper it was pointed out by Papenfuss
(1945) that Chantransia could not properly replace either Lemanea
or Batrachospermum. Of the two remaining genera, Oedogonium
has already been conserved; and Cladophora is thus left as the
only genus possibly requiring conservation against Chantransia.

1947 | PAPENFUSS: PROPOSED GENERA CONSERVANDA 9

Careful study of the text of DeCandolle has revealed that
under Chantransia vesicata he had listed among other synonyms
“Chantransia nodosa Decand. Bull. n. 51. p. 21.” This clue led to
the discovery that DeCandolle had actually described Chantransia
four years previously in the third volume of the Bulletin des Sci-
ences de la Société Philomatique (1801). In this publication, DeCan-
dolle had referred only two species, Conferva nodosa Vauch. and
Chantransia nigricans Decand., to his new genus. It has not been
possible to ascertain the subsequent fate of Chantransia nigricans,
but since DeCandolle excluded it from his genus as extended in
the Flore Frangaise (1805), it may be supposed that he regarded
it as belonging elsewhere. This would leave Chantransia nodosa
(= Conferva nodosa Vauch.) as the type of Chantransia; and inas-
much as this entity is a species of Oedogonium, viz., O. vesicatum
(Vauch.) Wittr., Chantransia becomes synonymous with Oedo-
gonium, which is a nomen conservandum.

PHAEOPHYCOPHYTA

Zonaria Agardh (Dictyotaceae), Syn. alg. scand. p. xx. 1817.
versus
Zonaria Roussel, Fl]. Calvados 99. 1806.

Type species: Zonaria Tourneforti (Lamour.) Mont.

As a genus of algae, Zonaria was apparently first described by
Agardh in 1817 in his Synopsis algarum scandinaviae. From this
and certain other early works, the impression is gained that the
name was originally given by Draparnaud. Usually, the place of
publication was not cited, but occasionally reference was made,
sometimes with a query, to Draparnaud’s (1801) Discours sur les
moeurs et la maniére de vivre des plantes. However, search through
this and other works of Draparnaud has failed to reveal the epi-
thet; and it seems safe to conclude that Zonaria was a manuscript
name of Draparnaud which he had never published. Further sup-
port for this assumption is furnished by the fact that the binomials
under Zonaria that are attributed to Draparnaud are usually cited
as “ined.” or “mscr.” or with a query, while at least two of them,
Z. Pavonia and Z. squamaria, are referred to by DeCandolle (in
DeLamarck et DeCandolle, 1805, p. 17) as occurring in “Herb.
Jussieu.”

As a generic epithet, the name Zonaria seems to have been first
published by Roussel (1806, p. 99). He applied it to a group of
fungi, typified by “Auricularia tremelloides B.” Zonaria Roussel
has, however, not been taken up by mycologists, whereas Zonaria
Agardh has for a long time received recognition among phy-
cologists.

The conservation of Zonaria Ag. would automatically reduce
Villania, which name was proposed by Nieuwland (1917) as a
substitute for Zonaria Ag. Nieuwland’s arguments for the rejec-

1 he MADRONO [Vol.9

tion of Zonaria have been severely criticized by Setchell and
Gardner (1925).

Recently, Tandy (1935) proposed Zonaria for conservation
against Villania. As lectotype, he suggested Z. variegata (Lamour. )
Ag. But J. Agardh had already, in 1894, made this species the
basis of his genus Gymnosorus. The entity is now known as Po-
cockiella variegata (Lamour.) Papenfuss (1943).

In this connection, attention may also be drawn to Chordaria
divaricata, which species Tandy (loc. cit.) has designated as the
type of the genus Chordaria Ag. This species had previously been
removed from Chordaria by both Kitzing (1843) and J. Agardh
(1880); and more recently Kylin (1940) made it the type of his
new genus Sphaerotrichia. Chordaria flagelliformis, which Kylin
(op. cit.) has selected as the type of Chordaria, would be a more
appropriate choice for this purpose.

RHODOPHYCOPHYTA

As is apparent from a casual glance at the synonymy of cer-
tain species of Lomentaria Lyngbye (1819), Chylocladia Greville
(an Hooker, 1833) and Gastroclonium Kitzing (1843), there has
existed in the past much confusion as to the exact limits of
these genera, and transfers from one to the other have repeat-
edly been made. Im addition, several other genera, such as
Champia, Laurencia, and Chondria, have at times also been drawn
into the confusion. This situation was not remedied until the
appearance of Kylin’s monograph of the Rhodymeniales. As
delimited by Kylin (1931), these genera are comparatively easily
separated.

If, however, the synonymy of Lomentaria, Chylocladia, and
Gastroclonium is checked, it is found that these names are invali-
dated by epithets given by Stackhouse (1809) in his ill-fated
“Tentamen marino-cryptogamicum,’ which was published in the
second volume of the Mémoires de la Société Impériale des natu-
ralistes de Moscou. <As is known, this paper of Stackhouse em-
bodies one of the earliest attempts towards a division of the
Linnean genera, Fucus, Ulva, and Conferva, into a greater number
of genera. But the work was overlooked for a long time with
the result that most of Stackhouse’s genera were never taken up.

In 1891, Kuntze revived a large number of generic names
which had been forgotten, including several of those of Stack-
house (in some instances with slight modification of spelling).
This restoration of old names by Kuntze was the prime factor
responsible for the passage of rules governing the conservation
of generic names by the International Botanical Congress.

In the list of Nomina Generica Conservanda, in the third (1935)
edition of the “International Rules of Botanical Nomenclature,”
is included, among others, a number of names of algal genera

1947 | PAPENFUSS: PROPOSED GENERA CONSERVANDA 11

which have been conserved against names given by Stackhouse.
Of the three genera mentioned above, viz., Lomentaria, Chylo-
cladia, and Gastroclonium, only Chylocladia has been conserved,
even though, as has been said, names given by Stackhouse invali-
date all three.

Lomentaria Lyngb. has been typified by both Schmitz (1889)
and Kylin (1981) with ZL. articulata (Huds.) Lyngb. and both
these authors have also designated Chylocladia kaliformis (Good.
and Woodw.) Grev. as the type of Chylocladia Grev. This species
should, however, be known as:

Chylocladia verticillata (Lightf.) Papenfuss, comb. nov.
(= Fucus verticillatus Lightfoot, Fl. scot. 2: 962, pl. 31. 1777.
Fucus kaliformis Goodenough and Woodward, Trans. Linn. Soc.
3: 106, 206, pl. 18. 1797).

As the type of Gastroclonium Kiitz., Kylin (1931) has chosen
G. ovale (Huds.) Kiitz., which species is now known as G. ovatum
(Huds.) Papenfuss (1944).

Although the conservation of Chylocladia Grev. is highly de-
sirable, in view of its invalidation by both Kaliformis Stackhouse
(1809, as represented by K. verticillatus) and Kaliformia Stack-
house (1816, as represented by K. verticillata and K. diaphana),
it is unfortunate that the name was conserved against Sedoidea
Stackhouse (1809). This genus of Stackhouse is based on two
entities, S. purpurea and S. olivacea, both of which are synonymous
with Gastroclonium ovatum; and, as pointed out above, this species
was designated as the type of Gastroclonium by Kylin in 1931. In
its conservation of Chylocladia against Sedoidea, the committee on
conservation was guided no doubt by the belief of Kuntze (1891)
that these names were synonymous.

In addition to Gastroclonium and Lomentaria, the following
genera of red algae have been found to require conservation:
Helminthora J. Ag., Laurencia Lamour., Catenella Grev., Gelidium
Lamour., Pterocladia J. Ag., Chondria Ag., Plumaria Schmitz,
Dasyphila Sonder, and Iridaea Bory. All of them, with the excep-
tion of Helminthora, are invalidated wholly or in part by names
given by Stackhouse.

Hetmintuora J. Agardh (Helminthocladiaceae), Sp. alg. 2(2):
415.1852.

versus
Helminthora Fries, Syst. orb. veg. 1: 341. 1825; Corp. flor. prov.
suecia. I. Flor. scan, 811. 1835.

Type species: Helminthora divaricata (Ag.) J. Ag.

Fries in 1825 erected a genus Helminthora without, however,
referring any particular species to it until 1835, when he assigned
to it Rivularia multifida Weber et Mohr. This species was subse-
quently found to belong to Nemalion Duby (1830).

12 MADRONO [Vol. 9

The later homonym of Helminthora J. Agardh dates from
1852, and was proposed by J. Agardh for a genus belonging to
the same family as Nemalion. Since that time Helminthora J. Ag.
has become well established in algological literature.

In regard to Nemalion, it may be remarked that this genus is
usually ascribed to Targioni-Tozzetti, who presumably published
it in Bertoloni (1819, p. 300). From this work of Bertoloni, it is
obvious, however, that Targioni-Tozzetti had used the name only
in manuscript. Bertoloni described the entity in question as
Fucus Nemalion. The genus Nemalion should consequently be
accredited to Duby, who formally described it in 1830 (p. 959).
In certain older works, the genus is, in fact, ascribed to Duby.
The type species is N. helminthoides (Velley) Batters (1902),
which is an older name for N. lubricum Duby (loc. cit.).

GeELipium Lamouroux (Gelidiaceae), Essai 128. 1813.
versus

Cornea Stackhouse, Tent. mar.-crypt. 57. 1809, in part (as repre-
sented by C. spinosa, op. cit., p. 83; C. filicina, C. sericea, C. pusilla,
and C. deformis, op. cit., p. 84).
Clavatula Stackhouse, Tent. mar.-crypt. 95, 97. 1809.
Kaliformia Stackhouse, Ner. brit., ed. 2, p. ix. 1816, in part (as
represented by K. pusilla, op. cit., pp. xii, 9, pl. 6).
Clavaria Stackhouse, Ner. brit., ed. 2, pp. x, xii, 28, pl. 12. 1816:

Type species: Gelidium corneum (Huds.) Lamour. (cf. Schmitz,
1889, p. 439).

Prerociapia J... Agardh (Gelidiaceae), Sp. ale. 22). 482) 1352:
versus
Cornea Stackhouse, Tent. mar.-crypt. 57. 1809, in part (as repre-
sented by C. capillacea, op. cit., p. 84).
Type species: Pterocladia lucida (Turn.) J. Ag.

CaTENELLA Greville (Rhabdoniaceae), Alg. brit. pp. xiii, 166, pl.
17. 1830.
versus

Kaliformis Stackhouse, Tent. mar.-crypt. 56. 1809, in part (as
represented by K. Opuntia, op. cit., p. 79).
Kaliformia Stackhouse, Ner. brit., ed. 2, p. ix. 1816, in part (as
represented by K. Opuntia, op. cit., pp. xii, 42, pl. 17).

Type species: Catenella repens (Lightf.) Batters (1902, p. 69),
which is the valid name for C. Opuntia (Good. and Woodw.) Grev.

Iriwaea [Iridea] Bory (Gigartinaceae), Dict. class. hist. nat. 9:
15. 1826; in Duperrey, Voy. autour du monde (Coquille), Bota-
nique 2Cl jz VOS..E8 28,

versus
Iridea Stackhouse, Ner. brit.,,ed. 2, pps ix, xi11816,

1947 | PAPENFUSS: PROPOSED GENERA CONSERVANDA 13

Mazzaella J. de Toni, Not. nomencel. alg. vii: [4]. 1936 (May).
Iridophycus Setchell and Gardner, Proc. Nat. Acad. Sci. 22: 469.
1986 (August). 7

Type species: Iridaea cordata (Turn.) Bory.

When describing his genus Jridaea, in 1826, Bory pointed out
that he used this name because the earlier Iridea of Stackhouse
(1816), which was synonymous with Desmarestia Lamouroux
(1813), could not be maintained. Originally, Bory spelled the
name both as Jridea and as Iridaea, but later he (Bory, 1828) and
others spelled it as Iridaea; and it is in this form that the name
has come down to the present. Technically, both versions of the
name are invalid, since Jridea is a later homonym of the genus of
Stackhouse and Iridaea is an orthographic variant of Iridea.

The validity of Iridaea was not questioned until 19386, when
Setchell and Gardner rejected it in favor of their newly erected
Iridophycus on the grounds that (1) it was invalidated by Iridea
Stackhouse, (2) the species, Fucus edulis Stackh., with which it
was first typified by Greville (1830, p. 157) was the type of
Dilsea Stackhouse (1809), and (3) the lectotype, Fucus cordatus
Turn., of J. Agardh (1876) seemingly had been lost and was
‘i . somewhat confused as to exact identity’ (Setchell and
Gardner, op. cit., p. 470). Later, Setchell (1940) succeeded in
tracing the type specimen of Fucus cordatus to the herbarium of
the Royal Botanic Garden of Edinburgh; and a study of it re-
vealed that it actually was a species of Iridaea or Iridophycus.

In recent times, some writers (Taylor, 1989; Smith, 1944)
have accepted Iridophycus while others (Kylin, 1941; Skottsberg,
1941; Baardseth, 1941; Levring, 1944) have retained Iridaea.

Inasmuch as Iridaea Bory has been an accepted genus for a
long time, it seems desirable to conserve it against the earlier
Iridea of Stackhouse. Such action would be especially advan-
tageous in view of the fact that Kylin (1941) has found that the
type and only species of Collinsia J. Ag., C. californica, was a
species of Iridaea. Collinsia had, however, previously been re-
named Mazzaella by J. de Toni (May, 1936), and this name ante-
dates Iridophycus Setchell and Gardner (August, 1936) by several
months. Unless Iridaea were conserved, a comparatively large
number of species would thus have to be transferred to the little-
known Mazzaella.

Lomentaria Lyngbye (Champiaceae), Tent. hydrophyt. dan. 101,
ple 30, He. A. 1819.

versus
Kaliformis Stackhouse, Tent. mar.-crypt. 56. 1809, in part (as
represented by K. articulatus, op. cit., p..78).
Dasyphylla Stackhouse, Ner. brit. ed. 2, p. ix. 1816, in part (as
represented by D. articulata, op. cit., pp. xi, 14, pl. 8).

Type species: Lomentaria articulata (Huds.) Lyngb.

14 MADRONO [Vol. 9

GasTROCLONIUM Kiitzing (Champiaceae), Phyc. general. 441. 1843.
versus
Sedoidea Stackhouse, Tent. mar.-crypt. 57, 88. 1809.
Dasyphylla Stackhouse, Ner. brit. ed. 2, p. ix. 1816, in part (as
represented by D. ovalis and D. sedoidis, op. cit., pp. xi, 25, pl. 12).
Type species: Gastroclonium ovatum (Huds.) Papenfuss (cf.
Kylin, 1931, p. 80).

Priumaria Schmitz (Ceramiaceae), Kleinere Beitr. Kennt. Florid.
vi, 5. 1896. .

versus
Plumaria Stackhouse, Tent. mar.-crypt. 58, 86. 1809.

Type species: Plumaria elegans (Bonnem.) Schmitz (1889, p.
450).

Stackhouse, in 1809, founded a genus Plumaria upon two
entities (at least as to synonymy), Fucus plumosus Hudson (1762,
p- 473) and F. pectinatus Gunnerus (1772, p. 122), which he
united under the name Plumaria pectinata.

Unaware of Stackhouse’s genus, Agardh (1817, pp. xix, 39)
established a genus Ptilota upon exactly the same two entities,
which he united under the name Ptilota plumosa, and an addi-
tional variety B Asplenioides (= Fucus Asplenioides Turner = Ptilota
Asplenioides (Turn.) Agardh).

Although both these genera are still accepted, and usually
accredited to Stackhouse and Agardh, respectively, they have
been the cause of much confusion.

Technically, Ptilota Ag. is invalid; and the species currently
assigned to it should be referred to Plumaria Stackh., and a new
generic name should be created for those species that are now
placed in Plumaria. However, such action would probably aggra-
vate the confusion instead of remedying it. Consequently, it
seems best to retain Ptilota and to reject Plumaria Stackhouse in
favor of Plumaria Schmitz. In 1896, Schmitz “amended” Plu-
maria; but in such a way that it had nothing but the name in
common with the genus of Stackhouse. Although Plumaria
Schmitz, like most old genera, originally contained certain species
that belong elsewhere, it nonetheless includes a group of plants
which is separable from Ptilota Ag. As now accepted, Plumaria
is essentially the genus as it was understood by Schmitz, although
the name is still accredited to Stackhouse. However, in a recent
work, Kylin (1944) broke away from this long-established
custom, and ascribed the genus to Schmitz.

Dasypuita O. G. Sonder (Ceramiaceae), Bot. Zeit. 3: 53. 1845.
versus
Dasyphylla Stackhouse, Ner. brit. ed. 2, pp. ix, xi. 1816.
Type species: Dasyphila Preiss O. G. Sonder.

1947 | PAPENFUSS: PROPOSED GENERA CONSERVANDA 15

Cuonpria Agardh (Rhodomelaceae), Syn. alg. scand. p. xviii.
P31 7.

versus
Kaliformis Stackhouse, Tent. mar.-crypt. 56. 1809, in part (as
represented by K. dasyphyllus, op. cit., p. 78).
Dasyphylla Stackhouse, Ner. brit., ed. 2, p. ix. 1816, in part (as
represented by D. Woodwardii, op. cit., pp. xi, 64, pl. 16; and D.
tenuissima, op. cit., pp. xii, 51, pl. 18).

Type species: Chondria tenuissima (Good. et Woodw.) Ag.

Laurencia Lamouroux (Rhodomelaceae), Essai 130. 18138.
versus
Osmundea Stackhouse, Tent. mar.-crypt. 56, 79. 1809.
Kaliformis Stackhouse, Tent. mar.-crypt. 56. 1809, in part (as
represented by K.'obtusus, op. cit., p. 79).
Pinnatifida Stackhouse, Ner. brit. ed. 2, pp. ix, xii. 1816.
Type species: Laurencia obtusa (Huds.) Lamour. (cf. Schmitz,

1889).
Department of Botany,
University of California, Berkeley.

LITERATURE CITED

Acarpu, C. A. 1817. Synopsis algarum scandinaviae. xl+135 pp. Lund.

AcarpHu, J. G. 1852. Species, genera et ordines algarum, 2 (2). 337-700
[+ addenda, 701-720] pp. Lund.

1876. Species, genera et ordines algarum, 3(1). Enpicrisis sys-
tematis floridearum. vii+724 pp. Lund.
1880. ‘Till algernes systematik, afd. 2. Lunds Univ. Arsskr. 17:
136 pp., 3 pls.
. 1894. Analecta algologica, continuatio 1. Actis Soc. Physiogr.
Lundensis 29: 144 pp., 2 pls.

ArescnHouG, J. E. 1866. Observationes phycologicae. I. Acta Reg. Soc. Sci.
Upsalienses, Ser. 3, 6: 1-26, 4 pls.

BaarDsEetH, E. 1941. The marine algae of Tristan da Cunha. Results Nor-
wegian Sci. Exp. Tristan da Cunha 1937-1938, No. 9. 174 pp., 75 figs.
Oslo.

Batters, E. A. L. 1902. A catalogue of the British marine algae. Jour. Bot.
40 (suppl.): 107 pp.

Bertotoni, A. 1819. Amoenitates italicae,... 472 pp., 6 pls. Bonn.

Bory bE SAInt VINCENT, J. B. 1826. TIridea, Dict. Class. Hist. Nat. 9: 15-16.

—_—————.. 1828. Cryptogamia, in L. I. Duperrey, Voyage autour du monde
... Coquille, 2(1). 301 pp. Paris.

DeCanpotte, A. P. 1801. Extrait d’un rapport sur les conferves. Bull. Sci.
Soc. Philom. Paris 3(51): 17-21, 1 pl.

De Toni, J. 1936. Noterelle di nomenclatura algologica. vii. Primo elenco di
Floridee omonime. [8] pp. Brescia.

Draparnavun, J. 1801. Discours sur les avantages de Vhistoire naturelle, et
discours sur les moeurs et la maniére de vivre des plantes. 41 pp.
Montpellier.

Dury, J. E. 1830. Botanicon gallicum ..., ed. 2,2. [2] + 545-1068 + i-Iviii pp.
Paris.

Fries, E. M. 1825. Systema orbis vegetabilis, pars I. Plantae homonemeae.
vii+374 pp. Lund.

1835. Corpus florarum provincialium sueciae. I. Floram scani-
cam, xxiv +394 pp. Uppsala.

16 MADRONO [Vol. 9

GoovENouUGH, S., and T. J. Woopwarp. 1797. Observations on the British fuci,
with particular descriptions of each species. Trans. Linn. Soc. London
3; 84-235, 4 pls.

GREVILLE, R. K. 1830. Algae britannicae. Ixxxviii+218 pp., 19 pls. Edin-

burgh.
Gunnervs, J. E. 1772. Flora norvegica, 2. [4] +viiit148+ [60] pp., 9 pls.
Copenhagen.

Hooker, W. J. 1833. The English flora of Sir James Edward Smith 5(1).
x +432 pp. London.

Hupson, G. 1762. Flora anglica, ed.1. viii+ [8] +506+ [22] pp. London.

Kuntze, O. 1891. Revisio generum plantarum,...,1. clv+[1]+1011 pp.
Leipzig, etc.

Kirzinc, F.T. 1843. Phycologia generalis. xvit+458+ [1] pp.,80 pls. Leipzig.

Kyun, H. 1931. Die Florideenordnung Rhodymeniales. Lunds Univ. Arsskr.
N. F. Avd. 2, 27(11). 48 pp., 8 figs., 20 pls.

1940. Die Phaeophyceenordnung Chordariales. Ibid. 36(9).
67 pp., 30 figs., 8 pls.
. 1941. Californische Rhodophyceen. Jbid. 37(2). 51 pp. 7
figs., 13 pls.
. 1944. Die Rhodophyceen der schwedischen Westkiiste. Ibid.
40(2). 104 pp., 53 figs., 32 pls.

Lamarck, J. B. Dre, and A. P. DeCanpotir. 1805. Flore francaise, ed. 3, 2.
xii+ 600 pp., map. Paris.

Lamovrovux, J. V. F. 1813. Essai sur les genres de la famille des thalassio-
phytes non articulées. Ann. Mus. Hist. Nat. Paris 20: 21-47, 115-139,
267-293, 7 pls.

Levrine, T. 1944. Meeresalgen von den Crozet-Inseln und Kerguelen. Arkiv.
t. Bot. 31YA(8). 31 pp; 19 figs:

Licutroot, J. 1777. Flora scotica 2: 531-1151 + [24] pp., 12 pls. London.

Link, H. F. 1820. Epistola...de algis aquaticis in genera disponendis
scripsit, in C. G. D. Nees von Esenbeck, Horae physicae berolinenses,
..<«, No.1. 8 pp., 1 pl. Bonn.

Linnaeus, C. 1753. Species plantarum, 2: 561-1200+ [31] pp. Stockholm.

Lynosyr, H. C. 1819. Tentamen hydrophytologiae danicae. xxxiit+248 pp.,
70 pls. Copenhagen.

NIEUWLAND, J. A. 1917. Critical notes on new and old genera of plants. X.
Am. Midl. Naturalist 5: 50-52.

Parenruss, G. F. 1943. Notes on algal nomenclature. II. Gymnosorus J.
Agardh. Am. Jour. Bot. 30: 463-468, 15 figs.

. 1944. Notes on algal nomenclature. III. Miscellaneous species
of Chlorophyceae, Phaeophyceae and Rhodophyceae. Farlowia 1: 337—
346.

. 1945. Review of the Acrochaetium-Rhodochorton complex of
the red algae. Univ. Calif. Publ. Bot. 18(14) : 299-334.

Rousset, H. F. A. 1806. Flore du Calvados et des terreins adjacens, ...
ed. 2. 371+[1] pp. Caen.

Scumitz, F. 1889. Systematische Ubersicht der bisher bekannten Gattungen
der Florideen. Flora 72: 435-456.

1896. Kleinere Beitrige zur Kenntniss der Florideen. VI.
Nuova Not. 7: 1-22.

SetcHeLL, W. A. 1940. Fucus cordatus Turner. Proc. Nat. Acad. Sci. 26:
643-651, 2 figs.

SETCHELL, W. A., and N. L. Garpner. 1925. The marine algae of the Pacific
coast of North America. Univ. Calif. Publ. Bot. 8(3): 383-898, incl.
73 pls.

1936. Jridophycus gen. nov. and its representation in South
America. Proc. Nat. Acad. Sci. 22: 469-473.

SxortsserG, C. 1941. Communities of marine algae in subantarctic and ant-
arctic waters. Kungl. Svenska Vetenskapsakad. Handl. Ser. 3, 19(4).
92 pp., 7 figs., 3 pls.

’

1947] QUICK: GERMINATION 17

Smitu, G. M. 1944. Marine algae of the Monterey Peninsula. vii+ [1] + 622
pp., incl. 98 pls.

Sonper, O. G. 1845. Nova algarum genera et species, quas in itinere ad oras
occidentales: Novae Hollandiae, collegit L. Preiss, Ph.Dr. Bot. Zeit. 3:
49-57,

SracknouseE, J. 1809. Tentamen marino-cryptogamicum, ... Mem. Soc. Imp.
Nat. Moscou, 2: 50-97.

1816. Nereis britannica, ed. 2. xiit+68 pp., 20 pls. Oxford.

Tanpy, G. 1935. Additions and amendments to the international rules of
botanical nomenclature, ed. 3. Appendix III. 1. Algae. Kew Bull.
1935: 82-83.

Taytorn, W. R. 1939. Algae collected by the “Hassler,” “Albatross,” and
Schmitt Expeditions. II. Marine algae from Uruguay, Argentina, the
Falkland Islands, and the Strait of Magellan. Papers Michigan Acad.
Sci., Arts and Letters 24: 127-164, 7 pls.

GERMINATION OF PHACELIA SEEDS
CLARENCE R. Quick

In the fall of 1942 the writer made some tests on the germi-
native reactions of Phacelia seeds. This work was carried on to
facilitate cyto-taxonomic studies in Phacelia being carried on by
Dr. Marion S. Cave and Dr. Lincoln Constance of the Department
of Botany, University of California, Berkeley, which have culmi-
nated in the publication of three papers on chromosome numbers
in the Hydrophyllaceae (Univ. Calif. Publ. Bot. 18: 205-216.
1942; 18: 2938-298. 1944; in press). The methods used were
based upon previous experience in the germination of seeds of
Ribes and Ceanothus, and of other plants native to California.
Ribes seeds, which are similar to those of Phacelia in size, appear-
ance, and ecologic relationships, have been studied extensively
by the writer during the past fifteen years.

Quick (1935) showed that germinative reactions of Ceanothus
species vary widely, and that variations tend to correlate with the
ecologic and taxonomic affinities of species and groups of species
concerned. Mirov (19386) summarized the results of a large num-
ber of germination tests on seeds of native California plants with
respect to taxonomic position, altitudinal distribution, and growth
form of the species concerned. In general he found no consistent
relation between systematic position and germinative behavior,
but he did find definite correlations between germinative behavior
and altitudinal distribution. He found that failure of germination
due to seed-coat dormancy was more common in plants from low
altitudes, and that failure due to embryo dormancy was more
often encountered in plants from high altitudes. Seeds of annuals
were less generally dormant than seeds of herbaceous perennials,
shrubs, and trees.

The low germination percentages for some samples of Phacelia
seeds, as reported by Mirov (1940) and Mirov and Kraebel
(1987), suggested that stratification might be necessary to satis-

‘spotted ply} pue puoddas a4} SUIMOT[OJ JUD Aod QJ JO S[e}O} JO 10 ‘UOT}BOYIVe.AYS YSAY IY
SUIMOT[OJ BLOW 10 JUID Lod OF JO SUOIZVUIULAIS [[V@ PUB UOTJDIT[OD YORI OJ ase JUIIAed YsoysIy oY} JUVSIIde4 9dAY VOVJ Plog UI Sons 4+
*xo[duioo propdAjod v7] fhydowazay
—DIWULO fypa—vaUun] abou *q IY} JO SLoquiowl ‘9oUL}SUOD U[OOUL]T 0} SUIpPlooo"e ‘a1e [[ JSP[ BY} ‘sTenuuP ae YYAIDY_ JO Saeds Gg ASAT x

[ Vol. 9

UMOUU | 22S "110 J, pyiydoona |

1g 19 G 69 0 0 SE 0 0) 0
8e Gov Go 9€ G&S GG VS SIT. V O QO G9 Nie IUVIAL) SUIQDINUE “|
GL GE GL | 8g O€ OG vv 8 0 0) OO GOiee nee sae dudaLH) 0JIVAWOdI *q
LS 6G 0 eV €P 6G €V al ) ) COTS Is + |Ps ee ‘QpAY vurdjn -g
€9 SP St | GL s € | 9s j j 0 OOPS. 2715 aes qusaI) DjaDAdWOd “gq
€9 SI ST | vi ey 1G OZ VS OT 9 0) OOS 0S ans a IUDIAL) SYIQDINU * J
LG OS Se €8 | S OG 09 SE 0 ) GOO GES alee eee JUIIAL) SUIQDINU * J
ZS TE Zl 9P O€ OL v8 Poe aie <0 ) OO GS el a see ee, ‘qpAy vurdjn *g
OG TT «61 69 0 9 0) feGcea): <0 | 0 OOO thee auIa1Q ny iydhsop gq
rASi 0% 0 Ze 8G sé | 08 4 a 6 QOO:S: Fl ees “ALOT, DI iydoona) “gq
S GE 91 | G | £1 CT PG L9 G9 9 0 OO SSO Fee OUBIAL) $UIQDINU "J
fo) 0 I 0 20 0 02 Lv 96 | 29 Z9 2000; Gy nl te te ABIy Vy nuowweaT ‘gq
oat L 13 0 03 L p} LG LV L9 09 OQOOLG "|e ees ‘Yue susp ‘gq
la 0 0 0 g I € Ge GZ ag G8 OOS |e IP*OH Lf UY9NO *d
< 0 0 0 jh 0 0 SI ue 9L €6 O008 ase Uopsu0) »1094]02 "gq
a 9 Z el i 0 3 Z 8 OT 49% OOO ah es ‘yuo suvysep ‘qf
skep sAep sfep sep shep shvep sep sAep skep ey (q203)
LVT PST PST OPT GIL raul cVl PIT 98 -yiyeays pape]][oo
O60 ~Or0 ot Ocak 6G | eo | Wel || OeoG. |” oO 6S oN sem yuryd
da +" V la d "#14 | OF" | 2 8TH | V "I | GPT | OPT | Pla Vv Ie Yor ye «Sotoods yueTg
epnyzy|e
Sane UOHEa nese puosves uoT}ROyIVe.A}s seeriae
-juLia8 [eUCHIppY 1d} Je UOI}PVUIUIII¢d [BUOITIPPpY }S1Y 1o}Je UOTJEUTWIID ,

(‘o1n}[ND pus WOLJ pos1oUla Jey} SSuUI[pses Jo sasejUs.Ied a1e SsainsI,y)
‘poJeoIpuUl Ssoin}zeroduls} 3e@ pue spotted 10} UOT}eOYI}VeI}s Ja}ye asnoyuseis sy} Ul Spaes vyaInYq JO UOT}eUTUIAIy

18

‘[ a1av yy,

1947] QUICK: GERMINATION 19

factory germination of many species of Phacelia seeds. Unpub-
lished data of the author indicated that seeds of Phacelia mutabilis
had dormant embryos, and suggested also that the seed coat was
somewhat impermeable to water. Quick (1943) outlined several
common methods of forcing germination of dormant seeds. Dor-
mant seeds, that is, seeds not immediately germinable, often grow
satisfactorily in the greenhouse if the germination period is im-
mediately preceded by one of stratification, that is, by a period of
refrigeration with an adequate supply of air, moisture, and
mineral nutrients. Difficulty in plumping seeds with slightly im-
permeable seed coats is sometimes obviated by repeated alternate
exposures at stratification and greenhouse temperatures. Another
treatment which is sometimes successful with hard-coated seeds,
and which may be called delayed stratification, subjects dormant
seeds to greenhouse temperatures under favorable germinative
conditions prior to any refrigeration treatment. This delayed
stratification treatment is attained if a control culture—a culture
placed in the greenhouse immediately after planting to see
whether untreated seeds will germinate—is subjected to treat-
ment in cold storage, following the exposure at the greenhouse
temperature, and is then retested for germination.

Table 1 presents results of germination tests upon seeds of
several species of Phacelia listed in the approximate order of their
germinative reactions. The sequence of treatments applied to flat
A may be outlined as follows: The seeds were planted in moist,
washed, autoclaved river sand on October 11, 1942. The cultures
were wet with a mineral nutrient solution and thereafter kept
moist at all times, treated with copper oxalate to prevent damp-
ing-off, and placed immediately in the greenhouse for germination
test. After 56 days in the greenhouse they were subjected to 112
days of stratification at 2.2° C. and were then returned to the green-
house. After a second germination period of 35 days they were
again placed in stratification, this time at 0° C. for 154 days. On
October 5, 1943, they were placed in the greenhouse for a third
germination test and after 35 days were discarded. Flats B, C,
and D were given a similar sequence of treatments. The tests
averaged about 50 seeds per culture.

From the author’s experience with many seeds, especially with
those of Ribes, the maximum effects from stratification can be
expected from the three low temperatures of 5°, 2:2°, and 0° C.,
for periods of approximately 84, 112, and 140 days, respectively.

It is apparent that seeds of the four annual species of Phacelia
were not dormant; that is, there was little or no interference to
immediate greenhouse germination. Seeds of these species were
usually collected at lower altitudes than the perennial species.
Seeds of many of the perennial species were best conditioned for
germination by stratification at 0° C. In germination tests follow-

20 MADRONO [Vol. 9

ing the first refrigeration treatment, one sample showed germina-
tion only after stratification at 0°. Several species failed to
germinate after stratification at 5°. Stratification at 0° may obvi-
ate seed-coat dormancy more effectively than stratification at
slightly higher temperatures because of the probable alternation
of temperatures above and below the freezing point of water.
In any event, stratification at 0° was generally much more satis-
factory for conditioning seed of the perennial species for germi-
nation. This result suggests that some of the special treatments
for making hard seeds permeable to water might be used to ad-
vantage on a number of Phacelia species. The appreciable germi-
nation obtained after second and even after third refrigeration
indicates that freezing-point stratification may be used advantage-
ously on many Phacelia species, and suggests again that the coats
of many Phacelia seeds are more or less impermeable.

Under the conditions of the experiment, retrial (second)
stratification at 0° C. caused many more samples to reach or
exceed 70 per cent germination than retrial stratification at 2.2°.
For example, in flat B retrial stratification at 2.2° after primary
stratification at 5° caused only one sample to reach 70 per cent
germination, whereas still another (third) stratification of the
same flat, at 0°, caused seven additional samples to equal or
exceed 70 per cent germination.

In conclusion, then, it appears that seeds of many annual
species of Phacelia will grow immediately and without special
treatments designed to obviate seed dormancy. Seeds of many
perennial species are satisfactorily conditioned for germination
by stratification treatment at 0° C. for 140 days. An alternation
of stratification periods at low temperature with germination
periods in the greenhouse is more generally productive of seed-
lings than a single stratification treatment followed by a germi-
nation period in the greenhouse.

Bureau of Entomology and Plant Quarantine,
Agricultural Research Administration,

United States Department of Agriculture,
Berkeley, California.

LITERATURE CITED

Mirov, N. T. 1936. Germination behavior of some California plants. Ecology
17(4) : 667-672.
1940. Additional data on collecting and propagating the seeds
of California wild plants. U.S. Forest Serv., Calif. Forest and Range
Expt. Sta., Res. Note 21, I7 pp. [Processed. |
Mirov, N. T., and C. J. Krarset. 1937. Collecting and propagating the seeds
of California wild plants. U.S. Forest Serv., Calif. Forest and Range
Expt. Sta., Res. Note 18, 27 pp. [Processed. |
Quick, C. R. 1935. Notes on the germination of Ceanothus seeds. Madrono
3: 135-140.
. 1943. Certain methods of forcing the germination of seeds.
Jour. Calif. Hort. Soc. 4: 95-102.

1947] GENTRY: MIMULUS 21

THE GENUS MIMULUS IN OR ADJACENT TO
SINALOA, MEXICO

Howarp Scorr GENTRY

In the course of naming my Sinaloa collections of Mimulus, it
was found that only two species, M. pallens and M. verbenaceus,
have been published as occurring in the Mexican state of Sinaloa.
These are represented by two collections cited by Grant in her
monograph of the genus (Mo. Bot. Gard. Ann. 11: 99-388. 1924).
This paucity of material in a large genus that is known to range
widely north and south of Sinaloa can also be detected in many
other montane genera, for little mountain collecting has yet been
done in that area.

The following list records the species now known to occur in
Sinaloa, as based upon recent collections. Also included are
several species which have been found adjacent to Sinaloa and
hence can be expected to be found there as explorations continue.
A provisional key is offered for ready determination of species
in the area and to orient two entities proposed as new species.
Following the key the species are annotated in alphabetical order.

Key ro Species or Mimuuus

Corolla red or orange-red, large, ca. 2 inches long (Sect.

Erythranthe).
Calyx teeth unequal; style included in calyx ............ 6. M. Nelsoni
Calyx teeth equal; style exserted ....:...2..222.0.:.05. 9. M. verbenaceus

Corolla yellow, sometimes red-spotted, less than 1 inch long.
Calyx teeth equal, acute; style twice length of capsule
(Sect. Paradanthus)) 0540404 .0¢a%os.444c8eas ge os 3. M. floribundus
Calyx teeth unequal, upper longer than others (Sect. :
Simiolus).
Calyx teeth three
Calyx teeth five.
Flowers terminal, racemose; corollas 15 mm. long or
TI) OG near ie ee ne eg al eee a are ees 5. M. guttatus
Flowers axillary, usually solitary; corollas less than
15 mm. long.
Corolla lobes laciniate ..............0... 000.005. 2. M. dentilobus
Corolla lobes not laciniate.
Calyx lobes broadly rounded, sometimes mucro-
nate; plants usually over 10 cm. high ...... 4, M. glabratus
Calyx lobes triangular-acute, mucronate; plants
less than 10 em. high.
Calyx 7-8 mm. long in fruit; style long ex-
serted; basal leaf blades all less than 1.5
cm. long; petioles narrow ............... 8. M. Pennellii
Calyx 9-10 mm. long in fruit; style included
in fruiting calyx; some basal leaf blades
2 cm. long or more; petioles broadly
winged, clasping the stem .............. 1. M. calciphilus

ee eee en trea des anh as ee 7. M. pallens

1. Mimulus calciphilus sp. nov. Annuus, ad basim multo-
ramosus, 7-14 cm. altus; foliis ad basim aggregatis, late ovatis,

22 MADRONO [Vol. 9

obtusis, undulati-dentatis, glandulosi-pilosis, 2.5-3.5 cm. longis,
2—3 cm. latis; petiolis brevidus alatis; pedunculis 2—4 cm. longis;
calyce angulati-urceolato, pubescenti, circo 1 cm. longo, dente
supremo 3-4 mm. longibus, dentibus lateralibus 1 mm, longis,
ventralibus 1.5-2 mm. longis; corolla 1 cm. longa, flava, lobis
inequalibus, eis lateralibus dorsali quam brevioribus; antheris
glabris, inclusis; stylo glabro, incluso; capsula ovati-subglobosa,
subsessili.

Small mesophytic annual 7-15 ecm. high with a basal rosette
of large, broadly ovate leaves and wide-winged petioles; stems
scapose, pubescent, shallowly sulcate, bearing 1-2 pairs of leaf-
like bracts; leaves radical, 2.5—-3.5 cm. long, 2-3 cm. wide, 5—7-
nerved, villous on both surfaces, undulate-denticulate, obtuse,
broadly decurrent making a winged petiole; pedicels 2—4 cm.
long, axillary from radical leaves or from the denticulate leaf-like
bracts on the scape, simple, glandular-villous; calyx accrescent,
8-11 mm. long, angulate-urceolate, sparsely glandular villous,
red-tinged in maturity, the lobes triangular-acute ; upper lobe 3—4
mm. long, lateral lobes 1 mm. long, ventral lobes 2 mm. long and
often folding over lateral lobes; corolla about 1 cm. long, yellow,
the limb short; lobes rounded, equal, 1 mm. long, 2 mm. broad;
stamens glabrous, included in corolla; stigma broadly lanceolate-
lobed, the lobes subequal; style glabrous, included in fruiting
calyx; ovary glabrous, short-stipitate ; capsule oblong, subsessile,
5 mm. long.

Type. Near Los Pucheros, at about 6500 feet elevation,
Sierra Surotato, Sinaloa, Mexico, March 17-24, 1945, Gentry 7217
(Univ. of Mich. Herbarium; isotypes to be distributed).

This species appears to have no close relative in North
America. It belongs in the section Simiolus and in Grant’s mono-
graph (op. cit., p. 145) keys to M. Whipplei. The radical leaves
with broadly winged petioles and five to seven veins, the larger
calyx, equally lobed corolla, and the included style and stigma
are only some of the more important characters distinguishing it
from M. Whipplei and other members of the genus. The flowers
are persistent for a Mimulus. With its basal rosette of leaves and
scapose stem flanked by rather sinuous pedicels holding the large
semi-nodding calyces, M. calciphilus presents a singular appear-
ance. The only observed occurrence of the species is the small
compact colony in rich humus soil of the talus slope footing the
calcareous crags (whence the specific name) known locally as
Los Pefiascos.

2. Mimuuus peENTILOBUS Rob. & Fern. Proc. Am. Acad. Sci. 30:
120. 1895.

Mimulus dentilobus is a small, delicate, finely cut plant, the
small heads not reaching more than 3 to 4 inches in height. If

1947 | GENTRY: MIMULUS 23

the plants are observed closely, the laciniate corollas are visible
to the naked eye and they serve at once to identify the species.

This is a rarely collected species known previously from
southern New Mexico, southern Arizona, Sonora, and from above
Puerto Escondido in the Sierra Giganta in the southern part of
Baja California. Recently it has been collected in the Sierra
Surotato, northern Sinaloa, in a cave near Los Pucheros at about
6500 feet elevation (Gentry 7220), growing in a tight mass on
bare igneous rock kept moist by dripping water. Cattle had
access to the cave and had stripped off the succulent moist carpet
as high as they could reach.

3. Mimutus FLtorisunpus Dougl. in Lindl. Bot. Reg. pl. 1125.
1828.

A cespitose plant, the stems slender, pilose, often decumbent
and rooting at the nodes, sometimes reaching 30 cm. in height but
usually shorter. A rather slimy or wet-viscous herb. Corolla
yellow with reddish spots in the open throat.

In western North America this species is known from British
Columbia to Jalisco, Mexico (Meaia 1853). In Sinaloa, as indi-
cated by the following collections, it appears to be common
through the cerro and valley region of the Sierra Madre pied-
mont: Varomena, Gentry 7147; Puerto a Tamiapa, Gentry 5&66a;
Cerro Colorado, Gentry 5469; San Blas, Rose, Standley & Russel
13395,

4. Mimuxus Guasratus H.B.K. Nov. Gen. & Sp. 2: 370. 1895.

This species as interpreted by Grant (op. cit.) is a wide-ranging
highly variable perennial herb. Grant cited it from several states
in Mexico north, south, and east of Sinaloa. There are, however,
no known collections from Sinaloa, but further explorations should
reveal it there.

5. Mimuxus eutrratus DC. Cat. Hort. Monsp. 127. 1813.

There are no known collections of this species from Sinaloa,
but since it has been found in adjacent Sonora (Gentry 1305, Car-
negie Inst. Publ. no. 527: 237. 1942) and Chihuahua (Gentry 2780,
loc. cit.), it doubtless occurs in northern Sinaloa.

6. Mimutus Netsoni Grant, Ann. Mo. Bot. Gard. 11: 144.
1924,

This species is known only from the adjacent state of Du-
rango, “30 miles north of Guanaceri [ Guanacevi], Sierra Madre
(Nelson 4775). This locality is close to the Sinaloa-Durango
boundary.

7. Mimuxus pauttens Greene, Leafl. Bot. Obs. & Crit. 2: 4.
1909.

This species is distinguished by the three-toothed calyx. The
yellow flowers are spotted red in the open throat. It forms showy

24 MADRONO [Vol. 9

colonies 15 to 25 cm. high along streams. It appears to be re-
stricted to a riparian habitat and is rather common through the
higher elevations, only rarely dropping below the lower limit of
oaks. Itis a Sierra Madre species known previously from Sonora,
Chihuahua, and Durango. The following Sinaloa collections have
been made: Los Pucheros, Sierra Surotato, Gentry 7213, 7232;
above La Jolla, Sierra Surotato, Gentry 7285; all of the collections
were made in March when the plant was in full bloom.

8. Mimulus Pennellii sp. nov. Annuus vel perennis; caulibus
procumbentibus, 6-12 mm. longis, sparsim villosis foliis late
ovatis, dentatis acutis, petiolatis, supra villosis, infra glabris, 6-10
mm. longis, 8-12 mm. latis; pedunculis ‘axillaribus, 1.5-3 cm.
longis; calyce 6-8 mm. longo, puberulenti, dentibus triangulari-
bus, acutis, inequalibus, dente supremo alteris plus quam duplo
longiore; corolla 8-9 mm. longa, flava; staminibus glabris, in-
clusis; stylo glabro, laciniis aequalibus, exsertis, ligulatis; capsula
ovata, attenuata, subsessili; seminibus levibus fuscis.

Small annual or perennial aquatic herbs 5—8 cm. high, cespi-
tose, rooting at the nodes, the leaves petioled and calyces narrow.
Stems procumbent, slender, sparsely pilose, glabrate, reddish, the
internodes 1-3 cm. long; leaves petiolate, the petioles 2-6 mm.
long, blades 6-10 mm. long, 8-12 mm. wide, broadly ovate to
triangular-ovate, dentate, 5-nerved, minutely scabrous, glabrate
or glabrous below and villous above; peduncles axillary, single,
1.5—3.0 cm. long, pubescent near the base; calyx weakly accres-
cent, 6—8 mm. long, narrowly funnelform or strictly campanulate,
ascending-strigillose within, sparsely pubescent outside and rather
lucid, the lobes triangular-acute, upper 2—3 mm. long, the others
1-2 mm. long, the ventral folding over weakly in maturity; corolla
8—9 mm. long, yellow with red-spotted throat, the lobes rounded,
spreading; stamens included, glabrous; stigma lobes subequal,
broadly lanceolate; style flat, minutely striate, glabrous, exserted
beyond fruiting calyx; ovary glabrous, subsessile; capsule nar-
rowly ovoid-terete, broadly lobate between the sutures, the mar-
gins of the lobes undulate; seeds light brown, ovoid-orbicular,
somewhat depressed.

Type. Rancho Africa, elevation 2000 to 3000 feet, Sierra
Tacuichamona, Sinaloa, Mexico, February 19, 1940, Gentry 5691
(Univ. of Mich. Herbarium; isotypes in the following herbaria:
Gray, Univ. Ariz., Univ. Calif., Stanford Univ., Mo. Bot. Gard.,
N. Y. Bot. Gard., Inst. Biol. Mex., Gentry).

This diminutive hydrophyte formed a loose mat on a large
water-covered rock in tropical montane forest. In Grant’s mono-
graph it keys to M. glabratus, but is distinguished from that broad
far-flung complex by the petiolate upper leaves and particularly
by the rather narrowly triangular acute calyx lobes. It is named
in honor of Francis W. Pennell, authority on the family.

1947] ISENBERG: REDWOOD 25

9. MiIMULUs VERBENACEUs Greene, Leafl. Bot. Obs. & Crit. 2: 2.
S09"

The natural habitat of this species is montane in the origins of
streamways, such as rocky seeps and waterfalls in canyons where
permanent moisture is available. In such situations it is often
found in low bushy clumps. Occasionally it occurs in a less cespi-
tose form along the lower river courses, as near Fuerte, but the
seasonal floods sweeping such rocky river beds alternately initi-
ate and destroy lowland adventives of the species.

Known from the Grand Canyon, Arizona, to northern Sinaloa.
Sinaloa, Mexico: above La Jolla, Sierra Surotato, Gentry 7254;
sandy soil along the river near Fuerte, Rose, Standley, 5 Russel
18074. Grant (op. cit., 144) cites Sierra de Alamos as in Sina-
loa. This is an error since that mountain is in Sonora.

Publication no. 838 from Botany Department,
University of Michigan, Ann Arbor.

LOCATION OF EXTRANEOUS MATERIALS
IN REDWOOD

Irvine H. IsENBERG

This paper is one of a series originating from the laboratories
of The Institute of Paper Chemistry, Appleton, Wisconsin, and
covering a fundamental study of the botanical, chemical and other
characteristics of the California redwood (Sequoia sempervirens) ;
this work has been sponsored by The Pacific Lumber Company,
Scotia, California.

In order to study the location of extraneous materials in red-
wood and the effect of certain treatments on these substances,
microsections (20 y thick) were made from the following loca-
tions in a redwood tree—namely, outer portion of heartwood,
sapwood, junction of sapwood and heartwood, and rootwood.
The major portion of the work was done on the heartwood sec-
tions. The treatments used on the various sections and the results
obtained are, for the most part, described briefly in Table 1.

HEARTWOOD

In the heartwood the extraneous material is located chiefly in
the cell cavities of the wood ray parenchyma and the longitudinal
parenchyma but may also be present in the walls of these cells as
well as in the walls of the tracheids, the only other type of cell
structure present in redweod (pl. 1, figs. 1,2). This extraneous
material includes mainly dead protoplasm, proteins, starch,
tannins, phlobaphenes, and fats, and is dark red in color, espe-
cially in the longitudinal wood parenchyma cells.

The reaction of the cell contents to the various reagents, as

[ Vol. 9

26 MADRONO
Taste 1. Errect or REAGENTS ON CELL CONTENTS
TREATMENT APPEARANCE AND RESULTS
Heartwood
None

Dilute ferric chloride
Cold alcohol for 5 minutes
Hot alcohol

Hot water for 3 hours
(following hot alcohol)

Hot dilute alkali
(1% NaOH) for 3 hours
(following alcohol and water)

Dilute ferric chloride

Cold alcohol

None
Dilute ferric chloride
Hot alcohol

Acidulated alcohol

(3% acetic acid) for 3 hours
Hot water for 3 hours
(following acidulated alcohol)
Hot dilute alkali

(1% NaOH) for 2 hours
(following alcohol and water)

Dark red (especially the. longitudinal wood
parenchyma).

Cell contents bluish black.
Considerable amount of cell contents dissolved.

Some cell contents still remain, located mainly
in ray parenchyma.

Small amount of cell contents still retained.

All extraneous material removed. Cell walls
greenish, slightly swollen.

Sapwood

Only a small amount of cell contents affected
in color, and then only slightly.

Considerable amount of material remains, yel-
low in color.

Rootwood

Light to dark brown in color.

Color darkened, but not uniformly.

Considerable amount of material of various
shades of brown persists.

Tannates converted to tannins; large amount
of extraneous material retained.

Some extraneous material remains (pl. 2,
figeS)),

All extraneous material removed; cell walls
greenish yellow (pl. 2, fig. 4).

well as the analytical data reported elsewhere by Lewis [ Chemical
Nature of Redwood, The Vortex, 7(5): 218. May, 1946], indi-
cate a considerable amount of tannin and phlobaphene in the
heartwood. For comparison, these data are presented in Table 2.

If the cross-sections, which have been previously extracted
with hot aleohol for three hours and hot water for three hours,
are treated alternately with chlorine gas and a 3 per cent alcoholic
solution of monoethanolamine in such a way as to dissolve the
lignin without hydrolysis of the non-cellulosic carbohydrates the
holocellulose is left. The visible effect (pl. 1, fig. 3) of the lignin
solvent is a removal of the intercellular material so that the cells
pull apart by themselves or at the touch of a dissecting needle.
The summerwood ring seems particularly susceptible to handling.
It is impossible to obtain other than very small sections of holo-
cellulose unless the treated section is handled on a microslide

1947 | ISENBERG: REDWOOD 27

throughout the operation. After the holocellulose is prepared, a
piece is treated with 72 per cent sulfuric acid; if it dissolves (leav-
ing only minute traces), it is considered that the lignin has been
removed during the preparation of the holocellulose. When the
holocellulose is treated with dilute sulfuric acid to remove the
easily hydrolyzable carbohydrates and leave a material closely
resembling Cross and Bevan cellulose, the structure does not
appear to change.

Taste 2. Amountv oF TANNIN AND PHLOBAPHENE IN REDWOOD
(Based on oven-dry wood)

Per cent of tannin Per cent of phlobaphene
PVCATLWOOR © si Me toe oe es 4.1 6.2
SapwoOod) 46.46 fies eens dv gee 0.2 0.7
ROOEWOOG! “ie oe ce oe das 7.8 22.1

When extractive-free cross sections are treated with 72 per
cent sulfuric acid to remove the carbohydrates and leave the
framework of lignin, the cell walls immediately turn green and
begin to swell. Since the walls of the summerwood cells are much
thicker than those of the springwood cells, the swelling is greater
and the summerwood is distorted considerably. This swelling
causes a loss of identity in some of the summerwood cells, espe-
cially if the acid is removed and the section is washed and dehy-
drated preparatory to making a permanent slide mount. Aftera
brief period of exposure to the concentrated acid the structure
turns brown but continued contact apparently causes no further
action (pl. 1, fig. 4).

SaPwoop

Although considerable extraneous material is present in the
cell cavities of parenchymatous cells in the sapwood, they do not
appear to contain much tannin. This observation is in agreement
with the low tannin content of the sapwood as determined ana-
lytically. .

Rootwoop

The rootwood contains large amounts of extraneous materials
and many of the tracheids are filled, in addition to the ray paren-
chyma and longitudinal parenchyma cells (pl. 2, figs. 1,2). This
extraneous material is light brown to dark brown in color.

Uses aND SUMMARY

Redwood makes an interesting plastic pulp because of the
tannin and phlobaphene present. These substances flow under
high pressures and serve as binders. Unfortunately, in rootwood
the fibers are not suitable and the resin content is too high for this

[ Vol. 9

~

MADRONO

z ‘ Se pee
%, Sees “: i
ses Re cS am
Leg Be ™ x
Be Sissies porn cis

S
cS,

‘

“a

Be

oes

Transverse sec-

Mies i:
ransverse section showing holocellulose.

OOD OF SEQUOIA SEMPERVIRENS.

WwW

Heart
Fic. 2. Radial section.

PLATE 1.

i)

lA

Fic. 3.

tion.

(All

sulfuric acid.

72 per cent

after treatment with

Fic. 4. Transverse section

xX 50.)

ISENBERG: REDWOOD 29

1947]

Transverse sec-

xtrac-
1 per

Fic. 1.

Fic. 3. Transverse section after 3 hours e
ent acetic acid) and 3 hours extraction

SEMPERVIRENS.

c
Fic. 4. Transverse section extracted for 2 hours with

xide in addition to treatment with hot acidulated alcohol and

5
46 2
5) op)
2 ~~
&
33
=
6.28
Ben
Sore ae
6 Se Sox
Bo <
es oS
ces
ee esp NS,
SSIS ee
ata sla
io) D> sy
SA Sasa)
Smee OSS
BGaSs
Ry, =
hom)
Sk&pe-r
eI ey ie anc
eo) Se

30 MADRONO [Vol. 9

purpose. However, it is of interest as a source of extractive
materials.

Tannin from redwood is a good tanning agent and anti-oxi-
dant; the phlobaphene is a source of catechol on vacuum destruc-
tive distillation. Roots and stumps on destructive distillation
yield a tar richer in phenol, creosol, guiacol, and p-cresol than
that obtained from sapwood and heartwood. Although the tannin
is not highly toxic it does help, with related materials, to make the
heartwood rot resistant.

It appears from this study that most of the tanniferous mate-
rial present in redwood is located in ray parenchyma and longi-
tudinal parenchyma cells. The cell walls must contain some also,
because it is unlikely that solutions would not diffuse into the
walls.

The Institute of Paper Chemistry,
Appleton, Wisconsin.

REVIEW

The Pacific Coast Ranges. Edited by Ropericx Pratrtir. Con-
tributors Archie Binns, John Walton Caughey, Lois Crisler,
Aubrey Drury, Idwal Jones, Donald Culross Peattie, Thomas
Emerson Ripley, Richard Joel Russell, Judy Van der Veer, and
Daniel C. Willard. 402 pp. 4 maps. 29 illustrations. $3.75.
The Vanguard Press, New York. 1946.

This volume is the fourth in a popular series of books written
about the American mountains. The three previous volumes have
been published under the titles The Rocky Mountains, The Great
Smokies and the Blue Ridge, and The Friendly Mountains (Green,
White, and Adirondacks). In this volume, The Pacific Coast
Ranges, a capable editor and equally capable contributors have
presented in non-technical language and in a narrative form an
exceptional amount of general scientific and historical informa-
tion. The scope and varied nature of the subject matter make a
detailed review difficult and impractical. However, an outline of
the thirteen chapters or narratives will give an idea of the exten-
sive subject matter treated. ;

The California missions and the first peoples of the Coast
Ranges furnish the subject material for the first two chapters,
“Father Serra’s Rosary,’ by Donald Culross Peattie and “The
First Inhabitants of the Coast Ranges,” by John Walton Caughey.
The two authors give slightly different interpretations of the fate
of the Indians upon secularization of the missions. In the first
chapter, page eleven, we read:

The missions were secularized, that is reduced to parish churches with

a single priest, and stripped of everything except the immediate build-

ings themselves. First, many of the pioneering padres who had been
men of education and high ideals, were supplanted by inferior friars,

1947] REVIEW 31

eee Then the lands, which the fathers held in trust for the Indians
and brought to high productivity, were taken over and given in immense
feudal tracts to settlers from Mexico, the rancheros and ranchmen.
The Indians who had given up their native life for the white man’s way,
were stripped of both at once, and so driven to beggary or to acts of
violence for which they could be punished.

In the second chapter, page forty-two, we read:

The missions, therefore, were secularized. As religious institutions they
were transformed into parish churches. The Indians were released
from the friars’ paternal care and were put on their own. From the
property that had been accumulated and held in trust for the Indians,
each released neophyte was allotted a reasonable amount. Secular
administrators controlled the oftentimes considerable residue, and the
surplus lands reverted to the state, which was ready to parcel it out
generously to prospective rancheros.

In the second account, written by Dr. John Caughey, the historian,
it appears that the Indians were given slightly more consideration
at the time the missions were secularized.

The third narrative, “Footsteps of Spring—a Wild Flower
Trail,’ by Donald Culross Peattie, gives an account of the native
vegetation of the Coast Ranges as it comes into flower, beginning
in the south and continuing to the north. The botanical informa-
tion here included and style of presentation could only be written
by one versed in botany and possessed of the gift of writing, as is
the good fortune of Mr. Peattie.

In chapter four, “Glimpses of Wild Life,” Aubrey Drury gives
brief but interesting accounts of the distribution and habits of
bears, various deer, cats, coyotes and other mammals, birds, and
fish of the stream and sea.

Judy Van der Veer in chapter five presents in an entertaining
manner some aspects of the human, animal, and plant life of the
foothills of the Coast Ranges of California as observed in summer,
fall, winter, and spring.

Further insight is given into some of the variety of wild life
and people living in the Coast Ranges of California in the next
chapter, “Farm, Rock, and Vine Folk,’ by Idwal Jones.

John Caughey in chapter seven, “Headlands in California
Writing,’ writes of the earliest manifestations of literary talent
in California, namely that of the primitive Indians. This is fol-
lowed by a brief account of Spanish annals, writings of early
explorers and travellers, writings stimulated by the gold rush
period, writers of the second generation, and the moderns.
Caughey’s excellent knowledge of California history adds greatly
to this narrative on California writing.

One of the lesser explored areas of the Pacific Coast Ranges,
the Olympics, furnishes the subject material of chapter eight.
This narrative, entitled ““‘The Wilderness Mountains,” was written
by Lois Crisler, former instructor of English at the University of
Washington and now an inhabitant of the Olympic area. Mrs.

32 MADRONO [Vol. 9

Crisler gives the reader many illuminating glimpses of the pio-
neers, the history, the mountains, and the life of the “dark realm
of the Olympics.”

Since lumber is one of the main products of the forested
regions of the Pacific Coast Ranges, particularly in the North-
west, the chapter on timber by Thomas Emerson Ripley comes as
no surprise. The surprise comes, however, in the amount of inter-
esting and instructive material—from the explorations of David
Douglas to modern logging and conservation—that is included in
the twenty-three pages of narrative comprising chapter nine.

Chapters ten and eleven, “People of the Oregon Coast Range”’
and ‘“‘People of the Washington Coast Range’”’ respectively, were
written by Archie Binns, the well-known writer of novels having
their background in the great Northwest. These two chapters
contain many human interest stories associated with the early
frontier, the hills, the coast line, lonely valleys, mountains, rivers,
and inlets. These stories are so well integrated with the history
of the region that one feels that he has been reading the history
of the Oregon and Washington Coast Ranges written by one not
only gifted as a writer but as an historian.

In most strictly scientific treatments of geographical areas the
geology and climatic conditions are usually given in the introduc-
tion or in the eary sections. The editor of Pacific Coast Ranges
has placed these subjects in the last two chapters. This seems to
the reviewer to be good practice, because, as the editor states,
“But until the lay reader has learned some place geography, some
variety of landscape, he or she is not ready to understand geology
(and climatology). “The Geological Story,” the title of chapter
twelve, was written by Professor Daniel E. Willard in a popular
yet a scientific vein, a style of writing of which so few scientists
are capable.

The last chapter, “Climatic Transitions and Contrasts,’ by
Professor Richard Joel Russell, contains the statistical data and
other more interesting facts about the climate and its effects upon
the Pacific Coast Ranges from Puget Sound to the Gulf of Mexico.

After reading the thirteen chapters, full of information, about
the first inhabitants of the Coast Ranges, the California missions,
the wild flowers, shrubs, trees, animals, the foothills, the wilder-
ness areas, the people of the Oregon and Washington Coast
Ranges, the geological story, and the climatic transitions and con-
trasts, one is inclined to agree with the editor when he says that
he does not have the answer to the question, ““Why do people live
in the East ?.”

The twenty-nine excellent photographs of varied and well-
selected subjects and the four maps add greatly to the enjoyment
and value of the book.—H. E. McMinn, Mills College, California.

MADRONO

A West American Journal of Botany

A quarterly journal devoted to important
and stimulating articles dealing with
plant morphology, physiology, taxonomy,
and botanical history. These volumes
should be a part of every botanist’s li-
brary and should be made accessible to
students of all universities and colleges.

Volume I, 1916-1929. . . $5.00
Volume II, 1930-1934 .. 5.00
Volume III, 1935-1936 . 5.00
Volume IV, 1937-1938 . 5.00
Volume V, 1939-1940 . . 5.00
Volume VI, 1941-1942 . 5.00
Volume VII, 1943-1944 . 5.00
Volume VII, 1945-1946 . 5.00
Single mumbers,) 6.7.4.2). 0.75

The subscription price of MADRONO
is $2.50 per year. We solicit your pat-
ronage,

Address all orders to:
Reed C. Rollins, Bus. Mer.

Natural History Museum,
Stanford University, California

VOLUME Ix NUMBER 2

MADRONO

A WEST AMERICAN JOURNAL OF
BOTANY

Contents
VEGETATION AND CLIMATE oF CoanutLA, Mexico, Cornelius H. Muller ...... 33
Viapimir L. Komarov, 1869-1946, NV. T. Mirov ............ 00.00.00 eee 57

Tue Present Status OF THE GENUS POLEMONIELLA HELLER, John F. Davidson 58
Wiis Linn Jepson, Herbert L. Mason ........... 00. cece ees 61

Notes anp News: Old World Plants Apparently Recently Introduced into
PN Ge Gps OE IAL C0) A AS EREMIDY A CMTE NERD OANID nU o L-R RP P a ean 64

Published at North Queen Street and McGovern Avenue,
Lancaster, Pennsylvania

April, 1947

MADRONO

A WEST AMERICAN JOURNAL OF BOTANY

Board of Editors

Herrert L. Mason, University of California, Berkeley, Chairman.

LeRoy Asrams, Stanford University, California.

Epcar ANvbeRson, Missouri Botanical Garden, St. Louis.

Lyman Benson, Pomona College, Claremont, California.

Herpert F. Copetanp, Sacramento Junior College, Sacramento, California.
Ivan M. Jounston, Arnold Arboretum, Jamaica Plain, Massachusetts.
Muipprep EK. Maruias, 2851 North Lake Avenue, Altadena, California.
Basserr Macuire, New York Botanical Garden, N. Y. C.

Maxzion Ownsey, State College of Washington, Pullman.

Secretary, Editorial Board—ANNetra Carter
Department of Botany, University of California, Berkeley

Business Manager—Reep C. Ro.iins
North Queen Street and McGovern Avenue, Lancaster, Pennsylvania
or
Natural History Museum,
Stanford University, California

Entered as second-class matter October 1, 1935, at the post office at
lancaster, Pa., under the act of March 3, 1879.

Kstablished 1916. Published quarterly. Subscription Price $2.50 per year.
Completed volumes I to VII inclusive, $35.00; each volume $5.00; single numbers
$0.75.

Papers up to 15 or 20 pages are acceptable. Longer contributions may
be accepted if the excess costs of printing and illustration are borne by the
contributor. Range extensions and similar notes will be published in con-
densed forma with a suitable title under the general heading “Notes and News.”
Articles may be submitted to any member of the editorial board. Manuscripts
may be included in the forthcoming issue provided that the contributor pay the
cost of the pages added to the issue to accomnmodate his article. Reprints of
any article are furnished at a cost of 4 pages, 50 copies $4.10; 100 copies $4.50;
additional 100’s $0.85; 8 pages, 50 copies $5.95; 100 copies $6.60; additional 100’s
$1.30; 16 pages, 50 copies $8.35; 100 copies $9.35; additional 100’s $2.00. Covers,
50 for $2.75; additional covers at $1.65 per hundred. Reprints should be
ordered when proofs are returned.

Published at North Queen Street and McGovern Avenue, Lancaster,
Pennsylvania, for the

CALIFORNIA BOTANICAL SOCIETY, INC.

President: Howard E. McMinn, Mills College, California. First Vice-
President: Milo S. Baker, Santa Rosa Junior College, California. Second
Vice-President: David D. Keck, Carnegie Institution of Washington, Stanford
University, California. Secretary, John Whitehead, University of California
Botanical Garden, Berkeley. ‘Treasurer, Reed C. Rollins, Natural History
Museum, Stanford University, California.

Annual membership dues of the California Botanical Society are $2.50,
which includes a year’s subscription to Madrono. For two members of the same
family the dues are $3.50, which includes one copy of Madrono and all other
privileges for both. Dues should be remitted to the Treasurer. General
correspondence and applications for membership should be addressed to the
Secretary.

1947] MULLER: VEGETATION OF COAHUILA, MEXICO 33

VEGETATION AND CLIMATE OF COAHUILA, MEXICO

Cornetius H. MuLuer

INTRODUCTION

Like much of Mexico, Coahuila has never received sufficient
attention to make known the diversity of its physiography, cli-
mate, and vegetation. The extreme physiographic diversities of
the state occupy a relatively small proportion of its total area.
Much of Coahuila’s surface is characterized by the Basin and
Range type of topography; that is, by relatively isolated mountain
ranges scattered on a nearly continuous series of undrained bol-
sones or desert plains. It is therefore possible to travel great
distances in Coahuila without ever climbing above the level of the
desert. Many biologists and climatologists passing within sight
of the mountains seem to have been not at all impressed by their
climatic and vegetational features. The principal travel routes
in Mexico are largely confined to the level plains, and this is in
large part responsible for the erroneous notion of vegetational
and climatic homogeneity in Northern Mexico. It is the purpose
of this paper to record observations within the mountains of
Coahuila as well as on the plains and to present evidences of the
existence of climatic diversity.

This study is based upon the division of the area into “natural
regions,’ using as criteria topography, elevation, substratum,
vegetation, and history. The important characters of a natural
region vary less than those between any two natural regions.
Essentially the same method was employed in a discussion of the
vegetation and climate of Nuevo Leon (Muller, 1939).

Homogeneity of climate is an important character of a natural
region, but the present state of our knowledge of the meteorology
of Coahuila is very poor. <A few scattered observatories, all of
them located on the desert plains, give no clues to the climatic
diversity of the mountainous areas. The close dependence of
vegetation upon climate, other factors being accounted for, per-
mits one to approximate an expression of the climatic types of
areas in which the vegetation is known. In Coahuila, for instance,
several meteorological observatories with relatively long records
give us a fairly accurate picture of the climates of the several
plains regions of the state. The vegetation of these plains is con-
sistently one phase or another of Chihuahuan Desert Shrub in the
Central Plateau Region or of Tamaulipan Thorn Shrub in the
Coastal Plain Region. Departures from these vegetation types
are confined to the high elevations and abrupt topographic fea-
tures that favor increased rainfall and reduced temperature. A
pine forest, for example, may be regarded as a definite indication
of a climate of greater precipitation and lower temperature than
that of the desert shrub.

Maprono, Vol. 9, No. 1, pp. 1-32. January 27, 1947.

mio 2.

34 MADRONO [Vol. 9

TREATMENTS OF VEGETATION AND CLIMATE OF COAHUILA

Until recently only incidental and fragmentary reference to
the vegetation of Coahuila could be found in the literature, al-
though several historic botanical collections were made within its
limits. Gregg and Wislizenus were the earliest, but neither left
any published record of the vegetation beyond casual mention of
the widespread desert shrub. Pringle (1888) described the desert
shrub in a few of its phases on the Central Plateau and the pine
forests and chaparral of the mountains in the same region. He
made no effort to indicate the distributions of these vegetation
types. Merriam (1898) in mapping the distributions of his life
zones credited northern and southeastern Coahuila with rather
extensive areas of the Transition Zone, but he did not indicate the
source of his information. The forests he intended to include in
the Transition Zone are actually not nearly so extensive as his
map indicates. Contreras (1942b) mapped the vegetational for-
mations of Mexico in a general way, recognizing steppe, desert
shrub, succulent desert shrub, and thorn shrub in Coahuila but not
distinguishing them from similar but quite unrelated vegetation
in other regions.

Treatments of the vegetation and climate of adjoining states
give some clues to the nature of Coahuilan vegetation and climate.
Palmer (1928) and Muller (1937) each gave brief descriptions
of the vegetation of the Chisos Mountains in western Texas, which
constitute one of the links between the Sierra Madre of Nuevo
Leon and the Rocky Mountains of New Mexico. The Chisos area
has proved to parallel the Sierra del Carmen of adjacent Coahuila
in many respects. The eastern border of Coahuila is character-
ized in the treatment of Nuevo Leén by Muller (1989), and the
western border is described in a paper on the vegetation of Chi-
huahua by Shreve (1939) and in one by LeSueur (1945). More
recently Shreve (1942a,b) has described the desert and grassland
vegetation of Coahuila in studies of more extensive areas.

The climates of Coahuila have been treated only incidentally
to more extensive classifications. Sanchez (1929) and Thorn-
thwaite (1931) both allowed only two climatic types within the
state. Sanchez named these “Clima de Estepas” and “Clima
Desertico” (steppe and desert climates). Thornthwaite’s divi-
sions are similar; he called them “semiarid, mesothermal, and
[with] deficiency of rainfall at all seasons” (DB’d)* and “arid,
mesothermal, and [with] deficiency of rainfall at all seasons”’
(EB’d). Page (1980) mapped the rainfall of Coahuila and indi-
cated an area of low precipitation conforming roughly to Sanchez’
and Thornthwaite’s desert climates. He also credited the extreme
southeastern and northeastern portions with a greater rainfall, a

1These formulae are those used by Thornthwaite and derived from his
system.

1947 | MULLER: VEGETATION OF COAHUILA, MEXICO 35

feature clearly indicated by the vegetation of those areas. Ward
and Brooks (1936) followed Page and Sanchez rather closely in
their work on the climates of northern Mexico. An anonymous
Mexican publication (1936) follows Thornthwaite closely. Con-
treras (1942a,b) recognized a temperate climate throughout the
state, but did not indicate a humidity greater than “semiarid.”
Shreve (1944), utilizing meteorological records and physiographic
diversity, mapped the rainfall of the northern half of the republic.
His results conform much better to the observable landscape than
do earlier attempts.

PHYSIOGRAPHY AND SoILs OF COAHUILA

As has been noted, the physiography of Coahuila is character-
ized by Basin and Range topography. This form occupies all the
area lying west of the Rocky Mountain—Sierra Madre axis that,
in Coahuila, is represented by the Sierra del Carmen. This axis
runs from the Sierra del Carmen in north-central Coahuila to the
Sierra Madre in southeastern Coahuila. Thayer (1916) regarded
these elevations as the escarpment of the Central Plateau, but
they are really a range of mountains, some of whose peaks reach
3,000 meters and rise considerably above the plateau, which itself
ranges from about 1,000 to 1,500 meters. There are two major
breaks in this mountain range, one east of Saltillo where the
plateau has the eroded remnant of a true escarpment and one east
of Monclova where the Rio Salado has eroded a wide pass through
the mountains and projected the head of its valley beyond Mon-
clova, forming an area of low elevation incising the plateau.

The mountains west of the Carmen axis are mostly isolated
and scarcely continuous, but some of them are oriented in definite
directional trends and are so nearly contiguous as to exercise a
positive influence on the climate. Just west of the Carmen axis
the Sierra de la Madera at Cuatro Cienegas, the Sierra del Fuste,
and the Sierra del Pino project northward in a nearly unbroken
range. Along the western border of Coahuila the Sierra Mojada,
Sierra de Almagre, and Sierra de Hechiceros run north almost
to the valley of the Rio Grande. These north-south trending
ranges together with the northern half of the Sierra del Carmen
axis effectively divert winds coming from the east coast of the
continent and are responsible for much of the rain that falls in
the mountains of northern Coahuila. Furthermore, the wide
basins separating these mountains are open to the north and allow
the free entrance of extra-tropical cyclonic storms during the
winter season. Across extreme southern Coahuila the Diamante-
Carneros and Parras ranges form an axis that links that region
with the highlands of both the Sierra Madre Oriental of Nuevo
Leon and the Sierra Madre Occidental of Durango. In addition
to these major ranges of the plateau, there are many scattered
minor ones which are much less effective as wind barriers or in
inducing precipitation and lower temperatures.

36 MADRONO } [Vol. 9

In summary, the northern three-fourths of the Central Plateau
portion of Coahuila is characterized by several relatively contigu-
ous north-south trending ranges as well as by numerous scattered
minor ranges. The southern portion, on the other hand, lacks
major north-south trending ranges, has only scattered minor
ranges, and is bounded to the south by discontinuous east-west
trending ranges. |

The northeastern quadrant of the state lies on the Atlantic
slope of the continent. It consists of a plain ranging from about
200 to 500 meters in elevation, lying east of the Sierra del Carmen.
It includes an extensive mountain mass, the Serrano del Burro,
whose highest peaks reach about 2,300 meters altitude. In the
north this eastern plain is essentially identical with the desert
plains of the Central Plateau and the upper Rio Grande Plain,
but toward the south it merges with the Eastern Coastal Plain and
the Piedmont areas of Nuevo Leon.

Soils in Coahuila fall roughly into four classes. These are
stony immature soils, dark loam soils, light desert soils, and alka-
line desert soils. Dune sands are of only local occurrence.
These groups and their effects on vegetation are easily recognized
and evaluated. The characteristics of the soil are a possible
source of error in evaluating the relationship of climate and vege-
tation. It is necessary to determine that soil is not responsible
for differences in vegetation ascribed to climate.

The stony soils sometimes are confusing in their effect upon
the vegetation. It is not difficult to recognize the effect of stony
outcrops in an area of soil derived from the same substratum.
However, where limestone and igneous stony soils alternate in a
semiarid region, the igneous areas take on a much more mesic
appearance than the limestone areas. It is often a temptation to
regard the well-developed grasslands of granitic areas as indica-
tive of a less arid climate than the adjacent desert shrub of lime-
stone areas. Only by careful consideration of alternating lime-
stone and igneous outcrops can this source of error be avoided.

The influences of chemical differences in soil, such as salinity
and presence of gypsum, are readily recognized by the presence of
indicator species, as described by Johnston (1941) in the case of
gypsum. Where such factors are operative over wide areas, they
may themselves serve in some measure as indicators of climate.

It is well known that heavily saline soils develop and persist
only under extremely arid conditions. In the mature soils, disre-
garding entirely the stony immature types, arid regions are charac-
terized by light gray soils and red soils lacking in humus, while
mesic regions develop dark loams with relatively high percentages
of humus. Some of the more densely vegetated plains east of the
mountains have dark loam soils, but on the plateau west of the
mountains this group is confined to the higher elevations where
greater precipitation permits the development of more mesic
vegetation.

1947] MULLER: VEGETATION OF COAHUILA, MEXICO 37

CiuimaTic INFLUENCES

Before attempting to classify the vegetation types of Coahuila
and to map their distributions, it is necessary to consider some of
the major physical influences that control climate and therefore
the distribution of vegetation. First is the position of our area
on the continent. Coahuila lies in the broader end of Mexico and
just east of the center of the continental mass. Its western border
(the Coahuila-Chihuahua line) marks roughly the continental
center at this latitude. The presence on the north of the desert
and semi-desert regions of Trans-Pecos Texas and southern New
Mexico further accentuates the continental influence upon the
climate of Coahuila. The continental influence is largely limited
to the Central Plateau. It is made possible by the presence of
the Sierra del Carmen axis which effectively excludes winds from
the Atlantic Ocean and Gulf of Mexico. On the west the Sierra
Madre Occidental similarly reduces the effectiveness of any winds
~ from the Pacific Ocean and the Gulf of California. This rain
shadow phenomenon was encountered in the Sierra Madre Ori-
ental of Nuevo Leén where it is strikingly clear. West of the
Sierra del Carmen in Coahuila and east of the Sierra Madre Occi-
dental in Chihuahua secondary and lesser ranges add to the effect
of the major ranges so that as one approaches the centrally located
Coahuila-Chihuahua line, one finds the climate increasingly arid.

West of Saltillo an east-west directed lowland crosses the
state. It is bounded on the north by the discontinuous mountains
of central Coahuila and on the south by the more nearly continu-
ous uplands of the Saltillo-Parras axis. The Eastern half drains
onto the Coastal Plain, while the western half is an undrained
basin. This basin, open more or less to the easterly winds, is not
less arid than the great protected basins of northwestern Coahuila.
In fact, rainfall records show that the southwestern basin is more
arid than the northerly ones. This is in part caused by a partial
blocking of the easterly winds by the Saltillo highlands and in
part by the low altitude of the basin. The effect is similar to that
of the low altitude of the valley of the Rio Grande in northern
Coahuila.

The trend toward an arid climate from east to west is not con-
fined only to the general climate (of the plains). The local
climates of higher elevations, such as the increased precipitation
and lower temperature induced by a mountain range, become less
and less extreme toward the west as general aridity increases.
The most extreme influence is seen in the Sierra del Carmen, the
first mountains encountered by the easterly winds. West of this
the succeeding ranges are not quite so high, but there are several
of comparable height, notably Sierra de la Madera and Sierra
Mojada. The former range lies in central Coahuila and bears a
well developed though depauperate pine forest. This forest,

38 MADRONO [Vol.9

though not quite so mesic, is essentially comparable to those of
the Sierra Madre Oriental in Nuevo Leén and the Sierra del
Carmen of northern Coahuila. The Sierra Mojada, on the other
hand, lies in western Coahuila and bears no pine forest. It is said
to have been forested at one time, but it has been denuded, and the
xeric climate does not permit reseeding of forest species as is so
evidently occurring in the Sierra de la Madera. The present vege-
tation of the Sierra Mojada is largely chaparral and clearly re-
flects the drastically reduced precipitation in the paucity of its
herbs, which do not readily survive the summer drought.

The latitudinal position of Coahuila places it partially within
the subtropical high-pressure belt or the belt of subtropical calms,
a fact to which it owes much of its aridity. The Coastal Plain and
Rio Grande Plain portions of the state lie in the path of the north-
ernmost trade winds, but these are much less effective here than
in Nuevo Leon. The principal evidence of their influence is seen
in the much more humid nature of the east slopes of the eastern
mountains than is noted on their western slopes.

Convectional storms which center about the scattered moun-
tain masses account for the greater part of the summer rainfall in
the interior portions of Coahuila.

VEGETATION TYPES

In the following discussion a general description of the major
vegetation types is presented. In those cases where intensive
local analytic studies have been made, the details are omitted in
order to avoid distortion of the picture of the whole area. The
vegetation types are mapped in text figure 1.

1. CurHuaHuAN Desert Surus. The plains and basins of the
southern, western, and northern three-fourths of the state of Coa-
huila are characterized by many variants of a general vegetation
formation which is strictly desert in all its attributes (pl. 3, fig. 1).
This was previously described in Nuevo Leon as Central Plateau
Desert Scrub, but Chihuahuan Desert Shrub, a variant of Shreve’s
term (1942b), seems preferable. Low, sparse perennials and
ephemeral annuals make up this polymorphic vegetation. The
most characteristic species is Larrea tridentata (DC.) Cov., and
the structure and composition of the variant types may best be
considered from the standpoint of the species that are associated
with Larrea or that occasionally replace it. Not all of the area
occupied by Larrea and its associates is true desert, for the various
species range more or less widely into adjacent areas of a more
mesic nature. In such cases, however, the function of the desert
species is not the same as on the true desert; they not only do not
dominate the site, but their physiognomy and spacing are much
changed.

Most widely associated with Larrea are Flourensia cernua DC.,
Acacia vernicosa Standl., Fouquieria splendens Engelm., and Prosopis

1947] MULLER: VEGETATION OF COAHUILA, MEXICO 39

velutina Wooton. In addition to these, the following species,
listed approximately in order of decreasing importance, are com-
mon associates:

Condalia lycioides (A. Gray) Yucca australis (Engelm. )

Weberb. Trel.
Koeberlinia spinosa Zucc. Yucea Torreyi Shafer
Coldenia Greggii ( Torr.) Acacia constricta Benth.
me Gray Rhus microphylla Engelm.
Parthenium incanum H.B.K. Citharexylum brachyanthum
Lycium Berlandieri Dunal A. Gray
Celtis pallida Torr. Microrhamnus ericoides
Condalia spathulata A. Gray A. Gray
Opuntia imbricata (Haw.) Sericodes Greggii A. Gray
DC. Hilaria mutica (Buckl.)
Opuntia leptocaulis DC. Benth.

Opuntia spp. (Platyopuntia)

The habitat type on which the typical phase and minor vari-
ants of this association develop is characteristically a plain,
bajada, or outwash plain of more or less gentle declination. The
soil is usually rather shallow and stony, but it might be quite deep
and covered by a “desert pavement” of small stones. Edaphic
variations are responsible for major changes in the vegetation
irrespective of homogeneous climatic conditions.

A very deep soil in well-drained situations is usually charac-
terized by an abundance of Flourensia which may even exclude
Larrea. Prosopis, Koeberlinia, Condalia spathulata, and Parthenium
are often more abundant in such situations. Small catchments of
deep soil are frequently considerably more moist than the sur-
rounding plain, and these usually develop a very dense stand of
the common shrubs in which Acacia vernicosa, Rhus microphylla, and
Yucca australis are apt to be quite important. Parts of such areas
are sometimes covered by a sod of Scleropogon brevifolius Phil. or
Hilaria mutica or both. Such a thicket, with or without the grass
sod, is locally termed a mogote (pronounced mo-gé-tay), a very
useful word meaning “island” that should be accorded recognition
in ecological literature along with chaparral, encinal, and other
words of Spanish and Mexican origin that are more expressive
than any corresponding English words.

The larger basins may have a more or less well-developed
laguna or lake bed surrounded by a nearly flat playa and the
gently sloping bajada. If accumulation of water is frequent in
the laguna, its center may be quite bare and may have a ring of
Sporobolus Wrightiti Munro encircling the bare center where occa-
sional flooding prevents the growth of other species. If the bolson
receives little runoff from the surrounding mountains and hills,
its playa may be very limited in extent. Such a playa is usually

40 MADRONO [Vol. 9

fore

103°

Fic. 1. Distribution of vegetation types in Coahuila. A, Chihuahuan
Desert Shrub; B, Tamaulipan Thorn Shrub; C, Piedmont Shrub; D, Grassland
(and Grassland Transition); E, Montane Low Forest; F, Montane Chaparral;
G, Montane Mesic Forest.

covered by a nearly pure stand of Hilaria mutica. These tobosa
flats, however, are not invariably of small extent.

More saline lagunas or salt lakes are surrounded by Atriplex
spp., Allenrolfea sp., Suaeda sp., Prosopis velutina, and Sporobolus
(pl. 3, figs. 2, 3). The larger playas may develop rather broad
zones of somewhat saline soil bearing Prosopis and Atriplex canes-
cens (Pursh) Nutt. These are especially prominent in south-
western Coahuila.

1947 | MULLER: VEGETATION OF COAHUILA, MEXICO Al

The undrained basins are confined to the southern and western
half of the state, the Basin and Range country west of the Sierra
del Carmen and Sierra Madre.

The low elevations of the bolson grassland and the fact that
it is always overlooked by the surrounding desert shrub was re-
garded by Shreve (1942a) as evidence of the desert nature of the
tobosa Nanos. He points out that this vegetation is a response
to soil conditions on the desert and is not to be regarded as part
of the climatic grassland formation.

A few of the basins have local beds of sandstone from which
small areas of dune sand have developed. Where these have be-
come more or less stabilized, they bear a sparse cover of Prosopis,
Yucca elata Engelm., Ephedra Torreyana S. Wats., Atriplex canes-
cens, Gutierrezia sp., and Heliotropium Greggu Torr.

Another and very important variation of the desert shrub
vegetation is the development in two poorly delimited areas of
succulent desert types. The extensive low limestone hills in north-
ern Coahuila, especially in the low altitude zone of the Rio Grande
Valley bordering northern Coahuila, exhibit a particularly rich
flora of succulent species, including Euphorbia antisyphilitica Zucc.,
Jatropha dioica Sesse, Agave lecheguilla Torr., Hechtia sp., Opuntia
spp., and various species of Echinocactus, Echinocereus, Mamillaria,
ete. In the central-southwestern quarter of the state, rocky hills
above the desert plain develop a similar variant of the desert in
which Grusonia Bradtiana (Coult.) Britt. & Rose is abundant.
Such succulent desert types are confined to rocky slopes of shal-
low soil (pl. 3, fig. 4).

2. Tamautipan THorn SuHrvus. East of the Sierra del Carmen
and Sierra Madre Oriental the foothills of these ranges gradually
give way to the Gulf Coastal Plain which, in Coahuila, is of rela-
tively high elevation, averaging perhaps 400 meters. This merges
in the north with the upper Rio Grande Plain, which is predomi-
nately true desert. East-central Coahuila, however, has little in
common with the desert of central and western Coahuila; it is
rather much more closely related to southern Texas, northern
Nuevo Leon, and northern Tamaulipas. The vegetation of this
area, including the well-known “brush country” of southern Texas,
is a formation equal in rank to the desert shrub vegetation of the
Chihuahuan Desert in western Chihuahua and eastern Coahuila.
Shreve (1917) recognized this vegetation as distinct from all other
types in the United States and separated it under the name “Texas
Semi-Desert.” The term chaparral is locally applied to this vege-
tation by the natives, and the term was adopted by Clover (19387)
in southern Texas. This use of the term seems objectionable from
both an etymological and practical standpoint. The original
Spanish meaning of chaparral is a growth of low, evergreen oaks.
The word was thus correctly applied to the oak-dominated scrubby

42 MADRONO [Vol. 9

broad-sclerophyll vegetation of California, and it is applied for
the same reason to the closely related broad-sclerophyll vegeta-
tion of Nuevo Leon and Coahuila. Its application, then, to a semi-
desert thorn shrub vegetation is unfortunate and misleading. This
thorn shrub was described in Nuevo Leon under the name Eastern
Coastal Plain Scrub. In order to describe and locate the center
of its development, it is here proposed to apply to the Eastern
Coastal Plain Scrub the term Tamaulipan Thorn Shrub. The
more luxuriant and arborescent developments of this general type
in extreme southern Texas and especially in Tamaulipas may be
termed Tamaulipan Thorn Forest.

The Tamaulipan Thorn Shrub occupies an area of plains and
low hills similar in many respects to the topography of the Chi-
huahuan Desert Shrub west of the mountains but with the three
important differences of lower elevation, greater rainfall, and
exposure to winds from the Gulf of Mexico. These habitat dif-
ferences are correlated with the development of a vegetation
characterized by a higher preponderance of thorny species, a
greater abundance of grasses, and a more luxuriant and denser
growth of shrubs. Characteristic species are much more numer-
ous than in the desert, the flora is much richer in total species, and
the number of variants or phases of the vegetational formation
are correspondingly higher. No attempt will be made to classify
these variants except in so far as this is required by the descrip-
tion of the most prominent Coahuilan phases. This should not be
regarded as a characterization of the entire Tamaulipan Thorn
Shrub formation, a task that will require a separate treatment.

Although many species characteristic of the Tamaulipan Thorn
Shrub and even locally dominant in that formation are also com-
mon in the desert shrub, the thorn shrub is distinguished clearly
by several very distinctive constituents. Among these are the
following:

Acacia amentacea DC. Cercidium floridum Benth.
A. Berlandieri Benth. Lippia ligustrina (Lag.)
Leucophyllum frutescens Britt.
(Berl.) Johnst. Parkinsonia aculeata L.
Porlieria angustifolia Acacia Farnesiana (L.)
(Engelm.) A. Gray Willd.
Karwynskia Humboldtiana Castela texana (Torr. &
(Roem. & Schult.) Zuce. Gray) Rose
Prosopis glandulosa Torr. Colubrina texensis (Torr. &
Cordia Boissieri DC. Gray) A. Gray
Schaefferia cuneifolia Lantana Camara L.
A. Gray

Outstanding among species important in both formations are
Celtis pallida, Opuntia leptocaulis, Condalia lycioides, Jatropha dioica,

1947 | MULLER: VEGETATION OF COAHUILA, MEXICO AS

Koeberlinia spinosa, Opuntia imbricata, Agave lecheguilla, and Micro-
rhamnus ericoides. Even Larrea and Flourensia filter onto the
Coastal Plain from the upper Rio Grande Plain and across the
mountains where these are broken by erosion systems. Also com-
monly associated with the Tamaulipan Thorn Shrub are the fol-
lowing species:

_Lycium Berlandieri Dunal Forestiera angustifolia Torr.
L. pallidum Miers Citharexylum Berlandieri
Eysenhardtia texana Scheele Robinson
Sophora secundiflora Salvia ballotaeflora Benth.

(Ortega) Lag. Leucophylum minus A. Gray
Bernardia myricifolia Viguiera stenoloba Blake

(Scheele) S. Wats. Yucca australis (Engelm.)
Condalia obovata Hook. rel.
Bumelia lanuginosa (Michx.) Y. rostrata Engelm.

Pers. Opuntia Lindheimeri Engelm.

Diospyros texana Scheele

Grasses are quite abundant, and often the vegetation assumes
characteristics of well-developed grassland, but this is not so
evident in Coahuila. Bouteloua trifida Thurb., Hilaria Belangeri
(Steud.) Nash, and Aristida purpurea Nutt. are most abundant.
Locally Andropogon scoparius Michx., A. saccharoides Swartz, Chloris
virgata Swartz, Buchlée dactyloides (Nutt.) Engelm., and others
are important. Perennial herbs are much more obvious here than
in the desert shrub, composites, Croton, and Verbena being particu-
larly abundant.

The phases of thorn shrub observed in eastern Coahuila include
the transition from desert shrub, in which both Larrea and Flou-
rensia are important on either rocky or deep-soiled areas, as well
as several phases of the true thorn shrub (pl. 4, figs. 1, 2, 3). On
shallow soils overlying limestone Acacia Berlandieri, A. amentacea,
Leucophyllum frutescens, Porlieria angustifolia, Opuntia spp., and
Jatropha dioica are amongst the outstandingly obvious species.
On deep clay soils a dispersed sod of Bouteloua trifida and Hilaria
Belangeri alternates with Prosopis, Opuntia, Castela, and some of
the species listed above. Deep alluvial soils, especially adjacent
to streams, may bear a richer grass flora forming a more dense
sod on which the shrub and small tree species appear as a savanna.
On very rocky sites the shrubs may be quite as stunted as those
of the desert, while at higher elevations the luxuriance of the
shrubs and the introduction of additional species tend to develop
the Piedmont Shrub described below.

3. PirpMont Surus. Along the eastern base of the mountains
that bound the western limit of the Tamaulipan Thorn Shrub is a
poorly defined zone of large shrubs and small trees that occupies
the irregular piedmont topography of this region as well as some

4A MADRONO [Vol.9

outlying small mountains on the Coastal Plain. Although this
type is best developed and most extensive in Nuevo Leén, where
is has been described as Piedmont Scrub, it is distinctly present in
southeastern Coahuila, and evidences of it may be seen north of
Muzquiz at the base of the Sierra del Carmen in north-central
Coahuila. Its northward extension covers too narrow a band to
be mapped on a small scale.

The first evidence of change from Tamaulipan Thorn Shrub
to Piedmont Shrub is a greater density and luxuriance of thorn
shrub species as they ascend to higher elevations. The eventual
introduction of non-thorny and more tree-like species accom-
plishes the transition. The principal characteristics of the Pied-
mont Shrub are the preponderance of this life form, the abun-
dance of low suffrutescent species, and the aestivation of her-
baceous species during the rigorous drought of midsummer. In
the following list the characteristic species are named approxi-
mately in the order of decreasing importance:

Quercus fusiformis Small Quercus sinuata var. brevi-

Diospyros texana Scheele loba (Torr.) C. H. Mull.

Bumelia lanuginosa (Michx.) Quercus Mohriana Buckl.
Pers. Rhus virens Lindh.

Sophora secundiflora Vauquelinia corymbosa
(Ortega) Lag. Correa

Bauhinia lunarioides A. Gray _Leucaena glauca (L.) Benth.
Quercus invaginata Trel. |

Common shrub associates are Colubrina macrocarpa (Cav.)
Don, C. Greggi S. Wats., Rhus trilobata Nutt., Eysenhardtia texana,
Pielea trifoliata L., and Amyris madrensis S. Wats. Along water-
ways in the hills Juglans rupestris Engelm., Celtis reticulata Torr.,
Acacia Farnesiana, and Ungnadia speciosa Endl. are characteristic,
while very mesic ravines often bear Quercus Muehlenbergu Engelm.
and Ulmus multinervosa C. H. Mull.

The evident part played by Quercus sinuata var. breviloba and
Q. Mohriana in the narrow strip of Piedmont Shrub in north-
central Coahuila is very suggestive of the Edwards Plateau of

EXPLANATION OF Figures, PLATE 3.

Piatre 3. VEGETATION oF CoanutLa.? Fic. 1. Chihuahuan Desert Shrub
with Larrea, Fouquieria, and Opuntia sp. Fic. 2. Laguna de Leche, an un-
drained desert lake bed intermittently inundated. Fic. 3. Saline phase of Chi-
huahuan Desert Shrub at the edge of an undrained basin, with Prosopis velutina,
Opuntia sp., and Atriplex spp. Fic. 4. Chihuahuan Desert Shrub, succulent
phase, with Agave lecheguilla, Hechtia sp., Grusonia Bradtiana, and Euphorbia
antisyphilitica associated with Larrea, Fouquieria, and Acacia vernicosa.

2 For the photographs in plates 3, 5, 6, and 7 the author is indebted to Mr.
B. Y. Morrison who permitted their use from the collection of the Division of
Plant Exploration and Introduction, Bureau of Plant Industry, Soils, and Agri-
cultural Engineering, U. S. Department of Agriculture, where they had been
filed by the author.

1947] MULLER: VEGETATION OF COAHUILA, MEXICO

PuatTe 3. VEGETATION OF COAHUILA.

4G MADRONO [Vol. 9

Texas as well as similar shrub zones beneath the lower or dry
limits of forest and woodland in western Texas.

In several broad canyons of the Piedmont on the east flank of
the Sierra del Carmen occur open forests of Brahea bella Bailey
associated with Piedmont Shrub. This arborescent palm is equally
at home on the flat, grassy valley floor or the steep cliffs of the
lower mountain slopes and canyon walls. It appears in Coahuila
principally about the headwaters of the Rio Sabinas, but at the
base of the Sierra Madre in Nuevo Leon it is more extensive.

The Piedmont Shrub is to some extent climatically coextensive
with grassland transition and replaces this transition over much
of the lower east slope of the mountains.

4. GrassLanp. True climatic grassland is not extensive in
Coahuila. The greater proportion of the grass-covered area never
develops much beyond a transition between desert shrub and
grassland, further development being terminated by failure of
elevation or by the development of chaparral (pl. 5, fig. 1). Since
grassland requires an area of relatively deep, flat soil and a con-
siderably more mesic climate than that of the desert plains, it is
to be sought only at higher elevations on gently sloping hills and
plateaus. Such situations in Coahuila are found only in the north-
western quarter of the state where the Sierra de las Cruces, Sierra
de Hechiceros, and (locally) the Sierra del Carmen present these
conditions. The latter locality is partially a granitic area on the
gentle east slope, and the development of a true grassland there
may be partly edaphic, but a neighboring grassy limestone slope
is indicative of climatic effect. Grassland Transition, by contrast,
is very widespread, occurring about the flanks of most of the
mountain ranges. A few of the broader canyons of the Sierra del
Carmen are covered by grassland, but these areas are too small
to be mapped on a small scale.

True Grassland is largely dominated by Bouteloua gracilis
(H.B.K.) Lag. with B. curtipendula (Michx.) Torr., Andropogon
saccharoides, Lycurus phleoides H.B.K., Stipa eminens Cav., Aristida
glauca (Nees) Walp., Buchlée dactyloides, and Muhlenbergia monti-
cola Buckl. locally abundant or even dominant. Associated with
the dominants are numerous species of perennial herbs, the Com-
positae, Asclepiadaceae, and Scrophulariaceae being especially
well represented. Rocky prominences in such an area are apt to

EXPLANATION OF Ficures, PLATE 4.

PLATE 4, VEGETATION oF CoaHutmnaA. Fic. 1. Tamaulipan Thorn Shrub
consisting of Condalia lycioides, Opuntia leptocaulis, O. Lindheimeri, Acacia
amentacea, and various grasses. Fic. 2. Tamaulipan Thorn Shrub with Kar-
winskya Humboldtiana, Leucophyllum frutescens, Acacia amentacea, Yucca
rostrata, Condalia spp., and various grasses. Fic. 3. Tamaulipan Thorn Shrub,
savanna phase, with Koeberlinia, Opuntia leptocaulis, Cercidium texanum, etc.
on a sod of Bouteloua trifida, Hilaria Belangeri, and Aristida purpurea.

1947] MULLER: VEGETATION OF COAHUILA, MEXICO AT

PLATE 4. VEGETATION OF COAHUILA.

48 MADRONO [Vol. 9

bear oaks, junipers, Dasylirion, Nolina, and Yucca, and these plants
also occur sparsely on the grassy slopes.

Grassland Transition subtends the upper limits of the desert
shrub as it ascends the foothills from the plains and bajadas. At
first the desert species become denser and larger, simultaneously
as the grassland species begin to appear. Larrea, Flourensia,
Fouquieria, Parthenium incanum, Viguiera stenoloba, Agave lecheguilla,
Yucca australis, Y. Torreyana, and Dasylirion sp. are amongst the
more common woody species of this transition zone. As the
grasses take over a greater percentage of the available space and
form a more continuous sod, the woody species become more
sparse but do not suffer diminution in size. _ Their eventual exclu-
sion from the Grassland, here as elsewhere, is purely the result of
competition under conditions more favorable to grasses and less
favorable to shrubs. Nolina spp. persist with Dasylirion and Yucca
after Larrea, Flourensia, and Fouquieria have been eliminated. At
this stage the vegetation is entering the phase of true Grassland.
Over much of the area of the Grassland Transition on limestone
hills there occur more or less dense stands of Parthenium argen-
tatum A. Gray (guayule), a species that competes poorly with both
the grasses and the desert shrubs. Of the several characteristic
transition species, it is the most obligate member of this vege-
tation type.

5. Montane Low Forest. Above the Piedmont Shrub on the
east slope of the Sierra del Carmen there occurs a vegetation type
essentially equal to that which has been described in Nuevo Leon
as Montane Low Forest. This vegetation type is the regional
equivalent of the Encinal described by Shreve (1939) in Chihua-
hua and the well-known oak-pinyon-juniper zone of the south-
western United States. It is not intended to suggest that these
widely separated vegetational zones are identical but only that
they are vicarious representatives of one another. The Montane
Low Forest of Nuevo Leon and southeastern Coahuila is the most
aberrant of the several types; that of northern Coahuila (which is
also identical with the equivalent zone in the Chisos and Davis
Mountains of Trans-Pecos Texas) is intermediate in character
between the Nuevo Leén and southeastern Coahuila type and the
oak-pinyon-juniper of farther north. The Coahuilan woodland is
therefore not homogeneous, some stands suggesting the Nuevo
Leén Montane Low Forest while others are very similar to the
oak-pinyon-juniper woodland. A third phase clearly indicates a
relationship with the low forest of the escarpment of the Edwards
Plateau in west-central Texas. In such transitions between diver-
gent vegetation types are clearly illustrated the dangers of naming
widespread plant associations and regarding them as identical
throughout. Perhaps even the claim of equivalence is too great
in such cases.

The principal characteristics of the Montane Low Forest in

1947 | MULLER: VEGETATION OF COAHUILA, MEXICO 49

Piate 5. VEGETATION oF CoanHuiLa. Fic. 1. Grassland Transition on the
eastern slope of the Sierra del Carmen with Andropogon scoparius, A. sac-
charoides, Bouteloua curtipendula, and B. gracilis. Dasylirion and Brahea are
prominent. Fic. 2. Montane Chaparral with Quercus intricata, Q. invaginata,
Dasylirion, Yucca carnerosana, Ceanothus Greggii, Cercocarpus mojadensis, and
various grasses.

50 MADRONO | [Vol.9

Nuevo Leon, Coahuila, and southern Trans-Pecos Texas are an
open stand of low trees with rounded crowns, usually gnarled
branches, and trunks less than one foot in diameter, the develop-
ment of a copious grass cover, and the presence of one or more
species of Agave of the Applanatae relationship. The tree layer
is usually dominated by oaks, while pinyon and junipers may be
prominent (although these are nearly absent from this zone in
Nuevo Leén). Several tree species characteristic of this zone are
also prominent in the next higher one, the Mesic Forest zone. In
the Montane Low Forest these adopt the characteristic low round
crowns and gnarled branches of that zone. The distinction be-
tween the two vegetation types is to some, extent based upon the
physiognomy of the vegetation and to some extent upon the flo-
ristic composition.

The principal tree species of the Coahuilan Montane Low
Forest are named approximately in the order of decreasing impor-
tance in the following list:

Quercus Gravesii Sudw. Quercus arizonica Sarg.

Pinus cembroides Zucc. Quercus sinuata var. brevi-

Juniperus pachyphloea Torr. loba (Torr.) C. H. Mull.

Quercus hypoleucoides Quercus Mohriana Buckl.
A. Camus

Quercus Laceyi Small Arbutus xalapensis H.B.K.

Juniperus flaccida Schlecht. Fraxinus cuspidata Torr.

Associated with these are many shrub species of which Salvia
regla Cav., Garrya ovata Benth., Rhus trilobata, Vitis arizonica
Engelm., Ptelea trifoliata, Bumelia lanuginosa, and Cercocarpus brevi-
florus A. Gray are the more prominent. Nowhere does one en-
counter all these species combined in a single forest.

The east slopes of the Sierra del Carmen, and to lesser extent
those of the Sierra Madre farther south, are dominated principally
by Quercus Laceyi, Q. Gravesii, Q. sinuata var. breviloba, and Q.
Mohriana with which Bumelia, Vitis, Ptelea, and Rhus are associ-
ated. The presence in arroyos in this zone of Cercis reniformis
Engelm., Smilax bona-nox L., Rhus virens, and Ungnadia speciosa,
in the light of the identity of the oak species, is particularly sug-
gestive of the woodland of the escarpment of the Edwards Plateau,
especially toward its southwestern quarter. Grasses in this phase
are rather sparse, partly perhaps because they are heavily grazed.

EXPLANATION OF Figures, PLATE 6.

PLATE 6. VEGETATION oF CoaHuILaA. Fic. 1. Montane Chaparral consist-
ing of Quercus intricata, Q. invaginata, Dasylirion, Yucca carnerosana, Ceano-
thus Greggti, Rhus virens, and Cercocarpus mojadensis. Fic. 2. Meadow of
Bouteloua hirsuta, B. gracilis, and Stipa eminens lying between pine forest and
the more elevated chaparral of the background.

1947 | MULLER: VEGETATION OF COAHUILA, MEXICO 51

PLATE 6. VEGETATION OF COAHUILA.

52 MADRONO [Vol.9

Lycurus phleoides, Stipa eminens, Muhlenbergia monticola, and Aris-
tida spp. are the principal species. Herbs, and particularly suf-
frutescent species, are strikingly abundant, legumes and com-
posites being the most evident.

At higher elevations in the Sierra del Carmen there occurs a
phase of the woodland much more suggestive of the oak-pinyon-
juniper association. The dominant tree species are Quercus hypo-
leucoides, Q. Laceyi, Q. arizonica, Pinus cembroides, Juniperus pachy-
phloea, and J. flaccida. Associated with these are Salvia regla,
Garrya ovata, Agave Havardiana Trel., and Arbutus xalapensis. The
open stand of trees permits a heavy cover of Muhlenbergia Emers-
leyi Vasey, a very course bunchgrass, with Lycurus phleoides, and
other lesser species. This vegetation is very similar to that of the
Chisos and Davis Mountains with the oustanding difference that
Quercus Emoryi Torr. and Q. grisea Liebm. are both absent from
the Coahuilan phase.

The arroyo forest at this elevation includes, in addition to the
above species, Pinus arizonica Engelm. and Acer brachypterum
Woot. & Standl. This, and the somewhat more closely spaced
trees with straighter and taller trunks, indicate a transition to the
Mesic Forest, the next higher zone.

6. Montane Cuaparrat. On the west slope of the Sierra del
Carmen and equivalent climatic zones of the more westerly moun-
tains, the Montane Low Forest is replaced by a shrubby vegeta-
tion type that has much in common with the broad-sclerophyll
vegetation of California. This has been described in Nuevo Leén
where it is well developed on the major west slopes of the Sierra
Madre. This zone may replace the grassland or it may lie between
a well-developed grassland and the pine forest of higher eleva-
tions or variously alternate with these. On deep-soiled sites grass-
land and pine forest meet and mingle, but shallow, stony soils at
the same general elevation are invariably covered by chaparral.
This dependence of chaparral upon shallow soils is closely paral-
leled in the Californian type where deep soils are grass covered
(ple Gh ties 2))e

The Coahuilan chaparral is very well developed on the upper
slopes of Sierra de la Madera and Sierra del Pino (pl. 5, fig. 2;
pl. 6, fig. 1). The type is dominated by several species of oak,
including Quercus intricata Trel., Q. invaginata Trel., Q. Pringlei
Seemen, Q. Laceyi, and Q. hypoxantha Trel. Of these only Q.
Pringlei and Q. Laceyi are deciduous. Associated with these are

EXPLANATION OF FicureEs, PLATE 7.

Puate 7. VEGETATION OF CoAHuILA. Fic. 1. Meadow of Andropogon
Gerardi, Monarda sp., and various Compositae with pine forest in the back-
ground. Fic. 2. Montane Mesic Forest of Pinus arizonica with Juniperus
pachyphloea and Quercus Gravesti as an understory and a sod of Andropogon
scoparius, A. saccharoides, and A. Gerardi.

Puate 7. VEGETATION OF COAHUILA.

5A MADRONO [Vol. 9

the following species listed approximately in the order of decreas-
ing importance:

Garrya ovata Benth. Arbutus xalapensis H.B.K.
Rhus virens Lindh. Fraxinus Greggii f. nummu-
Cercocarpus mojadensis laris (Jones) C. H. Mull.

C. Schneid. Ceanothus lanuginosus
Cercocarpus sp. (Jones) Rose
Microrhamnus ericoides Rhus trilobata Nutt.

A. Gray Rhus microphylla Engelm.
Berberis trifoliolata Moric. Nolina erumpens (Torr.)
Cowania plicata D. Don S. Wats.

Arctostaphylos pungens Dasylirion sp.

H.B.K. Yucca carnerosana (Trel.)

Ceanothus Greggii A. Gray McKelvey

Amelanchier denticulata

(H.B.K.) Koch

The very high preponderance of evergreen species and the
many genera common to both this chaparral and that of California
clearly indicate the close relationship of the two. Just as in Cali-
fornia and Arizona the more easterly stands of chaparral closer
to the desert influence show a paucity of true chaparral species
and an influx of desert species, so also the westerly mountains of
Coahuila, more subject to the continental influence, bear only a
depauperate chaparral and a high percentage of thorny shrubs.
Grasses are much reduced in this phase of the chaparral. Sierra
Mojada and Sierra Almagre, near the Coahuila-Chihuahua line,
both exhibit this depauperate association. Quercus intricata, Q.
Pringlei, Q. pungens Liebm., Lindleyella sp., Rhus virens, Cercocarpus
sp., Ceanothus Greggii, Arctostaphylos pungens, Acacia Greggii A.
Gray, and Mimosa biuncifera Benth. are the principal woody spe-
cies. Dasylirion sp. occurs but sparsely, and both Yucca and Nolina
are absent. Frazinus cuspidata, Rhus microphylla, Ptelea trifoliata,
Juniperus flaccida, and J. pachyphloea occur only along arrovyos.
This xeric chaparral is exceedingly dense. The only grasses of
any importance are Bouteloua gracilis, B. hirsuta Lag., and B. curti-
pendula, and these occur only in small clearings.

The upper slopes of Sierra Mojada are said once to have borne
timber, principally Juniperus pachyphloea and Pinus arizgonica,
which has since been cut clean for use in nearby mines. This
must undoubtedly have been a relic stand, for reseeding has not
progressed.

7. Montane Mesic Forest. Above the Montane Low Forest
and the Montane Chaparral in the higher and more massive moun-
tain ranges there occurs a forest type of definitely mesic nature.
The principal characteristics of this type are dominance by Pinus
arizonica and its associated species, close spacing, tall straight

1947] MULLER: VEGETATION OF COAHUILA, MEXICO 55

trunks, a closed canopy of crowns, and poor development of both
shrub and grass layers except in clearings (pl. 7, figs. 1 and 2).
Only strictly shade-tolerant species comprise the lower layers.
The surface soil of this zone, unlike that of the Montane Low
Forest, has well developed humus and duff layers, and its moisture
content remains high or moderate throughout the summer so that
aestivation does not occur.

In the better developed stands of Montane Mesic Forest, such
as the upper elevations of Sierra del Carmen, the following species
comprise the vegetation:

Pinus arizonica Engelm. Quercus Muehlenbergii
Pseudotsuga taxifolia (Lam.) Engelm.

Britt. Populus tremuloides Michx.
Cupressus arizonica Greene Acer brachypterum Woot. &
Quercus Gravesii Sudw. Standl.

Quercus hypoleucoides

A, Camus

The principal shrubs are Lonicera pilosa (H.B.K.) Willd. and
Ceanothus coeruleus Lag. Stipa tenuissima Trin., Piptochaetium
fimbriatum (H.B.K.) Hitche., and a few lesser species are the only
grasses. Dicot herbs are common but by no means so abundant
as in the lower zones.

In the Sierra Madre of southeastern Coahuila the same type
is dominated by Pinus arizonica, P. teocote Schlecht. & Cham., P.
montezumae Lamb., Quercus affinis Scheidw., Q. Endlichiana Trel.,
Cupressus arizonica, Pseudotsuga tazifolia, etc. Pinus ayacahuite
Ehrenb. occurs in the highest and most moist sites of both locali-
ties, and the Sierra de la Madera of Central Coahuila bears in
addition stands of Abies coahuilensis Johnst. <A similar forest
appears on the Sierra de Parras.

The lesser mountain ranges and those more subject to the
continental influence bear a much more depauperate pine forest.
In the Sierra del Pino, for instance, Quercus Gravesii, Juniperus
pachyphloea, Arbutus xalapensis, and Prunus capuli Cav. are the
principal tree species associated with dominant Pinus arizonica.
The ground cover consists principally of Andropogon scoparius
Michx., Bouteloua curtipendula, Elymus canadensis L., Muhlenbergia
Emersleyi, and several dicot herbs, of which Monarda sp., Erigeron
sp., Aster sp., and additional composites are the more prominent.
Although the entire flora is depauperate, the forest is not a relic
stand in these mountains, for regeneration is obviously taking
place. The climate here is not sufficiently favorable, however, for
the development of so rich a forest as that of the Sierra Madre of
Nuevo Leén, for instance.

Apparently no mountains in Coahuila very closely approach
the condition described in Nuevo Leén as subalpine. The rich

56 MADRONO [Vol.9

meadows and subalpine forests are only suggested by the occur-
rence of Pinus ayacahuite and Abies coahuilensis, and a timberline
is never reached.

CoNCLUSION

The obvious variations in climate throughout the mountainous
area of Coahuila are not reflected in the available data from the
few meteorological observatories on the lowlands and _ plains.
These data adequately represent the climates of their respective
vicinities and perhaps give an index to the climates of the plains
in general.

The area in Coahuila that clearly exhibits vegetational and
soils evidence of climates more mesic and cooler than those of the
lowlands and plains equals about ten percent of the total area of
the state. That this was not recognized in published maps of the
climatic types of Coahuila is illustrated in Table I. The climates
of Sanchez (1929) in the second column are essentially equal to
those of Thornthwaite (1931) and of Ward and Brooks (19386).
It will be noted that the five more mesic and temperate vegeta-
tion types have no counterparts in the existing classifications of
climates in Coahuila.

TABLE I. VEGETATION Types AND CiimatTic Types or CoAHUILA

Vegetation types Climatic types of Sanchez
Chihuahuan Desert Shrub Clima Desertico
Tamaulipan Thorn Shrub Clima de Estepas
Piedmont Shrub’. * +). & 4.) 0) wee ee
Grassland

Montane Low Forest
Montane Chaparral
Montane Mesic Forest

oe Ne

Biologists cannot rely upon existing climatic classifications for
significant geographic data. However, biologic data offer a valu-
able check upon the accuracy of climatic classifications, especially
where such gross areal discrepancies exist as here illustrated.
These discrepancies represent biologic, geographic, and economic
deviations of considerable importance.

University of California,
Santa Barbara.

LITERATURE CITED

Anonymous. 1936. Apéndice sobre un estudio preliminar de climas llevado a
cabo en el Servicio Meteorolégico, de acuerdo con la nueva classificacién
de C. Warren Thornthwaite, y publicado con la colaboracion del Insti-
tuto Panamericano de Geografia e Historia. Méx. Sec. Agr. y Fom.,
Dir. Geogr. Met. e Hidrol., Serv. Met. Publ. 37: 1-6. Illus.

Crover, E. U. 1937. Vegetational survey of the Lower Rio Grande Valley,
Texas. Madrono 4: 41-66, 77-100.

Contreras, Artas A. 1942a. Estudios climatolégicos, Areas geograficas de dis-
persion: Parthenium argentatum, Hevea brasiliensis, Castilloa elastica.
Méx. Sec. Agr. y Fom., Dir. Geogr. Met. e Hidrol. Pp. 1-112. Illus.

1942b. Mapa de las provincias climatolégicas de la Republica

1947] MIROV: KOMAROV 51

Mexicana. Méx. Sec. Agr. y Fom., Dir. Geogr. Met. e Hidrol., Inst.
Geogr. Pp. viit+ 54. Illus.
Jounston, I. M. 1941. Gypsophily among Mexican desert plants. Jour.
Arnold Arb. 22: 145-170.
LeSvueur, Harpe. 1945. The ecology of the vegetation of Chihuahua, Mexico,
north of parallel twenty-eight. Texas Univ. Publ. 4521. Pp. 1-92.
Merriam, C. H. 1898. Life zones and crop zones of the United States.
U.S.D.A., Div. Biol. Surv. Bull. 10: 1-79.
Mutter, C. H. 1937. Vegetation in Chisos Mountains, Texas. Texas Acad.
Sci. Trans. 20: 3-31.
1939. Relations of the vegetation and climatic types in Nuevo
Leon, Mexico. Am. Midland Nat. 21: 687-729.
Pace, J. L. 1930. Climate of Mexico. U.S. Monthly Weather Rev. Sup. 33:
1-30.
Patmer, E. J. 1928. A botanical trip through the Chisos Mountains of Texas.
Jour. Arnold Arb. 9: 153-173.
PrInGcLE, C. G. 1888. The forest vegetation of northern Mexico. I-X. Gard.
and For. 1: 70 et seq.
SancHez, P. C. 1929. Estudio de climatologia comparada con applicaciones
ala Republica Mexicana: classificacién provisional de sus climas. Méx.
Sec. Agr. y Fom., Dir. Estud. Geogr. y Climatol. Publ. 19: 1-13. Illus.
Sureve, Forrest. 1917. A map of the vegetation of the United States. Geogr.
Rev. 3: 119-125.
1939. Observations on the vegetation of Chihuahua. Madrono
5: 1-13.
. 1942a. Grassland and related vegetation in northern Mexico.
Madrono 6: 190-198.
1942b. The desert vegetation of North America. Bot. Rev. 8:
195-246.
| ———————. 1944. Rainfall of northern Mexico. Ecology 25: 105-111.
Tuayer, W.N. 1916. The physiography of Mexico. Jour. Geol. 24: 61-94.
THORNTHWAITE, C. W. 1931. The climates of North America according to a
new classification. Geogr. Rev. 21: 633-655.
Warp, R. pe C., and C. F. Brooxs. 1936. The climates of North America, in
W. Koéppen and R. Geiger, Handbuch der klimatologie 2 (pt. J):
J9-J79.

VLADIMIR L. KOMAROV, 1869-1946

Early in 1946, in Moscow, Russia, Dr. Vladimir L. Komarov,
President of the USSR Academy of Sciences and an eminent
botanist, died at the age of seventy-six. To American botanists
Komarov is known chiefly as the author of the “Flora of Man-
churia,’ published in three volumes in 1901-1907 by the St.
Petersburg Botanic Garden. Besides the “Flora of Manchuria,”
Komarov published many other botanical works. His “Flora of
Kamchatka,” which was the result of his travels in 1908-1909,
was almost ready for publication at the time of the Revolution of
1917, but most of the proof sheets were lost. The “Flora of Kam-
chatka” (918 pages) was finally published in 1927-1930.

In 1912 Komarov submitted to the Moscow State University
his doctor’s thesis, “An introduction to the flora of China and
- Mongolia.” He had in mind the gigantic task of publishing a
“Flora of China and Mongolia,’ and shortly after his doctoral

58 MADRONO [Vol.9

dissertation was written, he began to work on the voluminous
materials and collections accumulated by previous Russian ex-
plorers of Mongolia and China. How big the task was, might
be judged by the fact that Komarov’s monograph on Chinese
Caragana contains 22 new species. Although the “Flora of China
and Mongolia’? was never completed, Komarov did finish several
monographs intended for it. When in 1922 the Botanical Insti-
tute of the Academy undertook the publishing of the “Flora of
USSR,’ Komarov became its editor-in-chief. In addition to
occupying this supervisory position, he also contributed to the
“Flora” as a rank-and-file botanist. Komarov’s other botanical
contributions are too numerous to mention. More than one hun-
dred titles deal with the vegetation of the Russian Far East, and
many papers are devoted to the flora of central Asia and Siberia.

As a student in Komarov’s classes from 1914 through 1916
I remember him as being very quick in all that he did, and his
lectures were sometimes quite witty and sarcastic. Then and
throughout his life, his mind was occupied by the problem of what
constitutes a species. In the introduction to the “Flora of Man-
churia,” as well as in his later works, he devoted a great deal of
space to this problem.—N. T. Mirov, California Forest and Range
Experiment Station, Berkeley.

THE PRESENT STATUS OF THE GENUS
POLEMONIELLA HELLER

Joun F. Davipson

The status of the genus Polemoniella, in which has been in-
cluded P. micrantha (Benth.) Heller, P. antarcticum (Griseb.) Nel-
son & Macbride, and P. Gayanum (Wedd.) Nelson & Macbride,
has been doubtful ever since Heller (1904, p. 57) segregated it
from Polemonium. Brand (1907) reduced it to a synonym of
Polemonium, whereas Nelson and Macbride (1916, p. 35) and
Wherry (1941) accepted Heller’s genus.

Heller differentiated between Polemonium and Polemoniella as
follows:

CHARACTER Polemoniella Polemonium
Inflorescence scattered, solitary cymose panicles or
racemes |
Flower shape nearly rotate mostly funnelform
Filament base nearly naked dilated pilose
appendaged
Corolla length less than calyx several times calyx
Flower size small, inconspicuous showy

Habit annual perennial

1947] DAVIDSON: POLEMONIELLA 59

Close examination of the inflorescence of Polemoniella micrantha
(Benth.) Heller shows that the inflorescence is sympodial. In this
type of inflorescence, a flower is borne terminally, and from the
axil of the uppermost leaf a bud develops into an elongated branch
bearing a terminal flower and a subterminal leaf. From the axil
of the latter leaf another branch develops in a like manner, so that
under favorable conditions in this species five or more flowers may
be produced, apparently on a central axis, but with the leaf oppo-
site the pedicel, not as a subtending bract. Under unfavorable
conditions, of course, the plant may produce only a single flower,
as may occur in similar circumstances in many polemoniums.

In Polemonium pulcherrimum Hook. there are many individuals
in which the excessive elongation of the lateral branches of the
inflorescence suggests this sympodial condition. At the northern
limit of distribution some of the plants also exhibit greatly reduced
inflorescences, so that in Alaska and the Aleutian Islands it is not
uncommon to see only a single flower on the lateral branches. In
this condition, the inflorescence is identical with that of Polemoni-
ella micrantha, and is the only known occurrence of the sympodial
inflorescence in the perennial polemoniums. The unusual appear-
ance of these northern specimens at first suggested a specific or
subspecific separation from Polemonium pulcherrimum, but analysis
of the specimens throughout the range of that species shows the
tendency for the elongation of the lateral branches to be wide-
spread throughout the population, and the reduction of flowers is
gradual as the northern limit is approached. Thus the material
examined is regarded as at most an ecotype of P. pulcherrimum.

Regarding Heller’s conception of flower shape, it is difficult to
visualize a rotate or nearly rotate corolla included within a cam-
panulate calyx. The corolla of Polemoniella micrantha is more aptly
described as “‘short-campanulate.”’ While the corollas of such
polemoniums as Polemonium pauciflorum Wats. and P. confertum
A. Gray are certainly funnelform, those of P. occidentale Greene,
P. pulcherrimum Hook., and P. carneum A. Gray are at least as
nearly rotate as those of Polemoniella.

Many polemoniums have undilated filaments, similar to those
of Polemoniella micrantha, while other species of Polemonium have
dilated filaments like those of Polemoniella antarctica (Griseb.)
Nelson & Macbride. While the filaments of Polemoniella are
nearly naked at the bases, those of Polemonium Lemmoni Brand are
completely so. The filaments of the perennial species of Pole-
monium vary from glabrous to densely woolly.

In the Polemoniaceae both the annual and perennial habits are
common. In the primarily perennial genus Phloz, for example,
P. Drummondii Hook. is undoubtedly a Phlox, despite its annual
habit. Similarly in Collomia, the annuals, C. linearis Nutt. and
C. grandiflora Doug]. occur in a genus in which perennials are

60 MADRONO [Vol. 9

common. The writer has at present several species of Polemonium
growing in experimental plots at Berkeley, California, and Van-
couver, British Columbia. Of these plants, some perennial species
(P. caeruleum L., P. foliosissimum A. Gray, P. pauciflorum Wats., and
P. pulcherrimum Hook.) have produced viable seeds within six
months of germination. In Berkeley, some individuals of P. pauci-
florum and P. foliosissimum have died following seed production,
although they received the same care as the persistent individuals
of the species. While the latter were behaving as normal per-
ennial plants, the former, since the seeds produced were viable,
obviously were behaving as annuals. The possibility of such a
reaction occurring also in nature cannot be overlooked.

In other genera of the family the annual or perennial habit
apparently is not considered of generic significance, nor does the
writer so consider it here. The only remaining point separating
Polemoniella and Polemonium is the fact that the corolla is shorter
than the calyx, the flower being therefore inconspicuous. This
distinction appears to have been broken down, also. Ostenfeld
(1923), in discussing genetic experiments with Polemonium cae-
ruleum, mentions the occurrence of natural populations of this
species in which micropetalous flowers were common, including
some in which the corollas were shorter than the calyces.

In sharp contrast to the supposed differences between these
two groups discussed above are the following features which
Polemonium and Polemoniella have in common:

Leaves pinnately divided.

Leaflets predominantly ovate to elliptical.

Calyx herbaceous throughout, campanulate, accres-
cenit:

Corolla from rotate-campanulate to narrowly fun-
nelform, with no sharp distinction between tube
and throat.

Stamens of equal length, inserted equally on the
tube.

It would appear from the above evidence that Brand’s (1907)
treatment of Polemoniella as a synonym of Polemonium is correct.

Department of Botany,
University of California, Berkeley.

LITERATURE CITED

Brann, A. 1907. Polemoniaceae in Engler, Pflanzenreich 4° 1: 208.

Herter, A. A. 1904. Western Species, New and Old. II. Muhlenbergia 1:
47-62.

Newtson, A., and J. F. Macsripe. 1916. Western Plants. Bot. Gaz. 61: 30-47.

OsTENFELD, C. H. 1923. Genetic studies in Polemonium caeruleum. Hereditas
4: 17-26.

Wuerry, E. T. 1944. The minor genus Polemoniella. Am. Midland Nat. 31:
211-215.

1947] MASON: JEPSON 61

WILLIS LINN JEPSON

The passing of Willis Linn Jepson on November 6, 1946, closed
the career of one of the most colorful, dramatic, and stormy fig-
ures in the history of California botany. His detailed biography
(Science, in press, 1947) has been written by others; his botanical
work will stand on its merits. It is my purpose here to depict the
man as I knew him—as a friend, as a teacher, and as a colleague.

Jepson was the son of a pioneer California family that settled
in the Vaca Valley to the northeast of San Francisco Bay. He
was a native son who never quite forgave the march of progress
for its inroads on the pristine landscape that was early California.
He yearned for the return of the days of the “open range’ when
the Sacramento Valley stretched before him as a seemingly end-
less, open, fenceless plain, for the days when he had wandered
waist deep in its native flora. He resented the “foreigner from
beyond the Sierra” who was tempted into the region by the wealth
of its flora to poach upon what he regarded as his special heritage
by right of birth, the flora of California. This resentment in-
cluded those who preceded him as well as his contemporaries, save
for one early botanist, the Scotchman, David Douglas. To Jep-
son, Douglas, “The Indefatigable,’ along with Bentham and
Hooker, constituted the Saints and Apostles of Botany. They
were set up as ideals for his students to emulate: Douglas, because
of his tireless persistence and devotion to botany in the face of
seemingly insurmountable obstacles and dangers; Bentham, be-
cause of his attention to detail and his eagerness to dissect and
sketch all plants that passed through his hands; and Hooker,
because of his orderly mind, meticulous habits, and prodigious
energy. No student ever left Jepson’s laboratory without gaining
an appreciation of the virtues and accomplishments of these three
men. He all but ignored the continental schools of botanists from
Linnaeus to DeCandolle and Engler. His leanings were wholly
toward the British school and, except for his reverence for Doug-
las, he worshipped wholly at the altar of Kew.

As a teacher, Jepson was an exponent of the theory of self-
reliance and personal experience. He frankly discouraged stu-
dents who were not able. to pursue and solve their own problems
by themselves. He would give counsel and advice when asked,
but the student had to work out his own salvation. Rarely would
Jepson answer a question directly; often he would change the
subject and not answer the query at all. There was little of what
might be classed as formal teaching in his methods. Problems
were solved as they arose and never were they anticipated. In
effect, the student group was largely its own teacher through a
process of pooled experience, a group which soon learned to con-
sult the professor only when the problem was outside the experi-
ence of any of its members and could not be solved by consulting

62 . MADRONO [Vol.9

the literature. Although Jepson was genial and friendly with his
students, he was very formal and kept them at a distance. Each
student soon learned that the professor was never to be disturbed
when his door was closed and that he must never be waylaid when
passing to and from his quarters. However, when the door was
left open, or when the professor came into the laboratory, the
student was welcomed with a friendly smile. There were times
when the door remained closed for many weeks on end. The only
recourse the student had during such times was to exercise pa-
tience or to write the professor a letter which in due time might
be answered. Jepson cultivated an atmosphere of mystery as to
his comings and his goings. Usually no one knew when and if he
would come on any particular day, or where he had been or where
he was going. He chose his entries and his exits carefully. He
almost never took a student into the freld with him.

Jepson never married. His time was entirely his own to devote
as he chose—to himself, or to the flora of California. He lived a
singularly lonely life surrounded by friends whom he would not
accept. There were few, if any, of his fellow men whom he
trusted; he was suspicious of the motives of anyone who offered
unsolicited favors. Every word and every move was carefully
scrutinized lest it carry some imagined hidden meaning, construed,
as he thought, to undermine his prestige or to betray him to his
fancied “enemies.” He relied solely on his intuition in determin-
ing other’s motives. His decision was final and irrevocable even
unto himself. The countless hours spent in brooding over fancied
wrongs cost him dearly in time and in energy and robbed him of
the vitality that he needed to complete his life work. Of his
students and friends he demanded service, loyalty and extreme
forbearance such as he himself was constitutionally unable to
reciprocate. In spite of all this, he was socially very versatile.
On occasion he was a most genial and gracious host, and a charm-
ing and witty companion. He revelled in reading or reciting the
works of the poets of whom Kipling was most cherished. He
excelled in recounting in vivid language the incidents of his ex-
periences. These were always cloaked in dramatic settings.
Actually almost every incident of his stormy life was a drama, a
fact that was always in his consciousness even to the point of
histrionics.

His chief obsession was a fear that he would be robbed of a
rightful place in posterity by the wilful chicanery of his “‘ene-
mies.’’ His suspicions were always directed toward a succession
of individuals who usually had no idea of the bitterness and depth
of resentment which he felt for them, nor of the reason for his
resentment. Attempts to reconcile Jepson to reality were futile;
he irrevocably closed all avenues of approach. To overcome
these imagined situations, Jepson turned his attention early in life
to the preparations for his own commemoration, completely for-

1947]

MASON: JEPSON

drole SP tab ce j : cae .

a
‘
ee

Ge Yk } a ae

Puiate 8. Wiruis Linn Jepson

Portrait by Peter Van Valkenburgh, February, 1927.

63

64 MADRONO [Vol. 9

getting that posterity would judge him by the volume and charac-
ter of his botanical writings. It was his often expressed hope to
be remembered as the Hooker of American botany. To this end,
he chose to endow the symbols of the physical setting of a botanist
at Kew and pedestal upon them his own memorial. Here again,
time, energy, and money were diverted from his life’s work, which,
on his passing, remained unfinished—a memorial to a tragic life.

To those of us who knew him longest and best, it was obvious
that these outward incidents were but evidence of a struggle he
was having with himself. It was amply clear that the significant
storms of his life were wholly within himself—a one-sided strug-
gle between uncontrolled suspicions and the charming, friendly
Jepson.—Hersert L. Mason, Department of Botany, University
of California, Berkeley.

NOTES AND NEWS
Oxtp Wortp PLiants APPARENTLY RECENTLY INTRODUCED INTO
Texas. After twenty-one years of traveling around Texas and
studying the native vegetation, the following four Old World
plants were met for the first time in 1944. Certainly three of
these, and possibly all four, are herein reported for the first time
from Texas.

POLYGONUM ARGYROCOLEON Steud. ex Kunze. Railroad siding,
Rankin, Upton County, May 11, 1944, Cory 44176 (Herbarium of
Dr. J. F. Brenkle, Mellette, South Dakota). This Central Asian

species had previously been reported from California and Arizona.

ERYSIMUM REPANDUM L. (Cheirinia desertorum Woot. & Standl.).
Dry and gravelly bed of Four Mile Draw, a tributary of the Pecos
River in northeastern Pecos County (16 airline miles northwest
of Sheffield), April 6, 1944, Cory 44092, in flower; May 10, 1944,
Cory 44172, in fruit (Herbarium, Southern Methodist University,
Dallas, Texas).

SWAINSONA SALSULA (Pall.) Taub. In and along an irrigation
ditch, crossroads four miles southeast of Ysleta, El Paso County,
July 14, 1944, Cory 45017 (Gray Herbarium, Harvard University).
Swainsona has robust rootstocks as much as a foot or more under
the surface of the soil and forms dense stands. This plant, with
its numerous papery, inflated pods, reminds one of an Astragalus.
It is an Asiatic genus which has become established occasionally
in the Western states, the nearest locality to ours being Holbrook,
Arizona.

SENEcIo vuLtearis L. Winter Garden Experiment Station, near
Winter Haven, Dimmit County, March 27, 1944, Cory 43829
(Tracy Herbarium, A.M. College of Texas). V. L. Cory, Insti-
tute of Technology and Plant Industry, Southern Methodist Uni-
versity, Dallas, Texas.

MADRONO

A West American Journal of Botany

A quarterly journal devoted to important
and stimulating articles dealing with
plant morphology, physiology, taxonomy,
and botanical history. These volumes
should be a part of every botanist’s li-
brary and should be made accessible to
students of all universities and colleges.

Volume I, 1916-1929. . . $5.00
Volume II, 1930-1934 .. 5.00
Volume III, 1935-1936 . 5.00
Volume IV, 1937-1938 . 5.00
Volume V, 1939-1940 . . 5.00
Volume VI, 1941-1942 . 5.00
Volume VII, 1943-1944 . 5.00
Volume VIII, 1945-1946 . 5.00
Stnglejmumbers, 219... ¢ 0.75

The subscription price of MADRONO
is $2.50 per year. We solicit your pat-
ronage,

Address all orders to:
Reed C. Rollins, Bus. Mer.

Natural History Museum,
Stanford University, California

VOLUME IX _ NUMBER 3

MADRONO

A WEST AMERICAN JOURNAL OF
BOTANY

ve

Contents

OBSERVATIONS ON THE STRUCTURE AND CLASSIFICATION OF THE PyYROLEAE,
TEU OPA DOTA SYA CLUES AN RIS SON LO AI Oe PR eR Ie ee CO ae 65

A New SPEcIEs oF BLENNOSPERMA FROM CALirorNIiA, Charles B. Heiser, Jr. .. 103

“

UL 2 GA 7)

Yo

EONIAN INSTIFS
'f

Nv on ah
“Tone, muse?

Published at North Queen Street and McGovern Avenue,
Lancaster, Pennsylvania

July, 1947

a

MADRONO

A WEST AMERICAN JOURNAL OF BOTANY

Board of Editors

Hearsert L. Mason, University of California, Berkeley, Chairman.

LeRoy Asrams, Stanford University, California.

Epcar Anperson, Missouri Botanical Garden, St. Louis.

Lyman Benson, Pomona College, Claremont, California.

Hererrt F, Coretanp, Sacramento Junior College, Sacramento, California.
Ivan M. Jounston, Arnold Arboretum, Jamaica Plain, Massachusetts.
Mivprep E. Maruias, 2851 North Lake Avenue, Altadena, California.
Bassetr Macuirr, New York Botanical Garden, N. Y. C.

Marion Ownsey, State College of Washington, Pullman.

Secretary, Editorial Board—ANNETTA CaRTER
Department of Botany, University of California, Berkeley

Business Manager—ReEeEp C. Ro.uins
North Queen Street and McGovern Avenue, Lancaster, Pennsylvania
or
Natural History Museum,
Stanford University, California

Entered as second-class matter October 1, 1935, at the post office at
Lancaster, Pa., under the act of March 3, 1879.

Established 1916. Published quarterly. Subscription Price $2.50 per year.
Completed volumes I to VII inclusive, $35.00; each volume $5.00; single numbers
$0.75.

Papers up to 15 or 20 pages are acceptable. Longer contributions may
be accepted if the excess costs of printing and illustration are borne by the
contributor. Range extensions and similar notes will be published in con-
densed form with a suitable title under the general heading “Notes and News.”
Articles may be submitted to any member of the editorial board. Manuscripts
may be included in the forthcoming issue provided that the contributor pay the
cost of the pages added to the issue to accommodate his article. Reprints of
any article are furnished at a cost of 4 pages, 50 copies $4.10; 100 copies $4.50;
additional 100’s $0.85; 8 pages, 50 copies $5.95; 100 copies $6.60; additional 100’s
$1.30; 16 pages, 50 copies $8.35; 100 copies $9.35; additional 100’s $2.00. Covers,
50 for $2.75; additional covers at $1.65 per hundred. Reprints should be
ordered when proofs are returned.

Published at North Queen Street and McGovern Avenue, Lancaster,
Pennsylvania, for the

CALIFORNIA BOTANICAL SOCIETY, INC.

President: Howard E. McMinn, Mills College, California. First Vice-
President: Milo S. Baker, Santa Rosa Junior College, California. Second
Vice-President: David D. Keck, Carnegie Institution of Washington, Stanford
University, California. Secretary, John Whitehead, University of California
Botanical Garden, Berkeley. Treasurer, Reed C. Rollins, Natural History
Museum, Stanford University, California.

Annual membership dues of the California Botanical Society are $2.50,
which includes a year’s subscription to Madrono. For two members of the same
family the dues are $3.50, which includes one copy of Madrofio and all other
privileges for both. Dues should be remitted to the Treasurer. General
correspondence and applications for membership should be addressed to the
Secretary.

1947] COPELAND: PYROLEAE 65

OBSERVATIONS ON THE STRUCTURE AND
CLASSIFICATION OF THE PYROLEAE

Hersert F. Corpetanp

Having discussed the natural classification of the Monotropoi-
deae (1941) and Rhododendroideae (1944), as indicated by
studies of the microscopic structure, I now deal in the same
fashion with a third distinct group within the same order of
plants. I believe that it is right to call this order by the first name
applied to it as such, namely Bicornes L. The rule that a group
cannot be published by enumeration of the included groups seems
merely an excuse for breaches of priority and not entitled to
respect.

History

The tribe Pyroleae coincides with the genus Pyrola as delimited
by Linnaeus (1753). Radius (1821) tells us that the name Pyrola
was introduced by Brunfels, and cites those naturalists who first
discovered or recognized the six species named by Linnaeus: only
the names were original with the latter. These species are moder-
ately divergent, and most subsequent authorities have distributed
them among two or more genera. All agree, however, that none
of them is to be placed quite apart from the others, and that no
additional forms, belonging with them and conceivably represent-
ing additional genera, have been discovered. The names of the
species of Pyrola known to Linnaeus, and those of the genera
subsequently based upon them, are as follows.

1. Pyrola rotundifolia is the obvious type of Pyrola and has been
construed as such by all authorities except one. Alefeld (1856)
made the group typified by it a distinct genus Thelaia.

2. Pyrola minor is typical of the genera Eralebenia Opiz, 1852",
and Amelia Alefeld.

3. Pyrola secunda was treated by Alefeld as the typical Pyrola.
As a segregate from Pyrola, it has been named Ramischia Opiz and
Actinocyclus Klotzsch (1851).

4,5. Pyrola umbellata and P. maculata belong to the genus
Chimaphila Pursh (1814).

6. Pyrola uniflora typifies the genus Moneses Salisbury, 1821.

Linnaeus (1764) placed Pyrola in the natural order Bicornes;
Jussieu (1789) placed it in order Ericae. I have not had access
to various other early essays in natural classification, and draw a
part of the following information from a discussion by Domin
(1915). The Pyroleae were so named as a separate order or

1 Dates in parentheses are references to literature cited. Dates not in
parentheses are those of publications which I have not been able to consult,
and do not list.

Manprono, Vol. 9, No. 2, pp. 33-64. April 25, 1947.

WL 21 1947

66 MADRONO [Vol. 9

family by Lindley, 1821, and this group was renamed Pyrolaceae
by Agardh, 1825. In 18380, Lindley included the monotropoid
plants in his order Pyroleae. De Candolle (1889) admitted as
distinct orders both Monotropeae and Pyrolaceae. He misplaced
among the latter the genus Galax; this mistake gave him occasion
to apply to the typical examples the name Pyroleae as that of a
tribe. Asa Gray (1848) included both Pyroleae and Monotropeae
in Ericaceae; Bentham and Hooker (1876), while maintaining a
separate order Monotropeae, included Pyroleae in Ericaceae.
Klotzsch (1851) had followed Lindley to the extent of combining
the pyroloid and monotropoid plants in a separate order which he
called Hypopithieae. Drude (1889) maintained the same group
as the family Pirolaceae, placing it before Ericaceae as if defi-
nitely more primitive.

The most recent comprehensive treatment of the group is by
Andres (with whom, in happier times, I had the honor and benefit
of regular correspondence) in a series of papers (1909-19386) the
main item of which (1914) is in effect my point of departure and
primary object of criticism. In it, Andres followed in many re-
spects the usage of Drude. He conscientiously misspelled Pyrola
and the names derived from it; declared the family typified by it
to be primitive as compared with Ericaceae; and construed this
family as including the Monotropoideae. The system of the
pyroloid plants was in outline as follows:

Pirolaceae subfamily 1. Pirolotideae Dumortier. This is to be
understood as including a single tribe Piroleae.

Genus 1. Ramischia Opiz. 1. R. secunda (L.) Garcke, in all
northern continents; 2. R. truncata Andres in eastern North
America.

Genus 2. Pirola Salisb. [ !]

Subgenus 1. Amelia Hook. f. 1. P. minor L., in all
northern continents.
Subgenus 2. Thelaia Hook. f.

Section 1. Ampliosepala Andres. Sepals short, tri-
angular. The subsections are distinguished by
details of the texture and form of the leaves.

Subsection 1. Elliptica Andres. 2. P. elliptica
Nuttall in North America; 3. P. alpina Andres
in Japan and North America.

Subsection 2. Obscura Andres. 4. P. chlo-
rantha Swartz in all northern continents;
5. P. renifolia Maximowicz, 6. P. soldanelli-
folia Andres, 7. P. morrisoniana Hayata, 8. P.
gracilis Andres, 9. P. atropurpurea Franchet,
all in eastern Asia.

Subsection 8. Scotophylla Andres. 10. P.
spathulata (Alefeld) Andres; 11. P. aphylla

Smith; both from western North America.

1947] COPELAND: PYROLEAE 67

Subsection 4. Rotundoides Andres. 12. P.
uliginosa Torrey, in North America.

Anomalous species: 13. P. orypetala Austin, in
New York.

Section 2. Euthelaia (Alefeld) Andres.» Sepals
elongate.

Subsection 1. Eralebenia (Opiz) Andres.
Sepals tongue-shaped, more than one-third
and less than half as long as the corolla.
14. P. sororia Andres in eastern Asia; 15. P.
media Swartz in northern Europe and Asia;
16. P. Faurieana Andres, 17. P. Corbieri Le-
veillé, 18. P. nephrophylla Andres, in eastern
Asia; 19. P. Sartoriti (Alefeld) Hemsley in
Mexico; 20. P. paradoxa Andres in western
North America.

Subsection 2. Alefeldiana Andres. Sepals
lance-acuminate, at least half as long as the
petals. The three included groups are dis-
tinguished by details of the form of the
leaves.

Group 1. Genuina Andres. 21. P. For-
restiana Andres in eastern Asia; 22. P.
rotundifolia L. in all northern continents ;
23. P. japonica Siebold; 24. P. americana
Fernald [| actually of Sweet]; 25. P. sub-
aphylla Maximowicz in eastern Asia;
26. P. asarifolia Michaux in eastern
North America; 27. P. bracteata Hooker
in western North America.

Group 2. Amoena Andres. 28. P.decorata
Andres and 29. P. alba Andres in east-
ern Asia.

Group 8. Pictoides Andres. 30. P. septen-
trionalis Andres, 31. P. blanda Andres,
and 32. P. Conardiana Andres, in west-
ern North America.

Genus 3. Moneses Salisbury. 1. M. uniflora (L.) Salisb., in
all northern continents.
Genus 4. Chimaphila Pursh.
Sectionl. Aristata Andres. 1. C. japonica Miquel.
Section 2. Campanulata Andres. 2. C. umbellata
(L.) Nuttall [actually of Barton] in all northern
continents; 3. C. maculata (L.) Pursh in eastern
North America; 4. C. Menziesit Sprengel in west-
ern North America.

In Andres’ other papers one finds minor variations from the
framework of classification just set forth. In 1986 he added P.

68 MADRONO [Vol. 9

coreana (said to be a subspecies, though designated by a binomial)
and P. sumatrana, both placed next to P. japonica. His elaborate
system of categories was extended downward into an elaborate
subdivision of the collective species P. rotundifolia. This compli-
cation of classification is objectionable as a matter of taste and
expediency rather than of fact and hypothesis. One can express
the opinion that certain species are related by placing them under
a common key-heading and by listing them in succession; it is not
necessary to make of every such cluster a named subgeneric
group.

Rydberg (1914) referred duly to Andres when dealing with
the North American pyroloid plants in a work which goes to the
extreme in avoiding recognition of subsidiary categories. In his
treatment, these plants were the family Pyrolaceae, of five genera,
Pyrola, Eralebenia (Pyrola minor L.), Ramischia, Moneses, and
Chimaphila. Under Pyrola, eighteen species were recognized;
three of Andres’ species were reduced, two new species were
described, and six older ones reduced or overlooked by Andres
were restored.

Henderson (1919) studied extensively the macroscopic and
microscopic structure of the pyroloid and monotropoid plants, and
concluded that these groups are not primitive, but derived from
Ericaceae. 7

Fernald has discussed specific limits in the group of Pyrola ro-
tundifolia (1904) and has shown (1941) that P. virens Schweigger
is the right name of the species generally known as P. chlorantha.
Camp has dealt with specific limits in eet ie (1939) and
among the allies of Pyrola picta (1940).

Several morphological contributions, old and new, will be
cited below.

MaTERIAL AND METHODS

Alefeld opened his classic paper with the words, “Im Sommer
1845 fand ich einmal auf einem ganz kleinen Raume 6 deutsche
Arten von Pyrola L. beisammen. Dies veranlasste mich, diese
Pflanze naher zu untersuchen und vergleichen.” My introduction
to the group was by a quite similar experience: in the summer of
1920, I saw at Jonesville (a locality in Butte County, California,
at an altitude of about 1500 m.) seven distinct races of Pyroleae.

The plants growing at Jonesville have afforded the mass of my
material. This has been supplemented by the generous contribu-
tions of correspondents. On the present occasion as on former
ones, it is a pleasure to acknowledge a cordial obligation to the
Juneau Botanical Club, of Juneau, Alaska, and particularly to the
secretary, Mrs. Lucille Stonehouse; and to Dr. W. H. Camp of the
New York Botanical Garden. By their contributions, the avail-
able material has amounted to a fair representation of the range
of the group.

1947] COPELAND: PYROLEAE 69

ew

Pate 9. STrRucTURE OF THE PyrotedAr. Plants, approximately natural size,
photographed by E. B. Copeland at Jonesville, Butte County, California, July,
1920. Fic. 1. Chimaphila Menziesii. Fic. 2. Pyrola picta.

The races of which material in histological fixatives has been
available are listed below. As it seems expedient to designate
these races by the names which appear correct, and to place them
in the sequence which appears best as a representation of natural
classification, the listing implies some of the conclusions which
are to be stated explicitly in the sequel.

1. Ramischia secunda (L.) Garcke, collected at Jonesville; on

70 MADRONO [Vol. 9

Mount Rainier, Washington, by Camp; and on the Mendenhall
Flats, near Juneau, Alaska, by the Juneau Botanical Club.

2. Chimaphila umbellata (L.) Barton, collected at Jonesville;
and in Oregon and in the Alleghany National Park, New York, by
Camp.

3. Chimaphila maculata (L.) Nuttall, collected at Chilhowees,
Tennessee, by Camp.

4. Chimaphila Menziesu Sprengel, collected at Jonesville; and
on Mount Rainier by Camp.

5. Pyrola minor L., collected on the Mendenhall Flats by the
Juneau Botanical Club.

6. Pyrola virens Schweigger, collected on Mount Rainier by
Camp.

7. Pyrola picta Smith, collected at Jonesville; and on Mount
Rainier by Camp.

8. Pyrola dentata Smith, collected on Mount Rainier by Camp.

8a. Pyrola dentata var. integra Gray, collected at Jonesville.

8b. Pyrola dentata var. apophylla n. var.,? collected at Jonesville.

9. Pyrola americana Sweet, collected in Alleghany Park by
Camp.

10. Pyrola uliginosa Torrey and Gray, collected at Jonesville.

11. Pyrola bracteata Hooker, collected by Camp on Mount Rai-
nier. A plant collected by the Juneau Botanical Club on Menden-
hall Flats, and sent under the name of P. asarifolia Michaux,
appears to represent the same species.

12. Moneses uniflora (L.) Gray, collected at Lena Cove and the
Shelten Islands by the Juneau Botanical Club. The material from
the Shelten Islands is said to represent var. reticulata (Nuttall)
Blake.

The material listed has been used principally in study by
routine histological methods. The necessary herbarium and li-
brary study has been facilitated by the continued, unstinted, and

2 Pyrola dentata var. apophylla n. var., laminis foliorum reductis. Jones-
ville, Butte County, Calif.. H. F’. Copeland, sn., July 28, 1935; type in the
Herbarium of the University of California.

Aphyllous forms of Pyrola have repeatedly been collected in western North
America; they have generally been referred to P. aphylla Smith. Fernald (1920,
1941), however, has described and named one of these forms as a variety of
P. virens, and Camp (1940) has identified the classic P. aphylla as a form of
P. picta. The aphyllous plants of Jonesville appear, by their distribution in
the woods, to represent P. dentata var. integra. It is not possible from Hooker’s
(1834) plate of P. aphylla, and it would probably be impossible from Smith’s
type specimen, to decide whether this name belongs to a variant of P. virens,
P. picta, or P. dentata var. integra: the loss of leaves obliterates the distinctive
characters. Under these circumstances it appears best arbitrarily to treat
Camp’s action as sound, that is, to consider the classic P. aphylla as a form of
P. picta; and to give the new name here published to the leafless variant of
P. dentata var. integra. The plant which Holm (1898) described as P. aphylla
appears, by the characters of its sporadically occurring foliage leaves, to repre-
sent the present variety.

1947] COPELAND: PYROLEAE 71

cordially appreciated hospitality of the Herbarium and Biology
Library of the University of California.

The observations lead to conclusions as to the proper place of
these plants in the taxonomic system, and as to their expedient
arrangement in genera. Specific limits remain obscure in some
parts of the group, and the present observations are of little help
in clarifying them.

VEGETATIVE Gross STRUCTURE

Linnaeus noted certain Pyroleae as undershrubs, and others as
perennial herbs; all are in fact rather of the latter character, their
aerial shoots being less enduring than the underground structures.
The main underground structure, in all members of the group
except Moneses, is a rhizome. The rhizome is slender, not more
than a few millimeters in diameter, yellow to brown in color, bear-
ing distant small scales. The scales subtend scaly buds; associ-
ated with each of these, as Henderson noted, there is usually a
single root, springing from within the rhizome, and usually brief
and less than one millimeter in diameter. Most buds remain
dormant; the rhizome is sparsely branched, though often quite
elongate. I found one of Chimaphila umbellata which was approxi-
mately 2.5 meters long, connecting two leafy shoots near one end
with a single one at the other. Occasionally, the tip of a rhizome
turns toward the surface of the soil; produces, at the end of a
growing season, a scaly bud; and gives rise, during the following
growing season, to an aerial shoot.

In Moneses uniflora, the aerial shoots grow upward directly
from slender roots, within which they originate, of course, as
adventitious buds. This habit was first reported by Irmisch
(1855). Itis the same as that of the monotropoid plants. Irmisch
reported that in Ramischia secunda and Pyrola virens the aerial
shoots may arise either from rhizomes or from roots; Holm (1898)
states that this occurs also in P. aphylla (that is, as I suppose, P.
dentata var. apophylla), and in P. picta, Chimaphila umbellata, and
C. maculata. Henderson could not confirm this, nor have I done
so in digging up a moderate number of specimens. It seems safe,
nevertheless, to accept these observations as sound: it is inherently
probable that the roots are capable of the occasional production
of adventitious buds throughout the group.

Various observers, as Wydler (1860), Drude (1889), and
Henderson (1919), have noted the habit of the Pyroleae, of pro-
ducing a winter bud at the end of each season of growth; with the
effect (the scales being persistent) that the stem is found to bear,
alternately, series of scales and of leaves. In Ramischia, the leaves
of each year are distributed along a few centimeters of stem. In
Chimaphila, they are typically crowded near the summit of the
year’s growth of stem, and give a superficial appearance of being
whorled. In Pyrola and Moneses, the year’s growth of stem is

72 MADRONO [Vol. 9

usually very brief; all the leaves on one shoot are crowded in a
single rosette.

Growth as described continues until the year in which the
growing point is used up in inflorescence. This occurs sooner or
later, characteristically in different species. In Chimaphila umbel-
lata and Chimaphila Menziesit it may be delayed until after the pro-
duction of seven or eight annual pseudo-whorls of leaves. In the
group of Pyrola rotundifolia, the shoots seem most often to flower
in the fourth growing season, after bearing leaves during three.
Ramischia, Chimaphila maculata, Pyrola picta, P. dentata var. integra,
and Moneses uniflora flower most often in the third growing season.
P. virens flowers usually after producing a single series of foliage
leaves. Shoots of the various aphyllous races flower in the season
in which they come up.

Axillary buds do not usually open before the terminal bud
above them is ready to produce an inflorescence. If they open
in the same year, it is usually to produce additional inflorescences.
This they do rather commonly in P. virens and the group of P. picta.
In Ramischia, Chimaphila, and the group of P. rotundifolia, they
usually open the year after the stem bearing them has produced
flower and fruit, and produce leaves during two or three years
before bearing flowers inturn. In Moneses, the shoots usually die
after flowering and fruiting.

The leaves are alternate. JI examined a few plants and found
the leaves to form a regular spiral: starting from any given leaf,
the third leaf or bract above it stands somewhat to one side, say
the left, of directly above it; the fifth falls a smaller distance to
the right, the eighth a yet smaller distance to the left, and so
forth. Wydler (1860) has noted irregularities in the phyllotaxy
of this group: such irregularities are common in all groups of
plants.

The size, shape, and texture of the leaves, their characters as
petioled or sessile, entire or dentate, are duly set forth in the
manuals and need not be recounted. As noted, the stalked leaves
of Pyrola form rosettes. Usually, both leaves and bracts are per-
sistent ; Pease (1917) has found green leaves up to seven years old
on shoots of Chimaphila umbellata, and up to eight years old in
C. Menziesii. In this genus, the bracts, and later the leaves, are
finally allowed to fall by the action of a disjunction mechanism.
In Pyrola, the leaves wither in place. In the group of P. rotundi-
folia, this occurs usually after they have been green for two years.

EXPLANATION OF THE Ficures. PuLare 10.

PuatTe 10. Structure oF THE PyroreAE. Anatomical features of Chima-
phila Menziesii. Fic. 3. Longitudinal section of root tip, x 200. Fic. 4. Cross
section of young root, X 200. Fic. 5. Cross section of older root, X 200. Fic. 6.
Cross section of rhizome, X25. Fic. 7. The marked area of fig. 6, X 200.
Fic. 8. Cross section of young stem, X25. Fic. 9. The marked area of fig. 8,
x 200.

1947] COPELAND: PYROLEAE

coe

Onis
aE ESS
NA HY RS a
eat AA
a eS
REE

a
S&S
si
V p——

ev

b

pat

VW]
LY

(TRS

ja

—,

one
Fest
OFF

DB,
Le

SE,
[) > 3.85 cA LS

a PO EEOOES
By oa i

ry
“

(B)

&
sw:
o

>
ob3
aah
BA 1] ae 28@
el ew)
NGI.)
Ors

<a

Us
¥¢ a) {x Feat)
em teis

Piate 10. SrrRuctrure OF THE PYROLEAE.

73

4 MADRONO [Vol.9

As the shoot usually produces no leaves in the year in which
it flowers, there are bud scales above the highest foliage leaves
as well as below the lowest. Irmisch, Wydler, and Henderson
have reported Chimaphila umbellata as an exception. Of this spe-
cies and of C. Menziesu, I find that they may or may not bear
foliage leaves above the highest bud scales, in the same year as
the flowers. Wydler and Andres have reported the same varia-
bility in Pyrola virens and P. minor. ,

ANATOMY OF THE VEGETATIVE STRUCTURES

The structure of well-developed primary roots has never to
my knowledge been observed in any pyroloid plant. The roots
examined by Henderson, of Chimaphila umbellata, C. maculata,
Pyrola rotundifolia (that is, presumably, the scarcely distinct P.
americana), and P. elliptica, and by myself, of Ramischia secunda,
Chimaphila Menziesii, Pyrola minor, and P. uliginosa, were either
adventitious roots springing from the rhizome or secondary roots
springing from these. The elongate roots of Moneses uniflora,
seen by Henderson and myself, were not traced to their origin;
presumably they spring from other roots.

The structure observed presents no particular peculiarities.
In the root tips (pl. 10, fig. 3) there is a definite dermatogen as
distinct from an inner body of meristematic cells. The root cap
is very scant. The epidermal cells grow considerably in the radial
direction as soon as they emerge from the protection of the cap.
In almost all examples, they presently become beset, both exter-
nally and internally, with a mycorrhiza of fine hyphae. It is a
familiar and probable theory, that the mycorrhiza contributes sig-
nificantly to the nutrition of the plants. This has not been demon-
strated experimentally in the present group. Eventually, the epi-
dermal cells die and disappear; this may be true also of the outer
layers of the cortex.

The inner meristematic tissue produces, of course, a cortex and
a stele surrounded by it. In most examples, the cortex is of only
about three layers of cells in addition to the endodermis. In
Moneses, the root being the permanent organ of the plant, the
cortex is of several layers of cells. The cortical cells are gener-
ally thin-walled; in Chimaphila Menziesu, the innermost ones are
thick-walled. The cells of the endodermis are thin-walled. In
some specimens, the radial walls resist staining, being apparently
of the nature of Casparian strips.

There is no pith. Henderson described the primary xylem as

EXPLANATION OF THE Ficures. Puarte 11.

Puate 11. Srrucrure or THE Pyrorear. Fic. 10. Chimaphila Menziesia,
cross section of mature stem, X 200. Fic. 11. Ramischia secunda, cross section
of leaf, x 400. Fic. 12. Chimaphila Menziesii, cross section of leaf, x 400. Fic.
13. Pyrola picta, cross section of leaf, X 400. Fic. 14. Chimaphila Menziesii,
cross section of mature peduncle, x 25. Fic. 15. Part of same, X 200.

1947 | COPELAND: PYROLEAE 15

er)

Ct

am we

s

:

KX)

a 4 C
x0)
wal

aig
iat
ie
a0.)
é);

or ))
Cs cA

Ses. : Faw
AW Mt f
AW Ek Oa

Puate 11. STRUCTURE OF THE PYROLEAE.

16 MADRONO [Vol. 9

usually 4- or 5-arch, but found a diarch specimen of Moneses. So
far as my own observations go, roots are always diarch at their
origin (pl. 10, fig. 4). Usually they remain diarch, but my speci-
men of Moneses is triarch. By the production of secondary xylem,
a cylinder of wood is formed (pl. 10, fig. 5). Phloem, and a peri-
cycle of thin-walled tanniniferous cells, were noted between xylem
and endodermis. As lateral growth goes forward, the endodermis
becomes divided by radial walls. No cork is formed.

The rhizome and the aerial stem are essentially alike in struc-
ture. They do not attain great thickness, nor develop a consider-
able mass of wood: further differences from the stems of typical
Bicornes are noted below. During the annual brief period of
active growth, the cortex, procambial cylinder, and pith become
differentiated immediately back of the growing point. Scattered
spiral tracheids become differentiated at the inner boundary of
the procambial cylinder (pl. 10, fig. 9). These may be few, or
may become fairly numerous; in the latter case, there is a radial
transition from tracheids with an extended spiral, scattered in
parenchyma, to others with a compact spiral and scant associated
parenchyma. A typical cambium begins to function. The second-
ary xylem produced by it consists, as to the tracheary elements, of
cells whose walls are marked by compact spiral striations and
bordered pits with elliptic included openings, horizontal or
oblique. These are the characters both of the numerous fiber-
tracheids and of the fewer and only slightly broader vessels; in
the latter, the spiral ridges are less prominent, and there are com-
monly two alternate columns of pits on each face. The vessels
have strongly oblique ends with numerous scalariform perfora-
tions. The numerous rays are essentially uniseriate. The cells
of which they consist are vertically elongate, and there may be
several tiers to the ray; the tiers may overlap at the margins, pro-
ducing a biseriate appearance in cross sections. Except in the
rays, no parenchyma was found in the wood.

In rhizomes (pl. 10, fig. 7), spring wood and summer wood are
scarcely distinguishable. In aerial stems (pl. 11, fig. 10), the
annual rings are evident, though neither prominent nor numerous.

These rhizomes and stems differ notably from those of typical
Bicornes in two anatomical features. No pericycle, that is, no
layer of fibers at the outer margin of the phloem, is differentiated,
and no cylinder of cork is formed.

The cortex is mostly of thin-walled cells, but on aerial stems
the outermost layers are thick-walled, constituting a differentiated
hypodermis. The epidermis also is thick-walled and bears a longi-
tudinally striate cuticle. The growth of the cylinder of wood, so
long as it continues, results in a crushing of the thin-walled inner
cells of the cortex. The epidermis and hypodermis are neither
ruptured nor reinforced by a layer of cork: they persist as long
as the shoot does. The plants are not totally incapable of pro-
ducing cork. This tissue may be formed in response to wounding,

1947] ' COPELAND: PYROLEAE 17

and is formed regularly as a disjunction layer at the bases of the
bracts and leaves of Chimaphila.

The nodes are unilacunar. A single trough-shaped bundle,
unaccompanied by fibers, enters each petiole. The bundle which
enters a bud or branch is formed by a junction of two bundles
springing from the sides of the leaf gap.

As to the leaves, Table 1 gives, for the species of which I have
seen cross sections (pl. 11, figs. 11-13), the thickness and the
numbers of layers of palisade tissue observed.

TABLE 1. COMPARISON OF CERTAIN LEAF CHARACTERS IN THE PYROLEAE

Shadi Approximate thick- Number of layers

pecies ness of leaves of palisade
in microns

Ramischia secunda ............. 0

Chimaphila umbellata 1

CRIMOCUIOLA © oie isso tsetse 2

GRIVENZIESU, Fo ssn oo ee oe. 1-2

Pyrola Minor... 8s. kc be. 0

PROTOS RN REE hes cite the glen 1

[Es TOG TR SSA ee ea eee 2

P. dentata var. integra 2

WPPOMETICONG iru cise yp bee ls ' 0

PE ULOULOSO hs hn sits hans hades 0

PEO TOCUC CUO wo carr. big dass oo ohw Rh 0

Moneses uniflora .............. 0

Henderson found three layers of palisade tissue in Chimaphila
umbellata and cited European authorities who had made the same
observation. As to several other species, her observations agree
with mine. She cited European authority for the absence of pali-
sade tissue in Pyrola rotundifolia; her own agreeing observation
applies, I assume, to the race here called P. americana. She found
the same condition in P. elliptica. Holm found about two layers
of palisade tissue in the foliage leaves of his “Pyrola aphylla”: this
would be true whether the plant in question was related to P.
virens, P. picta, or P. dentata.

On thick leaves with palisade tissue, the epidermis bears a
thick cuticle; the stomata are confined to the lower epidermis and
open at the level of the outer surface of the cuticle. Thin leaves
without palisade tissue bear a thin cuticle and the stomata project
moderately from the lower surface. As Henderson noted, sto-
mata in small numbers may be found on the upper surfaces of
these leaves. The outer opening of the stomatal pore is marked
by a prominent knife-like ridge; a less prominent ridge was found
at the inner opening in some specimens.

There is usually an accumulation of tannin in the uppermost
cells of the mesophyll, whether or not these are of the character
of a palisade, and a less considerable accumulation in the lowest

78 MADRONO [Vol. 9

cell-layers. In many cells of the less tanniniferous intermediate
layers, there are star-shaped crystals.

The leaf contains no fibers, though there are more or less con-
siderable bodies of collenchyma associated with the veins and in
the margins.

The thin leaves, lacking palisade tissue and bearing projecting
stomata, are evidently adapted to life in moist and shaded places.
According to my understanding of the relationships of the species,
I would suppose that this set of characters has developed more
than once within this limited group.

INFLORESCENCE

The typical inflorescence of the group is a bracted raceme
embraced at the base by the outer bud scales of the winter bud
from which it grew. The peduncle bears usually a few “scales,”
that is, scattered lower bracts which do not subtend flowers.
Abortive buds can usually be found in their axils. Bud scales,
scales on the peduncle, and bracts are all of quite the same charac-
ter. The pedicels are without bractlets.

The typical inflorescence thus described occurs in Ramischia
and in Pyrola (pl. 9, fig. 2). The occasional presence of foliage
leaves above the highest bud scales in some species of Pyrola was
noted above. Rydberg and Henderson took careful note of the
numbers of flowerless lower bracts in the species with which they
dealt. In Ramischia secunda the number is variable, from one to
four; in Pyrola elliptica and P. virens, it is often just one; aphyllous
shoots bear several; in the remaining American species, the num-
ber is two or three, the species not being distinguishable by this
character.

The occasional presence of foliage leaves above the highest
bud scales in Chimaphila has been noted. In this genus, the
peduncle bears no scales and the bracts are adnate to the pedicels.
The inflorescence of Chimaphila umbellata is no umbel, but a con-
densed raceme, a corymb; Pursh (1814), in introducing the genus
Chimaphila, undertook to change the epithet of this species, calling
it C. corymbosa. In C. maculata and C. Menziesiu (pl. 9, fig. 1), the
inflorescence is a cyme-like cluster of two or three flowers. The
bracts of C. Menziesii are suborbicular, pale, fleshy, with a dentate
margin,

EXPLANATION OF THE Ficures. PLate 12.

PLatTe 12. STRUCTURE OF THE PyROLEAE. Flowers, X 5; stamens and anthers,
x10. Fic. 16. Flower of Ramischia secunda. Fic. 17. Flower of Chimaphila
Menziesii. Fic. 18. Flower of Pyrola minor. Fic. 19. Flower of Pyrola ameri-
cana. Fies. 20, 21,22. Stamens and anthers of Ramischia secunda, respectively
in early bud, older bud, and in open flower. Fies. 23, 24. Stamens of Chima-
phila Menziesii, respectively in older bud and in open flower. Fics. 25, 26.
Younger and older stamen of Pyrola minor. Fics. 27, 28. Pyrola americana,
mature anther seen from within and stamen seen laterally. Fics. 29, 30. Sta-
mens of Moneses uniflora.

1947 |

COPELAND: PYROLEAE

~ 19

PLATE 12. STRUCTURE OF THE PYROLEAE.

79

80 MADRONO [Vol. 9

The inflorescence of Moneses consists of a solitary flower ter-
minal on a peduncle which may bear no scales or one or two.

The peduncle differs anatomically from vegetative stems in
showing scant secondary tissue or none, and in the presence of a
sheath of fibers outside the phloem (pl. 11, figs. 14, 15). In
Ramischia secunda, Chimaphila Menziestui, and Pyrola minor, the de-
velopment of this sheath is scant and tardy, so that it may not be
recognizable in young peduncles. The sheath does not extend
into the pedicels.

The nodal anatomy of the peduncle is in Ramischia the same
as in the vegetative stems: each bract is supplied by a single
bundle leaving a single gap, and each flower by a vascular cylinder
formed in the base of the pedicel by the junction of two bundles
running from the sides of the gap.

In Chimaphila, each pedicel is entered by a broad and trough-
like band of vascular tissue, like a leaf trace, but differing in the
fact that the margins of the trough coalesce, not far above the
base of the pedicel, to form a cylinder. Some distance up the
pedicel, the cylinder emits the single scant bundle which supplies
the bract. In Pyrola (P. minor and P. uliginosa were studied) the
structure is similar except that each bract is supplied by three
small bundles. One might say that in these genera the physio-
logical dominance of the axillary structure, the flower, has resulted
in reduction of the subtending leaf to an accessory status, so far
as the vascular supply is concerned.

FLOWERS: THE PERIANTH

The flower is choripetalous, pentacyclic, pentamerous, actino-
morphic or essentially so in Ramischia, Chimaphila, and Pyrola
minor, zygomorphic in most species of Pyrola and in Moneses.
Perianth, stamens, and the disk if any is present, are hypogynous.
The median sepal is on the upper or adaxial side of the flower
(pl. 12, fig. 19; pl. 17, fig. 56). The petals are alternate with the
sepals, and the median one is accordingly on the lower or abaxial
side. The stamens of the outer cycle are opposite the petals.
Stamens of the inner cycle and carpels are alternate, each with
the leaves of the preceding cycle, so that the carpels lie in the
radii of the petals.

The aestivation of the petals is imbricate. Roeper (1852)
studied the phyllotactic sequence of the petals, and found it
variable.

The sepals are ovate to lanceolate. In Chimaphila they are
irregularly finely dentate. In Ramischia and Moneses they are
finely ciliate. In Pyrola they are strictly entire and glabrous;
shape of sepals is a character of subgeneric grouns in this genus.

The petals are glabrous, rotund to obovate, generally with
obtuse or rounded apices. American floras recognize a single
local species, Pyrola orypetala, with acute petals. In color, they

1947] COPELAND: PYROLEAE 81

are white, cream, greenish, pink, or red, the main variations dis-
tinguishing subgeneric groups in Pyrola. In Ramischia only, each
petals bears a pair of minute tubercles on the inner surface near
the base. This obscure distinction was discovered by Alefeld.
The tubercles project from each petal into the spaces between the
three filaments lying within the petal. Sectioned and stained, they
are found not to be of the microscopic character of glands. Their
function is not evident; possibly it is that of the lodicules of
grasses, namely to effect the opening of the flower.

STAMENS

The lower half of the filament of Chimaphila is laterally ex-
panded and beset, particularly on the margins, with rather coarse
white hairs which are outgrowths of epidermal cells (pl. 12, figs.
23, 24). In the remaining genera, the filaments taper smoothly
and are glabrous (pl. 12, figs. 20-22, 25-30).

The stamens of Ramischia, Chimaphila, and of a few of the
species of Pyrola, are nearly uniform in length and dorsiventrally
symmetrical. In most of the species of Pyrola, the petalad sta-
mens are perceptibly the shorter, and all of the filaments are so
bent as to gather the anthers in a cluster on the upper side of the
flower. A similar inclination is present in Moneses, but the cluster-
ing of the anthers is less pronounced; in many specimens, the fila-
ments of the sepalad stamens are so bent that the anthers are
born in pairs opposite the petals.

Throughout the group, the summits of the filaments in bud are
sigmoidally bent, outward and upward. The filament merges into
the lower part of the inner side of the anther. The two areas
which are to become the pores of the anther are located on the
outer side near the base. At anthesis, the summit of the filament
bends inward and inverts the anther, so that instead of standing
erect on the filament it hangs from it, into the interior of the
flower, with the pores at the top. Undoubtedly, the juvenile posi-
tion of the anther expresses its true morphology, this in spite of
the fact that Asa Gray (1846) once stated the contrary opinion.
It is not true that the stamens are formed in an inverted position,
nor that they right themselves at maturity: they are formed right
side up and later turn upside down.

The pores are of various shapes, and may or may not be born
at the ends of tubes of various shapes.

In Ramischia, the pores are not circular, but elongate, crossing
the proximal ends of the anthers in the direction of the radial
planes of the flower; these ends are scarcely extended as tubes
Goly 12, fie. 22). .

In Chimaphila the circular pores are terminal on brief conical
tubes (pl. 12, figs. 23, 24).

In Pyrola minor (pl. 12, figs. 25, 26) the pores are widely open
and tubes are scarcely developed. Similar pores were seen in her-

82 MADRONO | [Vol.9

barium specimens of P. media and P. grandiflora, and are reported
in an Alaskan species, P. occidentalis.

In other species of Pyrola which are available to me in pre-
servative, the small pores are on the dorsal (at maturity, inner)
surfaces of brief tubes (pl. 12, figs. 27, 28). In two Mexican
species, P. Sartorit and P. angustifolia, Alefeld described the pores
as almond-shaped (amygdaliformes), that is, elongate, rounded
above, narrowed to a slit below. This character is perceptible in
herbarium specimens of the latter species, together with another
peculiarity: there is a slight protuberance on the ventral side of
each tube. It is as though the tubes bore the rudiments of horns
like those of the Andromedeae.

In Moneses (pl. 12, figs. 29, 30) the pores terminate definite
tubes which are so curved as to diverge below and converge above.

The internal structure of the young anthers is in most respects
that of flowering plants in general. There are four masses of
pollen mother cells surrounded by tapetum, inner wall cells, and
epidermis. In Ramischia, some of the epidermal cells project as
papillae, principally in vertical bands on the front and back of
each lobe of the anther. In Chimaphila, the microscope reveals a
tendency to this papillose character; it is so obscure as not to be
evident when the anthers are examined under the dissecting lens.
In Pyrola and Moneses, the epidermis of the anthers is essentially
smooth.

The epidermis of the areas of dehiscence is of small, thin-
walled, darkly-staining cells, and is underlain by further cells of
the same character, extending to the tapetum at the proximal ends
of both the pollen sacs in the same lobe of the anther. The entire
body of darkly-staining cells, conceived in space, is roughly a tri-
angle several cells thick, with one angle at the area of dehiscence,
the other two respectively at the proximal ends of the pollen sacs.
Because the two pollen sacs of the lobe are not in the same radial
or tangential plane of the flower, longitudinal sections of the
flower do not usually show the whole extent of this body of cells,
but merely sections through it (pl. 13, figs. 32-35), which may
appear to extend from the area of the pore to one pollen sac or
the other. The entire structure appears to be homologous with
the body of resorption tissue described in Arbutus and Arcto-
staphylos by Matthews and Knox (1926), and again in the latter
genus by Doyel (1942); but differs in being more than a single

EXPLANATION OF THE Figures. PuLare 13.

Priate 13. Srrucrure or THE Pyrorear. Sections of the proximal ends of
anthers, x 125. Fic. 31. Chimaphila Menziesii, cross section through the tubes
of the young anther with the filament ascending between them. Fics. 32-35:
Sagittal sections of young anthers. Fic. 32. Ramischia secunda. Fic. 33.
Chimaphila umbellata. Fie. 34. Pyrola picta. Fic. 35. Moneses uniflora.
Fic. 36. Sagittal section of open tube of anther of Chimaphila umbellata.

1947 ] COPELAND: PYROLEAE 83

ae
iis

Gt OND

23}
Lhe}
aes,
mee
eat

es
x

a,
acne,

TE
Prue
ase oe
PS

@S

See

Bec
ae

ty (23
en

Or

(2

sss as

Tere
os
ag
ted
Sig
Wr Ag)
~

LS
Lo 24524
ToC/
ses

ai
Bet

Puate 13. Structure oF THE PYROLEAE.

84 MADRONO [Vol. 9

layer of cells, and in being extended to include a broad area of
epidermis.

Stages in the development of the pollen grains have been seen
in the anthers of Chimaphila umbellata, C. Menziesii, and Pyrola
picta (pl. 14, figs. 37-42). They occur, as Hagerup (1928) noted,
in spring (in the mountains of California, in June) not long before
anthesis. They are perfectly normal. The haploid chromosome
number in both species of Chimaphila is 13; that of Pyrola picta
is 23. The latter number had been reported by Hagerup in P.
grandiflora, P. rotundifolia, and P. minor; but Samuelsson (1913)
had reported 16 in P. wirens, P. rotundifolia, and Moneses uniflora.
I confess to having held both reports in doubt, expecting to find
either 13 or 26, but have unmistakably found in P. picta the num-
ber reported by Hagerup. |

The pollen grains of Ramischia are solitary; those of Chimaphila
occur in easily disrupted tetrads; those of the remaining Pyroleae
are firmly united into permanent tetrads, as in most Bicornes. In
Ramischia, the »ollen grains are tricolpate. In Pyrola, as in most
Bicornes, the wall of each individual grain is marked by three
half-grooves, continued as half-grooves on the three associated
grains. Mature pollen grains are binucleate, with one of the
nuclei lying in a clear area, the generative cell.

The tapetal cells become binucleate and then shrivel to noth-
ing. The inner wall cells disappear as far as, but not including,
the layer next within the epidermis. The septum between the
pollen sacs of the same lobe breaks down. The connective—the
septum between the lobes, traversed by a bundle—persists.

Artopoeus (1903) and Matthews and Knox (1926) distin-
guished two tissues involved in the opening of the anthers of
Bicornes. Both tissues originate as small, thin-walled, darkly-
staining cells, as already described. In some circumstances, these
cells undergo disappearance by collapsing successively, each
against the one which is to disappear next. The tissue which
behaves in this fashion may be called collapse tissue. It is essen-
tially the same thing as the inner wall tissue of the anthers of
flowering plants in general, and of the present group. It may be
held to be present in the anthers of all Bicornes.

Under other conditions, the small cells do not collapse. They
become granular and undergo deliquescence, forming a granular,
apparently slimy mass, which gradually disappears. The cells
which undergo this process constitute resorption tissue. It is
present in the anthers of some Bicornes but not others.

As to the Pyroleae:

In Ramischia, the epidermis where the pores are to form, to-
gether with several layers of the underlying cells, consist of
resorption tissue (pl. 13, fig. 32); but the inner part of the mass
of differentiated tissue which extends to the tapeta at the proxi-
mal ends of the pollen sacs is of collapse tissue. This is to say
that the outermost plugs of the anther pores disappear by deli-

1947] COPELAND: PYROLEAE 85
fel

L,
ri of

42

ee

PuiaTe 14. StTrRuctuRE oF THE PyroredAr. Details of microsporogenesis.
Fic. 37. Chimaphila umbellata, diakinesis, x 900. Fic. 38. Chimaphila umbel-
lata, heterotypic metaphase, x 900. Fics. 39, 40. Chimaphila Menziesii, hetero-
typic metaphases, X 1200. Fic. 41. Pyrola picta, heterotypic metaphase, X< 1000.
Fic. 42. Pyrola picta, chromosomes in heterotypic anaphase, the two groups
from a single spindle, 1000. Fic. 43. Chimaphila Menziesii, pistil, x5. Fie.
44. Pyrola minor, stigma, X10. Fic. 45. Moneses uniflora, pistil, x5. Fuies.
46-48: Models of the vascular system in the floral receptacle, 50. Fic. 46.

Ramischia secunda. Fic. 47. Chimaphila umbellata. Fie. 48. Chimaphila
Menziesti, |

86 MADRONO [Vol. 9

quescence; the further disappearance of tissue within the anther,
as already described, is by collapse.

In Chimaphila, the entire body of differentiated tissue, from
the area of dehiscence to the tapeta, is resorption tissue. The
disappearance of further tissue within the anthers is by collapse.

In various species of Pyrola, and in Moneses, I have searched
material, less abundant than the available material of Ramischia
and Chimaphila, and have found no resorption tissue. In Pyrola
picta I found anthers in which dehiscence was barely beginning
to take place, by a process of collapse forming a cleft within the
differentiated tissue. I am convinced that resorption tissue does
not occur in these genera. .

The persistence of the hypodermal cell layer of the anthers
has been mentioned. In the proximal end of the anther but not
in the distal, and excluding, of course, the area of the pores, the
cell walls of the hypodermal layer develop reticulate thickenings
and constitute a rigid lining maintaining the form of the anther
tubes (pl. 18, fig. 36). This mechanical layer occupies a part of
the position of the endothecium in the anthers of typical flowering
plants, but it is not necessarily homologous with it. It is a rigid
structure, not a dynamic one, and the reticulate thickening of the
cell walls is distinctly different from the ribbing of a typical
endothecium.

In Ramischia, a rather scant extent of hypodermis develops
reticulate thickenings, but this genus has the peculiarity that
reticulate thickenings are developed in the epidermis throughout
its extent.

Disk. PisTin

Respected authorities have differed as to the presence of a
disk in the flowers of certain Pyroleae. The facts were set forth
correctly by Irmisch (1856), being as follows.

In Ramischia there is a rather massive disk of glandular tissue
forming a complete ring below the base of the pistil and project-
ing between the bases of the filaments.

In Chimaphila there is a well-developed disk like a collar about
the base of the pistil (pl. 14, fig. 43). Its margin is nearly entire,
scarcely impressed by the filaments nor projecting between them.

Neither in Pyrola minor, nor in the typical species of Pyrola,
nor in Moneses, is there any disk whatever; there is no projection
at the base of the pistil nor any glandular tissue in this region.

The ovary (pl. 14, figs. 48, 45) is subglobular with ten distinct
vertical grooves in the planes both of the sepals and of the petals.
It is deeply impressed at the summit, the style springing from
within the impression.

The style and stigma of Pyrola minor (pl. 14, fig. 44) are much
like those of the Rhododendroideae. An extended cylindrical
style flares at the summit to form a circular platform whose mar-

1947] COPELAND: PYROLEAE 87

Puate 15. Structure oF THE PyroteAr. Models of the vascular system in
the floral receptacles, x 50. Fic. 49. Pyrola minor. Fie. 50. Pyrola bracteata.

88 MADRONO [Vol. 9

gin is describable as a collar; upon the surface there are five pro-
jecting knobs in the planes of the sepals; between them, in the
planes of the petals, there are five clefts, meeting in the center,
all leading into the open style passage. By what circumstance
some authorities have been led to describe the stigma of P. minor
as having no collar, I do not know. The ares of collar between
the knobs are the extremities of the carpels; each knob represents
the end of the coalescent margins of two adjacent carpels.

The style and stigma of Moneses are similar, except that the
knobs are remarkably prominent (pl. 14, fig. 45).

Those of the typical species of Pyrola (pl. 12, fig. 19) are of
the same structure but on a finer scale; and the style is bent to the
shape of an old-fashioned italic J, or an integral sign: the flower
standing nearly horizontally, the style is bent downward, toward
the median or abaxial petal, and then outward.

The style of Chimaphila is brief and broadly flaring (pl. 14,
fig. 43). The stigma is of the same structure as before, but with
the proportions greatly modified. It is bordered by an obscure
collar. The coalescent margins of adjacent carpels do not form
projecting knobs, but merely sectors, separated by clefts, of a
domed surface.

The extended style of Ramischia flares to a domed stigma
divided into five sectors by radiating clefts which lie, of course,
in the planes of the petals. The margin of the stigma at the ends
of the clefts does not project and form lateral extensions coales-
cent below the knobs; hence it is descriptively correct to say that
the stigma is without a collar. This stigma is simpler than that
which is typical of the Bicornes, as if primitive; more probably,
it is reduced.

As in most Bicornes, the style is traversed by an open channel.
Five lengthwise flanges of tissue project into this channel. These
represent the coalescent margins of adjacent carpels; they are
continuous with the knobs on the stigma above and with the septa
in the ovary below. The grooves between the flanges are con-
tinuous with the locules.

There are, of course, five locules, located oposite the petals.
They are nearly filled by massive placentae each of which is
radially divided into two by a vertical cleft. Above the middle of
the ovary, the clefts in the placentae meet in the center of the
ovary. Thus, in the upper half of the ovary, the placentation is
parietal. The structure thus described is the same as the usual
structure of the ovaries of Monotropoideae; Henderson has duly
noted its occurrence in both groups. ‘Typical Bicornes are differ-
ent in the fact that the locules are in communication with the
style channel only at the summit of the ovary.

The placentae are densely beset with numerous minute ovules.
The inner surfaces of the ovary walls are clad with about two
layers of slender fiber-like cells with thick walls.

1947] COPELAND: PYROLEAE 89

THE RECEPTACULAR VASCULAR SYSTEM

The vascular cylinder which ascends from the pedicel into the
receptacle of Ramischia (pl. 14, fig. 46) emits there, in succession
from proximal to distal, the following whorls each of five bundles:
(1) sepal bundles, each soon forking into three; (2) petal bundles,
each forking into three some distance from its origin; (3) petalad
stamen bundles, close above the petal bundles; (4) sepalad stamen
bundles, some distance above the sepal bundles; (5) carpel dor-
sals, in the radial planes of the petals, in a horizontal plane close
above the two whorls of sepal bundles; (6) carpel laterals, at a
considerably higher level: each of these belongs, it is clear, jointly
to two adjacent carpels. At levels between the carpel dorsals
and the carpel laterals, sporadic small bundles run out into the
disk. Columnar placental bundles ascend the central column of
the ovary beyond the carpel laterals, reaching approximately the
level at which the central column is broken up by radial clefts.
There each placental bundle splits into two which depart radially
to supply two half-placentae lying in different locules but attached
to the same septum. These ultimate bundles which supply the
placentae are carpel ventrals, and the description just given
means that each carpel ventral is fused in the central column with
that of the adjacent carpel, and, lower down, with the carpel
laterals of both. The style is supplied by the carpel dorsals,
which dip under the groove about its base and ascend it in the
tissue between the flanges. The carpel laterals ascend the ovary
wall in the planes of the septa and fade out before reaching the
base of the style. .

The gaps above the petal bundles are rather profound and
divide the vascular tissue which ascends past them into five almost
distinct parts. Petalad stamen bundles and carpel dorsals spring
often from one side of a gap instead of being formed by strands
from both sides. Except in this respect, and in the fusion of the
lateral and ventral bundles of adjacent carpels, the vascular sys-
tem just described is almost precisely the theoretically ideal or
primitive vascular system of a pentamerous pentacyclic flower.

In Chimaphila (pl. 14, figs. 47, 48), the vascular supply of each
sepal and petal is normally of three bundles. The origin of these
bundles is quite variable. Each may spring from a separate gap,
or more than one from a single gap; or an originally single bundle
may split into two or three, supplying the same organ or different
ones; frequently rather than usually, bundles both of a sepal and
a petal may be of the same origin. Above the median bundles of
the petals, the ascending tissue is more or less definitely divided
into five bands in the planes of the sepals. From these bands
spring the following whorls of bundles in fairly regular fashion:
petalad stamen bundles, sometimes from the petal bundles but
characteristically from paired branches from the margins of ad-

90 MADRONO [Vol.9

jacent bands; sepalad stamen bundles, from the faces of the
bands; carpel dorsals from the margins of adjacent bands; and
carpel laterals from the faces of the bands. The remnants of the
bands ascend the central column of the ovary, as in Ramischia,
approximately to where it breaks up; then each of them supplies
two half-placentae in adjacent locules. Both the carpel dorsals
and the carpel laterals ascend the ovary wall, dip under the groove
about the base of the style, and ascend the latter.

To speculate as to things of which we know very little, one
would say that the genetic and physiological system which deter-
mines the receptacular vascular system in Bicornes in general has
become modified in Chimaphila; and that in becoming modified it
has lost precision. One may imagine that the tendency of evo-
lution would be to fix it in a definite pattern different from that
of the related plants; but that this has not yet happened in the
present genus.

In Pyrola, including P. minor, and in Moneses, the vascular
cylinder flares within the receptacle and emits ten large bundles
which lie in the median planes of the perianth parts but are not
the definitive bundles of the respective perianth parts (pl. 15, figs.
49, 50; pl. 16, fig. 51). Characteristically, though with many
exceptions, the petal laterals arise from the sepal dorsals and the
sepal laterals from the petal dorsals; in general, then, each petal
lateral passes diagonally above a sepal lateral, and each sepal
lateral passes diagonally below a petal lateral. Stamen bundles,
and carpel dorsals and laterals, arise in fairly regular fashion,
either from the upper sides of main perianth bundles or from the
vascular tissue which ascends above their departure. This tissue
becomes organized, finally, as five bands ascending the central
column of the ovary in the planes of the sepals.

The vascular supply of the style of Pyrola is peculiar in con-
sisting of five bundles lying in the flanges projecting into the style
passage. These bundles are upward continuations of the pla-
cental bundles beyond the level where each of these emits two
branches to half-placentae in adjacent locules. It is as though
the distal ends of the carpel laterals had lost connection with their
bases, and had established connection with the placental bundles.
This peculiarity, illustrated in P. bracteata (pl. 15, fig. 50), was
observed also in P. minor, P. virens, P. picta, P. dentata, and P.
americana.

In species of Pyrola other than P. minor, the vascular system
of the flower is bent so that it is of dorsiventral symmetry, in
keeping with the slight zygomorphy of the flower (pl. 15, fig. 50).

In Moneses the style is supplied as in Chimaphila by ten bundles,
being both the carpel dorsals and the carpel laterals.

Thus, in Pyrola and Moneses the receptacular vascular system
shows specializations, some of which, however, are not so firmly
established as to exclude many exceptions.

1947 | COPELAND: PYROLEAE 91

Puate 16. STRucTuRE oF THE PyroLeAE. Fic. 51. Moneses uniflora, model

of the vascular system in the floral receptacle, < 50.

EMBRYOGENY

The embryogeny of the Pyroleae, as of the Bicornes in general,
has been studied by Peltrisot (1904) and by Samuelsson (1913).

The anatropous ovules have a single integument of only two layers

of cells. The nucellus consists of a single ephemeral superficial

layer of cells and of a single megaspore mother cell enclosed by it.

Reduction division takes place later than in the pollen mother

cells, and results as usual in three ephemeral non-functional mega-

spores in the micropylar end, together with a single large func-
tional megaspore. This develops through normal stages into a
slender embryo sac of normal structure. The inner cell-layer of
the integument, where it lies against the embryo sac, is perceptibly
of the character of a jacket layer.

After fertilization, the triploid endosperm mother nucleus
divides once and again. A transverse cell wall is formed after
each nuclear division, with the result that the developing endo-

92 MADRONO [Vol. 9

sperm passes through a stage in which it consists of four cells in
arow. The terminal cells of this row do not enlarge considerably,
and do not divide; the two middle cells divide and produce a
globular mass of several cells. The zygote does not begin to grow
until after the endosperm has passed the four-celled stage. It
then becomes elongate and penetrates into the mass of endosperm
cells (pl. 17, fig. 52). It becomes divided into a suspensor and
a terminal cell, and the latter develops into a mass of smaller cells
enclosed in the endosperm (pl. 17, fig. 53). The embryo develops
no distinct parts, but eventually absorbs all of the endosperm
except a single layer of cells (pl. 17, fig. 54).

The external cells of the integument develop moderately thick
walls marked by elliptical pits on their inner and lateral surfaces;
their external cell walls remain thin. Along the sides of the seed,
the inner cells of the integument are absorbed by the endosperm;
in the ends of the seed, the inner integumental cells die and remain
in existence only as more or less shrivelled empty spaces. The
undivided terminal cells of the endosperm are called haustoria.
They become darkly-staining, and remain prominent for some
time, but eventually shrivel. Thus, in the ripe seed, there is a
central ellipsoid mass, consisting of an embryo covered by a single
layer of cells of endosperm. This ellipsoid body lies within a
cylindrical epidermis of cells with partially thickened walls, which
contains, for the rest, the collapsed remains of the haustoria and
the empty walls of a few other cells.

The features described are in most respects those of typical
Bicornes such as the Rhododendroideae. The present group is
distinguished by the integument of only two layers of cells; by the
failure of those cells of the endosperm which le next to the
haustoria to become converted into “plugs”; by the rudimentary
state of the embryo in the mature seed: in short, by small size,
fewness of cells, and scant differentiation in every part. Every
peculiarity of the embryogeny of the Pyroleae is shared by some
or all of the Monotropoideae.

THe Fruit. GERMINATION

The locules of the fruit are jacketed on the inner surface, as
in many other Bicornes, by about two layers of fibers constituting
the mechanical layer which effects dehiscence. As a marked dif-
ference from the Rhododendroideae, the mechanical tissue of each
locule does not act as a single body producing septicidal dehis-
cence: the adjacent fibrous layers in each septum maintain their

EXPLANATION OF THE Figures. Puate 17.

Puate 17. Srrucrure or THE Pyrorear. Fics. 52-54: developing seeds,
x 400. Fic. 52. Moneses uniflora. Fie. 53. Chimaphila Menziesii. Fic. 54.
Ramischia secunda. Fies. 55, 56: fruits, x5. Fie. 55. Chimaphila umbellata.
Fic. 56. Pyrola minor.

1947]

56 Swot

COPELAND: PYROLEAE

55

Puate 17. STRUCTURE OF THE PYROLEAE.

93

94 MADRONO [Vol. 9

connection, and the wall of each locule is ruptured in the median
radial plane, so that dehiscence is loculicidal.

Rydberg (1914) distinguished Ramischia, Eralebenia, and Pyrola.
as having capsules splitting from below upward; and Chimaphila
and Moneses as having capsules splitting from above downward.
I cannot persuade myself, from a study of herbarium specimens,
that this difference exists. Certainly, none of these plants have
capsule valves which tear loose at the base as they do in Ledum.
One rarely sees capsules in the act of splitting; herbarium speci-
mens which appear to be in this state seem usually to show
splitting both from above and from below toward the middle.

Rydberg further distinguished the three former genera as
having cobwebby hairs between the separating valves, and the two
latter as lacking these. This distinction is valid. The scant wefts
of white or tawny hairs seen between the separating valves con-
sist of torn fragments of the fibrous layer. Their absence in
Chimaphila and Moneses is a derived character, a matter of the
more definite establishment of loculicidal dehiscence by the for-
mation of a more definite line of splitting.

To the best of my knowledge, the process of germination has.
never been followed through in this group. Christoph (1921)
induced seeds of Pyrola rotundifolia to germinate, to the extent that
the micropylar end of the embryo grew forth as a root bearing a
cap. He could induce no further development. His figures leave
no doubt as to the validity of his observation, though Andres
(1929) was unable to repeat it. Firth (1920) observed, in a pot
seeded with Moneses, a body resembling a brief length of root with
secondary roots and an adventitious bud. Holm (1898) identified
certain young plants as seedlings of Chimaphila.

Velenovsky is said to have affirmed, in works which I have not
seen, that the germinating seed gives rise to an underground
cylindrical structure, neither stem nor root, but more primitive
than either, and to be known as the procaulon. I affirm that there
is in the Pyroleae no such thing as a procaulon; there is a defi-
nitely cauline rhizome, except in Moneses, whose permanent mem-
ber is a perfectly definite root. The Monotropoideae likewise
have no procaulon, but definite roots.

THE CLASSIFICATION OF THE PYROLEAE

As noted, the generally accepted classification of plants, being
that of Engler and Prantl, or, as to the Bicornes, that of Drude,
combines the Pyroleae and the Monotropoideae in a separate
family Pyrolaceae which is listed before most of the other families
of the order, as if primitive. Andres, the leading authority of the
Pyroleae, has accepted this arrangement. My own observations
as just set forth, together with previous studies of the Mono-
tropoideae (1941), have led me to quite other conclusions. These
are in general the same as those of Henderson (1919).

1947] COPELAND: PYROLEAE 95

The Pyroleae, many of the Monotropoideae, and a third group,
the Clethraceae, have choripetalous flowers and loculicidally de-
hiscent fruits. All three were placed as primitive Bicornes on the
basis of the floral character, the character of the fruit being over-
looked. The further characters of the Clethraceae—that they are
woody plants, with a true endothecium in the anthers and primi-
tive features in the stigma—substantiate this placement of this
group.

Choripetalous flowers are not, however, necessarily a primitive
character among Bicornes. Costerus and Smith (1916) and Camp
and Gilly (1943) have shown that choripetaly occurs occasionally
in Vaccinium as a teratological phenomenon, presumably by a
mutation. I am convinced that the choripetalous flowers of Hy-
popitys and Monotropa are derived from a sympetalous ancestry
represented among living plants by Monotropsis, and that the chori-
petalous flowers of Ledum represent a deviation from the sym-
petalous flowers of Rhododendron.

If among the Bicornes choripetalous flowers are frequently
derived from sympetalous, we may consider the possibility of
associating the Pyroleae and Monotropoideae with that group of
Bicornes which is particularly characterized by loculicidal cap-
sules, namely the tribe Andromedeae. The Andromedeae have
not been subjected to a thorough survey of the microscopic charac-
ters, and I am not able to show that they agree with the Pyroleae
and Monotropoideae in any of those striking positive details which
are accepted as convincing evidence of relationship. Nothing in
the gross characters of the groups contradicts it. Mention was
made above of the characteristic dehiscence mechanism of the
anthers of the Arbuteae, a group related to, and presumably
derived from, the Andromedeae. This mechanism appears quite
probably to be related to the corresponding mechanism of the
Pyroleae, as though it were a specialized derivative of the same
original type.

Henderson was inclined to regard the Pyroleae as represent-
ing a stage in the evolution of the Monotropoideae (she was aware
that some of the genera listed in the latter group are quite iso-
lated, as though of independent origin). The most striking com-
mon characters of the Pyroleae and Monotropoideae are the
peculiar placentation, axile in the lower half of the ovary and
parietal in the upper, and the numerous, minute, and delicately
constructed ovules and seeds. These characters are much more
readily interpreted as derived than as primitive, and one must
recognize the possibility that they have had a repeated independ-
ent origin. In fact, the Pyroleae and Monotropoideae exhibit
only a general similarity: there are no striking identities in posi-
tive details. The genera demonstrably related to Monotropa are
marked by a pairing of the lobes of the disk, by petals with saccate
bases, and by obsolescence of the petal dorsal bundles. Nothing

96 MADRONO [Vol.9

of these characters occurs in the Pyroleae; these characters can
be traced back to Monotropsis, a genus with sympetalous flowers,
obviously not descended from the Pyroleae. So, likewise, Ptero-
spora and Sarcodes have sympetalous flowers, and the latter genus
has ovules with an integument of several layers of cells. As to
Allotropa alone among the accepted genera of Monotropoideae, a
derivation from the Pyroleae is a reasonable possibility; and even
as to this genus, an independent origin it at least equally probable.

The Pyroleae and Monotropoideae are accordingly not to be
combined in one group. I abide by my former suggestion, that
the Monotropoideae be construed as a subfamily of Ericaceae and
placed after subfamily Arbutoideae, as presumed descendants of
Andromedeae. The strong probability that the group includes
more than one line of descent from Andromedeae must be kept
in mind. The group is maintained pending further knowledge,
for the accommodation of various genera, Allotropa, Sarcodes,
Pterospora, and Pleuricospora, whose detailed relationships remain
obscure.

The Pyroleae are to be construed as a rather inconsiderable
but thoroughly natural tribe of Ericaceae, to be placed in sub-
family Arbutoideae after tribe Andromedeae as presumably de-
rived from the latter.

We have seen that five genera have been distinguished within
the original genus Pyrola of Linnaeus. Among these, Chimaphila
(Pyrola umbellata and P. maculata of Linnaeus) and Ramischia (Py-
rola secunda L.) are quite decidedly distinct. Whether Eralebenia
(Pyrola minor L.) and Moneses (Pyrola uniflora L.) are tenable is
more questionable. Andres has decided to maintain the latter but
not the former. This decision is tenable, P. minor being distinct
merely in the fact that the flowers are not definitely zygomorphic,
whereas Moneses has a different type of underground structure as
well as a different inflorescence. Accordingly, I maintain Andres’
list of genera.

It appears that Ramischia exhibits many primitive characters,
and may be interpreted as a moderate modification of what one
would postulate as the original form of the group. Chimaphila
likewise exhibits primitive characters, but is in other respects
specialized, more highly so than Ramischia, and in different fea-
tures. I would differ from Andres by listing it second rather than
last among the genera of the tribe. Pyrola minor represents the
transition from our hypothetical original form to the typical spe-
cies of Pyrola. Moneses appears to be a derivative of something
much like Pyrola minor.

The following is the taxonomic system of the Pyroleae which
best represents the natural system of the group as far as I am able
to infer it:

Order Bicornes L. Gen. Pl. ed. 6 (1764).
Family Ericaceae de Candolle in Lamarck and de 'Candolle,

1947] COPELAND: PYROLEAE 97

Fl. Franc. ed. 2, 3: 675 (1805).

Subfamily Arbutoideae Drude in Engler and Prantl, Nat.
Pflanzenfam. 4 (1): 82 (1889).

Tribe Pyroleae (Lindley) DC. Prodr. 7: 772 (1839).° Plants
essentially perennial herbs with permanent underground parts,
mostly with green leaves; stems without pericycle or a continuous
layer of cork; the inflorescences racemes or reduced racemes, the
peduncle with a pericycle, bracts present, bractlets absent; flowers
choripetalous; anthers without horns, with brief tubes or none,
opening through pores actually basal and external, appearing by
inversion of the anthers to be terminal and internal, the lobes lined
near the pores with a hypodermis of cells with reticulately thick-
ened walls; ovary with the placentation axile below, parietal
above; ovules minute and numerous, with an integument of two
layers of cells; embryo without differentiated parts; fruit a locu-
licidal capsule.

I. Disk present. Dehiscence of anthers effected, at least in part,
by resorption tissue. Permanent underground structure of
the plant a rhizome.

1. Ramiscuia Opiz. Plants with ovate, finely-toothed leaves
lacking palisade tissue; inflorescence racemose, the vascular sup-
ply of the flowers originating in the same manner as that of
axillary buds; flowers small; each perianth part supplied by a
separate bundle which forks into three; sepals ciliate; petals con-
cave, ascending, each with two ventral basal tubercles, collec-
tively forming an ovate corolla; anthers sparsely papillose, the
epidermal cell walls with reticulate thickenings, tubes essentially
absent, pores elliptic, opening to some distance from the surface
by resorption tissue, for the rest by collapse tissue; style elongate,
traversed by the five carpel dorsal bundles; stigma without a
collar; capsule valves connected by cobwebby hairs.

Ramischia secunda (L.) Garcke (R. secundiflora Opiz), occurring
around the world in the northern part of the north temperate
zone, is perhaps the only tenable species. I have been unable to
locate a formal description of R. truncata Andres.

2. CuimaPuHita Pursh, Fl. Am. Sept. 1: 279 (1814). Leaves
ovate to oblanceolate, serrate, with palisade tissue; inflorescence
a condensed raceme, the bracts adnate to the pedicels, each bract
supplied by a single bundle springing from the stele in the pedicel;
flowers larger than those of Ramischia; each perianth segment
supplied by three bundles, the origin of which is quite variable;

3 Some botanists declare a new combination upon transferring a subfamily
to a different family or a tribe to a different subfamily or family. This practice
is unsound because the names of subfamilies and tribes are not combinations at
all. De Candolle first applied to the present group a name in -eae as that of a
tribe; and he is authority for this name as that of a tribe in whatever subfamily
or family it may be included.

98 MADRONO [Vol.9

sepals finely dentate; petals spreading, so that the corolla is
saucer-shaped; anthers obscurely papillose, the epidermal cell
walls not thickened, the tubes brief, tapering, opening from the
pores to the tapetum by resorption tissue; style brief, obconic,
supplied by both the carpel dorsal bundles and the carpel laterals;
stigma dome-like with an obscure collar; capsule valves without
cobwebby hairs.

There are four generally recognized species, Chimaphila ja-
ponica Miquel in Japan, C. Menziesi Sprengel in western North
America, C. maculata (L.) Pursh in eastern North America, and
C. umbellata (L.) Barton general in the northern north temperate
zone. The latter occurs respectively in Europe, in eastern North
America, and in western North America as obscurely distinguish-
able races which have been treated as species but are scarcely
tenable as such. Camp (1939) has reduced a number of other
proposed species; whether any further species, as C. dominguensis
Blake in Jour. of Bot. 52: 169 (1914), are tenable, I do not know.

II. Disk absent; anthers without resorption tissue, their epi-
dermis not papillose, its cell walls not thickened. Perianth
supplied by ten bundles, each typically forked into three
which are respectively the dorsal bundle of one perianth
part and the laterals of the two adjacent ones. Style
elongate, stigma with a collar.

3. Pyrota L. Sp. Pl. 396 (1753). Underground permanent
member of the plant a rhizome; flowers racemose, the bract sub-
tending each supplied by three bundles from the stele in the
pedicel; sepals glabrous, entire; style supplied by five bundles
springing from the placental bundles; capsule valves connected
by cobwebby hairs.

This genus is widely distributed in the north temperate zone,
with outliers in Sumatra and Mexico. There are rich arrays of
distinguishable forms in northern North America and in eastern
Asia, and the list of tenable species may exceed forty. Among
these, a small minority typified by P. minor are distinguished by
essentially actinomorphic flowers. The majority, with zygo-
morphic flowers, were subdivided by Alefeld with emphasis on
the shape of the anther pores; by Andres, with emphasis on the
shape of the sepals; and by Rydberg with emphasis on the color
of the flowers. The following arrangement is merely tentative.
In recognizing three sections, it is a simplification of that of

Andres.

A. Flowers essentially actinomorphic. ©

Section 1. Amelia (Alefeld) Bentham and Hooker, Gen. Pl.
2: 603 (1876). Andres cites Hooker as authority for this group
as a subgenus; but in the Genera Plantarum the name is applied
definitely to a section.

1947] COPELAND: PYROLEAE 99

Pyrola minor L. and various races in eastern Asia, together,
probably, with P. media Swartz.

B. Flowers definitely zygomorphic.
1. Sepals essentially ovate, scarcely longer than broad.

Section 2. Scotophila Nuttall in Trans. Am. Phil. Soc. n. s.
8: 271 (1843), based on P. aphylla. Section Ampliosepala Andres.
Often producing aphyllous forms.

a. Leaf blades if present bright green, containing one
layer of palisade tissue or none.

Pyrola virens Schweigger (P. chlorantha Swartz), supposedly
around the world, though with variations; P. renifolia Maximowicz
Prim. Fl. Amur. 190 (1859), and other species in eastern Asia;
P. elliptica Nuttall in eastern North America.

b. Leaf blades if present dull green or with dull green
mottling, containing two layers of palisade tissue.

Pyrola picta Smith, P. dentata Smith, and variations (P. aphylla

Smith is treated as one of these), in western North America.

Alefeld combined the two species here accepted under the name

of Thelaia spathulata. His action was justified by the practice of

the times, but Andres was not justified in using Alefeld’s epithet
under Pyrola.

2. Sepals essentially lanceolate, distinctly longer
than broad. Leaves without palisade tissue.

Section 3. Thelaia (Alefeld) Bentham and Hooker, l. c.
Subgenus Euthelaia Alefeld, in part; section Euthelaia (Alefeld)
Andres. In the usage of everyone except Alefeld, this is the
type group of the genus Pyrola.

a. Flowers white.
Pyrola rotundifolia L.; P. americana Sweet, scarcely distinguish-
able from the foregoing; other races in Canada and eastern Asia.

b. Flowers with more or less considerable red pigment.

The first species to be distinguished in this group was P. asari-
folia Michaux, Fl. Bor. Am. 1: 251 (1803). As to P. uliginosa
Torrey and Gray in Torrey Fl. New York 1: 453 (1843), Fernald
(1904) has found it to intergrade with the preceding and has
reduced it; Rydberg, on the other hand, has maintained it and has
reduced to it P. elata Nuttall in Trans. Am. Phil. Soc. n. s. 8: 270
(1843), a perfectly definite race in western North America. I
have applied the name in Rydberg’s sense: the plant here called
P. uliginosa is not positively representative of that species; it is
positively representative of P. elata Nuttall. Pyrola bracteata
Hooker, Fl. Bor, Am. 2: 47 (1834) is a definite though not pro-
foundly distinct race in western North America. Further races
of this group occur in Mexico and in eastern Asia; the oldest name
for the latter is P. incarnata Fischer apud DC. Prodr. 7: 773
(1839).

100 MADRONO [Vol.9

4. Monesss Salisbury ex S. F. Gray. Permanent underground
structure a root; leaves without palisade tissue; flowers solitary,
terminal on an extended peduncle; sepals ciliate ; stamens inclined
toward the upper side of the flower, as in the species of Pyrola
with zygomorphic flowers; anthers with the tubes curved, diverg-
ing and converging; style supplied both by carpel dorsal bundles
and by carpel laterals; capsule valves without cobwebby hairs.

Moneses uniflora (L.) A. Gray, around the world in northern
regions, is presumably the only species. Moneses reticulata Nuttall
in Trans. Am. Phil. Soc. n. s. 8: 271 (1843), described from west-
ern North America, was reduced by Piper (1906) to complete
synonymy and restored as a variety by Blake (1915); it is very
feebly if at all distinct.

SUMMARY

The Pyroleae and the Monotropoideae agree in having chori-
petalous flowers, parietal placentation in the upper part of the
ovary, the ovules and seeds numerous, minute, and delicately con-
structed, and septicidally dehiscent capsules (actually, most of
these characters extend only to a part of the Monotropoideae).
All of these characters appear to be derived, not primitive, and
the two groups appear not properly to be united into one; rather,
they represent parallel lines of descent from a common origin,
presumably the tribe Andromedeae.

The Pyroleae are an undoubtedly natural small group best
construed as a tribe of Ericaceae to be placed in subfamily
Arbutoideae after tribe Andromedeae.

It is expedient to recognize four genera, as Andres did. Ra-
mischia (Pyrola secunda) is definitely distinct from Pyrola proper.
It is the most primitive genus of the group. Chimaphila is also
primitive in most respects. These genera are distinguished by a
disk below the ovary and by the presence of resorption tissue in
the anthers. Pyrola minor has actinomorphic flowers but agrees
in all other respects with Pyrola proper, and may properly be left
among the moderately numerous species of that genus. Moneses,
with shoots springing from roots instead of from rhizomes, soli-
tary flowers, and styles supplied by ten bundles, is tenable as a
distinct genus.

Sacramento College,
Sacramento, California.

LITERATURE CITED

ALEFELD, F, 1856. Ueber die Familie der Pyrolaceen, insbesondere die Unter-
familie der Pyroleen (Gen. Pyrola L.). Linnaea 28: 1-88.
Awnopres, H. 1909. Die Pirolaceen des rheinischen Schiefergebirges, der angre-
zender Tieflander des Rheins und des mainzer Beckens. Verhandl.
naturh. Ver. preussischen Rheinlande u. Westfalens 66: 99-151.
. 1910. Die Pirolaceae des aschersonischen Herbariums. Abh.
bot. Ver. Prov. Brandenburg 52: 90-95.
. 1910. Beitrage zur Pirolaceen-Flora Asiens. Deutsche bot.
Monatschr. 22: 4-7, 18-22.

1947] COPELAND: PYROLEAE 101

1912. Zwei neue Pirolaceae aus der Subsection Erxlebenia
(Opiz) H. Andres nebst einigen Bemerkungen zur Systematik der
heimischen Arten. Verh. bot. Ver. Prov. Brandenburg 54: 218-227.
1912. Zusiétze und Verbesserungen zur “Monographie der rheini-
schen Pirolaceae.” Sitzungsber. naturh. Ver. preussischen Rheinlande
u. Westfalens 1911 E: 6-10.
1913. Zusiatze und Verbesserungen zur “Monographie der rheini-
schen Pirolaceen” II Teil. Sitzungsber. naturh. Ver. preussischen
Rheinlande u. Westfalens 1912 E: 70-92.
1913. Zusétze und Verbesserungen zur “Monographie der rheini-
schen Pirolaceae” III Teil. Berichte bot. zool. Ver. Rheinland-West-
falen. 83-91.
1913. Pictoides H. Andres, eine neue Subsektion der Eu-
Thelaia-Gruppe aus dem Genus Pirola Salisb. Osterreichische bot. Zeit.
63: 68-75.
1913-1914. Studien zur speciellen Systematik der Pirolaceae.
Osterreichische bot. Zeit. 63: 445-450; 64: 45-50, 232-254.
1914. Piroleen-Studien. Beitrige zur Kenntnis der Morphologie,
Phytogeographie und allgemeinen Systematik der Pirolaceae. Verh.
bot. Ver. Prov. Brandenburg 56: 1—76.
1929. Weitere Zusitze zur “Monographie der rheinischen Pirola-
ceae.” Sitzungsber. bot. zool. Ver. Rheinland-Westfalen 1928 D: 36-46.
1931. Herbarstudien zur bulgarischen Flora. I. Pirolaceae.
Mitth. kgl. naturw. Inst. Sofia 4: 61-64.
1936. Pirolaceae. In H. Handel-Mazetti. Symbolae Sinicae 7:
762-768.
1936. Pirola sumatrana, species nova. Bull. Jard. Bot. Buiten-
zorg 3e sér. 14: 4-7.
Artoproeus, A. 1903. Uber den Bau und die Offnungsweise der Antheren und
die Entwicklung der Samen der Erikaceen. Flora 92: 309-345.
BENTHAM, G., and J. D. Hooker. 1862-1883. Genera Plantarum. ... 3 vols.
London.

Buiake, S. F. 1914. A new Chimaphila from San Domingo. Jour. Bot. 52: 169.

———_———. 1915. Moneses uniflora var. reticulata. Rhodora 17: 28-29.

Camp, W.H. 1939. Studies in the Ericales. IV. Notes on Chimaphila, Gaul-
theria, and Pernettya in Mexico and adjacent regions. Bull. Torrey
Bot. Club 66: 7-28.

. 1940. Aphyllous forms of Pyrola. Bull. Torrey Bot. Club 67:
453-465.

and C. L. Gritty. 1943. Polypetalous forms of Vaccinium. Tor-

reya 42: 168-173.

DE CANDOLLE, A. P. 1839. Pyrolaceae. In de Candolle. Prodr. 7: 772-776.

CuristopH, H. 1921. Untersuchungen iiber die mykotrophen Verhiltnisse der
“Hricales” und die Keimung der Pyrolaceae. Bot. Centralbl. Beih. 38
(Abt. 1): 115-157.

CoreLtann, H. F. 1941. Further studies on Monotropoideae. Madrofo 6:
97-119.

. 1944. A study, anatomical and taxonomic, of the genera of

Rhododendroideae. Am. Midland Nat. 30: 533-625.

Costerus, J. C, and J. J. Smiru. 1916. Studies on tropical teratology. Ann.
Jard. Bot. Buitenzorg 2e sér. 14: 83-94.

Domin, K. 1915. Vergleichende Studien iiber den Fichtenspargel. Sitzungsber.
kgl. Bohm. Gess. Wiss. II Cl., I Stiick: 1-111.

Doyet, B. E. 1942. Some features of the structure of Arctostaphylos viscida.
Am. Jour. Bot. 29: 254-259.

Drupr, O. 1889. Pirolaceae. In Engler und Prantl Nat. Pflanzenfam. 4
(I Abt.) : 3-11.

Fernatp, M. L. 1904. Pyrola asarifolia Michx., var. incarnata, n. comb.
Rhodora 6: 178-179.

102 MADRONO [Vol. 9

. 1920. The American varieties of Pyrola chlorantha. Rhodora
22: 49-53.
. 1941. Transfers in Pyrola. Rhodora 43: 167.
Firtu, P. 1920. Zur Biologie und Mikrochemie einiger Pyrola-Arten. Sitz-
ungsber. Akad. Wiss. Wien Math.-nat. K]. Abt. I 129: 559-588.

Gray, A. 1846. Chloris Boreali-Americana. ... Mem. Am. Acad. 3: 1-56.
———. 1848. A manual of the botany of the northern United States. . .
Boston.

Hacerup, O. 1928. Morphological and cytological studies of the Bicornes.
Dansk Bot. Arkiv 6 (part 1): 1-26.

Henverson, M. W. 1919. A comparative study of the structure and sapro-
phytism of the Pyrolaceae and Monotropaceae with reference to their
derivation from the Ericaceae. Contrib. Bot. Lab. Univ. Pennsylvania

5: 42-109.
Horm, T. 1898. Pyrola aphylla: a morphological study. Bot. Gaz. 25: 246-254.
Hooxer, W. J. 1829-1840. Flora Boreali-Americana.... 2 vols. London.

IrmiscH, T. 1855. Bemerkungen iiber einige Pflanzen der deutschen Flora.

Flora 38: 625-638.
. 1856. Einige Bemerkungen iiber die einheimische Pyrola-Arten.
Bot. Zeit. 14: 585-591, 601-606.
1859. Kurze Mittheilungen iiber einige Pyrolaceen. Flora 42:
497-501.

Jussteu, A. L. 1789. Genera plantarum secundum ordines naturales disposita.
Paris.

Kuorzscn, J. F. 1851. Studien iiber die natiirliche Klasse Bicornes Linné.
Linnaea 24: 1-88.

pE Lamarck, J. B., and A. P. pe CAanpoutie. 1805. Flore Francaise. 2d ed.
5 vols. Paris.

Linnaeus, C. 1753. Species plantarum. 2 vols. Stockholm.

. 1764. Genera plantarum. 6th ed. Stockholm.

MatruHews, J. R., and KE. M. Knox. 1926. The comparative morphology of the
stamen in the Ericaceae. Trans. and Proc. Roy. Soc. Edinburgh 29:
243-281.

Maximowicz, C. J. 1859. Primitiae florae Amurensis. St. Petersburg.

MicHavux, A. 1803. Flora Boreali-Americana. ... 2 vols. Paris.

Nourtatt, T. 1843. Description and notices of new or rare plants in the natu-
ral orders Lobeliaceae, Campanulaceae, Vacciniaceae, Ericaceae... .
Trans. Am. Phil. Soc. n. s. 8: 251-272.

Pease, V. A. 1917. Duration of leaves in evergreens. Am. Jour. Bot. 4:
145-160.

Petrrisot, C. N. 1904. Développement et structure de la graine chez les
Ericacéees. Jour. de Bot. 18: 309-367, 386-402.

Pirer, C. V. 1906. Flora of the State of Washington. Contrib. U. S. Nat.
Herb. 11: 1-367.

Pursn, F. 1814. Flora Americae Septentrionalis. ... 2 vols. London.

Raptivus, J. 1821. De Pyrola et Chimophila. . . . Leipzig.

Roeper, J. 1852. Normales und Abnormes. Bot. Zeit. 10: 425-434, 441-448,
457-464,

RypzerG, P. A. 1914. Pyrolaceae. In North American Flora 29: 21-32.

SAMUELSSON, G. 1913. Studien iiber die Entwicklungsgeschichte der Bliite
einiger Bicornes-Typen. ... Svensk Bot. Tidskr. 7: 97-188.

Torrey, J. 1843. A flora of the state of New York (Natural history of New
York, Part II). 2 vols. Albany.

Wyoter, H. 1860. Kleinere Beitrage zur Kenntnis einheimischer Gewichse.
Flora 43: 613-617.

1947] HEISER: BLENNOSPERMA 103

A NEW SPECIES OF BLENNOSPERMA FROM
CALIFORNIA

Cuartes B. HEIser, JR.

Blennosperma Bakeri sp. nov. Herba annua, aliquid succu-
lenta, 15-30 cm. alta; foliis inferioribus integris vel 3-lobatis, ad -
15 cm. longis, foliis superioribus 3-5 (raro 7-) lobatis, lobis ca.
1 mm. latis; involucri lobis 6—8, 6-8 mm. longis, 3—4 mm. latis, non
reflexis; ligulis 12-14, 5-7 mm. longis, 2-3 mm. latis, luteis supra
et saepe fulvis infra; stigmatis ramis rubris; disci floribus 35-50;
achaeniis 8-4 mm. longis, 1-2 mm. latis, 4-6 angulatis.

Fic. 1. California species of Blennosperma: A, B. Bakeri; B, B. cali-
fornicum. (Growing in four-inch pots.)

Annual herb, somewhat succulent, 15-30 cm. tall; lowermost
leaves entire or 3-lobed up to 15 ecm. long; upper leaves 3-5
(rarely 7-) lobed, lobes about 1 mm. wide; lobes of the involucre
6-8, 6-8 mm. long, 3-4 mm. wide, curved over fruits at maturity,
not at all reflexed; ligules 12-14, 5-7 mm. long, 2-3 mm. wide,
yellow above and somewhat brown below, drying pinkish at times;
branches of the stigma red, dark purple on drying; disk-flowers
35-50; achenes 3—4 mm. long, 1-2 mm. wide, 4—6 angled.

Type. Western outskirts of Sonoma in “hog wallow” about
0.25 mile south of Napa Street in field on east side of street,
Sonoma County, California, April 2, 1946, M. S. Baker 11307 (Her-

104 MADRONO [Vol. 9

barium of the University of California no. 725276; isotypes are
to be widely distributed).

The genus Blennosperma was previously considered to consist
of only two species: our local B. californicum Torr. and Gray
[B. nanum (Hook.) Blake] and the South American B. chilense
Less. A new species in this genus, then, is of considerable inter-
est. It is a great pleasure to name it after its collector, Milo S.
Baker, who has made many contributions to our knowledge of the
flora of California.

The new species is

readily distinguished from

B. californicum first by its

much larger size (text fig.

1). In addition to this the

leaves of B. Bakeri are

generally three-parted

A B into large lobes, whereas

those of B. californicum are

Fic. 2. Chromosomes of Blennosperma: more finely divided. The

A, B. Bakeri; B, B. californicum. X 1200. branches of the stigma are

red in B. Bakeri, yellow in B. californicum. The achenes of B. cali-

fornicum are, as a rule, less conspicuously angled. Mr. Baker has

also called my attention to another difference which he noted dur-

ing his field observations. The involucral bracts of B. californicum

are reflexed at maturity and those of B. Bakeri are curved over the

mature achenes. In all probability there are also differences in

ecological preferences between the two species inasmuch as B.

Bakeri was found growing in standing water in a low marshy

pasture and B. californicum commonly grows on moist hillsides.

The new species because of its apparently restricted distribution

and its occurrence in aquatic habitats is another of the vernal pool

endemics of California alluded to by Mason (Madrofio 8: 241-257.
1946).

Cytological investigation of Blennosperma Bakeri revealed the
haploid chromosome number from the microsporocytes to be nine
(text fig. 2, A). Dr. G. L. Stebbins, Jr. (oral communication) has
found that B. californicum has the haploid chromosome number of
seven and the diploid number fourteen. This haploid number for
B. californicum has also been obtained by the writer (text fig. 2, B).
Specimens on which these chromosome numbers are based are to
be deposited in the Herbarium of the University of California.
No filled achenes have resulted from the attempts to cross the two
species. No chromosome number has as yet been reported for
B. chilense which, on the basis of morphology, is closely related

to, if not conspecific with, B. californicum.
College of Agriculture,
University of California, Davis.
Division of Botany,

MADRONO
A West American Journal of Botany

A quarterly journal devoted to important
and stimulating articles dealing with
plant morphology, physiology, taxonomy,
and botanical history. These volumes
should be a part of every botanist’s li-
brary and should be made accessible to
students of all universities and colleges.

Volume I, 1916-1929. . . $5.00
Volume II, 1930-1934 .. 5.00
Volume III, 1935-1936 . 5.00
Volume IV, 1937-1938 . 5.00
Volume V, 1939-1940 . . 5.00
Volume VI, 1941-1942 . 5.00
Volume VII, 1943-1944 . 5.00
Volume VIII, 1945-1946 . 5.00
Single numbers’... .... 0.75

The subscription price of MADRONO
is $2.50 per year. We solicit your pat-
ronage,

Address all orders to:
Reed C. Rollins, Bus. Mer.

Natural History Museum,
Stanford University, California

As

&

i

‘ ;
Laie

|
;
|

|
{
}

VOLUME IX NUMBER 4

MADRONO

A WEST AMERICAN JOURNAL OF
BOTANY

Contents

Tue Potyconat GRAPH FOR SIMULTANEOUS PoRTRAYAL OF SEVERAL VARIABLES

in Porputation Anatysis, John F. Davidson ......................... 105
MarTuRATION OF THE GAMETES AND FERTILIzATION IN Nicotiana, 7. H. Good-

STA ahs. EN a as ig Se MEMES Aa FEW Ge ae AE ny Ae ea om mY 110
NoMENCLATORIAL CHANGES IN ELymus witH a Key To THE CALIFORNIAN

SPECIES NET OI IG Vi. (GOULGN ys Chee Lg ee Shc. ote da loons al tke te Peale Uae 120
Two New Varieties or Conpatia From Texas, V. DL. Cory ................ 128
A New Vioter rrom Mexico, Milo S. Baker ............. 0... 0c. c cee eee 131
A New Spectes or Oxyrropis rrom THE CEN'rRAL Rocky Mountains, C. L.

TEP EAE SLT EIEN GAGA NPA iy SOROS st A URN er 133

Review: Maximino Martinez, Los Juniperus Mexicanos (Ira L. Wiggins) .. 135

Published at North Queen Street and McGovern Avenue,
Lancaster, Pennsylvania

October, 1947

MADRONO

A WEST AMERICAN JOURNAL OF BOTANY

Board of Editors

Hersert L. Mason, University of California, Berkeley, Chairman.

LeRoy Axsramgs, Stanford University, California.

Epcar AnpbeRson, Missouri Botanical Garden, St. Louis.

Lyman Benson, Pomona College, Claremont, California.

Herzert F’. Copeuanp, Sacramento Junior College, Sacramento, California.
Ivan M. Jounston, Arnold Arboretum, Jamaica Plain, Massachusetts.
Miuprep E. Matutas, 2851 North Lake Avenue, Altadena, California.
Bassert Macuire, New York Botanical Garden, N. Y. C.

Marion Ownsey, State College of Washington, Pullman.

Secretary, Editorial Board—ANNeETTA CARTER
Department of Botany, University of California, Berkeley

Business Manager—Reep C. RoLiins
North Queen Street and McGovern Avenue, Lancaster, Pennsylvania
or
Natural History Museum,
Stanford University, California

Entered as second-class matter October 1, 1935, at the post office at
Lancaster, Pa., under the act of March 3, 1879.

Established 1916. Published quarterly. Subscription Price $2.50 per year.
Completed volumes I to VII inclusive, $35.00; each volume $5.00; single numbers
$0.75.

Papers up to 15 or 20 pages are acceptable. Longer contributions may

be accepted if the excess costs of printing and illustration are borne by the
contributor. Range extensions and similar notes will be published in con-
densed form with a suitable title under the general heading “Notes and News.”
Articles may be submitted to any member of the editorial board. Manuscripts ~
may be included in the forthcoming issue provided that the contributor pay the
cost of the pages added to the issue to accommodate his article. Reprints of —

any article are furnished at a cost of 4 pages, 50 copies $4.10; 100 copies $4.50;

additional 100’s $0.85; 8 pages, 50 copies $5.95; 100 copies $6.60; additional 100’s
$1.30; 16 pages, 50 copies $8.35; 100 copies $9.35; additional 100’s $2.00. Covers, —

50 for $2.75; additional covers at $1.65 per hundred. Reprints should be
ordered when proofs are returned.

Published at North Queen Street and McGovern Avenue, Lancaster,
Pennsylvania, for the

CALIFORNIA BOTANICAL SOCIETY, INC.

President: Howard E. McMinn, Mills College, California. First Vice-
President: Milo S. Baker, Santa Rosa Junior College, California. Second
Vice-President: David D. Keck, Carnegie Institution of Washington, Stanford
University, California. Secretary, John Whitehead, University of California
Botanical Garden, Berkeley. Treasurer, Reed C. Rollins, Natural History
Museum, Stanford University, California.

Annual membership dues of the California Botanical Society are $2.50,
which includes a year’s subscription to Madrono. For two members of the same
family the dues are $3.50, which includes one copy of Madrofio and all other
privileges for both. Dues should be remitted to the Treasurer. General
correspondence and applications for membership should be addressed to the
Secretary.

1947] DAVIDSON: POLYGONAL GRAPHS 105

THE POLYGONAL GRAPH FOR SIMULTANEOUS
PORTRAYAL OF SEVERAL VARIABLES
IN POPULATION ANALYSIS

Joun F. Davipson

Hutchinson (19386) described the polygonal graph as a means
of portraying various related data. While he used this method to
depict relationships in a few angiospermous families, no analytical
use was made of the method. A later paper (1940) was restricted
to the ecological applications of the technique, and the method
of comparing climatic zones could well be adapted to the com-
parison of plant populations (figs. 4, 5,6). The polygonal graph
may serve also as a convenient and useful taxonomic technique in
population analysis.

CALYX LENGTH

STYLE LENGTH COROLLA LOBE

COROLLA TUBE

Fic. 1. A base for the polygonal graphing of four characters, using abso-
lute measurements. The radii have been calibrated so that the expected range
of variation covers most of each radius.

In population analysis, a taxonomist is generally concerned
with determining whether a population contains two or more
segregates or merely one variable entity. This may be determined
by a fairly large series of two-dimensional graphs, such as curves,

bar graphs, or scatter graphs, or by a single polygonal analysis.

MavroXo, Vol. 9, No. 3, pp. 65-104. July 17, 1947. NOV 25 i847

106 MADRONO | [Vol.9

The advantage of polygonal graphing is that members of an
entire population may be compared in a number of varying charac-
ters at a glance. The completed figure in a polygonal graph
readily supplies the following data:

(a) the range of variation in any one of three to ten characters.

(b) the mean of any character.

(c) the occurrence of a bimodal distribution in any character.

(d) the correlation between any two of the characters plotted.

(e) the validity of a character in delimiting proposed segre-

LEAFLET WIOTH
LEAFLET LENGTH

CALYX LOBE

S———

CALYX TUBE

S
>

STAMEN INSERTION
COROLLA TUBE

COROLLA LOBE

LEAFLET NUMBER COROLLA TUBE

COROLLA LENGTH MM.

Fic. 2. A base for graphing six characters, some being absolute measure-
ments, some being relative. The ratio leaflet width over leaflet length is an
approximation of the more qualitative concept of leaflet shape. The single
specimen plotted shows the following characters:

Calyx lobe subequalling the calyx tube.

Corolla lobe 1.3 times the corolla tube.

Corolla 13 mm. long.

Stamens inserted slightly below the middle of the corolla tube.
Leaflets 19.

Leaflets 0.8 as broad as long.

Basically the graph consists of a circle, with as many radii as
there are characters to be compared (fig. 1). The characters,
measured along each radius, are assigned absolute, relative, or
arbitrary values. The characters possessed by each specimen are
plotted along each radius, and the points so plotted are joined.
Thus each specimen is represented by a polygon (fig. 2), and the
entire population is represented by the polygons plotted on the
same or similar bases.

1947] DAVIDSON: POLYGONAL GRAPHS 107

The choice of characters is important, and normally those
which are thought to possess ‘“‘keying possibilities” are chosen.
However, if many characters are used, separation of entities may
be accomplished on one or more of the criteria utilized. In prac-
tise, using a circle 10 cm. or more in radius, fifty or more speci-

WIDTH
LENGTH

LEAFLET

LEAFLET NUMBER

CALYX LOBE
CALtYx TUBE

PISTIL LENGTH
COROLLA LENGTH

'
ro LIE
JSS Se Vi

INV,

COROLLA LOBE
COROLLA TUBE

COROLLA - mm.

STAMEN INSERTION
COROLLA TUBE

Fic. 3. A mixed population has been graphed for seven characters. In the
interest of simplicity only eight specimens are shown. ‘The presence of two
entities is demonstrated by the bimodal distribution on three of the radii. The
other radii show normal distributions.
stamen insertion

corolla tube
other characters of possible diagnostic value would be selected. If these all
showed normal distributions, the conclusion could be drawn that the population
consisted of one variable entity.

If all radii showed a normal distribution, similar to that of

mens may be compared on one base sheet. In larger populations,
the use of thin paper or tracing paper makes superposition of two
or more sheets possible, provided of course, that the same base
measurement is retained. In some problems, particularly those
dealing with the correlation of color with other characters, it may
be expedient to use appropriately colored pencils to form the
polygons.

After plotting a few members of a population, random varia-
bility is commonly apparent on some of the radii representing the
selected characters. In practise it has been found convenient to

108 MADRONO [Vol.9

replace such characters with others, in order to test as many as
possible for diagnostic value.

The completed figure will show a normal frequency distribu-
tion on any radius which represents a character which varies at
random (fig. 3, corolla length, stamen insertion, pistil length).
From such radii, the customary frequency curve may be con-
structed, if desired, by utilizing the number of specimens occur-
ring in any desired interval.

LEAFLET WIDTH

CALYX TUBE
LEAFLET LENGTH
oe COROLLA LOBE
en COROLLA TUBE
NUMBER OF L\.
CAULINE LEAFLETS
STAMEN INSERTION

2S SENS ERTION!
COROLLA TUBE

NUMBER OF
BASAL LEAFLETS

Sa

RECEPTACLE To ANTHER
hrc COR OL Ag ann

Fic. 4. A synopsis of a polygonal graph of a Polemonium pulcherrimum
population from the Rocky Mountains of British Columbia and Montana. Com-
pare the range of variation in each character in this population with the corre-
sponding characters in the southern populations (figs. 5 and 6).

The occurrence of a bimodal distribution on any radius shows
the presence of two entities, and if the distributions around each
mode are mutually exclusive, the character would be valid for
keying purposes. If several diagnostic characters have been
graphed, the segregation into the two or more entities may be
seen on the radius for each character (fig. 3, leaflet, calyx, and
corolla proportions).

Positive correlation of characters is shown between adjacent
radii by essentially parallel lines, as would be shown in figure 3,
between calyx proportions and corolla proportions, were these
two radii adjacent. Negative correlation is demonstrated in fig-
ure 3 between leaflet and calyx proportions, by the intersecting
lines. The narrower leaflets show negative correlation with the

1947] DAVIDSON: POLYGONAL GRAPHS 109

short calyx lobes, or conversely, positive correlation with the long
calyx lobes. The fact that this correlation is not 100 per cent is
shown in the specimen with the longest calyx lobe proportion.
To show 100 per cent correlation, this specimen should have a
leaflet proportion value of less than 0.3.

Figures 4, 5, and 6 show the use of the polygonal graph for

CALYX LOBE

LEAFLET WIOTH

LEAFLET LENGTH
COROLLA LOBE
COROLLA TUBE
NUMBER OF
CAULINE LEAFLETS kK
NUMBER OF

BASAL LEAFLETS

STAMEN INSERTION

CORSICA TUBE

RECEPTACLE TO ANTHER
COROLLA

Fic. 5. A synopsis of a Polemonium population from the Sierra Nevada
of California which appears to be, on the basis of the graphed characters as well
as other characters that were selected, a less variable form of Polemonium
pulcherrimum.

the simultaneous comparison of seven characters in three distinct
populations. In each figure, only the minimum, mean (heavy
line) and maximum for each character has been plotted. The
similarity of polygon shape in figures 4 and 5 is evidence of close
relationship, considered here to be conspecific. Conversely, the
difference in shape between figures 4 and 6 was the basis for rejec-
tion of conspecific status for these two entities. A further point:
shown in these figures is the reduced variability in both the
southern populations as compared with the northern population.

The figures will serve to illustrate the method used, as well as
to show some of the results obtainable.

LITERATURE CITED

Hurcuinson, A. H. 1936. “The polygonal presentation of polyphase phe-
nomena.” Trans. Roy. Soc. Can., Ser. 3, Sect. V, 30: 19-26.

110 MADRONO [Vol. 9

CALYX LOBE
CALYX TUBE

CEAFLET WIOTH
LEAFLET LENGTH

COROLLA LOBE
COROLLA TUBE

NUMBER OF
CAULINE LEAFLETS

STAMEN INSERTION
COROLLA TUBE

NUMBER OF
BASAL LEAFLETS

RECEPTACLE TO ANTHER
COROLLA

Fic. 6. A synopsis of another Polemoniwm population from the Sierra
Nevada of California. The variation from P. pulcherrimum is evident in a
number of respects in the mean (heavy line), but the calyx proportion appears
to be most diagnostic.

. 1940. Polygonal graphing of ecological data. Ecology 31:
475-487.
Department of Botany,
University of California, Berkeley.

MATURATION OF THE GAMETES AND FERTILIZATION
IN NICOTIANA?

T. H. GoopsrreEep

Examination of megasporo- and megagametogenesis in a num-
ber of species of Nicotiana was undertaken with special reference
to the extent to which details in the development of the female
gametophyte might contribute evidence concerning species origins
and relationships. The investigation was later extended to de-
termination of the development, structure and behavior of the
sperms. On this latter point no detailed reports have been pub-
lished in the case of Nicotiana and relatively few references to
megasporo- or megagametogenesis appear in the literature deal-—
ing with the genus.

1 Contribution no. 118 from the University of California Botanical Garden.
Investigations aided by grants of the Committee on Research, University of
California.

1947] GOODSPEED: NICOTIANA Ill

For N. tabacum the complete sequence in megasporogenesis
and embryo sac development was investigated with comparative
studies, particularly of megagametogenesis, in a number of other
species. In N. alata early stages were emphasized, in N. sylvestris,
N. glutinosa and N. rustica 2- and 4-nucleate embryo sacs were
studied while the character of the 8-nucleate condition in the
same three species and in N. rotundifolia was compared. In addi-
tion, megasporogenesis in certain F, hybrids between more and
less distantly related species was examined in comparison with
that of the species above mentioned.

In N. tabacum the archesporial cell differentiating in the sub-
epidermal layer of the nucellus at the apex of the ovule is initially
distinguished by its large nucleus, deeply staining cytoplasm and,
later, by its increase in volume as contrasted with the surrounding
tissue. The archesporial cell becomes the megaspore mother cell
directly, without division into parietal and sporogenous cells.
Although in N. tabacum not more than one megaspore mother cell
in an ovule has been seen, in N. alata twin megaspore mother cells
(pl. 18, fig. 6), and later twin embryo sacs, each covered by its
own nucellus, sometimes occur (cf. however, Satina, Blakeslee and
Avery, 1934; Rees-Leonard, 1985; Cooper, 1943). In F, inter-
specific hybrids more than one megaspore mother cell has often
been observed and the same is true of chromosomal variants, par-
ticularly of N. tabacum, derived from treatment of sporogenous or
vegetative cells with high frequency radiation which, in addition,
produced abnormalities in organization of nucellar and other tis-
sues of the ovule.

At the time of differentiation of the megaspore mother cell in
N. tabacum the ovule is erect but becomes completely anatropous
at early meiotic stages. During this inversion epidermal cells at
a level just below the lower end of the megaspore mother cell
begin by periclinal divisions to form the single integument (pl. 18,
fig. 1). It develops rapidly, two or three layers in thickness, and
reaches the level of the apex of the nucellus before the end of
pachytene, and almost completely covers it by diakinesis-MI (pl.
nS) fio, 2).

During meiotic prophases the megaspore mother cell shows
rapid increase in size, particularly in length (pl. 18, figs. 1, 2),
while the cells of the nucellar covering become exceedingly nar-
row and elongated (pl. 18, figs. 2-4). At the same time the inner-
most layer of the integument begins differentiation to form the
integumentary tapetum. The two meiotic divisions produce a
quartet of megaspores—MII frequently occurring earlier in the
chalazal than in the micropylar member of the dyad (pl. 18,
fig. 4). More frequently the MII spindles are at right angles to
each other producing a T-shaped quartet (pl. 18, fig. 3) although
a linear quartet (pl. 18, fig. 4) is not uncommon. The chalazal
megaspore becomes the 1l-nucleate embryo sac and the other three
megaspores soon degenerate (pl. 18, fig. 5).

112 MADRONO [Vol. 9

In N. tabacum, therefore, the embryo sac is monosporic (cf.
Maheshwari, 1941), as was also reported by Modilewski (1935).
However, he found in N. glauca “‘a disporial eight-nucleate embryo
sac according to the type of Scilla,” and in Nicotiana ditagla (amphi-
diploid N. tabacum x N. glauca) that ‘‘the monosporial process of
formation of the embryo sac by means of forming a triad, occu-
pies an intermediate position between the bisporial type which is
proper to N. glauca and the monosporial type in form of a tetrad
distinctive of [N. tabacum] Dubeck.” In N. rustica, Persidski and
Modilewski (1934) report that the development of the embryo sac
“proceeds according to the disporial or Scilla type.” Although
complete analysis of the early sequence in Nicotiana glauca and
N. rustica has not been made here, it is doubtful whether two dis-
tinct types of embryo sac development occur in a single genus.
Furthermore, the illustrations and additional comments of the
above named authors suggest that the two species in question are
not typically bisporic.

Characteristic of the Tubiflorae, nucellar degeneration begins
during the period of the extremely rapid growth in volume of the
increasingly vacuolate 1-nucleate embryo sac while the remnants
of the three non-functional megaspores are still present (pl. 18,
fig. 5). This degeneration proceeds rapidly so that little indica-
tion of nucellar tissue is seen at the 2- and 4-nucleate stages, and
almost none at the 8-nucleate stage (pl. 18, fig. 7).

Following the establishment of the 8-nucleate condition, one
each of the four nuclei at the antipodal and at the micropylar end
moves toward the center of the embryo sac. During this period
the egg apparatus and the antipodal cells are matured. The
former consists of triangular or pear-shaped synergids provided
with basal vacuoles and a somewhat more spherical egg which
continues to enlarge as a result of increase in size of an upper
vacuole.

While in N. tabacum and N. glutinosa the mature embryo sac
is pointed at the micropylar end and rounded at the chalazal end

EXPLANATION OF THE Ficures. PLatTe 18

Piate 18. Somer STAGEs In MEGASPOROGENESIS AND EMBRYO SAc ForMATION
in Nicotiana. Figs. 1-4. Megasporogenesis in Nicotiana tabacum. Fig. 1.
Megaspore mother cell, early prophase; origin of integument. Fig. 2. Same,
diakinesis. Fig. 3. T-shaped quartet of megaspores. Fig. 4. Linear quartet;
chalazal megaspores already formed, MII in upper dyad cell. Fig. 5. WN. taba-
cum. 1-nucleate embryo sac; degeneration of other three megaspores advanced,
of nucellus beginning. 375. Fig. 6. MN. alata. Twin nucelli, each with a mega-
spore mother cell. Fig. 7. NW. tabacum. Mature embryo sac, normal organi-
zation of 8-nucleate condition, remnants of nucellus, chalazal end rounded.
xX375. Fig. 8. MN. rotundifolia. Same, chalazal end pocketed. 375. Fig. 9.
F, NV. tabacum x N. glauca. Dyad degenerating; enlargement of cells of nucellar
epidermis within integumentary tapetum (cf. fig. 4). All figures drawn with
the camera lucida by a special carbon pencil technique from paraffin sections
of ovules—longitudinal; reproduced X 290 unless otherwise indicated.

Empryo Sac KorMATION

AND

rE NESIS

POROG

N Mercas

I

ES

TAC

Somr S

Phare 1s.

In NICOTIANA.

1947] GOODSPEED: NICOTIANA 113

(pl. 18, fig. 7), in N. rustica both ends are pointed; in N. rotund:-
folia (pl. 18, fig. 8)—to a lesser degree also in N. sylvestris—the
chalazal end is distinctly pocketed. Although in the other species
examined distinctions in the structure of the included cells serve
to distinguish the two ends of the embryo sac, in N. rotundifolia
two antipodals are similar to the synergids and the third antipodal
suggests the egg. Attention is called to the fact that the species
just referred to as differing in shape of embryo sac are members
of three distinct taxonomic subdivisions of the genus. The extent
to which such variations possess phylogenetic significance must,
however, await further investigation.

Examination of megasporogenesis was made in a number of Fy
interspecific hybrids. In most cases the megaspore mother cell
degenerates during meiosis, in some after the first and in others
during the second division; the 4-megaspore stage was observed
only in F, N. bigeloviw x N. suaveolens. As degeneration proceeds
the surrounding nucellar cells increase in volume (pl. 18, fig. 9)
for a considerable period to produce a condition in striking con-
trast to that in species where, as already noted, the nucellar cover-
ing shows a progressive decrease in cell volume (cf. Greenleaf,
1941).

The sequence leading to the maturation of the female gameto-
phyte, as summarized above, was determined from paraffin sec-
tions of N. tabacum? prepared by conventional techniques. Sec-
tions of entire ovaries from which the wall had been removed were
cut at 15-20 » and stained in iron haematoxylin. The following
discussion of the origin, morphology and behavior in fertilization
of the male gametes is based upon analysis of squash and smear
preparations of N. tabacum® and of N. longiflora produced by a
technique described elsewhere by Dr. Muriel V. Bradley (1948).
By the use of this remarkably effective technique which permits
examination of entire embryo sacs preceding and following fertili-
zation many of the foregoing observations have been confirmed
(cef., also, photomicrographs, Bradley, 1948).

In Nicotiana cytokinesis by furrowing originates the quartet of
immature microspores which become the elliptical pollen grains.
The mitosis producing the vegetative and generative nuclei is
approximately central rather than near the wall of the microspore
(Brumfield, 1941), apparently because of the absence of a central
vacuole (Sax, 1935).

Normally in Nicotiana the division of the generative nucleus
occurs in the pollen tube although under conditions of artificially
induced germination this mitosis is sometimes observed in the
pollen grain itself. Pollen tube mitoses have been studied in N.

2 A variety collected, as an escape from cultivation, near Sartimbamba,
northern Peru. The taxonomic status of other species referred to here is com-
mented upon elsewhere (Goodspeed, 1945a).

3 A variety collected, as an escape from cultivation, near Huaras, Peru.

114 MADRONO [Vol.9

tabacum, N. longiflora and N. otophora. Distinctions in chromosome
morphology within the genoms of these species are clearly seen
in these mitoses. Thus, for example, in N. otophora the five long
st and the seven short m chromosomes (Goodspeed, 1945b) can be
distinguished (pl. 19, fig. 5). Examination of pollen tubes of
N. tabacum at increasing distances from the stigma indicates that
the division of the generative nucleus characteristically occurs in
the upper third of the style.

From dissections and smears of the stigma and style of N.
tabacum it is clear that before pollen tube development the vege-
tative nucleus may exhibit marked alteration in form and sub-
stance. At first, its outline becomes irregular and its contents
somewhat diffuse and weakly staining (pl. 19, fig. 1). Later, as
it advances in the tube its substance is greatly extended. As
shown in plate 19, figure 3, its material becomes thrown into folds
or loops and even, at times, a long, twisted ribbon commonly
terminating, at one or both ends, in a more condensed region. Its
extension may be rather extraordinary (pl. 19, fig. 2). Following

EXPLANATION OF THE Figures. PLate 19

PuaTe 19. MarturaTIon oF THE MALe GAMETES AND FERTILIZATION IN NIco-
TIANA. Figs. 1-4, 6-16. N. tabacum, n= 24; Fig. 5. N. otophora, n=12; Figs.
17-22. WN. longiflora,n=10. Fig.1. Pollen grain from stigma, early alteration
in form of vegetative nucleus. X<180. Figs. 2-4. Portions of pollen tubes from
stylar canal. X290. Fig. 2. Extreme thread-like form of vegetative nucleus.
Fig. 3. More usual appearance of thread-like vegetative nucleus; late prophase
of generative nucleus. Fig. 4. Later condensation of vegetative nucleus; ana-
phase of generative nucleus. Fig. 5. Division of generative nucleus in pollen
tube, 5 large subterminal and 7 smaller median chromosomes. 850. Fig. 6.
Portion of pollen tube in micropyle; further condensation of vegetative nucleus
(cf. figs. 2-4); sperms. Figs. 7-22. Studies of embryo sacs during and follow-
ing fertilization (for details, cf. text). Figs. 7-9. Sperms, discharged from
pollen tube, before contact with egg and larger fusion nucleus, sperms earlier
elongated (figs. 7, 8), later spherical (fig. 9); remnant of vegetative nucleus,
tapering, pycnotic; degenerating nucleus of disrupted synergid ring-like; other
synergid below pollen tube cytoplasm; sperm cytoplasm apparent (fig. 9).
Figs. 10-12, 18, 19, 21. Early, mid- and late fertilization stages, sperms under-
going alteration in form and structure; incorporation of sperm in fusion nucleus
more rapid; tapering vegetative nucleus and one synergid pycnotic, other
synergid intact; cytoplasm of early zygote differentiating (figs. 12, 21); cyto-
plasm of sperms apparent (fig. 11). Figs. 13, 14. Early and late fertilization,
before fusion of polar nuclei; differentiation of zygote cytoplasm conspicuous.
Fig. 15. Entire embryo sac, late fertilization stage with incorporation of sperms
almost complete; cellular character of antipodals. Fig. 16. Metaphase of first
division of primary endosperm nucleus, V. tabacum, 72 chromosomes. Fig. 17.
Contents of two pollen tubes in embryo sac, post fertilization; sperms from
second pollen tube near zygote and primary endosperm nucleus; pycnotic de-
generation products of two vegetative nuclei (one above zygote) and two syner-
gids. Figs. 20, 22. Post fertilization, persistence of vestiges of vegetative
nucleus and one synergid nucleus, other synergid intact; differentiation of zygote
cytoplasm; metaphase and telophase of first division of primary endosperm
nucleus, V. longiflora, 30 chromosomes (fig. 20). All figures drawn with the
camera lucida by a special carbon pencil technique from squash or smear prepa-
rations; reproduced X 375 unless otherwise indicated.

Puare 19.
NICOTIANA,

MATURATION OF

ror Maur

GAMETES

AND FERTILIZATION

IN

1947] GOODSPEED: NICOTIANA 115

the division of the generative a certain condensation of the mate-
rial of the vegetative nucleus occurs (pl. 19, fig. 4), which con-
tinues with increasing evidence of pycnosis as the tube enters the
micropyle (pl. 19, fig. 6). A somewhat tailed, deeply staining
structure represents an advanced stage in the degeneration of
the vegetative nucleus. Variations in its appearance at fertiliza-
tion and somewhat later are shown in plate 19, figures 7 to 15, and
the fact that it persists at least until the primary endosperm
nucleus is in division appears in plate 19, figures 20 and 22.

Numerous investigators (cf. Schnarf, 1941) have noted in
other genera, and variously interpreted, similar changes in the
form of the vegetative nucleus from the large, irregular, amoe-
boid, weakly staining condition characteristic of the pollen grain
to the much extended, often almost thread-like form assumed in
the tube. The change from an amoeboid to an extended form
may be a response to protoplasmic streaming and the autonomous
movement of the vegetative nucleus in the narrow confines of the
tube (Tischler, 1925; O’Mara, 1933). Earlier literature attributes
to the vegetative nucleus initiation of tube development followed
by degeneration. More recent interpretations agree in question-
ing early degeneration and tentatively assign to the vegetative
nucleus a function related to the continued growth of the tube
(Schnarf, 1941; ecf., however, Poddubnaja-Arnoldi, 19386). In
Nicotiana degeneration as indicated by pycnosis is conspicuous
only after growth of the pollen tube is at an end and, certainly,
as suggested by Wulff (1933), if the vegetative nucleus is the
bearer of growth-promoting substances any increase in its sur-
face would have causal significance.

Studies of smears and squashes of styles and ovules of an
horticultural race of diploid Petunia show, in this genus closely
related to Nicotiana, a sequence of events in the development and
degeneration of the vegetative nucleus corresponding to that in
Nicotiana. However, in Petunia this nucleus is commonly more
weakly staining and its irregularity and elongation are less con-
spicuous than in Nicotiana. Correspondingly, evidence of its de-
generation is not conspicuous until after fertilization, by contrast
with a strikingly pyenotic degeneration product of this nucleus
which in Nicotiana is found along with the two sperms, in the tube
cytoplasm previous to its contact with the female nuclei.

The proper squash technique applied to ovules provides an
abundance of material for study of the sequence beginning with
the penetration of the micropyle by the pollen tube and continu-
ing through the divisions of the endosperm and zygote nuclei.
Although the evidence thus obtained corresponds in general to
that reported by Guignard (1902) for N. tabacum, a considerable
amount of additional information is now available. Also, various
stages in fertilization have for the first time been seen in species
of Nicotiana other than N. tabacum.

116 MADRONO [Vol. 9

Normally and as noted above, the 8-nucleate embryo sac of
the species of Nicotiana investigated becomes the seven celled
megagametophyte; the differentiation of synergids, egg and anti-
podals is accompanied by wall formation (cf. Bradley, 1948).
Vacuolation produces a broad band of cytoplasm connecting egg,
fusion nucleus and antipodals, with strands extending to the walls
of the embryo sac. In at least one race of N. tabacum numerous
deviations from such normal development occur (cf. Bradley,
1948; Persidski and Modilewski, 1934, in N. rustica). Thus, 9-
to 16-nucleate embryo sacs have been seen, obviously the result
of division of from one to all of the normal eight nuclei (cf., also,
Korotkevich, 1940, in N. rustica). Frequently, alterations in nor-
mal position of nuclei, and particularly antipodal ones, occur.
From one to three antipodals may wander to a position near the
egg so that in some instances all nuclei are found in the micropylar
end of the embryo sac (cf. Guignard, 1902).

Although polar fusion is usually complete before fertilization,
it is in progress thereafter too frequently to be classed as an ab-
normality. Indeed, Guignard (1902) considered post fertilization
fusion of the two polar nuclei normal behavior in N. tabacum.
Fusion of three nuclei in 8-nucleate embryo sacs and of from three
to five in the multinucleate embryo sacs above described has been
observed. Twin embryo sacs have been found in squash prepa-
rations, with the corresponding nuclei similarly disposed and the
two embryo sacs in identical stage of development.

In N. longiflora, and presumably in other species also, two
pollen tubes may penetrate the micropyle. The contained cyto-
plasm of two tubes each with two sperms, a pycnotic vegetative
nucleus and a degenerating synergid nucleus may be seen in con-
tact with the female nuclei. On the other hand, penetration may
not be simultaneous for, as shown in plate 19, figure 17, fertiliza-
tion of both the egg (seen under the mass of tube cytoplasm) and
fusion nucleus has apparently been effected by the sperms of the
tube first penetrating the embryo sac while those of the second
tube appear in the combined cytoplasm and presumably will later
degenerate there.

In those species of Nicotiana studied the male gametes undergo
little structural alteration from the time of their origin in the
pollen tube until their contact with the female nuclei, and appar-
ently the same is true of Petunia. Throughout they are somewhat
elongated, ovoid bodies (pl. 19, figs. 6, 7; cf. Bradley, 1948,
fig. 8). However, it is not possible to confirm the conclusions of
Poddubnaja-Arnoldi (1936) and Sarana (1934) for Nicotiana nor
the suggestion of Cooper (1946) for Petunia that the sperms in
the pollen tube are cells. On the other hand, in both genera, and
particularly in the latter, in squash preparations one sperm appar-
ently within the egg cell and the other near or in contact with the
fusion nucleus have been observed, each surrounded by a cyto-

1947] GOODSPEED: NICOTIANA 117

plasmic accumulation limited by a membrane (pl. 19, figs. 9, 11),
something quite distinct from the nimbus (frequently seen about
other nuclei as well) interpreted by Guignard (1902) as a cyto-
plasmic sheath but probably better interpreted as artefact.

The evidence both in Nicotiana and in Petunia of a cytoplasmic
sheath about the sperms at the time of fertilization argues against
the suggestion of Cooper (1946) that the x-bodies in the pollen
tube of Petunia may represent remnants of cytoplasm originally
surrounding and peculiar to the male gametes in the pollen tube.
Apparently, in the two genera, structures interpreted as x-bodies
by Cooper (1946) in Petunta—and perhaps also those interpreted
as sperms by Ferguson (1927)—are actually remnants or de-
generation products of vegetative and synergid nuclei (cf. also,
Bradley, 1948).

In reports on fertilization in other genera the question whether
the pollen tube discharges into a synergid or near it has been dis-
cussed (Gerassimova, 1933; Warmke, 1943; Swamy, 1945). In
Nicotiana just previous to fertilization a bulb-like protuberance of
greater density than the cytoplasm of the embryo sac and con-
taining the sperms and remnant of the vegetative nucleus is seen
near or in contact with the egg and fusion nucleus. It may be
interpreted as the swollen apex of the pollen tube after its impact
has disrupted a synergid. On the other hand, it may represent a
synergid into which the apical contents of the tube have been dis-
charged (Guignard, 1902; Schnarf, 1941). At times, however,
it appears that both the membrane of one synergid and that of
the tube may have become ruptured at a point of contact, and that
thereafter their combined contents may flow in bulb-like configu-
ration deeper into the embryo sac. In any event, one synergid
nucleus undergoes degeneration and is usually present in the cyto-
plasm surrounding the sperms and remnant of the vegetative
nucleus, while the second synergid commonly remains intact.

During fertilization the remnants of the vegetative and syner-
gid nuclei are conspicuous. The former becomes increasingly
pycnotic with its deeply staining material frequently tapering to
produce a more or less extended, tail-like appendage (pl. 19, figs.
9,11, 12) which is often dual or forked (pl. 19, fig. 15). In other
cases—usually later stages—it is more contracted (pl. 19, figs. 10,
13,14). At the same time the nucleus of the disrupted synergid
also becomes pycnotic. It is distinguishable from the vegetative
nucleus by form and position, usually appearing ring- or open
ring-shaped and on a plane with the normal synergid (pl. 19, figs.
10-15, 20).

Although union of nuclei appears to follow almost immedi-
ately upon the presence of the bulb-like mass of cytoplasm in the
embryo sac, free sperms and the earliest stages of fertilization
can be seen. The sperms are indistinguishable morphologically
(pl. 19, figs. 6, 7,9). At first somewhat elongated and oval, they

118 MADRONO [Vol. 9

become spherical just previous to fertilization (pl. 19, fig.9). In
contact with the female nuclei they expand and become structu-
rally more diffuse and loosely granular (pl. 19, figs. 10, 11) as
they flatten against the nuclear membranes and gradually merge
with the female nuclei along the lines of contact. Although initial
contact of one sperm with the egg usually precedes that of the
other sperm with the conspicuously larger fusion nucleus (cf.
however, pl. 19, fig. 8), the fertilization process appears to be
completed more rapidly in the latter case (pl. 19, fig. 12). Thus,,
the outline of the sperm nucleus with its included nucleolus is
visible in the fertilized egg at a time when the only vestige of the
sperm in the fusion nucleus is the presence of one or more small
nucleoli in addition to the large nucleolus which, in the unfertil-
ized fusion nucleus, represents the fusion product of the nucleoli
of the polar nuclei (pl. 19, figs. 12, 14, 15, 18, 19). Indeed, the
vestige of the sperm in the egg nucleus can be seen even after the
division of the primary endosperm nucleus is in progress (pl. 19,
figs. 20,22). The egg cytoplasm undergoes a certain alteration in
appearance which is sufficiently consistent as well as conspicuous
to become diagnostic evidence that fertilization has occurred (cf.
Schnarf, 1928). Before fertilization, the cytoplasm of the egg
cell is concentrated at its base with a large vacuole occupying the
micropylar end. After fertilization the egg cytoplasm appears to
increase in volume, to be extended into the micropylar extremity
of the cell and, in particular, to be denser and of a uniform con-
sistency, becoming particulate or subdivided into more or less uni-
form aggregations, presumably by numerous small vacuoles (pl.
19, figs. 12, 14, 15, 20, 21, 22; cf. Bradley, 1948, fig. 9; cf. Schnarf,,
1928).

Although post fertilization stages were not studied in detail
it is clear that the first division of the zygote does not occur until.
after two or more divisions of the endosperm nuclei (cf. pl. 19,
figs. 20, 22), as shown by the chromosome numbers involved (cf.
pl. 19, figs. 16, 20).

| SUMMARY

The following features in megasporogenesis, megagameto-
genesis and fertilization in Nicotiana (with some comparative evi-
dence in Petunia) are described and illustrated:

1. Distinctions in nucellar development and degeneration dur-.
ing embryo sac formation in species and F, interspecific hybrids ;.
monosporic embryo sac.

2. Character of chalazal end of embryo sac in various species:
in relation to their taxonomic position.

3. Duplication of megaspore mother cells and embryo sacs;,
multinucleate embryo sacs; multiple pollen tubes.

4. Morphology of vegetative nucleus in pollen grain and tube;
‘“x-bodies” as degeneration products of vegetative and synergid.

1947] GOODSPEED: NICOTIANA 119

nuclei; penetration of pollen tube via disruption of or discharge
into a synergid.

5. Cellular character of sperms; their morphology before and
their structural alteration during fertilization; alteration in cyto-
plasm of egg cell during and following fertilization; variations in
time of polar fusion in relation to fertilization; rate of fertiliza-
tion of egg and fusion nuclei and of subsequent development of
zygote and endosperm.

The author is indebted to Dr. Muriel V. Bradley and Mildred
C. Thompson for assistance in the studies reported upon here and
in the preparation of the manuscript.

Department of Botany,
University of California, Berkeley.

LITERATURE CITED

BrapDtey, Muriet V. 1948. An aceto-carmine squash technique for mature
embryo sacs. Stain Tech. In press.

BruMFIieLpD, R. T. 1941. Asymmetrical spindles in the first microspore division
of certain angiosperms. Am. Jour. Bot. 28: 713-722.

Coorrer, D. C. 1943. Haploid-diploid twin embryos in Lilium and Nicotiana.
Am. Jour. Bot. 30: 408-413.

. 1946. Double fertilization in Petunia. Am. Jour. Bot. 33: 54-57.

Frercuson, M.C. 1927. A cytological and genetical study of Petunia. I. Bull.
Torrey Bot. Club 54: 657-664.

Gerasstmova, H. 1933. Fertilization in Crepis capillaris (L.) Wall. La Cellule
42: 103-148.

GoopsPEED, T.H. 1945a. Studies in Nicotiana. III. A taxonomic organization
of the genus. Univ. Calif. Publ. Bot. 18: 335-344.

1945b. Chromosome number and morphology in Nicotiana. VII.
Karyotypes of fifty-five species in relation to a taxonomic revision of
the genus. Univ. Calif. Publ. Bot. 18: 345-370.

GREENLEAF, W. H. 1941. Sterile and fertile amphidiploids: their possible rela-
tion to the origin of Nicotiana tabacum. Genetics 26: 301-324.
GuienarD, M. L. 1902. La double fécondation chez les Solanées. Jour. de

Botanique 16: 145-167.

Kororkevicu, A. 1940. Phases of the development of the generative appa-
ratus in Nicotiana rustica L. Jour. Bot. Acad. Sci. RSS Ukraine 1:
215-236.

MauesHuwari, P. 1941. Recent work on the types of embryo-sacs in angio-
sperms—a critical review. Jour. Indian Bot. Soc. 20: 229-261.
Mopitewski, J. 1935. Cytogenetical investigation of the genus Nicotiana. Cy-
tology and embryology of the amphidiploid Nicotiana ditagla. Vseukr.

Akad. Nauk Inst. Bot. Zhurn. No. 7 (15): 7-29.

O’Mara, J. 1933. Division of the generative nucleus in the pollen tube of
Lilium. Bot. Gaz. 94: 567-578.

Persipski, D. anp J. Mopitewski. 1934. Cytological and embryological studies
of the chief varieties of Nicotiana rustica L. Wseukr. Akad. Nauk
Inst. Bot. Zhurn. No. 3 (11): 33-49.

PoppuBNAJA-ARNOLDI, V. 1936. Beobachtungen iiber die Keimung des Pollens
einiger Pflanzen auf kiinstlichem Nahrboden. Planta 25: 502-529.

Rees-Leonarp, O. L. 1935. Macrosporogenesis and development of the macro-
gametophyte of Solanum tuberosum. Bot. Gaz. 96: 734-750.

Sarana, M. O. 1934. Interspecific hybrids of tobacco. Report 1. Cyto-
genetics of interspecific hybrids Nicotiana Tabacum X N. glauca and N.
Tabacum X N. sylvestris. Krasnodar. Vses. n-i Inst. Tab. Prom. Trudy
No. 110: 191-220.

120 MADRONO [Vol.9

SaTina, S.. A. F. BuaKkester anp A. Avery. 1934. Twins in the jimson weed,
Datura stramonium. Am. Nat. 68: 162. (Abstract.)

Sax, K. 1935. The effects of temperature on nuclear differentiation in micro-
spore development. Jour. Arnold Arb. 16: 301-310.

ScunarF, K. 1928. Embryologie der Angiospermen. Handbuch der Pflanzen-
anatomie 10: 2: 2.

1941. Vergleichende Cytologie des Geschlechtsapparates der

Kormophyten. Monographien zur vergleichenden Cytologie. Bd. 1.
Berlin.

Swamy, B.G.L. 1945. Embryo sac and fertilization in Cypripedium spectabile.
Bot. Gaz. 107: 291-295.

TiscHLER, G. 1925. Studien iiber die Kern-Plasmarelation in Pollenkérnern.
Jahrb. wiss. Bot. 64: 121-168.

Warmke, H. KE. 1943. Macrosporogenesis, fertilization and early embryology
of Taraxacum kok-saghyz. Bull. Torrey Bot. Club 70: 164-173.

Wutirr, H. 1933. Beitrage zur Kenntnis des miinnlichen Gametophyten der
Angiospermen. Planta 21: 12-50.

NOMENCLATORIAL CHANGES IN ELYMUS WITH A KEY
TO THE CALIFORNIAN SPECIES

Frank W. GouLtp

In the preparation of a systematic treatment of the genus
Elymus and the related Agropyron, Sitanion, and Hystriz groups
for the ‘““Manual of California Grasses” which Dr. Alan A. Beetle
is preparing, the writer has been compelled by a considerable
amount of evidence to view this complex of the Hordeae as a
single genus. Previously proposed nomenclatorial combinations
show that in the past other systematists have shared this concept
at least partially. Admittedly there is still much to be learned
concerning specific and subspecific relationships, but the evidence
at hand indicates that the groups of species involved cannot be
segregated satisfactorily on a generic plane. It is probable that
further submergence of genera in the tribe Hordeae will result
from current investigations, especially in the fields of cytogenetics
and plant breeding. Fertile hybrid Triticum x Agropyron genera-
tions are reported by Veruschkine (1935) and Tzitzin (1936).

Linnaeus recognized five genera in the Hordeae complex, in
the following order: Lolium, Elymus, Secale, Hordeum, and Triticum.
He referred the two known species of Agropyron to Triticum, the
one known species of Hystriz to Elymus, and indicated no disposi-
tion of the Sitanion group. Bentham and Hooker (1883) listed
twelve genera in the Hordeae, treating Agropyron, Triticum,
Elymus, and Asperella (Hystriz) as separate genera and including
Sitanion as a section of Elymus. In North American grass treat-
ments, Beal (1896), and Hitchcock (1935), follow essentially the
Bentham and Hooker classification, but Hitchcock recognizes
Sitanion as a genus distinct from Elymus.

Classically, Elymus and Agropyron are distinguished on the
basis of the number of spikelets at each node of the rachis, the

1947] GOULD: ELYMUS 121

former with two or more spikelets per node and the latter with
one. This results in the rather arbitrary separation of species that
are Obviously closely related, as in the Elymus triticoides, E. cine-
reus, FE. condensatus, E. salina, Agropyron Smithi, A. arenicola group,
and the Elymus glaucus, Agropyron subsecundum, A. pauciflorum com-
plex. The weakness of this basis for distinction is shown also by
the following series of Elymus species in which the characteristic
number of spikelets per node is: FE. salina, one spikelet at a node;
E. triticoides and E. glaucus two spikelets at a node but in forms
of both species only one spikelet at a node; EF. cinereus three spike-
lets at a node; E. condensatus eight or more spikelets at a node,
counting those on pedicels.

It has been noted that in Elymus the florets are oriented more-
or-less dorso-ventrally to the rachis while in Agropyron they are
lateral. This tendency is recognized readily in some species but
is not uniformly evident throughout the two groups. In some
spikes of E. tritacoides both conditions can be observed.

The type species of Agropyron, A. triticeum Gaertn., is an
annual, very unlike the American agropyrons, all of which are
perennial, and more similar to species of T'riticum. This and the
annual Elymus caput-medusae L., both sparingly introduced into
North America from Europe, probably should be excluded from
the genus Elymus as here interpreted.

Elymus and Sitanion probably have been treated more gen-
erally as sections of one genus than as separate genera. When
retained as distinct they are separated on the basis of the readily
disarticulating rachis and the usually narrow, setaceous glumes
of the latter. If this distinction were followed rigidly, Elymus
aristatus, as known in California, would appear more Sitanion-like
than the classically recognized species Sitanion Hansenii.

Hystriz, a genus of about four species, has been split off from
Elymus primarily on the basis of glume reduction, one or both
being completely lacking in some cases. Plants of the North
American species are very similar to species of Elymus, especially
E. interruptus which also has irregularly reduced glumes.

Stebbins, Valencia, and Valencia in their recent papers on arti-
ficial and natural hybrids in the Hordeae (1946) present numer-
ous points in agreement with the writer’s independent conclusions.
They give cytological evidence for assuming that Elymus glaucus
and Sitanion Hystriz are even more closely related than some spe-
cies of Elymus as previously delimited. They report the occur-
rence of Elymus-Sitanion and Elymus-Agropyron hybrids in nature
and describe artificially produced Sitanion-Agropyron hybrids.
Evidence is presented for the belief that all plants that can be
classified as Sitanion Hansenii are sterile F, hybrids between
Elymus glaucus and either Sitanion Hystrix or Sitanion jubatum, and
that Agropyron Saundersii probably is composed of a series of F,
hybrids between Agropyron pauciflorum and Sitanion Hystria or

122 MADRONO [Vol. 9

Sitanion jubatum. Reference is made to a colony of hybrid Elymus
glaucus x Agropyron paucifiorum plants growing with the parent
species near the Carnegie Institution experimental garden at
Mather, Tuolumne County, California. This hybrid is discussed
further by Hartung (1946).

The writer has noted an Elymus in the Sierra Ancha Mountains,
Gila County, Arizona, which is morphologically intermediate be-
tween Elymus glaucus and Agropyron subsecundum. Plants of this
type are abundant in the oak association at 5500 feet elevation,
and no other forms of these two species occur in the vicinity.

Exrymus L. Species Pl. 83. 1753. Agropyron Gaertn. Nov.
Comm. Petrop. 14: 589.1770. Asperella Willd. Roem. and Ust.,
Mag. Bot. 7: 5. 1790. Hystrix Moench. Meth. Pl. 295. 1794.
Sitanion Raf. Journ. Phys. 89: 103. 1819. Clinelymus Griseb.
Ledebour, Fl. Ross. 4: 330. 1853.

Annuals or perennials, many rhizomatous; blades linear or
lanceolate, flat or involute, frequently glaucous, glabrous or vari-
ously pubescent; inflorescence basically spicate with 1 to 38 or
occasionally 4 to 6 spikelets at a node, when more than 2 at one
node one or more spikelets often short-pedicelled, in E. con-
densatus the inflorescence is a dense panicle; spikes disarticulating
in the rachis or rachilla or both; glumes mostly subequal, reduced
or absent in a few species, broadly lanceolate to attenuate or
subulate, awnless, with a single principal awn, or with 2 to 4 awns
or aristate teeth; lemmas mostly lanceolate, rounded on back,
obtuse, acute, or aristate, usually inconspicuously nerved except
near the apex; paleas mostly obtuse or truncate, about as long as
and somewhat infolded by the lemmas.

Type species, Elymus sibiricus L. Species Pl. 83. 1753. (Con-
cerning choice of type species see Hitchcock, 1936.)

KEY TO THE CALIFORNIAN SPECIES OF ELymMus

A. Lemmas awned, the awns mostly 1 to 3 cm. long;
plants typically without rhizomes

Plants annual; lemma awns 3 to 8 cm. long; intro-
@uced Weed y Species vei. 2 tee ee eee 1. EH. caput-medusae.
Plants perennial.
Glumes absent or setaceous and scarcely reaching
the. first: lemmas. cos Guo8 bites eee eee 26. EH. californicus.
Glumes present and at least half as long as the first
lemma.

I. Awns of lemmas curving outward at maturity

Rachis not disarticulating at maturity; internodes of
spikes usually 1 cm. long or longer, the spikelets
rather distant and distinct from each other;
glumes mostly broad, acute or short-awned (occa-
sionally long-awned in E. arizonicus).
Spikes with mostly 2 or 3 spikelets at a node ...... 12b. E. glaucus
subsp. Jepsoni.

1947] GOULD: ELYMUS 123

Spikes with mostly 1 spikelet at a node.
Culms erect at base, usually 40 cm. long or longer;
blades, at least some, longer than 10 cm.
Culms slender; blades narrow, usually involute;
spikes erect; spikelets usually closely ap-
pressed; awns slender, sharply divergent .. 2. E. spicatus.
Culms stout; blades 4 to 6 mm. or more broad,
flat; spikes flexuous; spikelets usually
spreading; awns stout, not sharply diver-
OMe ae oe tee: SOR eed acy as ew Maca 3. E. arizonicus.
Culms usually decumbent at base, mostly 15 to 35
cm. long; blades usually flat and short, 10 cm.
or less long, mostly tufted at base of culms .. 4. E. sierrus.
Rachis readily disarticulating at maturity; internodes
of spike usually 4 to 6 or 8 mm. long, the spikelets
closely imbricated and rather crowded; glumes
narrow, attenuate to setaceous, long-awned.
Spikelets*mostly 1 ata node ......... 0.0.2.6. 0040. 5. EH. saxicolus.
Spikelets mostly 2 at a node.
Spikes, including awns, almost as broad as long;
glumes bristle-like or cleft into bristle-like
divisions, the body scarcely apparent.

Glumes cleft into at least 3 divisions .......... 6. E. multisetus.
Glumes entire or 2-cleft ...................... 7. E. elymoides.
Spikes much longer than broad; glumes lanceo-
late; the body apparent .....3. 066 css e aes 8. E. Hansenii.

IT. Awns of the lemmas straight or undulate, not curving
outward at maturity
Spikelets mostly 1 at a node.
Rachis readily disarticulating at maturity; glumes
mostly attenuate, with awns 4to10mm.long.. 9. EH. Saundersii.
Rachis not readily disarticulating; glumes acute or
abruptly short-awned, the awns seldom over 4
mm. long.
Spikes relatively dense, the spikelets overlapping
3 to % their length; rachis internodes mostly
AREOUS SINE: LOMO ou are eu Net idan Pht cae ene tk tes 13a. EB. pauciflorus
Spikes not dense, spikelets overlapping the one subsp. subsecundus
above on the opposite side of the rachis + or
less of their length; rachis internodes aver-
aging 10 mm. or more long.
Culm nodes glabrous; lemmas usually long-
awned; florets 3 to 5 per spikelet ......... 13b. E. pauciflorus
Culm nodes finely pubescent; lemmas_ short- subsp. laeve.
awned; florets mostly 6 to 8 per spikelet .. 14. EH. Stebbinsii.
Spikelets mostly 2 at a node.
Rachis not disarticulating at maturity; glumes usu-
ally broadly lanceolate, 3 to 5 nerved; culms
usually in small clusters; common in California... 12. EH. glaucus.
Rachis disarticulating at maturity; glumes narrowly
lanceolate or subulate, 1 to 3 nerved; culms
usually in dense clumps; rare or infrequent in
California.
Spikes slender, about 5 mm. broad, dense, the
spikelets small, closely placed; lemmas 6 to 8
mm. long excluding the awns; glumes lanceo-
te wletO. SeMELRVER Mo. soe eo ee hee ll. E. Macounii.

124 MADRONO [Vol.9

Spikes stouter, mostly 8 to 10 mm. or more broad;
lemmas 8 to 10 mm. long.
Glume with awn mostly 1 to 1.5 cm. long; spikes

usually 8 em. or less long.:............... 10. H. aristatus.
Glume with awn mostly 2.5 cm. long or longer;
spikes usually more than 8 cm. long ...... 8. EH. Hansenii.

AA. Lemmas awnless or with awns 6 mm. or less long

Glumes broadly lanceolate, strongly 3 to 9 nerved,
thin, or if thickened then the apex obtuse.
Plants without rhizomes.
Spikelets mostly 1 at a node.
Culm nodes glabrous; florets 3 to 5 per spikelet.. 13. HE. pauciflorus

Culm nodes pubescent; florets mostly 6 to 8 per
Spikeleby wiht vie ade epee tl Sega ee ee 14, EH. Stebbinsii.
Spikelets mostly 2. at a node .......2..77..0.0 12a. E. glaucus
subsp. virescens.
Plants with rhizomes.
Spikelets mostly 1 at a node.
Culm internodes 1 to 3 cm. long; rachis disar-
ticulating at maturity; seashore .......... 15. EH. multinodus.
Culm internodes mostly more than 4 cm. long;
rachis not disarticulating.
Lemmas glabrous or scabrous.
Blades flat, thin and lax, bright green,

rarely :laucous. 7)... 20.8 oe eee 18. E. repens.
Blades usually involute, stiff, mostly glau-
COUS Hohe ck on eh ae OR ee ie he eee 20. E. riparious
Lemmas finely pubescent: 2... 1.0.20: .22.%% 19. H. subvillosus.
Spikelets mostly 2 at a node ................... 16. H. mollis.
Glumes subulate, or if lanceolate then inconspicuously
nerved, hard or tough in texture, and awn-tipped
or acute.
Spikelets mostly 2 to many at a node.
Culms finely pubescent below the spike; glumes
lanceolate; plants rhizomatous; seashore .... 17. H. vancouverensis.

Culms glabrous below the inflorescence; glumes
subulate or narrowly lanceolate.
Spikelets 6 to 40 per node of the rachis includ-
ing those on branches; culms usually 6 to
10 mm. in diameter at base; blades 15 to 35
em: broads coastal’. 40 tee ee ee ee 24. HEH. condensatus.
Spikelets 1 to 6 at a node, rarely more; culms
usually less than 6 mm. in diameter; blades
3 to 15 mm. broad.
Culm nodes or vicinity of nodes with fine,
usually dense pubescence; plants typi-
cally non-rhizomatous .................. 23. EH. cinereus.
Culm nodes glabrous; plants rhizomatous.
Blades mostly 3 to 6 mm. broad; spikes
with 1 or 2, occasionally 3, spikelets at
a node; spikelets 8 to 15 mm. long with
3 to*6 florets: 205. ee 22. E. triticoides.
Blades mostly 6 to 15 mm. broad; at least
some nodes of the spike with 3 to 6
spikelets, or spikelets 17 to 25 mm. long
and with 6 to 9 florets ............... 22a. E. triticoides

subsp. multiflorus.

1947 | GOULD: ELYMUS 125

Spikelets mostly 1 at a node.
Culms mostly 25 to 80 cm. long; spikes well ex-
serted.
Glumes narrow, usually awn-like; florets usually
twisted so that the back of the lower lemma
is centered between the glumes ........... 22. EH. triticoides.
Glumes narrowly lanceolate but mostly broader
than in E. triticoides; lowermost lemma of
spikelet lateral to the rachis, the back not
centered between the glumes ............. 21. EH. Smithii.
Culms 10 to 20 cm. long; spikes little exserted,
often exceeded by the blades; seashore ...... 25. EH. pacificus.

The following species of Elymus occur in California.
1. Exymus caput-MEbvuSAE L. Sp. Pl. 84. 1753.

2. Elymus spicatus (Pursh) comb. nov. Festuca spicata Pursh,
Fl. Am. Sept. 88.1814. Agropyron spicatum Scribn. & Smith, Bull.
U.S. Div. Agrost. 4: 83. 1897.

3. Elymus arizonicus (Scribn. & Smith) comb nov. Agro-
pyron arizonicum Scribn. & Smith, Bull. U. S. Div. Agrost. 4: 27.
1897. <A. spicatum var. arizonicum M. E. Jones, Contr. West. Bot.
i 19. 1912.

4. Elymus sierrus nom. nov. Agropyron Gmelini var. Pringlei
Scribn. & Smith, Bull. U. S. Div. Agrost. 4: 81.1897. A. Pringlei
Hitchcock ex Jepson, Fl. Calif. 1: 188.1912. Not Elymus Pringlei
Scribn. & Merr., 1901.

5. Exymus saxicouus Scribn. & Smith, Bull. U. S. Div. Agrost.
11: 56. 1898. Sitanion flexuosum Piper, Erythea 7: 99. 1899.
S. lanceolatum J. G. Smith, Bull. U. S. Div. Agrost. 18: 20. 1899.
Agropyron saxicola Piper, Contr. U.S. Nat. Herb. 11: 148. 1906.

6. Exymus mutisetus (J. G. Smith) Davy, Univ. Calif. Publ.
Bot. 1: 57. 1902. Sitanion jubatum J. G. Smith, Bull. U. S. Div.
Agrost. 18: 10. 1899. Not Elymus jubatus Link, 1827. Sitanion
multisetum J. G. Smith, Bull. U. S. Div. Agrost. 18: 11. 1899.

7. Ex,ymus retymoies (Raf.) Swezey, Nebr. Pl. 15. 1891.
Aegilops Hystriz Nutt., Gen. Pl. 1: 86.1818. Not Elymus Hystri«
L. 1753. Sitanion elymoides Raf., Jour. Phys. 89: 103. 1819.
Elymus Sitanion Schult., Mant. 2: 426. 1824. Sitanion Hystriz
J. G. Smith, Bull. U. S. Div. Agrost. 18: 15, pl. 2. 1899.

8. Erymus Hanseni Scribn., Bull. U. S. Div. Agrost. 11: 56,
fig. 12. 1898. Sitanion Hansen J. G. Smith, Bull. U. S. Div.
Agrost. 18: 20. 1899.

9. E.ymus Saunpersit Vasey, Bull. Torrey Bot. Club 11: 126.
1884. Agropyron Saundersii Hitchcock, Proc. Biol. Soc. Wash. 41:
159.1928. Elymus Saundersii var. californicus Hoover, Leafl. West.
Bot. 3: 254. 1943.

126 MADRONO [Vol. 9

10. Exymus aristatus Merrill, Rhodora 4: 147. 1902. E.
glaucus aristatus Hitchcock ex Abrams, Illus. Fl. Pacific States 1:
252. 1923.

11. Exymus Macouni Vasey, Bull. Torrey Bot. Club 18: 119.
1886. Terellia Macouni Lunell, Am. Midl. Nat. 4: 228. 1915.

12. Exymus cGiaucus Buckley, Proc. Acad. Nat. Sci. Phila.
1862: 99.1862. Clinelymus glaucus Nevski, Bull. Jard. Bot. Acad.
Sci. U. R.S.S. 30: 648. 1932.

12a. E. auaucus Buckley subsp. virescens (Piper) comb. nov.
E. virescens Piper, Erythea 7: 101. 1899.

12b. E. euaucus Buckley subsp. Jepsonii (Davy) comb. nov.
E. glaucus var. Jepsonii Davy ex Jepson, Fl. West. Mid. Calif. 79.
1901. LE. glaucus f, Jepsoni St. John, Fl. S. E. Wash. & Adj.
Idaho 42. 1937.

18. Elymus pauciflorus (Schwein.) comb. nov. Triticum pauci-
florum Schwein., in Keating, Narr. Exped. Winnipeg 2: 383. 1824.
T. trachycaulum Link, Hort. Berol. 2: 189. 1833. Agropyron
tenerum Vasey, Bot. Gaz. 10: 258.1885. A. pauciflorum Hitchcock,
Am. Jour. Bot. 21: 132. 1934.

13a. E. paucirLorus (Schwein.) Gould subsp. subsecundus
(Link) comb. nov. Triticum subsecundus Link, Hort. Berol. 2:
190. 1833. YT. Richardsoni Schrad. Linnaea 12: 467. 1838. Agro-
pyron subsecundum Hitchcock, Am. Jour. Bot. 21: 181. 19384.

13b. E. paucirLorus (Schwein) Gould subsp. laeve (Scribn. &
Smith) comb. nov. Agropyron Parishiw Scribn. & Smith var. laeve
Scribn. Smith, Bull. U. S. Div. Agrost. 4: 28. 1897. A. laeve
Hitchcock ex Jepson, FI. Calif. 1: 181. 1912.

14. Elymus Stebbinsii nom. nov. Agropyron Parishii Scribn.
& Smith, Bull. U.S. Div. Agrost. 4: 28.1897. Not Elymus Parishiu
Davy & Merrill, 1902.

This species is named in honor of Dr. G. Ledyard Stebbins, Jr.
of the University of California. For the past several years Dr.
Stebbins has made cytogenetical investigations of species of the
Hordeae tribe, and has contributed substantially to our knowledge
of phylogenetic relationships in this group. Dr. Stebbins has
worked specifically with the Elymus complex to which the species
named in his honor belongs.

15. Elymus multinodus nom. nov. Triticum junceum L., Mant.
Pl. 2: 827. 1771. Not Elymus junceus Fisch. 1811. Agropyron
junceum Beauv., Ess. Agrost. 102. 1812.

16. Exymus mouuis Trin. ex Spreng., Neue Entdeck. 2: 72.
1821.

17. E.ymus vaNcouverRENs!Is Vasey, Bull. Torrey Bot. Club 15:
48.1888.

1947] GOULD: ELYMUS 127

18. Elymus repens (L.) comb. nov. Triticum repens L. Sp. Pl.
86.1753. Agropyron repens Beauv., Ess. Agrost. 102. 1812.

19. Elymus subvillosus (Hook.) comb. nov. Triticum repens
var. dasystachum Hook. FI. Bor. Am, 2: 254. 1840. Not Elymus
dasystachys Trin. ex. Ledeb. 1829. Triticum repens var. subvillosum
Hook. Fl. Bor. Am. 2: 254. 1840. J. dasystachum A. Gray, Man.
602.1848. Agropyron dasystachum Scribn., Bull. Torrey Bot. Club
10: 78. 18838.

20. Elymus riparius (Scribn. & Smith) comb. nov. Agropyron
riparium Scribn. & Smith, Bull. U. S. Div. Agrost. 4: 35. 1897.
A. Smithi var. riparium Jones, Contr. West. Bot. 14: 19. 1912.

21. Elymus Smithii comb. nov. Agropyron Smithi Rydberg,
Mem. N. Y. Bot. Gard. 1: 64. 1900.

22. ExLymus Triticoipes Buckley, Proc. Acad. Nat. Sci. Phila.
1862: 99. 1862. EE. Orcuttianus Vasey, Bot. Gaz. 10: 258. 1885.
£. simplex Scribn. & Williams, Bull. U. S. Div. Agrost. 11: 57.
pl. 17. 1898.

22a. Erymus rriticoiEes Buckley subsp. MuLtTirLtorus Gould,
Madrofio 8: 46. 1945.

23. Exymus ciNneREus Scribn. & Merrill, Bull. Torrey Bot. Club
29: 467.1902. £.condensatus pubens Piper, Erythea 7: 101. 1899.
E. condensatus f. pubens St. John, Fl. S. E. Wash. & Adj. Idaho 42.
1937.

24. ELymus conpENSATUs Presl. Rel. Haenk. 1: 265. 1830.

25. Elymus pacificus nom. nov. Agropyron arenicola Davy ex
Jepson, Fl. West. Mid. Calif. 76. 1901. Not Elymus arenicolus
Scribn. & Smith, 1899.

26. Elymus californicus (Bolander) comb. nov. Gymno-
stichum californicum Bolander, Thurber ex Brewer & Wats. Bot.
Calif. 2: 327. 1880. Hystrix californica Kuntze, Rev. Gen. Pl. 2:
778. 1891. Asperella californica Beal, Grasses N. Am. 2: 657.

1896.
University of Arizona,
Tucson, Arizona.

LITERATURE CITED

Beat, W. J. 1896. Grasses of North America. vol. II.
BEenTHAM, G., and Hooxer, J. D. 1883. Genera Plantarum. 3: 1093-1094.
Goutp, F. W. 1945. Notes on the genus Elymus. Madrofio 8: 42-47.
Hartune, M. E. 1946. Chromosome numbers in Poa, Agropyron, and Elymus.
Am. Jour. Bot. 33: 516-531.
Hircucock, A. S. 1935. Manual of the grasses of the United States. U. S.
Dept. Agric. Misc. Publ. No. 200.
1936. The genera of grasses of the United States. U.S. Dept.
Agric. Bull. No. 772. Revised ed.

128 | MADRONO [Vol.9

Hoover, R. F. 1943. Observations on Californian plants—III. Leafl. West.
Bot. 3: 254-256.

Srepsins, G. L. Jr., J. 1. Vatencia, and R. M. Vatencia. 1946a. Artificial and .
natural hybrids in the Gramineae, tribe Hordeae I. Elymus, Sitanion,
and Agropyron. Am. Jour. Bot. 33: 338-351.

> ———————_-, and ——————_. 1946b._ Artificial and natural

hybrids in the Gramineae, tribe Hordeae II. Agropyron, Elymus, and
Hordeum. Am. Jour. Bot. 33: 579-586.

Tzirzin, N. V. 1936. [The problem of perennial wheat.] Selektsija i Semeno-
vodstvo [Breeding and seed growing] No. 2: 21-27. Review in Plant
Breed. Abst. 7: 184. 1937.

VERUSCHKINE, S. M. 1935. [On the way towards perennial wheat.] Socialistic
grain farming. Saratov. No. 4: 77-83. Review in Plant Breed. Abst.
6: 258. 1936.

TWO NEW VARIETIES OF CONDALIA FROM TEXAS
V. L. Cory

The small pasture, or horse trap, in which the horses are
grazed at the Texas Agricultural Experiment Station, Substation
No. 14, contains 118 acres. The pasture is at the summit of the
Edwards Plateau at an elevation of 2400 feet, and has a surface
comparatively level except for the heads of two small drainage
courses. A gently rounded, highly calcareous knoll in the south-
central portion of the pasture covers several acres and bears an
almost pure stand of Juniperus Pinchoti Sudw. with a slight admix-
ture of Quercus Vaseyana Buckl. Below the knoll on the west
occurs a variety of shrubby vegetation; farther on, in the upper
part of a little valley, the shrubs give way to grassland. In this
shrubby vegetation occur four kinds of Condalia, all growing
within twenty-five feet of each other, a circumstance which I do
not recall having observed elsewhere. One of these forms of
Condalia occurs as a close colony and appears to merit varietal
recognition.

Conpaia opovata Hook. var. edwardsiana var. nov. A specie
differt foliis longioribus angustioribusque, spatulatis nec obovatis.
This differs from the typical form of the species in its longer and
narrower leaves, which are spatulate instead of obovate.

Type. Twenty-nine airline miles northwest of Rocksprings,
Edwards County, Texas, altitude approximately 2400 feet, May
27, 1943, Cory 41784 (Arnold Arboretum, Harvard University).

This variety is markedly different in appearance from other
members of the genus in this area because of its greater height
and lighter-colored foliage. Even after long and diligent search,
I have been unable to find it anywhere save in this single, isolated
thicket. It is closely related to the typical phase of the species,
which inhabits the Rio Grande Plains of Texas and northern
Mexico, but does not reach the Edwards Plateau or even the
escarpment area.

1947] CORY: CONDALIA ; 129

The colony of Condalia obovata var. edwardsiana is essentially a
pure consociation. The thicket is irregularly ovate, with a long
diameter of 45 feet and a short diameter of 30 feet, comprising a
calculated area of 900 square feet. It contains about 100 trunks,
or 20 to 25 individual plants. The tallest is 3.1 meters high, and
the average height is about 2.5 meters. Along the margins of
this pure stand occur the following woody plants: Quercus virgini-
ana, Berberis trifoliata, Rhus microphylla, R. virens, Prosopis juliflora
glandulosa, Diospyros texana, Opuntia leptocaulis, Cissus incisa, Colu-
brina texensis, and Condalia obtusifolia. A plant of Condalia viridis
Johnston is growing twenty feet south of the west end of the
thicket, and there are a few others nearby on the south and west,
some of them in thickets of C. spathulata Gray. ‘Twenty-two feet
southwest of the southwest corner of the pure stand of C. obovata
var. edwardsiana is the nearest plant of C. spathulata, a species which
is abundant southwest of this area. There are two plants of C.
obtusifolia at the southwestern margin of the consociation. The
plants of C. viridis and C. obtusifolia are clearly distinguishable
from those of C. spathulata by their greater height. On May 27,
1943, only C. obtusifolia showed any stage of inflorescence, and it
presented all the stages from bud to mature fruit, frequently on a
single branchlet. The fruit of C. obtusifolia, which is much larger
than that of the other condalias, changes from green to reddish
and finally to an intense bluish-black upon maturing.

In 19387, Mr. Hiram Reed, a former associate of mine in the
United States Department of Agriculture, retired and came to
Sonora, Texas, to live. He retained his interest in plants, particu-
larly those of some economic significance, and from time to time
turned his collections over to me. Among the first lot of these
collections was a Condalia from the hills of the Garner State Park
in northern Uvalde County, Texas. At first, this specimen was
referred to C. obovata, but subsequent detailed field study, espe-
cially on C. obovata and on C. viridis showed that it was more
closely related to the latter species. In my field experience, I
found that C. obovata is of common occurrence from sea level up
to the base of the Escarpment of the Edwards Plateau, or up to
elevations of approximately 600 feet, while C. viridis is common
on the Edwards Plateau and west and south of that area at ele-
vations of 2000 feet or more. In between these two elevations,
and occupying the Escarpment area of the Edwards Plateau, is the
plant under discussion. I wish to dedicate this newly recognized
variety to the man who first called it to my attention, to my friend
of former days as well as of the present, Mr. Hiram R. Reed.

Conpaia viripis Jtn. var. Reedii var. nov. <A specie differt
vulgo statura duplo majore foliisque duplo majoribus (i.e., ad 13
mm. longis, 6 mm. latis).

This variety differs from the species in being (on the average )

130 MADRONO [Vol.9

twice as tall with at least twice its spread, and with leaves about
twice as large, or up to 13 mm. long and 6 mm. broad.

Type. Northeastern or eastern brow of hills along the Frio
River, Garner State Park, Uvalde County, Texas, altitude 1500
to 1600 feet, June 4, 1944, Cory 44496 (Arnold Arboretum, Har-
vard University).

I know of this variety only in the Escarpment Area of the
Edwards Plateau at elevations below 2000 feet. It is my opinion
that two of the specimens cited by Johnston for his species, C.
viridis (Palmer 164, Eagle Pass, Val Verde County, 1880, and
Harvard 61, Eagle Pass, 1882), should instead be referred to var.
Reedu. Eagle Pass in on the Rio Grande in Maverick County; the
highest elevation in Maverick County is 956 feet.

Rather commonly, on account of its obovate leaves, var. Reedii
has been confused with C. obovata, which it does resemble in its
growth and general appearance. The light-green foliage of C.
obovata, however, serves to separate it at once from var. Reediu
and our other species of Condalia. On the other hand, C. viridis
typically in the field might be easily confused with C. obtusifolia

but never with C. obovata. —

ConpDaALia oBTUSIFOLIA (Hook.) Weberb. R(hamnus) ? obtusi-
folia Hook. ex Torr. & Gray, Fl. N. Am. 1: 685. 1840. Zizyphus
obtusifolia (Hook. ex Torr. & Gray) A. Gray, Genera 2: 170. 1849.
Condalia obtusifolia (Hook.) Weberb. in Engler & Prantl, Pflanzenf.
3: 404.1850. Zizyphus lycioides A. Gray, Boston Jour. Nat. Hist.
6: 168. 1850. Condalia lycioides (A. Gray) Weberb. in Engler &
Prantl, Pflanzenf. 3: 404. 1850.

For twenty years I have searched unsuccessfully for the enti-
ties described as Condalia obtusifolia (Hook.) Weberb. and C. lycio-
ides (A. Gray) Weberb., and I have finally reached the conclusions
that the two are synonymous. At my request, Dr. Ivar Tidestrom
checked the material of these two alleged species at the United
States National Museum and agreed with the above disposition of
C. lycioides. In young growth or in new growth of old plants the
foliage is typically that of C. obtusifolia, and very frequently the
young growth contrasts markedly with the more abundant older
growth of the same plant which has foliage typically that of C.
lycioides. A plant with the linear, oblong, or elliptic leaves charac-
teristic of C. lycioides will, upon being cut off at the surface of the
ground, send up vigorous sprouts with orbicular leaves as much as
2 cm. in diameter. At and toward the eastern limit of its range
in Texas the plant more commonly has foliage typical of C. obtusi-
folia, while a few hundred miles farther west, the more common
type of foliage is that of C. lycioides. Complete intergradation of
the two extremes, however, is evidence of their conspecifity.
Specimens verifying this conclusion have been deposited at the
Arnold Arboretum, Harvard University. The common species of

1947] : BAKER: MEXICAN VIOLET 131

Condalia occurring from central Texas westward to southern Cali-
fornia and south into northern Mexico should be known, there-
fore, as Condalia obtusifolia (Hook.) Weberb.

I am indebted to Dr. I. M. Johnston and to Mr. Ernest J.
Palmer for critical study of material and for suggestions as to
treatment, to Dr. Lloyd Shinners for valuable assistance with the
Latin diagnoses, and to Dr. Bassett Maguire for the above
synonymy.

Texas Agricultural Experiment Station.
Sonora, Texas.

A NEW VIOLET FROM MEXICO
Mito S. Baker

Viola galeanaensis sp. nov. Geophytum perspicuum, foliis
partibusque caulis tandrum supra solum; glabrum praeter granulis
dispersis superficierum superiarum venarum mediarum propeque
ad marginibus aliquot foliorum; rhizoma erecta, 2—4 cm. longa,
multis radiculis crassis; petioli foliorum radicitorum usque ad 8
cm. longi, foliorum caudicitorum breviores sed semper plures
longitudines laminae folii multiplicata, laminae late ovatae, ad
basim tenuiter cordatae, apiculatae, leviter crenatae, prope longi-
ores quam latiores, usque ad 18 mm. longae; stipulae oblongae-
lanceolatae, prope integrae, scariosae, usque ad 9 mm. longae;
caulis 1-3, fere subterraneae, 2-5 mm. longae; flores axillares,
9-13 mm. longitudinibus; peduncula 3-5 cm. longi, bracteolae fili-
formes, propinquae ad florem; sepala 4-5 mm., inferiori fere
maiores, lanceolati, marginibus scariosis, acuti, auriculae parvae,
rotunditatae; corolla pallida purpurea usque ad alba, pars media
flava nervis et dorso atris, petala laterales leviter clavata-barbati,
calear brevissimum; vagina staminis apertione-collare angusto;
stylus vix 2 mm. longus, abrupte inflectus prope ab ovario, abrupte
amplificatus ad caput clavatum, leviter barbatus ad lateribus,
stigma rostrum minimum ad superficiem ventralem capitis; cap-
sula globosa, glabra, semina incognita.

A pronounced geophyte with only leaves and parts of the
stems above the soil; glabrous except for scattered granulations
along the upper surface of the midveins and near the margins of
some leaves; rootstock erect, 2 to 4 cm. long, with many coarse
roots; petioles of radical leaves up to 8 cm. long, those of cauline
leaves shorter but always several times length of blade, leaf blades
broadly ovate, shallowly cordate at base, apiculate, remotely ser-
rate, approximately as wide as long, up to 18 mm. long; stipules
oblong-lanceolate, nearly entire, scarious, up to 9 mm. long; stems
1 to 3 mostly subterranean, 2 to 5 cm. long: flowers axillary 9 to
13 mm. in length; peduncles 3 to 5 cm. long, bractlets filiform,
near the flower; sepals 4 to 5 mm., the lower ones much larger,

132 MADRONO [Vol. 9

lanceolate, scarious margined, acute, auricles small, rounded;
corolla pale lilac to white with yellow center with purple veins
and back, lateral petals lightly clavate-bearded, spur very short;
stamen sheath with narrow collar-opening; style scarcely 2 mm.
long sharply bent near ovary, abruptly widened into a clavate-
head, lightly bearded on the sides, the stigma a minute beak on
the ventral surface of head: capsule spherical, glabrous, seeds
unknown.

Type. Near the peak of Cerro Potosi, 11,900 feet, Municipio
de Galeana, Nuevo Leon, Mexico, 1938, Mezican Biological Expe-
dition of Students of the University of Illinois, 923 (Gray Her-
barium). A specimen in the United States National Herbarium
labelled only Cerro Potosi, Nuevo Leon, Mexico, R. Schneider 923,
is doubtless an isotype.

Other collections. Altitude 9,000—10,000 feet, Mount Infer-
nillo, 15 miles southwest of Galeana: common on peak, June 16,
1934, C. H. and M. T. Mueller 828 (Gray Herbarium), June 29,
1934, F. W. Pennell 17126 (United States National Herbarium).

Character

Rootstock

V. galeanaensis

V. flagelliformis

V. pedunculata

more or less erect, | more or less erect, erect or horizontal,
but deep-seated not deeply buried; often as wide as long,
with many coarse | same type of roots buried 10 to 15 cm.;
roots numerous coarse
roots
Stems one-half subter- not at all or only one- to two-thirds
ranean; decum- slightly subterra- subterranean; erect
bent to erect nean; weak, decum- to decumbent; never
bent to trailing weak and trailing
Height up to 11 cm. up to 50 cm. up to 25 cm.
including
rootstock
Width up to 10 cm. up to 22 cm. up to 38 cm.
Pubescence | glabrous through- | conspicuously hirsute | microscopically
out | throughout except puberulent except
glabrous seed pods glabrous seed pods
Leaves broadly ovate, _ reniform to broadly ovate, subcordate to
shallowly cordate, | ovate, deeply cordate, | truncate, serration
remotely and regularly and con- irregular and remote;
faintly serrate; spicuously serrate; thick
thick not conspicuously
thickened
Stipules small and scarious | small, thin, green much larger, green
Flower small, pale laven- | small, yellow large, orange
der or white, with
yellow center
Pistil Chamaemelanium Chamaemelanium Chamaemelanium
type type type
Life-Zone Upper Sonoran to | Lower Sonoran Upper Sonoran
Transition ?

1947 | PORTER: OXYTROPIS 133

This violet is of interest to botanists of California because of
its affinity to Viola pedunculata T. & G. Although V. pedunculata
isa very much larger plant than V. galeanaensis, and its large yellow
flowers present a very different appearance, yet in vital characters,
as shown in the table below, these two species are much alike.
The closest relative of V.. galeanaensis is unquestionably the Mexi-
can species V. flagelliformis Hemsley, from which it is doubtless
derived. However, the characters of V. galeanaensis suggest that
it may be the original progenitor of the far distant Pacific Coast
species, V. pedunculata T. & G. The principal characters of these
three species are summarized in the above table.

Santa Rosa Junior College,
Santa Rosa, California.

A NEW SPECIES OF OXYTROPIS FROM THE CENTRAL
ROCKY MOUNTAINS

C. L. Porter

Oxytropis obnapiformis sp. nov. Subscaposis, sericeis, argen-
teis, erectis, perennis, 1-3 dm. altitudine; foliis pinnatis, 11—25-
foliolatis, foliolis oblongo-lanceolatis, 5-830 mm. longitudine, 2—4
mm. latitudine; stipulis adnatis petiolis; scapis foliis subaequalibus
vel longioribus, ca. 10—20 floris; corollis purpurascentibus, 15—20
mm. longis, leguminibus ovatis, inflatis, ad rostrum vehementer
contractis, villosis albis, subcoriaceis, basi ad rostrum 8-12 mm.
longa, 5-8 mm. lata, rostrum 5—8 mm. longum, sectione transversa
cordata, sutura superiore introflexa fere ad medium; semina com-
plura, reniforma, 1-2 mm. longa.

Grayish strigose or villous subscapose erect perennials, 1-3
dm. high, from a silky multicipital caudex surmounting a slender
taproot; leaves pinnate, mostly 11—25-foliolate, the leaflets oblong-
lanceolate, 5-30 mm. long, 2-4 mm. wide, grayish with somewhat
appressed silky pubescence; stipules adnate to the petiole, scari-
ous and villous-pubescent; scapes equaling or exceeding the
leaves, about 10—20-flowered, the inflorescence a spicate raceme ;
bracts about 5 mm. long, lanceolate; calyx cylindrical at time of
blooming, about 10 mm. long, grayish-strigose or villous, the teeth
lanceolate to oblong, 2-3 mm. long; corolla purplish, the banner
pale to white in the center with purple margin, 15-20 mm. long;
mature fruit splitting the calyx and exserted, ovoid and inflated,
with an abrupt slender beak, softly white-villous, somewhat trans-
versely wrinkled when dry, the texture thin-coriaceous, the body
8-12 mm. long, 5-8 mm. broad, the beak 5—8 mm. long, the cross
section cordate in outline with the ventral suture intruded about
to the middle; seeds several in each pod, reniform, 1-2 mm. long.

Type. Sand hills 8-9 miles west of Maybell, on U.S. 40,
elevation 5,900 ft., Moffat County, Colorado, Porter 3864, June

134 MADRONO [Vol. 9

19, 1946 (Rocky Mountain Herbarium, University of Wyoming,
Laramie. Isotypes, Gray Herbarium, Harvard University ; United
States National Museum; New York Botanical Garden; Missouri
Botanical Garden; Colorado Agricultural & Mechanical College;
University of California; and herbarium of R. C. Barneby, Wap-
pingers Falls, New York).

Cotypes. From type locality, July 6, 1945, Porter 3616 (Rocky
Mountain Herbarium, Gray Herbarium, New York Botanical

M1
#
/

WE SES

ie -S_S-= ~ + N
<= N SS SN
oN Se

1%
hig
CZ
NE
Ven
\

Fic. 1. Oxytropis obnapiformis Porter: 7, flower with wing removed;
2, fruit; 3, cross section of fruit. All approximately x 3.75.

Garden, Philadelphia Academy of Sciences, University of Wash-
ington, University of Texas, and Southern Methodist University) ;
from type locality, June 19, 1946, Harrington 1906 (Colorado
Agricultural & Mechanical College); from dry rocky hillsides
near Five Springs Falls, elevation 7,500 ft., Big Horn County,
Wyoming, July 11, 1936, Williams § Williams 3314 (Rocky Moun-
tain Herbarium) ; from the mouth of Shell Canyon, elevation 4350
ft., Big Horn County, Wyoming, Ripley § Barneby 8010 (Rocky
Mountain Herbarium, and the herbarium of R. C. Barneby).
This interesting species, common to the western slope of the
Big Horn Mountains of Wyoming and the dry sandy hills of north-
western Colorado, is named in allusion to the somewhat inverted
turnip-shaped pods, these being its most unique distinguishing
feature. It is undoubtedly a member of the section Campestres,
and in foliage and flower characteristics closely resembles mem-
bers of the O. Lambertit complex. It blooms early in June or even
in May in the type locality, since it was well past most of its

1947] REVIEW 135

blooming period when collected June 19, only a few flowers re-
maining on occasional plants. The specimen collected by Wil-
liams has both flowers and mature fruit; that collected by Barneby
has only fruits. The plants from Wyoming appear to have a
tendency toward a more exserted inflorescence than those from
Colorado, but this may be due to a habitat difference since there
is little else to distinguish them.

The writer is indebted to Mr. R. C. Barneby for making his
collection available for study, thus adding to our knowledge of
the known range of the species.

Department of Botany and
Rocky Mountain Herbarium,

University of Wyoming, Laramie,
(Contribution No. 203).

REVIEW

Los Juniperus Mexicanos. By Maximtno Martinez. Tom 17,
Anales del Instituto de Biologia de la Universidad Nacional de
Mexico, Mexico, D. F., 1946. 128 pp., 108 figs., paper cover.

A few months ago, in reviewing Professor Martinez’s book on
the genus Pinus in Mexico, I expressed the hope that he would
continue to produce papers of comparable excellence. In issuing
the above paper on the genus Juniperus in Mexico he has fulfilled
that hope. |

The first twenty-four pages of the paper are utilized in dis-
cussing the general characteristics of the genus and its representa-
tives in Mexico; a few paragraphs on the qualities of the lumber
produced by Mexican junipers; the vegetational zones in different
parts of the country; the subgeneric classification; and several
lists of species and lower entities based on such characters as size
of fruit, number of seeds and nature of the bark. This is followed
by a key to the species and full descriptions of the entities recog-
nized, together with citations of references and specimens ex-
amined.

Prior to 1944 only four species of Juniperus had been recog-
nized as occurring in Mexico. In this paper Professor Martinez
accepts twelve species, six varieties, and three formas. All of these
fit into Spach’s section Sabina, and Martinez has distributed them
among five subsections, the Flaccidae, Deppeanae, Jaliscanae, Monti-
colae and Monospermae. Of this number, four species (J. jaliscana,
Blancoi, durangensis, and Patoniana) ; six varieties (J. jaliscana var.
typica and var. poblana, J. Deppeana var. robusta and var. zacatecen-
sis, J. monosperma var. gracilis, and J. erythrocarpa var. coahuilensis ) ;
and three formas (J. Patoniana forma obscura, J. monticola forma
compactum and J. monticola forma orizabensis) are described as
new. One new combination and one new name also are proposed.
Two other species, J. Gamboana and J. comitans had been described
by Professor Martinez in 1944.

136 MADRONO [Vol. 9

He cleared up a puzzle centering around Cupressus thurifera
H. B. K. This species was known only from the type collection.
Professor Martinez reasoned that since no specimens, other than
the type, existed in any herbaria, and since no Cupressus was
known to grow in the area from which the type of C. thurifera had
come but “an abundance” of J. flaccida var. poblana occurred in
this region, probably C. thurifera was J. flaccida var. poblana. This
shrewd guess was completely confirmed when he examined the
type of C. thurifera! The latter, therefore, becomes a synonym
under J. flaccida var. poblana.

The paper is nicely illustrated with 108 “figures,” many of
which could easily have been classified as “‘plates’’ for they are
made up of a number of line drawings, each separate drawing
bearing a subnumber that corresponds with the numbered expla-
nations of the figures. Over one-third of the figures are half-tone
reproductions of photographs showing the general habit, charac-
ter of the twigs, leaves and fruit or the nature of the bark on the
mature trunks. The line drawings accurately depict the shape,
arrangement and size of the leaves and fruit and the presence or
absence of glands on the leaves. These drawings are refreshingly
simple and accurate, without shading and clearly show the critical
character.

Under each subsection Professor Martinez includes a schematic
chart representing the “‘supposed relations” among the entities
included. These charts express his ideas of the probable relation-
ships, based wholly upon his field observations and his herbarium
studies. Limits of time and funds precluded the tedious and very
slow experimental type of nursery work that could give more defi-
nite criteria on the phylogenetic lines along which the Mexican
junipers have developed.

The paper constitutes a substantial contribution to our knowl-
edge of the genus Juniperus as it occurs in Mexico. The format
and typography are good. The paper used is glazed so both the
half-tones and the line drawings are reproduced clearly. This is
another paper written by Professor Martinez for which we may
sincerely thank him. It is one that will be valuable to any bota-
nist interested in the flora, and particularly one interested in the
gymnosperms of Mexico.—Ira L. Wiaerns, Stanford University,
California.

, ot

MADRONO
A West American Journal of Botany

A quarterly journal devoted to important
and stimulating articles dealing with
plant morphology, physiology, taxonomy,
and botanical history. These volumes
should be a part of every botanist’s li-
brary and should be made accessible to
students of all universities and colleges.

Volume I, 1916-1929... $5.00
Volume II, 1930-1934 .. 5.00
Volume III, 1935-1936 . 5.00
Volume IV, 1937-1938 . 5.00
Volume V, 1939-1940 . . 5.00
Volume VI, 1941-1942 . 5.00
Volume VII, 1943-1944 . 5.00
Volume VIII, 1945-1946 . 5.00
Single numbers....... 0.75

The subscription price of MADRONO
is $2.50 per year. We solicit your pat-
ronage,

Address all orders to:
Reed C. Rollins, Bus. Mer.

Natural History Museum,
Stanford University, California

w

VOLUME Ix NUMBER 5

MADRONO
A WEST AMERICAN JOURNAL OF
BOTANY

ué

Contents
ENVIRONMENTAL EXTREMES AND Enpemism, LeRoy E. Detling .......... 137
NoTewortHy SoutH AMERICAN Puants, I ann II, Paul C. Standley and
MS ha EV atag al el CATED ERIE ll es ea a Re RS AA re A a Ee 149
Nores on Paciric Coast Marine Ancas, G. J. Hollenberg ................. 155
On Two Perenniart Carspitose Lepiptums oF WESTERN NortH AMERICA,
CCU OU LCOLRIGY NEED (Mier Bigle caialcls Waiere ahd we wide ae Perea A eg 162
Review: Ernest Brown Babcock, The Genus Crepis (Marion Ownbey) .... 165

Nores ann News: A Possible Record of Quercus Morehus im Oregon,
MrT. MAULEWS's) IVGZ08 oc cle em ets Wauseon DEG he. 168

Published at North Queen Street and McGovern Avenue,
Laneaster, Pennsylvania

January, 1948

MADRONO

A WEST AMERICAN JOURNAL OF BOTANY
Board of Editors

Herpert L. Mason, University of California, Berkeley, Chairman.

LeRoy Asrams, Stanford University, California.

Epcar ANDERSON, Missouri Botanical Garden, St. Louis.

Lyman Benson, Pomona College, Claremont, California.

Hersert F’. Copetann, Sacramento College, Sacramento, California.

Ivan M. Jounston, Arnold Arboretum, Jamaica Plain, Massachusetts.

Mitprep EF. Martutas, Dept. of Botany, University of California, Los Angeles 24.
Bassett Macuire, New York Botanical Garden, N. Y. C.

Marion Ownsey, State College of Washington, Pullman.

Secretary, Editorial Board—ANNETTA CARTER
Department of Botany, University of California, Berkeley

Business Manager—Rimo BaciGaLuri
North Queen Street and McGovern Avenue, Lancaster, Pennsylvania

or
Natural History Museum,
Stanford University, California

Entered as second-class matter October 1, 1935, at the post office at
Lancaster, Pa., under the act of March 3, 1879.

Established 1916. Published quarterly. Subscription price $3.50 per year.
Completed volumes I to VIII $5.00 each; single numbers $1.00.

Papers up to 15 or 20 pages are acceptable. Longer contributions may
be accepted if the excess costs of printing and illustration are borne by the
contributor. Range extensions and similar notes will be published in con-
densed form with a suitable title under the general heading ‘Notes and News.”
Articles may be submitted to any member of the editorial board. Manuscripts
may be included in the forthcoming issue provided that the contributor pay the
cost of the pages added to the issue to accommodate his article. Reprints of
any article are furnished at a cost of 4 pages, 50 copies $5.90; 100 copies $6.48;
additional 100’s $1.23; 8 pages, 50 copies $8.55; 100 copies $9.49; additional 100’s
$1.88; 16 pages, 50 copies $12.00; 100 copies $13.44; additional 100’s $2.88.
Covers, 50 for $3.95; additional covers at $2.38 per hundred. Reprints should
be ordered when proofs are returned.

Published at North Queen Street and McGovern Avenue, Lancaster,
Pennsylvania, for the

CALIFORNIA BOTANICAL SOCIETY, INC.

President: Adriance S. Foster, University of California, Berkeley. First
Vice-President: G. Ledyard Stebbins, Jr., University of California, Berkeley.
Second Vice-President: Herbert F. Copeland, Sacramento College, Sacramento,
California. Secretary: John Whitehead, University of California Botanical
Garden, Berkeley. Treasurer: Rimo Bacigalupi, Natural History Museum,
Stanford University, California.

Annual membership dues of the California Botanical Society are $3.50,
which includes a year’s subscription to Madrofo. For two members of the same
family the dues are $4.00, which includes one copy of Madrofo and all other
privileges for both. Dues should be remitted to the Treasurer. General
correspondence and applications for membership should be addressed to the
Secretary.

1948] DETLING: ENDEMISM 137

ENVIRONMENTAL EXTREMES AND ENDEMISM
LeRoy E. DEetTLIneG

One of the striking features of the distribution of plant species
is the effect produced by varying intensities of the major environ-
mental factors, mostly climatic, upon the composition of the plant
association, and suggests one basis for the delimitation of geo-
graphical areas in studies of plant speciation and distribution.
This effect is the subject of the present paper.

Several years ago a transect through western and central
Oregon was selected for special investigation. This transect is
fourteen miles wide and extends from the mouth of the Siuslaw
River on the Oregon coast in a general easterly direction to a
point between Redmond and Prineville, a distance of about 160
miles. In its wide extent from the coastal dunes and headlands,
through the humid slopes and high peaks of two mountain ranges,
to the continental plateau of central Oregon, it probably includes
as many and as varied environments as can be found within the
same distance anywhere in the Pacific Northwest.

A profile of the transect (fig. 1) shows first, beginning at the
western end, a narrow strip composed of dunes in various stages
of maturity cut through on the north by Heceta Head, a massive

Fic. 1. Profile of the transect, showing relative positions of the vegetation
zones.

igneous intrusion forming a rocky headland. This strip is flanked
on the east by the relatively low Coast Range, whose highest
points in this region do not exceed 3000 feet. Beyond this is the
Willamette Valley, its floor averaging perhaps 450 feet in eleva-
tion. To the east of this valley rises the Cascade Range, deeply
dissected by the valleys of the McKenzie River and its tributary,
White Branch. In general, its peaks do not surpass 5000 or
6000 feet in altitude, with the highest points, the Three Sisters,
slightly over 10,000 feet. East of the Cascades the transect
extends out over a plateau of about 3200 feet elevation, from
which rise several prominences. One of these, Powell Butte,
reaches approximately 5600 feet.

Manprono, Vol. 9, No. 4, pp. 105-136. November 20, 1947.

138 MADRONO [Vou. 9

A striking variation in climate is due to the prevailing westerly
winds from the Pacific Ocean which pass over two parallel moun-
tain ranges, giving rise to an oceanic climate west of the Cascades
with abundant rainfall and moderate temperatures at the lower
altitudes and heavy snowfall in the higher mountains, and a con-
tinental climate on the plateau east of the Cascades with much
reduced precipitation and greater extremes of temperature.

Cuimatic Factors

Source or THE Data. The climatic data used in this investiga-
tion are based upon the annual summaries issued by the United
States Weather Bureau. The factors which have been considered
are: (1) average annual precipitation, (2) average precipitation
for the combined months of June, July and August, (3) average
minimum temperatures for the month of January, and (4) average
of the maximum temperatures from April to September, inclusive.

Soil texture, slope of the terrain, conditions of light and shade,
and other environmental factors have an important bearing on the
constitution of the flora, but their effect in this area is largely local
and they are left for later investigation. Some exception has
been made in the case of soil salinity in the Littoral Belt, for
reasons to be made clear later.

Owing to the large area included in the transect, the relative
inaccessibility of the higher elevations during the winter months,
and the lack of assistance in carrying on the project, it has not
been feasible as yet to set up recording instruments throughout
the transect for the collection of climatic data.

Climatic records for a number of years are available for the
following points on the transect: Florence, Deadwood, Eugene,
Vida, McKenzie Bridge, Sisters and Redmond. These stations
are so located that it is possible to construct from their data
curves for precipitation and temperatures for the whole transect.
For some points in the higher elevations between McKenzie
Bridge and Sisters, where local temperature differentials are im-
portant, it has been impossible to obtain these data directly. For
this area temperatures have been computed on the basis of the
general formula of a decrease of 3.3° F. for each increase of 1000
feet in altitude (Finch and Trewartha, 1942, p. 48). The gradi-
ents so computed are found to be in harmony with temperatures
recorded from stations at comparable altitudes and locations in
the Central Cascade area.

Precipitation. Annual rainfall (fig. 2) at the coast is about
70 inches; ascending the west slope of the Coast Range this in-
creases to a maximum of over 90 inches, decreasing rather
abruptly over the east or leeward slope to about 30 inches. On
the west slope of the Cascade Range precipitation again increases
from 37 inches at Eugene to 71 at McKenzie Bridge. East of the
Cascade summit we again find a rapid decrease in annual precipi-

1948 ] DETLING: ENDEMISM 139

tation, with 16 inches at Sisters, a few miles from the base of the
mountains, and less than 8 inches at Redmond, 19 miles farther
out on the plateau.

10 IN

Fic. 2. Annual precipitation (transect profile in broken line).

From the standpoint of the maintenance of a particular flora
the seasonal distribution of precipitation is extremely important.
Throughout the transect most of the precipitation occurs during
the months from October to April, but the amount of soil moisture
probably does not become critical for any plant species before
June. The term “summer rainfall” as used in this paper refers
to the total for the months from June through August, by which
time most of the vegetative growth and seed production of plants
in this region have been completed.

The curve of summer rainfall for the transect (fig. 3) is
similar to that of the total annual amount in that it shows two
maxima, one on the west slope of each of the mountain ranges.
The chief difference lies in the fact that the greater maximum of
summer rainfall occurs in the Cascades rather than in the Coast

Range.
100 IN

Fic. 3. Summer precipitation.

TEMPERATURE. The usual effect of low temperature in limit-
ing the distribution of a plant species is winter-killing, and it is
obvious in this case that it is the absolute low temperature which
is actually the limiting factor. However, it would have been
difficult, if not impossible, to obtain these particular data for most
of the transect. For the purposes of comparison for which it is
to be used in this study the average of the lowest temperatures for
the coldest month of the year should reflect sufficiently well the
extremes to which the winter temperatures fall in the various
parts of the transect. The minimum temperatures referred to in
the following pages are the averages for a period of years of the
minimum temperatures recorded for the month of January, the

140 MADRONO [Vox. 9

month in which the lowest minima are recorded for all stations on
the transect.

The importance of high temperatures for plant growth lies in
the total amount of heat supplied in a given length of time.
Here again, due to lack of recorded data, it has been impossible
to determine this summation of temperature for any of the
stations along the transect. Records are available, however, for
the average maximum temperature for every month of the year,
and the total of these temperatures during the growing season
should reflect with sufficient accuracy for our purpose the vari-
ation in total solar heat among the vegetation areas under con-
sideration. We have considered the months from April to
September, inclusive, as the growing season. ‘Instead of at-
tempting to approximate a true summation of temperatures for
this period we have reduced them to a more readily obtainable
average of the maximum daily temperatures for the six months.
$00°E

Fic. 4. Average January minimum temperature.

Disregarding local variations, we may make the following
general statements regarding temperatures on the transect:

Average January minima (fig. 4) are highest at the coast,
gradually falling as we go inland and as we ascend to the summit
of the Cascade Range. East of the summit winter temperatures
rise as we descend to the plateau, although even here they do not
reach the high level prevalent over so much of the area west of
the Cascades.

Maximum summer temperatures (fig. 5) are relatively low
(64.1° F.) at the coast. These increase eastward over the Coast
Range until a first maximum (75.6° F.) is reached in the McKen-
zie Valley, after which they decrease as one ascends the Cascade
Range. On the east slope of this range summer temperatures
again rise until the highest maximum for the transect (76.3° F.)
is reached near Redmond.

Puiant ASsociATIONS

Data on the distribution of plant species are based largly upon
collections and field observations made by the writer over a period
of several years. This has been supplemented by study of other
collections in the herbarium of the University of Oregon when the
collectors’ ecological data have been sufficiently full and definite
to warrant their use. Distributional data drawn from collec-

1948] DETLING: ENDEMISM 141

tions made on the transect have been checked frequently by
observations made in field work in adjacent localities in western
and central Oregon.

The listing of species to be found on the transect is by no
means complete. However, while continued field work will un-
doubtedly reveal more species, they will be scattered through all
the associations, and their number will not be so great as to affect
appreciably the relative value of the data or the conclusions to
be drawn from them. This preliminary work is based upon 634
species and subspecies of spermatophytes, representing 275
genera.
100°F. -

Fic. 5. Average summer maximum temperature.

The plant associations used in this study are the major ones of
the region. They are the ones which have evolved in response to
combinations of the major climatic factors. They sometimes
coincide with forest types, where these mark the more important
environmental differences. However, these larger units have not
always been found useful as a basis for zone delimitation in the
present study. In some cases a certain group of plant species is
quite uniformly associated with a single dominant tree species,
such as the western juniper. In the Douglas fir forest, on the
other hand, the dominant species is accompanied by a well-
marked association in its upper limits which is not found in its
lower, each of these associations being clearly dependent upon
certain climatic factors for its existence. The zones used here,
then, have been determined not by the presence of any one indi-
cator species, but rather by an association of species due to major
climatic factors. Smaller associations, dependent upon local
conditions, are not recognized.

In the following brief summary of the vegetation zones as
treated in this study we have listed the “indicator association,”
i.e., the association of species characterizing that particular zone.
It is, of course, to be understood that all the species here listed for
any one indicator association may not always be present in that
association. On the other hand, the presence in a given area of
only one or two members of an indicator association cannot
always be considered sufficient to mark the existence of the zone
in which they ordinarily occur.

Lirrorat Beit. This belt is restricted to the coastal dunes,
headlands and tide flats, where the vegetation is subjected to the

142 MADRONO [ Von. 9

constant winds, shifting sands, brackish water, salt spray and
other direct influences of the ocean. Its indicator association is
composed of Carex macrocephala, Convolvulus soldanella, Lupinus
littoralis, Abronia latifolia, Lathyrus littoralis, L. maritimus, Glehnia
leiocarpa, Fragaria chilensis and Potentilla anserina.

Coastat Forest. In that section of the Oregon coast covered
by the transect this association extends but a short distance inland
from the previous one. It is better developed and more extensive
farther northward. On the whole, it is a dense forest. Climatic
conditions are favorable to the formation of sphagnum bogs.
The zone is marked by the presence of Picea sitchensis, Pinus
contorta, Myrica californica, Salix Hookeriana, Vaccinium ovatum, V.
uliginosum, Chrysamphora californica and Ledum columbiananum.

Oak Wooptanp. This type occurs in relatively small areas in
the Willamette Valley and on the lower west slopes of the Cas-
cades. It frequently occurs as a successional stage preceding the
Lower Montane Forest, but at other times is apparently the
climax association, and as such constitutes a northern extension of
the deciduous woodland of the Umpqua and Rogue River Valleys
of southern Oregon. The zone is here taken to include the open
meadows and grasslands and exposed rocky hillsides, whose flora
includes largely the same species as those found in the woodland
proper. Indicator association: Quercus Garryana, Q. Kelloggit,
Arbutus Menziesii, Rhus diversiloba, Rosa nutkana, Potentilla gracilis,
Dichelostemma pulchellum and Fritillaria lanceolata.

Lower Montane Forest. This comprises the Douglas fir
forest from its lower limit up to an altitude of about 4000 feet.
It does not occur on the east slope of the Cascade Range. The
indicator association is made up of Pseudotsuga taxifolia, Acer cir-
cinatum, Cornus Nuttallii, Vaccinium parvifolium and _ Berberis
nervosa.

Mippte Montane Forest. The Douglas fir forest from about
4000 feet to its upper limit, while containing many of the species
found in the preceding zone, is characterized by the presence of a
number of species not found in its lower elevations. This zone
occurs on both slopes of the Cascades, but the areas occupied by
it on the two slopes are separated from one another by the Sub-
alpine Forest above them. On the east slope Pseudotsuga is
largely replaced by Abies grandis. Characteristic species of the
Middle Montane Forest are Pseudotsuga tazifolia, Abies grandis, A.
procera, Acer glabrum, Garrya Fremontiu, Gaultheria ovatifolia, Vac-
cintum membranaceum, Calochortus Lobbii, Eriogonum compositum
and Cornus canadensis.

SuBALPINE Forest. The Subalpine Forest occupies a rela-
tively narrow strip along the summit of the Cascades, extending
up to timberline, and occurs on the summits of some of the more
isolated peaks that rise up out of the lower zones, such as Horse-
pasture and Carpenter Mountains. The association marking the

1948 | DETLING: ENDEMISM 143

zone is made up of T'suga Mertensiana, Abies lasiocarpa, A. amabilis,
Pinus contorta Murrayana, P. albicaulis, Vaccinium scoparium, Phyl-
lodoce empetriformis, Cassiope Mertensiana, Luetkea pectinata and
Vaccinium deliciosum.

Aupine Tunpra. All vegetation above timberline may well
be included in a single association. On this transect, tundra is
restricted to a few peaks which reach about 7000 feet elevation
ormore. The highest of these are the North and Middle Sisters,
both of which slightly exceed 10,000 feet. The indicator as-
sociation of this zone consists of Hulsea nana, Collomia Larsenii,
Polemonium elegans and Oxyria digyna.

Ponverosa Pine Forest. The whole belt dominated by ponderosa
pine is included in this zone. It occupies the lower east slopes of the
Cascades and the adjacent plateau, extending approximately from
3200 to 4000 feet. Its indicator association is composed of Pinus
ponderosa, Purshia tridentata, Ceanothus velutinus and Arctostaphylos
patula.

JunipER Woopianp. This association occupies the plateau
adjoining the Ponderosa Pine Forest. Its flora is essentially the
same as that of the sagebrush plains or desert shrub formation to
the east, but includes a notable stand of western juniper. The
zone is characterized by the presence of Juniperus occidentalis,
Artemisia tridentata, Chrysothamnus spp., Astragalus spp. and Lu-
pinus corymbosus.

Hereafter, the names of these zones may be abbreviated as
follows: L, Littoral Belt; C, Coastal Forest; O, Oak Woodland;
LM, Lower Montane Forest; MM, Middle Montane Forest; S,
Subalpine Forest; A, Alpine Tundra; P, Ponderosa Pine Forest;
J, Juniper Woodland.

In any consideration of the effectiveness of the environment in
determining the constitution of a given plant association we must
have some acceptable criterion for judging the distinctness of
that association, i.e., the degree of difference between it and any
other association. The most reasonable basis for such judgment
would seem to be the number of species endemic in any given as-
sociation or zone. If environmental selection has been rigorous
we should expect to find more highly specialized species com-
prising the flora of an area, in other words, more species which
can grow only in that set of environmental conditions. The
actual number of species endemic in an association is, of course,
less significant than the percentage which these comprise of the
total number occurring in it. The term “endemic,” as employed
here, simply means restricted to the area in question, with no
reference to breadth or narrowness of range, or any other pecu-
liarities of distribution. The number of endemics for each zone
of our transect, and the percentage which this number represents
in the total flora of that zone are shown in the following table:

144 MADRONO [Vot. 9

Taste 1. PercentTaGe oF ENDEMISM IN THE VEGETATION ZONES

| No. endemic species % of total
Littoral’ Belt}: 45) see 27 51
Coastal. Forest: 4)...7.... 2. 73.. 14 17
Oak Woodland ............:.. 61 41
Lower Montane Forest ....... 74 27
Middle Montane Forest ....... 12 if
Subalpine Forest ............. 69 43
Alpine: Tundra: 02.2) ue. 9 53
Ponderosa Pine Forest ....... 20 27
Juniper Woodland ........... 57 64

On the basis of these figures the Junipér Woodland is the most
distinctly marked of all the associations, followed by the Alpine
Tundra and the Littoral Belt, while the Middle Montane Forest
is the least so.

Tue AssociaATION-ENVIRONMENT RELATION

This relationship between the vegetation zones and _ the
various environmental factors is best observed when the east and
west slopes of the Cascade Range are treated separately.* The
following two tables (tables 2 and 3) present for each vegetation
belt the range of each of the four climatic factors under consid- |
eration here. Since the Middle Montane is a discontinuous zone,
the ranges of its climatic factors are treated separately for its east
and west slopes. In the case of the Subalpine Forest and Alpine
Tundra, since their areas are continuous over the summit, the
ranges of the climatic factors include the extremes for both
slopes. Precipitation is given in inches, temperatures in degrees
Fahrenheit.

The data of the foregoing tables can be translated into graphs
so constructed as to show the relative position of each zone in the
whole range of any given climatic factor (figs. 6 and 7). The
vegetation zones on each slope are arranged in ascending order
from the lowest in elevation to the highest. The left-hand
margin of the graph represents in each case one extreme of each
of the factors, and the right-hand margin the other extreme.
Each bar of the graph represents that portion of the total range
of a factor which is included within the zone indicated, as well as
its relative distance from each of the extremes.

Bearing in mind the figures of Table 1, on the degree of en-
demism in the flora of the various associations, an examination of
the foregoing graphs reveals some interesting and significant
facts.

Considering first the east slope, it will be noted that the high-
est percentages of endemism occur at the extremes of the slope,
viz., 64% in the Juniper Woodland and 53% in the Alpine

1948] DETLING: ENDEMISM 145

Tundra. Endemism decreases as one approaches the middle
altitudes, with 43% in the Subalpine Forest and 27% _ in the
Ponderosa Pine Forest. The lowest degree of endemism, 7%,
is found in the zone which occupies the middle position on the
slope, i.e., in the Middle Montane.

TasBie 2. RANGE OF CxiImMATIC Factors IN THE East Store ZONES

Annual Summer January Summer
precipitation precipitation min. temp. max. temp.

7.7-16.7 1.7-2.0 20.0—20.8 72.9-76.3

‘Pp 16.7-24.5 1.7-2.5 17.4-20.0 71.1-72.9
MM 24.5-45.0 2.5-3.2 15.5-17.4 69.8—71.1
S 45.0—71.2 3.04.6 6.2-15.9 63.7-70.0
A 36.0-64.5 3.0-3.9 —1.8- 7.2 57.3-63.7

TasLe 3. Rance or Curmatic Factors 1x THE West Store ZONES
Annual Summer January Summer
precipitation precipitation min. temp. max. temp.
L 69.4 3.5 38.5 64.1

Cc 69.4—72.0 3.5-4.0 38.0-38.5 64.1-64.5
O 32.0—40.0 1.5-2.5 32.1-34.5 72.0—72.5
L 32.0-92.0 1.5-4.6 20.9-38.0 64.5-75.6
MM 66.0—72.0 3.9-4.6 15.9-20.9 70.0—72.5
S 45.0—71.2 3.0-4.6 6.2-15.9 63.7—70.0
A 36.0-64.5 3.0-3.9 —1.8— 7.2 57.3-63.7

Now it may be noted that the climatic factors also form a
gradient on the slope, with the greater number of extremes oc-
curring in the Juniper Woodland and Alpine Tundra, the least
extreme conditions being found in the Middle Montane. More
specifically, all four climatic factors reach one of their extremes
in the Juniper Woodland. In the Alpine Tundra two factors,
January minimum and summer maximum temperatures, reach one
extreme, and the amount of precipitation is exceeded in only one
other zone. In the Subalpine Forest the two precipitation factors
are at one of their extremes while the summer and winter temper-
atures are roughly about midway of their extremes. The general
average of extremeness for all factors is less in the Subalpine
Forest than in the Tundra. In the Ponderosa Pine Forest only
one factor reaches an extreme. In the middle Montane Forest
no factor is at its extreme; the graph shows all the bars for this
association to be grouped fairly well toward the mid-points of
the factor ranges for the whole gradient.

On the west side of the Cascade divide, if we leave for later
consideration the Littoral Belt and Coastal Forest, we find a
situation quite comparable with that on the east slope, in that we

146 MADRONO [ Vox. 9

16.7 LEGEND

Kl ANNUAL PRECIP. (/NCHES)
WON SUMMER PRECIP. (/NCHES)

Pz] JANUARY MIN - TENP. ( °F.)

[Tl] SUMMER MAX: TEMP. (°F)

20.8
76.3 TATHTOUGATAOTAOENEUINTNUTIT 72.9

P
24. SWIMM, KLM LiL *5.0
2.5 BSS
mn 17. 4G IS
Tel ‘(oo 69.8
EE KK Lhd li KEELE R71 .2
S 15.9 [Erg nee en preteen p bene a

eres ene soe oS 7 sites eH hs 8
63. T TTATTATHTAAHUATUVUTAUVTOTTVGAUIAUITAURATTTATL 5T.3>

Fic. 6. Range of climatic factors in the east slope zones. (Based on
Table 2). J—Juniper Woodland. P—Ponderosa Pine Forest. MM—Middle
Montane Forest. S—Subalpine Forest. A—Alpine Tundra. .

774 69.4
a W>-5
LT sas
mm.
69.4 2.0
4.0
C} 38.5[£3 38.0 32
64.5] 64.)
32.0 WITT ZZ SO. 0
ol] 15 SSR IASS SASS 2.5
72.5 (i) 72.0
aH OLZZZ’LLL LLL LL LLL LLL LLL LLL LLL
LM] 1.5 RIS MEMS NAMA MAAS 4.6
28.9 é (tuto TUTELTHITEAMLOTOVATSATOTCAUAUOATOUETAULGATACAUGTUULITITVOIIN 64.5
66. PR AAL, TA 72.0
ni a S. DRYLAND nnn 4.6
MM 20.9 FR:
72.5 (TMT 70 - O
45.0 WZILLLLLLLLLLLLLLLL LL 7.2
5 Sa SSS SS SS 4.6.
70.0 IOULATOATOEBAATATASEOAIONCTUTTAUTTAATUATITI <i
3©.OL 777A LILI LL LL LLL LLL LLL LL OAS
, SSSA 15-9

Fic. 7. Range of climatic factors in the west slope zones. (Based on
Table 3). L—Littoral Belt. C—Coastal Forest. O—Oak Woodland. LM—
Lower Montane Forest. MM-—Middle Montane Forest. S—Subalpine Forest.
A—Alpine Tundra.

have a gradient of climatic factors from the Oak Woodland up to
the Alpine Tundra, coupled with a high degree of endemism in
those plant associations where most of the climatic extremes
occur and a low degree of endemism where climatic conditions
are for the most part moderate. 'The Tundra combines the highest
degree of endemism, 53%, with extremes in both summer and
winter temperatures and moderate precipitation. The Subalpine
Forest and Oak Woodland have almost the same degree of

]

1948] DETLING: ENDEMISM 147

endemism, 43% and 41% respectively. In the former it will be
noted that one factor, summer rainfall, reaches one of its ex-
tremes, while the lower extremes of both summer and winter
temperatures in this zone are surpassed only in the adjacent
Tundra. The graph shows how in the Oak Woodland all factors
are grouped at or near one of their extremes. The Lower Mon-
tane Forest is moderate in degree of endemism (27%) ; two of the
climatic factors, annual and summer precipitation, cover in this
one association the entire range for the west slope, while temper-
atures range from an extreme at one end well up into the middle
portion of the total gradient. The Middle Montane Forest, with
only 7% endemism, the lowest on the slope, has an extreme in
only one factor, viz., summer rainfall, the other three being
scattered through the middle portion of their ranges in this zone.

The foregoing data lend support to an hypothesis which, in-
so-far as this transect and the four climatic factors under consid-
eration are concerned, might be stated thus: The degree of en-
demism in the flora of a given area is directly correlated with the
total number of environmental factors reaching their extremes in
the area. The greater the number of factors which reach or ap-
proach one of their extremes in an area, the greater will be the
degree of endemism in the flora of the area.

Two questions regarding the west slope associations present
themselves. The first and undoubtedly the most obvious one is:
Why does the Lower Montane Forest have only 27% endemism
when the association embraces not only one extreme of summer
maximum temperature and practically one extreme of January
minimum temperature, but also both extremes of the two pre-
cipitation factors? The answer lies in the very fact that climatic
conditions in the association are so varied that they do reach both
extremes or at least have a very wide range. It will be noted
that in the graph the bars for the Lower Montane Forest overlap
all or many of the corresponding bars in adjacent associations.
This means that in the various parts of its wide range where the
climate approaches that of other associations there will be an ex-
change of species between the zones with a resultant decrease in
the number of endemic species.

The second question arises when it is noted that in the Oak
Woodland the range of each climatic factor is entirely overlapped
by the range of the same factor in the Lower Montane Forest.
Since the climate of the Oak Woodland is thus reproduced in cer-
tain parts of the Lower Montane Forest why does not this area
support such a forest? The answer may be sought in the fact
that all the factors in the woodland are so near to the extremes
present in the Lower Montane that any other environmental con-
dition that increases the effect of one of these extremes will be
sufficient to change the whole plant association. Such a con-
dition is the shallow soil frequently found on the low hills of the

148 MADRONO [ Vou. 9

Willamette Valley, as well as increased insolation due to slope
of the terrain. Early drying of these soils decreases the effec-
tiveness of annual and summer precipitation and shifts an already
marginal Lower Montane Forest into an Oak Woodland. It is
worthy of note that in the Umpqua and Rogue Valleys, where the
Oak Woodland occupies a much deeper soil, the annual and sum-
mer precipitation both fall below the lowest extremes for these
factors in the Willamette Valley.

Consideration of the Coastal Forest and the Littoral Belt has
been left for this time since the climatic factors in these as-
sociations do not form a part of the regular climatic gradient on
the west slope of the Cascades. In the case of the Coastal Forest,
however, as the graph shows, only one of its climatic factors, viz.,
January temperature, approaches an extreme, the others being
definitely median in their position. It is therefore in keeping
with the hypothesis advanced above that the distinctness of the
flora in this forest is represented by only 17% of endemism.

At first glance the Littoral Belt would seem to offer an excep-
tion to our hypothesis. Its climatic factors have practically the
same positions as do those of the Coastal Forest, yet the Littoral
flora has an endemism of 51%, second only to that of the Alpine
Tundra among the west slope-associations. However, bearing in
mind that one of the main features of the environment of the
Littoral Belt is soil salinity of such degree as to create a condition
of extreme physiological drought, we see that actually the avail-
ability of both annual and summer precipitation is so modified
that we could well consider them to be at their lower extreme here.
This low extreme of available soil moisture, coupled with the
upper extreme of winter temperature, and the fact that summer
temperatures are lower here than in any association except the
Subalpine Forest and the Tundra, may easily explain the high
degree of endemism in the Littoral association.

Since the summation of environmental extremes is so clearly
reflected in the distinctiveness of floras in various geographical
areas, this should be taken into account in our studies on the evolu-
tion of floras. To one familiar with areas of endemism, a survey
of climatic maps shows that the centers of floral areas richest in
endemic species coincide with areas in which occur the end points
of a number of climatic gradients. It is clear that our investiga-
tions on species distribution and the evolution of species and sub-
species should be based upon climatic areas so constituted. As
a specific example, taxonomic and distributional studies on one of
the sections of the genus Lupinus, well represented and wide-
spread in the Pacific Northwest, are being carried on by the writer.
These studies have this type of area as their basis, and it has been
found that the geographic pattern of species and subspecies fol-
lows closely the pattern of summations of environmental extremes.

1948] STANDLEY AND BARKLEY: SOUTH AMERICAN PLANTS 149

SUMMARY

The degree of endemism in the nine major plant associations
on a transect through western and central Oregon is compared
with annual precipitation, summer precipitation, January mini-
mum temperature and summer maximum temperature in the same
associations. It is found that the higher percentages of endemism
occur in those associations in which the greater number of climatic
factors reach one of their extremes. The hypothesis is advanced
that the degree of endemism in the flora of a given area is directly
correlated with the total number of environmental factors reach-
ing their extremes in the area. It is pointed out that the centers
of floral areas richest in endemic species coincide with areas in
which occur the end points of a number of climatic gradients, and
it is suggested that taxonomic and distributional studies be based
upon geographic areas constituted along these lines.

Museum of Natural History,
University of Oregon, Eugene

LITERATURE CITED

Fincnu, V. C., and G. T. TrewartHa. 1942. Elements of geography, physical
and cultural. 2nd Ed.

NOTEWORTHY SOUTH AMERICAN PLANTS, I AND II
Paut C. STANDLEY AND FrepD A. BarkKLey

In working over portions of recent collections of plants from
Peru and Bolivia collected by H. E. Stork, O. B. Horton, W. J.
Eyerdam, J. Soukup S.S., C. Vargas C., and T. H. Goodspeed, five
specimens have appeared which are sufficiently distinct from the
known species of that area that they seem worthy of description.

The types of these are all deposited in the Herbarium of the
Chicago Natural History Museum.

Phoradendron Storkii Barkley, sp. nov. Frutex glaber, para-
siticus, dioicus; foliis coriaceis, lanceolatis, oppositis, 130 mm.
longis, 20 mm. latis, basi cuneatis, decurrentibus, subsessilibus,
apice obtusis, 3—5-nerviis; cataphyllis 2, late deltoideis, ad nodum
primum ramis primarii; internodiis basi tetrangularibus (infimis
exceptis), apice planis et late alatis (alis caulies latitudine
aequantibus) ; inflorescentiis erectis et in axillis foliorum soli-
tariis; pedunculis circa 5 mm. longis; bracteis 2 inflorescentiam
subtendentibus; inflorescentiis 1 ad 3 cm. longis; petalis 3.

Shrubby, dioecious parasite on trees with stems to 60 cm. long,
nodes of stem yellow and the remainder of stem and leaves chloro-
phyll green; leaves thinly coriaceous, lanceolate, opposite, 180

150 MADRONO [Vou. 9

mm. long, 20 + broad, base cuneate, decurrent, subsessile, taper-
ing gradually to the narrowly obtuse apex, 3 to 5 nerved from
near the base; a pair of very broadly deltoid cataphylls at the first
node of each main branch; all but the lowest internodes obscurely
tetrangular at the base becoming flattened and broadly alate at
the apex (the wings equal to the stem in width) ; inflorescences
erect and solitary in the axils of the leaves; peduncles about 5 mm,
long; a pair of broadly deltoid bracts subtending the inflores-
cence; inflorescence 1 to 3 cm. long; petals 3; fruit white, cylin-
dric-ovoid, 2.5 mm. long, 1.5 mm. broad.

Type. Altitude 3200 m., 10 km. northwest of Socota, Depto.
Cajamarca, Peru, December 10, 19388, H. E. Stork § O. B. Horton
10133. :

This species apparently is closely related to Phoradendron
Mathewsi Trel. and P. semiteres Trel., but it differs in leaf, stem,
and inflorescence characters.

Cleome Eyerdamii Standley & Barkley, sp.nov. Suffrutex, 10

ad 30 cm. altus; foliis trifoliolatis; foliolis 3 (—5?), obovatis, 1-3
em. longis, cuneatis, sessilibus, ad apicem subacutis; pedicellis
circa 1.5 cm. longis, puberulentis glandulosis; capsulis glabris,
circa 2 cm. longis, 3-5 mm. latis; stipitibus circa 1 mm. longis.

Bush or perennial herb 10 to 30 cm. tall with woody stem and
peculiar odor; leaves trifoliolate, petiole slightly exceeding the
lamina in length; leaflets 3 (—5?), obovate, 1-3 cm. long, half as
wide, cuneate, sessile, at apex subacute; a pair of short stipular
spines at the base of the petiole; stems, leaves, bracts, and pedi-
cels densely glandular puberulent; bracts conspicuous, persistent,
ovate, about 1 cm. long, subsessile; sepals broadly lanceolate, half
as long as the petals, glandular puberulent; petals white, clawed,
obovate, about 6 mm. long; pedicels in fruit about 1.5 cm. long,
glandular puberulent; fruit glabrous, stipe about 1 cm. long, cap-
sule about 2 cm. long, 8-5 mm. broad.

Type. About 5 km. southeast of Cochabamba, Depto. Cocha-
bamba, Bolivia, March 20, 1939, W. J. Eyerdam 24917.

This species is similar to Cleome psoraleaefolia DC., but differs
most notably in the glabrous fruit, in its capsules which are only
about half as long, in its dense glandular puberulence in contrast
to the sparse long pubescence, and in its more broadly ovate
bracts which are less acute at the apex.

Psittacanthus Hortonii Standley & Barkley, sp. nov. Frutex
parasiticus ad 60 cm. altus, glaber, paulum ramosus; ramis longis,
laevibus, sparse striatis; foliis alternatis, cuneato-obovatis, 25 mm.
longis, 9-16 mm. latis, viridibus aliquid glaucis, subsessilibus,
truncatis, ad apicem emarginatis, mucrone nigro parvo; floribus
1-8, axillaribus, pedicellis erectis, circa 5 mm. longis; petalis 6,
rubris, linearibus, acutis, primo coalescentibus, deinde parte

SOUTH AMERICAN PLANTS 151

1948] STANDLEY AND BARKLEY

(‘ummasny AIOJSIPT [PANZEN OSvdIYD Jo uNIIeqJa}y 94} Ul poysodsaq)

[pues wuvplaig awoajlp “|

Ped pue ‘[pueys wuozwo pF snyjuvan7708 J
Ss

Peg pue ‘[pueys udnynog auosadojag “a

Vv

‘SENVId NVOIWAWY HLAOG dO SadA], ‘0Z

JicsetauNtiaie el BB eS Sc

‘peg pue

ALVIg

152 MADRONO [Vot. 9

supera valde reflexa; filamentis filiformibus, subulatis, petala
aequantibus; antheris flavis; fructu fulgente, nigro, globoso, circa
7 mm. longo.

Parasitic shrub to 60 cm., glabrate, little branched, branches
long, smooth, scarcely striate; leaves alternate, cuneate-obovate,
25 mm. long, 9 to 16 mm. broad, green but somewhat glaucous,
coriaceous, three-nerved from the base, subsessile, truncate, emar-
ginate at apex with a small black mucro about 0.5 mm. long;
flowers scarlet, axillary, 1 to 3, the pedicels erect, about 5 mm.
long; cup about 2.5 mm. deep, with three deltoid teeth; calyx 4
mm. long, 2 mm. broad, margin entire; petals 6, scarlet, linear,
acute, at first coalescent, later with upper third strongly reflexed;
filaments filiform, subulate, adnate to the petals two-thirds of
their length, equalling the petals in length; anthers yellow, versa-
tile, two celled, opening by two longitudinal slits, 4 mm. long;
stigma capitate, style filiform, 25 mm. long; fruit shiny, purplish
black, globose, about 7 mm. long.

Type. Huacho, 8 km. north of Huanuco, Depto. and Prov.
Huanuco, Peru, October 15, 1938, H. E. Stork §& O. B. Horton 9400.

This plant is closely related to Psittacanthus cuneifolius (R. &
P.) G. Don, from which it differs most markedly in the leaves
which are three to five times as broad as in that species. There
is no indication of the swelling at the nodes which is usually seen
in P. cuneifolius.

Beloperone Soukupii Standley & Barkley, sp. nov. Herba
perennis, ad basem decumbens; foliis 2—4 cm. latis, 4.5-10 cm.
longis, anguste ovatis, subrevolutis subintegrisque, apice sub-
acuminata, basi cuneata, in petiolum decurrentia; petiolis 2-3 cm.
longis, setoso-puberulentis; spicis 2-6 cm. longis, circa 1 cm. latis;
bracteis lanceolatis vel alte oblanceolatis, 2-3 cm. latis, 8-9 mm.
longis, acuminatis; calycis lobis viridibus, linearibus, 1 mm. latis,
8 mm. longis, acuminatis; corolla puberulenta, 17 mm. longa;
staminibus corollae subaequalibus; capsulis puberulentis, circa 5
mm. longis, recurvis.

Perennial herb, somewhat trailing, decumbent at base, mi-
nutely subappressed setose; stem slender, branched, striate, quad-
rangular above; internodes 5 to 7 cm. long; leaves petiolate, 2 to
4 cm. broad, 4.5 to 10 em. long, narrowly ovate, subrevolute and
subentire at margin, tapering at both ends, apex acute to sub-
acuminate, base cuneate, long decurrent on the petiole, finely and
sparsely setose-puberulent above and below, pale green and with
an almost metallic lustre when dry; flowering branches from the
axils of the uppermost leaves, with a single short-peduncled spike
or with several subsessile spikes; spikes 2 to 6 cm. long, about
1 cm. broad; bracts lanceolate to broadly oblanceolate, 2 to 3 mm.
broad, 8 to 9 mm. long, acuminate, sparsely pilose on the outer
surface, glabrous on the inner surface, ciliate, pinnate-veined with

: SOUTH AMERICAN PLANTS 153

1948] STANDLEY AND BARKLEY

pue °

Tpueys

sh

peadspooy vipunjaydp

a

peg

yy

(

°

uinasny, ALOSIFT [VANJEN OSvoIyD Jo WiniieqJazy 9Y} UT psysodaq)

PUP

YLOIG UOupuapnsoyd “YW ‘SINVId NvoOIwaWYyY HLNOG dO Sada], “[Z ALVIG

ee

7

154 MADRONO [Vou. 9

about 38 pairs of veins; calyx-tube essentially obsolete, calyx lobes
green, linear, 1 mm. broad, 8 mm. long, acuminate, glabrous on
inner surface, pilose on outer surface, ciliate; corolla puberulent,
17 mm. long, white tinged with pink, tube 11 mm. long, 1.5 mm.
broad, lobes unequal; stamens nearly as long as the corolla,
anthers narrow, 1.5 mm. long;-capsule puberulent, greenish, about
5 mm. long, recurved.

Type. Altitude 630 m., Depto. Hudnuco, 10 km. downstream
from Tingo Maria, Peru, October 28, 1938, H. E. Stork & O. B.
Horton 9582. Weedy perennial herb, partly trailing, along edges
of forest in partial sun, in forest mold.

This species closely resembles Beloperone cochabambense Rusby
in general habit and appearance, even to the metallic lustre of its
leaves ; however, its flowers are only half as long, the leaves some-
what broader, with longer petioles, longer and more slender
spikes, and it has narrower and less pilose bracts, as compared to
B. cochabambense.

Aphelandra Goodspeedii Standley & Barkley, sp.nov. Herba;
caulis simplex, subprocumbens, nodis infirmis radicans, setosus;
folia spatulato-lanceolata, apice obtusa, basi cuneata, integra vel
subintegra, utrinque laxe setosa; spicae terminales, pedunculatae ;
bracteae imbricatae, adpressae, serratae, pilosae; bracteolae
lanceolatae, acuminatae; sepalis lanceolatis, striatis.

Prostrate perennial herb up to 15 em. high, the stem simple,
subascending, rooting at the lower nodes, setose, the hairs sub-
appressed ; petioles 2 to 4 mm. long, densely pilose; lamina spatu-
late to lanceolate, 30 to 80 mm. long, 9 to 20 mm. wide, obtusely
rounded at apex, cuneate and decurrent at base, entire or sub-
undulate, when dry bluish above, grayish below, except whitish
along costa and lateral veins, entire or subundulate, setose on both
surfaces but more abundantly so below; flowers borne in one or
more terminal spikes about 4.5 cm. long; the peduncle about 8
mm. long, pilose like the rachis, stem, and petioles; bracts green,
lanceolate, acuminate, repand-serrate toward summit with about
five pairs of teeth, 3 mm. wide, 12 mm. long, in fruit ovate-lanceo-
late, about 4 mm. wide, 12 mm. long, hirsute on the outer surface ;
bractlets lanceolate, acuminate, entire, puberulent, membrana-
ceous-subhyaline, about 4.5 mm. long; calyx segments lanceolate,
acuminate, entire, minutely ciliate on the margins, subhyaline,
delicately striate-nerved, about 4.5 mm. long, 1 mm. wide; cap-
sules about 7.5 mm. long, glabrous, seed four, the valves of the
capsule recurved after dehiscence.

Type. Altitude 630 meters, Depto. and Prov. Huanuco, 10
km. downstream from Tingo Maria, Peru, November 2, 1938, H. E.
Stork & O. B. Horton 9597.

This species appears to be closely related to Aphelandra ornata
T. Anders. and A. Seibertii Leonard. A. ornata is a more upright

1948 | HOLLENBERG: MARINE ALGAE 155

plant, the leaves broadly lanceolate, subacute, glabrous except on
the veins and with petioles about 15 mm. long, and the bracts are
ovate with more and finer serrations. A. Seiberti has longer
pilose stems, petioles, and inflorescence rachis, much larger ovate
leaves, with finer and more abundant pubescence, and the bracts
are much broader.
Chicago Natural History Museum Herbarium,
Chicago, Illinois.
The University of Texas Herbarium,
Austin, Texas.

NOTES ON PACIFIC COAST MARINE ALGAE
G. J. HoLteNBERG

Data concerning the distribution of marine algae on the Pacific
Coast of North America are found chiefly in the various papers of
Setchell and Gardner and in the more recent works of Smith
(1944) and Dawson (1946). In the present paper the known
range of a number of species is extended as a result of collections
by the writer extending over a period of years. Unless otherwise
indicated, the place records are from the coast of southern Cali-
fornia and collection numbers are those of the writer.

CHLOROPHYCEAE

ULvEeLia SETCHELLU Dangeard, south of Redondo Beach, Janu-
ary, 1936, 1249; Corona del Mar, March, 19388, 2243b. This spe-
cies, previously reported from Pacific Grove on the coast of central
California (see Smith, 1944), is probably widely distributed along
the coast. In both of the above collections as well as at various
other times, this plant was found endophytic in the outer wall of
Amplisiphonia pacifica Hollenberg, although it is apparently not
limited to this host. Dangeard (1931) reported the Californian
material of this species as growing in or on the wall of Laurencia
sp. Observations on living material and parafine sections indicate
that the Pacific Coast plant is endophytic, except in the very early
stages of its development. The endophytic habit might be con-
sidered as reason for questioning the propriety of this generic
designation.

RHIZOCLONIUM RIPARIUM (Roth) Harvey, Corona del Mar, May,
1936, 1386; Balboa Harbor, December, 1936, 1928.5. Previous
records are from Carmel along the coast of central California and
northward (Setchell and Gardner, 1920). In the case of the first
of these collections the plant formed extensive brownish masses
on rocks in the midlittoral zone.

156 MADRONO [ Vou. 9

PHAEOPHYCEAE

EcTocarPus CHANTRANSIOIDES S. & G., Corona del Mar, October,
1942, 3288. This plant, which was found at least twice in this
general locality, seems not to have been recorded since first dis-
covered by Setchell and Gardner (1925). It forms abundant
brown hemispherical cushions several millimeters in diameter on

boulders in the middle littoral zone. Among 30 cushions exam- —

ined, nearly all bore plurangia, but no unilocular sporangia were
found.

EctTocarrus CYLINDRICcUs Saunders, on Halidrys, Corona del
Mar, November, 1936, 1600; on Gelidium and Eisenia, La Jolla,
December, 1936, 2028a, 2028b; on the stipe of Eisenia, Corona del
Mar, December, 1939, 3051. Although fairly common along the
coast of southern California, this plant was previously known only
from the type collection obtained by Saunders (1898) at Pacific
Grove, along the coast of central California. However, E. cylin-
dricus f. codiophilus S. & G. has been collected by Dawson (1945b)
at La Jolla. Hisenia seems to be the commonest host. In one
instance the Ectocarpus formed a patch two centimeters across.

DictyopTeris JOHNSTONE! Gardner, from a large tide pool at
medium high tide level on exposed rocky shore several miles south
of Redondo Beach, October, 1935, 1021. This specimen is 37 cm.
high. Gardner’s fragmentary specimen, the only one previously
known, was only 7 cm. high. The present specimen has a stupose
base as anticipated by Gardner’s account (1940). The stipe is
more prominent (5 cm. long) than in the type specimen and the
main axes are much less branched and decidedly percurrent. A
collection made from a high tide pool at Corona del Mar, Novem-
ber, 1934, 583, seems to belong here also, although a few of the
frond divisions are as much as 5 mm. wide.

RHODOPHYCEAE

GoNIOTRICHUM CORNU-CcERVI (Reinsch) Hauck, on Herposiphonia
sp., Santa Cruz Is., April, 1936, 1339. Previously reported for the
Pacific Coast by Kylin (1925) from the vicinity of the Biological
Station at Friday Harbor, Washington.

ERYTHROTRICHIA PULVINATA Gardner, on the usual host, Codium
fragile Suhr., Emerald Bay, Laguna Beach, February, 1935, 649.5.
Previously known only from the Monterey Peninsula (Smith,
1944).

ERYTHROCLADIA SUBINTEGRA Rosenvinge, from Santa Cruz Island,
April, 1986, 1367. Previously reported as far south as Monterey
(Kylin, 1941), recently from Baja California, Mexico (Dawson,
1945c), and probably the plant referred to without name by Daw-
son (1944a, p. 251) as occurring in the Gulf of California. The
plant is locally common on various algae along with E. irregularis
Rosenvinge.

1948 ] HOLLENBERG: MARINE ALGAE 157

-RHODODERMIS ELEGANS Crouan, was found growing on the stipe
of Cystoseira osmundacea (Menz.) Ag. south of Redondo Beach,
October, 1935, 1049; and on the same host at White’s Point, San
Pedro, November, 1933, 330. This plant was not previously
known south of Monterey, where it is reported as occurring by
Smith (1944). It is of frequent occurrence along the coast of
southern California.

Cruoria PaciFicaA Kjellman (?), on shells of turban snails,
Tegula sp., Corona del Mar, October, 1935, 1065; on rocks, Laguna
Beach, December, 1935, 1170. This species was previously re-
ported only from Alaska (Kjellman, 1889), however, it is very
imperfectly known and might well prove to belong to another
genus. The plants from southern California have zonate tetra-
sporangia and clearly belong to the genus Cruoria, but cannot at
present be identified with Kjellman’s species with certainty.

CHOREONEMA THURETI (Bornet) Schmitz, collected at Corona
del Mar, December, 1937, 2231. The writer has collected this
plant a number of times along the coast of southern California.
It has been previously reported from Laguna Beach by Guernsey
(1912), from Guadalupe Island off the coast of Mexico by Setchell
and Gardner (19380), and from the coast of Ecuador by Taylor
(1945).

GRATELOUPIA ABBREVIATA Kylin, Santa Catalina Island, April,
1935, 762; Point Vicente, near Redondo Beach, July, 1938, 2339;
Laguna Beach, May, 1941, 3176. This plant was previously re-
ported from La Jolla by Kylin (1941). It seems to be relatively
common at certain seasons near high tide level on large wave-
swept rocks near Laguna Beach, where it has been collected a
number of times.

PETROCELIS FRANCISCANA S. & G., White’s Point, San Pedro,
October, 1938, 2376. This species was previously reported as far
south as San Luis Obispo County (Smith, 1944).

GARDNERIELLA TUBIFERA Kylin, near Santa Monica, July, 1935,
1279. Until Dawson (1945c) reported this species from two
points on the coast of Baja California, it had been recorded only
from the Monterey Peninsula (Smith, 1944).

HyYpnea caLirornica Kylin, cast ashore near Santa Monica,
July, 1935, 882. This plant was previously reported only from
the type locality at La Jolla (Kylin, 1941), and from Guerrero,
Mexico (Taylor, 1945).

GIGARTINA PAPILLATA (Ag.) J. Ag., White’s Point, San Pedro,
October, 1938, 2390. Until Dawson (1945c) reported this species
from Punta Descanso, Baja California, Mexico, it had been re-
corded only as far south as Carmel Bay (Smith, 1944).

RuHoDoGLossuM PARVUM Smith and Hollenberg, from Laguna
Beach, March, 1936, 1295 and March, 1987, 2065. This species
was known previously only from the Monterey Peninsula (Smith
and Hollenberg, 1943).

158 MADRONO [Vot. 9

BINGHAMIELLA Forku Dawson, on corallines, Laguna Beach,
March, 1935, 658 (tetrasporic) ; on Gelidium sp., cast ashore near
the isthmus, Santa Catalina Island, April, 1935, 787 (cystocarpic) ;
on corallines, Laguna Beach, December, 1936, 1935a. This plant
was previously known only from the type locality, La Jolla (Daw-
son, 1944b).

Faucuea MEpDIA Kylin, Santa Cruz Island, April, 1936, 1365;
Laguna Beach, March, 1936, 1293. This plant was previously
known only from the Monterey Peninsula (Smith, 1944),

CALLITHAMNION PIKEANUM var. PaciFicum (Harv.) S. & G.,
Point Magu, Ventura County, March, 1935, 738. This variety
was reported by Gardner (1927) as ecenmnas from the Straits of
Juan de Fuca northward.

CERAMIUM GRACILLIMUM Grifhths and Harvey, on corallines,
Laguna Beach, December, 1936, 1943. This plant has been col-
lected frequently at this locality on large wave-swept rocks near
high tide level. It is similar to if not identical with a plant re-
ported by Setchell and Gardner (1930) from the Revilla Gigedo
Islands under the name C. transversale Collins & Hervey. Dawson
(1944a) identified as C. gracillimum a plant which he collected in
the Gulf of California, but on account of the different habitat of
the local plant the writer suspects that it may ultimately prove
to be distinct from the plant described by Griffiths and Harvey.

CERAMIUM SINICOLA var, INTERRUPTUM (S. & G.) Dawson, La-
guna Beach, November, 1935, 1103; Laguna Beach, December,
1936, 1932; on oyster shells, upper Balboa Harbor, December,
1936, 1927. This species was not previously known north of the

type locality, Ensenada, Baja California, where it was first re-

ported by Setchell and Gardner (1924).

CERAMIUM PROCUMBENS S. & G., Redondo Beach, August, 1934,
468.5; from near Scripps Institution, La Jolla, December, 1936,
1972, 2035. This species has not been reported since originally
described by Setchell and Gardner (1924) from material collected
in the Gulf of California. It seems common as an epiphyte along
the coast of southern California.

GYMNOTHAMNION ELEGANS (Schousboe) J. Ag., Laguna Beach,
November, 1935, 1101; Corona del Mar, January, 19438, 3305.
This plant was previously reported from the Revilla Gigedo
Islands by Setchell and Gardner (1930). It is of frequent occur-
rence on rocky coasts along the coast of southern California.

Pritota FiLicina (Farl.) J. Ag. This plant occurs in the Dim-
mick collection from Santa Barbara. The Monterey Peninsula is
given by Smith (1944) as the southern limit of its known range.

ANISOCLADELLA PACIFICA Kylin, from Laguna Beach, October,
1935, 1055 (male and sterile specimens) ; from near Santa Monica,
December, 1935, 1197, 1198 (male and cystocarpic) ; Corona del
Mar, February, 1940, 3089 (tetrasporic) ; Laguna Beach, Novem-
ber, 1942, 3290 (cystocarpic and tetrasporic). Until Dawson

1948 | HOLLENBERG: MARINE ALGAE 159

(1945c) reported this plant from the coast of Baja California,
Mexico, it had been known only from the Monterey Peninsula
(Smith, 1944). Examination of isotype material from the Monte-
rey region leaves no doubt that Kylin’s plant is the same as those
encountered more frequently farther south. The development of
the secondary cell rows at the tips of the blades, with alternate
emphasis in growth of the secondary cell rows on either side of
the midrib, as described by Kylin (1941), is very characteristic.
Cystocarpic and antheridial plants seem not to have been previ-
ously described. The cystocarps are commonly in pairs on either
side of the midrib near the middle of the blade as far as length is
concerned, but may also occur singly, and less frequently three
may occur together on one blade. Antheridial areas usually occur
as a single and relatively large pair near the tip of the blade.

Dasyopsis pensAa G. M. Smith. <A specimen in the collection of
Dr. L. N. Dimmick, which is housed in the Santa Barbara Museum
of Natural History, is probably this plant. It was previously
known only from the Monterey Peninsula (Smith, 1944).

PocgonoPHora caLirornica J. Ag., Laguna Beach, February,
1935, 644. This plant was described by J. Agardh (1890) from
Santa Barbara. It was collected by Gardner (no. 2506) as far
north as Bodega Bay in Marin County. It was also collected by
Dawson (1945b) at La Jolla and by Dawson (1945c) from Punta
Descanso, Baja California, Mexico. It is of frequent occurrence
along the coast of southern California.

STROMATOCARPUS GARDNERI Setchell, on Pterosiphonia dendroidea
(Mont.) Falk., Laguna Beach, March, 1936, 1289 (male, tetra-
sporic, and cystocarpic). The type collection was from near
Santa Monica, where the plant was parasitic on Pterosiphonia
Baileyi (Harv.) Falk. (see Setchell, 1923). Until Dawson (1945a)
reported this plant from La Jolla, it was not recorded as having
been collected since the type collection was made.

CuHoREOCOLAX Po.tysIPHONIA4E Reinsch, parasitic on Lophosiphonia
villum (J. Ag.) S. & G., Laguna Beach, December, 1937, 2208.
Until Dawson (1945c) reported this plant from Baja California,
Mexico, it had been known on the Pacific Coast of North America
only from Sitka, Alaska (Saunders, 1901), and from the Monterey
Peninsula (Smith, 1944). The writer has collected it several
times on the coast of southern California.

HERPOSIPHONIA TENELLA (Ag.) Naeg., south of Redondo Beach,
July, 1935, 981 (tetrasporic) ; Santa Catalina Island, April, 1935,
834; Corona del Mar, October, 1937, 2146 (cystocarpic) ; Laguna
Beach, December, 19386, 1929.5; Redondo Beach, October, 1935,
1050. This species is widely distributed. It was previously re-
ported for the Pacific Coast of North America by Dawson (1944a)
from the Gulf of California. It is locally quite common among
other mat-forming algae in the upper tide level. Sometimes on
spray-covered rocks in shallow tide pools it occurs by itself, in

160 MADRONO | [ Vor. 9

which case branching is sparse and it spreads horizontally on
the rocks, forming thin patches several centimeters in diameter.
The cystocarps are terminal or subterminal on the determinate
branches as described by Bgrgesen (1918) but the branch is not
continued as a slender projection as shown in his figure. In one
case the plant seemed to lack trichoblasts entirely, but usually
there is a brief but distinct tuft of widely branching trichoblasts
at the tips of the determinate branches. Rhizoids arise at the
distal ends of the pericentral cells as described by Bgrgesen. In
the writer’s experience, this seems to be a characteristic feature
of the genus.

HERpPOSIPHONIA SEcUNDA (Ag.) Naeg., from near Santa Monica,
April, 1934, 443.1. This species seems not to have been previ-
ously reported from the Pacific Coast of North America. Some
investigators are of the opinion that this plant should be con-
sidered as a variety under the preceding species. Bgrgesen
(1918), who considers it as a separate species, depends chiefly
on certain differences in the structure of the antheridial branches.
The latter structures were not found in the local plants but the
sequence of origin of determinate branches indicates that the
plants belong to this species.

PTEROSIPHONIA CALIFORNICA Kylin, Santa Catalina Island, April,
1935, 808; near Ferry Wharf, Oakland, Alameda County, Cali-
fornia, April, 1937, 2074; Corona del Mar, November, 1935, 1110.
This plant was previously reported only from San Diego and La
Jolla by Kylin (1941), who did not observe tetrasporic plants.
The tetrasporangia are of the usual sort, occurring one per seg-
ment in the ultimate branchlets. This plant was collected by
Gardner (no. 2480) from rocks three miles west of Santa Monica
as early as March, 1912, as indicated by specimens in the her-
barium of the University of California at Berkeley.

PotysipHonia EastwoopakE S. & G., Cabrillo Beach, San Pedro,
July, 1940, 3103. This collection seems to belong with P. East-
woodae, an entity which is possibly too close to P. Snyderae Kylin,
but which must be kept distinct at present on account of the differ-
ence in the basal attachment. P. Eastwoodae was inadvertently
omitted from the list of tetrasiphonous species treated in a previ-
ous paper (Hollenberg, 1942).

PoLystPpHoNIA SNYDERAE var. inTRIcCATA Hollenberg, from rocks
in the middle littoral zone, Corona del Mar, October, 1942, 3284.
This plant is a dwarf form of the variety sre was found associ-
ated with other mat-forming algae. The variety was previously
known only from Sonora, Mexico.

LAURENCIA DIEGOENSIS Dawson, Corona del Mar, October, 1934,
496. The writer has collected this plant a number of times along
the coast of southern California. In his original description Daw-
son (1944c) points out that the plant occurs farther north than
the type locality, La Jolla. It is a very distinctive plant as much

1948] HOLLENBERG: MARINE ALGAE 161

as 22 cm. high. A characteristic feature not pointed out by Daw-

son is the unilateral position of the male conceptacles. There may
be as many as four of these conceptacles bulging prominently on
the adaxial side of the ultimate ramuli.

University of Redlands,
Redlands, California.

LITERATURE CITED

Acarpu, J.G. 1890. Till algernas systematik, Afd. 6 Lunds Univ. Arsskr. 26.
Bgrcesen, F. 1918. The marine algae of the Danish West Indies. Part 3,
Rhodophyceae. Dansk. Bot. Ark. 3: 1-498.
DancearpD, P. 1931. L’Ulvella lens de Crouan et )’Ulvella Setchellii sp. nov.
Bull. Soc. Bot. France 78: 312-318.
Dawson, E. Y. 1944a. The marine algae of the Gulf of California. Allan
Hancock Pacific Expeditions 3: 189-453.
1944b. New and unreported marine algae from southern Cali-
fornia and Northwestern Mexico. Bull. So. Calif. Acad. Sci. 44: 75-91.
1944c. Some new Laurenciae from southern California. Ma-
drono 7: 233-240.
1945a. Notes on Pacific Coast marine algae, II. Bull. So. Calif.
Acad. Sci. 44: 22-27.
1945b. An annotated list of the marine algae and marine
grasses of San Diego County, California. Occas. Papers San Diego
Soc. Nat. Hist. 7: 1-87.
1945c. Marine algae associated with upwellings along the north-
western coast of Baja California, Mexico. Bull. So. Calif. Acad. Sci.
44; 57-71.
1946. A guide to the literature and distributions of the marine
algae of the Pacific Coast of North America. Mem. So. Calif. Acad.
Sci. 3 (No. 1): 1-134.
GarpNnerR, N. L. 1927. New Rhodophyceae from the Pacific Coast of North
America. V. Univ. Calif. Publ. Bot. 13: 403-434.
1940. New species of Melanophyceae from the Pacific Coast of
North America. Jbid. 19: 267-286.
GUERNSEY, JOHN. 1912. Notes on the marine algae of Laguna Beach. First
Ann. Report Laguna Marine Laboratory. 195-218.
Hoiienserc, G. J. 1942. An account of the species of Polysiphonia on the
Pacific Coast of North America. I. Oligosiphonia. Am. Jour. Bot.
29: 772-785.
KsELLMAN, F. R. 1889. Om Beringhafvets Algflora. Kongl. Sv. Vet. Akad.
Handl. 23: 1-58.
Kyun, H. 1925. The marine red algae in the vicinity of the biological station
at Friday Harbor, Washington. Lunds Univ. Arsskr., N. F. 21: 1-87.
1941. Californische Rhodophyceen. Ibid. 37: 1-51.
Saunvers, Dre A. 1898. Phycological memoirs. Proc. Calif. Acad. Sci., ser. 3.
Bot. 1: 147-168.
. 1901. Papers from the Harriman Alaska expedition. XXV.
The algae. Proc. Wash. Acad. Sci. 3: 391-486.
SETCHELL, W. A. 1923. Parasitic Florideae, II. Univ. Calif. Publ. Bot. 10:
393-401.
SETCHELL, W. A., and N. L. Garpner. 1920. The marine algae of the Pacific
Coast of North America. Part II. Chlorophyceae. Ibid. 8: 139-374.

, and ———————-. 1924. New marine algae from the Gulf of
California. Proc. Calif. Acad. Sci. 12: 695-949.
, and ———————_. 1925. The marine algae of the Pacific Coast

of North America. III. Melanophyceae. Univ. Calif. Publ. Bot. 8:
383-898.

162 MADRONO [ Von. 9

» and ————————_. 1930. Marine algae of the Revillagigedo
Islands expedition in 1925. Proc. Calif. Acad. Sci. Fourth Ser. 19:
109-215.

SmitH, G. M. 1944. Marine algae of the Monterey Peninsula. Stanford Uni-
versity Press.

Situ, G. M., and G. J. Hottenserc. 1943. On some Rhodophyceae from the
Monterey Peninsula. Am. Jour. Bot. 30: 211-222.

Taytor, W. R. 1945. Pacific marine algae of the Allan Hancock expeditions
to the Galapagos Islands. Allan Hancock Pacific Exped. 12: 1-528.

ON TWO PERENNIAL CAESPITOSE LEPIDIUMS OF
WESTERN NORTH AMERICA

Reep C. Roiiins

Among the known North American species of Lepidium, L.
nanum S. Wats. stands out as strikingly different in habit from all
others. It is a low, matted plant with many caudex branches
each terminated by a dense rosulate cluster of tiny leaves. The
species has a well-developed taproot system that often penetrates
the soil to depths of a foot or more. It grows at relatively low
elevations in northern Nevada. As Hitchcock (1936) remarked,
plants of this species are “more suggestive of Draba than of
Lepidium.” In the past, speculation as to the genetic relation-
ships of L. nanum inevitably lead to a comparison of it with
species of similar habit of the high Andes of South America. For
example, Thellung (1906) expressed the opinion that the Andean
L. Meyeni Walp. was probably a relative of L. nanum.

As long as LZ. nanum remained the only species of its type
known from North America, it was justifiable to try to link it up
with such remotely situated species as those of the South Ameri-
can continent. In fact, no obvious alternative existed because
there has been no basis for connecting L. nanum with any other
North American species of the genus. However, the recent dis-
covery of a new species from Idaho’ has supplied a probable con-
necting link between L. nanum and other perennial North Ameri-
can species. The new species is caespitose and possesses a very
thick taproot (pl. 22, fig. 1). It stands in an intermediate
position morphologically between ZL. nanum and such other

1 Professor Ray J. Davis discovered fruiting material of the new species
in June, 1946, returning to the same location in May, 1947, for flowering
specimens.

EXPLANATION OF Figures. PLatTe 22.

Piate 22. Lepmwrum Davisu. Fic. 1, a plant from the type sheet. It is
16 cm. wide. Fic. 2, enlarged infructescence. The siliques are 3-3.5 mm. long.

: LEPIDIUM

ROLLINS

1948]

Lepiium DavlIsit.

PLATE 22.

164 MADRONO [ Vor. 9

perennial species of Lepidium as L. montanum. Therefore, it is
now plausible to suggest that L. nanum may be related to other
North American species of Lepidium. Formerly such a hypo-
thesis would have seemed completely untenable. It is the prin-
cipal purpose of the present paper to present a description of the
new plant.

Lepidium Davisii sp. nov. Herba perennis caespitosa mul-
ticaulis; caudicibus ramosis, ramis crassis; caulibus erectis sim-
plicibus vel rare ramosis 4—8 cm. altis pubescentibus; foliis ses-
silibus integris spathulatis sparse pubescentibus 1—2.5 cm. longis,
2-5 mm. latis; inflorecentiis subcorymbosis; sepalis oblongis ca.
1.5 mm. longis, ca. 1 mm. latis; petalis albis spathulatis 2-3 mm.
longis ; pedicellis pubescentibus teretibus 83-4 mm. longis; siliquis
ovatis glabris vel sparse pubescentibus 3—3.5 mm. longis; stylis ca.
0.5 mm. longis; cotyledonibus accumbentibus.

Caespitose deep-rooted perennial forming clumps; roots
thick and fleshy, 0.5-2 cm. across, expanded at summit and
divided into numerous caudex branches; caudex corymbose, the
apex of the branches partially invested in old leaf-bases; stems
slender, numerous, each terminated by an inflorescence, simple or
rarely branched, leafy, 4-8 cm. high, densely pubescent with
minute whitish simple trichomes; leaves sessile, basal and cauline
similar, spatulate, obtuse, greenish, sparsely pubescent, much-
exceeding the internodes, 1—-2.5 cm. long, 2-5 mm. wide; in-
florescence subcorymbose, slightly elongated; sepals broadly ob-
long, not persistent, greenish with a broad hyaline margin, ca.
1.5 mm. long and 1 mm. wide, glabrous to very sparsely pubescent
near the base; petals white, spatulate, 2-3 mm. long, ca. 1 mm.
wide above; paired stamens only slightly longer than single
stamens; infructescence subcorymbose; pedicels divaricate, pu-
bescent, terete, 3-4 mm. long; siliques crowded, ovate (pl. 22,
fig. 2), glabrous to sparsely pubescent, slightly winged above,
flattened contrary to replum, slightly notched at apex, 3—3.5 mm.
long, 2—2.5 mm. wide; styles ca. 0.5 mm. long; seeds wingless,
one in each locule; cotyledons accumbent.

Type in the Dudley Herbarium of Stanford University, col-
lected June 27, 1946, in a dried-up pond of the lava plain, 14
mile north of the Snake River Canyon, about 14 miles south of
Mountain Home, Elmore County, Idaho, R. J. Davis 4670. A
second collection from the same station was taken May 9, 1947,
R. J. Davis 4745 (DS).

The siliques of LZ. Davisit are similar in shape to those of L.
nanum. The seeds are of similar size and shape, and the
cotyledons are accumbent in both species. As pointed out above,
the caespitose habit and thick perennial root further suggest a
possible relationship between these species. These and many
other kinds of plants of similar growth-habit are found in arid
areas at relatively low elevations in the intercordilleran region

1948 ] REVIEW 165

of western North America. Such plants are often mistakenly
assumed to be part of a high mountain arctic-alpine flora. Thel-
lung (1906) made such a mistake when he referred to L. nanum as
an alpine plant.

Stanford University, California.

LirerRATURE CITED

Hircucock, C. Leo. 1936. The Genus Lepidium in the United States.
Madrono 3: 314.

THELLUNG, A. 1906. Die Gattung Lepidium (L.) R. Br. Denks. schweiz.
Gesellsch. Naturwiss. 41: 204.

REVIEW

~The Genus Crepis. Part I. The Taxonomy, Phylogeny, Distri-
bution, and Evolution of Crepis. Part II. Systematic Treatment.
By Ernest Brown Bascock. University of California Publications
in Botany, vol. 21, pp. xii + 1-198, frontispiece, plate 1, figs. 1-11,
tables 1-12; vol. 22, pp. x + 199-1031, plates 2-36, figs. 12-305,
tables 13-19. 1947.

Hieracium excepted, Crepis is the largest genus of the tribe
Cichorieae, family Compositae. Its one hundred and ninety-six
species are distributed widely over much of Eurasia, Africa, and
western North America, with their greatest concentration in the
Mediterranean region. The cytology, genetics, cytogenetics, evo-
lution, and systematics of these species have been subjected to
lifelong study by Professor Babcock with the assistance of many
co-workers. The results of these studies are effectively summa-
rized in the present outstanding contribution. Space does not
allow, in a short review, even mention of more than a few of the
important and fundamental discoveries. Those who desire to
know more will have to consult the original source which will
surely become one of the classics of systematic botany.

Professor Babcock is one of the pioneers of the modern idea
that systematics should correlate and integrate all evidence per-
tinent to the interrelationships of organisms, regardless of its
source, and the present study fully substantiates that thesis. The
primary sources of evidence are the classic criteria of comparative
morphology and geographical distribution, yet these in themselves
may not be always reliable, and the instances in which evidence
from comparative cytology, genetics, cytogenetics, or geology has
proved critical are legion. Once all of the facts are assembled,
the evidence from each of these sources strengthens that from
each other, and as a whole points to a definitive and unequivocal
conclusion. On this broad basis of fact, gleaned from many
diverse approaches, though still incomplete in places, Professor
Babcock’s conclusions rest, and it probably is no overstatement

166 MADRONO [ Vo. 9

that no systematic work of comparable magnitude has ever rested
on a firmer foundation.

On the basis of comparative morphology, Crepis is divided into
primitive, intermediate, and advanced groups of species. These
groups, though often referred to, are not given taxonomic status,
their relationship being horizontal rather than vertical. The
most primitive species are those with perennial rhizomes, large
lyrate leaves, and few large heads with very large florets and
achenes, and an involucre consisting of many large bracts which
remain unchanged at maturity. These are obviously adapted to
a mesophytic climate, and are presumed to have persisted with
little change since their origin, perhaps in the middle of the Ter-
tiary period. The most advanced species include those of short-
annual duration, with tap roots, smaller dissected leaves, and
numerous small heads with small florets and achenes, and an
involucre of few small bracts which become thickened at maturity.
These have become specialized to xeric conditions, and are much
more recent in origin. Between these two extremes are found
most of the species of Crepis, specialized in one way or another
along several different lines of development.

One hundred and thirteen species have been studied cytologi-
cally. These, excluding ten North American species of the sec-
tion Psilochaenia, which are polyploid, «= 11, and six polyploid
Old World species, are all diploid, n = 7, 6, 5, 4, or 8, with 4-
paired species greatly predominating. Six is held to have been
the primitive basic number of chromosomes, this number being
possessed by some of the most primitive rhizomatous species,
with 5, 4, and 3 being derived through a progressive series of
reductions. This is correlated with a decrease in chromosome
symmetry, and a decrease in chromosome size, and each step is
thought to have occurred independently on several occasions.
The group of species with n = 7, the section Ixeridopsis, is not the
most primitive, and because of morphological and cytological
resemblances is assumed to have arisen through intergeneric hy-
bridization with Jzeris. Another group, the section Pyrimachos,
which has not been studied cytologically, and a single aberrant
species, C. paludosa, n= 6, may also have arisen by intergeneric
hybridization. Aside from these three possible exceptions, the
morphological and cytological evidence supports the generaliza-
tion that Crepis is monophyletic.

The evidence from genetics and cytogenetics also is limited
by the number of species which it has been possible to bring into
cultivation. This evidence is extensive, however, many species
having been crossed, and the resulting hybrids studied. In gen-
eral, the more closely related the species, the more readily they
may be crossed, the more vigorous and fertile the offspring, and
the more nearly homologous the chromosomes as revealed by
meiotic pairing. From this and the occurrence of similar genic

1948] REVIEW 167

mutations in widely separated species, it is inferred that the
genetic background is similar in all of the species investigated.
This in turn supports the hypothesis of monophylesis. It follows
that evolution in Crepis has taken place on the diploid level
through the accumulation of genic mutations, and a rearrangement
of these genic materials through chromosome changes. Poly-
ploidy and intergeneric hybridization have been relatively un-
important as far as the genus as a whole is concerned.

On the basis of comparative morphology, geographical
distribution, and chromosome numbers, it is inferred that the
twenty-seven genera of the subtribe Crepidinae including Crepis,
originated from the primitive genus Dubyaea or now extinct
Dubyaea-like species. That the genus Crepis originated in central
Asia near the middle of the Tertiary period is supported by geo-
logical, geographical, and biological evidence. It is from this
area that the natural phyletic lines in the genus radiate. Relics
of the putatively ancestral genus are preserved nearby, as are
primitive members of related genera. These primitive members
of Crepis and those of related genera are mesophytes which in
itself suggests a temperate mid-Tertiary development. That this
is a likely hypothesis is shown by a review of Tertiary history.

Matthew’s principle is called on for additional support of the
central Asiatic origin of Crepis. It is true that the distributions
of some of the primitive species do conform to this dictum, but the
dictum itself rests on the unsupported premise that evolutionary
forces are most effective at the center of origin, hence each suc-
cessive wave of outward migration would be more advanced than
those that have gone before. Thus, assuming the distribution of
each to be a function of time, uninfluenced by other factors, this
would result in the occurrence of the primitive members at the
periphery of the range of the group. That the conformity with
Matthew’s principle is coincidence only is evident from a similar
examination of the Crepidinae, the most primitive members of
which are found nearer the supposed center of origin than are
many of the obviously derived members, or the observation that
modern genetic theory, as clearly outlined in the treatise, lends
no support whatsoever to any such principle of distribution. It
is most unfortunate that the author, who must be fully aware of
these conflicts, does not resolve them, for it is certain that less
profound thinkers will seize upon this coincidental conformance
as confirmation of the principle of Matthew by evidence from
- modern genetics.

The second volume is a model of systematic conception and
execution. First, it includes the history and description of the
genus, followed by diagnoses of and keys to the twenty-seven
natural sections. These are then taken up, one by one, in order
from primitive to intermediate to advanced. This order admit-
tedly is progressive rather than phyletic, for phyletic relation-

168 MADRONO [Vot. 9

ships may be other than linear. With some exceptions, each
sectional treatment is introduced by a discussion of relationships
and distribution of the species, usually illustrated with a map and
accompanied by a key to the species. Then follow the treatments
of individual species, again arranged in a progressive sequence.
The descriptions are detailed and comparable point by point.
Each species is illustrated with a figure, usually showing habit as
well as other diagnostic features including the chromosome set
when this is known for the species. The figures are drawn to the
same scale and, again, are directly comparable. Following the
specific description is a statement of geographical distribution and
ecological relationships. Then comes the citation of critical
specimens, followed by notes on the variability of the species,
often including a numbered list of variants held to be minor.
Concluding is a statement of specific relationships based on a
synthesis of all available evidence. In those cases in which there
is more than one subspecies, slight departure is made from the
above sequence, but the treatment again is consistent.

Professor Babcock, whose achievements in genetics and evo-
lution are widely recognized, is to be congratulated on this out-
standing triumph in systematics. Likewise, the wise and far-
sighted policy of the administration of the California Experiment
Station, in supporting this excursion into pure science for more
than a quarter of a century, is to be highly commended. It is
only from unqualified support such as this that truly great ad-
vances in science may be expected.—Marion Ownsey, State Col-
lege of Washington.

NOTES AND NEWS

A PossisLte REcorp or QuERcus MorEHUsS IN OREGON. An inter-
esting oak thought to be Quercus Morehus, has been found along
the Lower Grave Creek County Road about 6.2 miles west of
Leland, northern Josephine County (Sect. 32 or 33 S., T. 33S., R.
7 W., Willamette Meridian). To date, only leaf specimens have
been collected, but if the tree proves to be Q. Morehus, it will be
the first record of the occurrence of this species in Oregon.—
Ouiver V. Matruews, Salem, Oregon.

A small residue of back numbers of the biological journal,
“Zoe, published by T. S. and Katharine Brandegee from 1890
until 1908, has come to light. From volume I, numbers 1, 2, 3, 6,
and 8 are missing; from volume II, numbers 1 and 2 are missing;
volumes III and IV are complete; volume 5 is complete except for
number 1. Information concerning this material may be obtained
from the secretary of the California Botanical Society.

MADRONO

A West American Journal of Botany

A quarterly journal devoted to important
and stimulating articles dealing with
plant morphology, physiology, taxonomy,
and botanical history. These volumes
should be a part of every botanist’s li-
brary and should be made accessible to
students of all universities and colleges.

Volume I, 1916-1929. . . $5.00
Volume II, 1930-1934 .. 5.00
Volume III, 1935-1936 . 5.00
Volume IV, 1937-1938 . 5.00
Volume V, 1939-1940 . . 5.00
Volume VI, 1941-1942 . 5.00
Volume VII, 1943-1944 . 5.00
Volume VIII, 1945-1946 . 5.00
Single numbers...:... 1.00

The subscription price of MADRONO
is $3.50 per year. We solicit your pat-
ronage,

Address all orders to:

Rimo Bacigalupi, Bus. Mer.
Natural History Museum,
Stanford University, California

a

PPR 24,

VOLUME 1X NUMBER 6

MADRONO

A WEST AMERICAN JOURNAL OF
BOTANY

wi

Contents

CONCENTRATION OF ENVIRONMENTAL E;XTREMES AS THE BASIS FOR VEGETATION

PAMEAS. EC LUOU Ne. CPVOTINIDE i 15). \ggee Ste k ti che ede WB see tic grate eee le leh es 169
; Tue Genus HELIANTHELLA IN Orecon, William A. Weber .............. 186
A New Poremonium From Mexico, John F. Davidson .................. 187

Nores ON THE TAXONOMY OF SOME EasTeRN ASIATIC FERNS OF THE GENERA
Prorowoopsia AND Preretis, Clyde F. Reed

Review: F. C. Stern, d Study of the Genus Paeonia (G. Ledyard Stebbins,

ERE HON Neer eg OREN STN tks SON EHR e lol uaa alt dau baer Ure NEM LO bs 498
Nores Ann News: Range Extensions of Grasses into Colorado, H. D. &
La Vey er Sit Re) MEd Son ale Me AE DTN Sarina PT REORDER RRA VORRON Cr ARDR OTL ides it 199

Published at North Queen Street and McGovern Avenue,
Laneaster, Pennsylvania

April, 1948

MADRONO

A WEST AMERICAN JOURNAL OF BOTANY

Board of Editors

Hersert L. Mason, University of California, Berkeley, Chairman.

LeRoy Asrams, Stanford University, California.

Epcar ANpDERSON, Missouri Botanical Garden, St. Louis.

Lyman Benson, Pomona College, Claremont, California.

Hersert F.. Copetann, Sacramento College, Sacramento, California.

ivan M. JouHnston, Arnold Arboretum, Jamaica Plain, Massachusetts.

Mitprep E. Marutas, Dept. of Botany, University of California, Los Angeles 24.
Bassett Macuire, New York Botanical Garden, N. Y. C.

Marion Ownsey, State College of Washington, Pullman.

Secretary, Editorial Board—ANwNeEtTTA CaRTEeR
Department of Botany, University of California, Berkeley

Business Manager—Rimo BaAcicaLuPi
North Queen Street and McGovern Avenue, Lancaster, Pennsylvania
or
Natural History Museum,
Stanford University, California

Entered as second-class matter October 1, 1935, at the post office at
Lancaster, Pa., under the act of March 3, 1879.

Established 1916. Published quarterly. Subscription price $3.50 per year.
Completed volumes I to VIII $5.00 each; single numbers $1.00.

Papers up to 15 or 20 pages are’ acceptable. Longer contributions may
_be accepted if the excess costs of printing and illustration are borne by the
contributor. Range extensions and similar notes will be published in con-
densed form with a suitable title under the general heading “Notes and News.”
Articles may be submitted to any member of the editorial board. Manuscripts

may be included in the forthcoming issue provided that the contributor pay the -

cost of the pages added to the issue to accommodate his article. Reprints of
any article are furnished at a cost of 4 pages, 50 copies $5.90; 100 copies $6.48;
additional 100’s $1.23; 8 pages, 50 copies $8.55; 100 copies $9.49; additional 100’s

$1.88; 16 pages, 50 copies $12.00; 100 copies $13.44; additional 100’s $2.88.

Covers, 50 for $3.95; additional covers at $2.38 per hundred. Reprints should
be ordered when proofs are returned.

Published at North Queen Street and McGovern Avenue, Lancaster,
Pennsylvania, for the

CALIFORNIA BOTANICAL SOCIETY, INC.

President: Adriance S. Foster, University of California, Berkeley. First
Vice-President: G. Ledyard Stebbins, Jr., University of California, Berkeley.
Second Vice-President: Herbert F. Copeland, Sacramento College, Sacramento,
California. Secretary: John Whitehead, University of California Botanical
Garden, Berkeley. Treasurer: Rimo Bacigalupi, Natural History Museum,
Stanford University, California.

Annual membership dues of the California Botanical Society are $3.50,
which includes a year’s subscription to Madrono. For two members of the same
family the dues are $4.00, which includes one copy of Madrofo and all other
privileges for both. Dues should be remitted to the Treasurer. General
correspondence and applications for membership should be addressed to the
Secretary.

a
aie

1948 | DETLING: ENVIRONMENTAL EXTREMES 169

CONCENTRATION OF ENVIRONMENTAL EXTREMES
AS THE BASIS.,FOR VEGETATION AREAS

LeRoy E. DETLING

Concepts of vegetation areas are ordinarily based upon some
concept of endemism; that is, a geographical area whose flora con-
tains a noteworthy number of species endemic to that area is
thought of as a vegetational unit. These areas are usually given
boundaries coinciding with major physiographic features, but the
environmental features of the region enter only secondarily into
the picture.

In a recent publication based upon a transect study in western
and central Oregon (Detling, 1948) attention was called to the
correlation existing between the percentage of endemism in the
various vegetation belts and the degree of extremeness of environ-
mental factors in those same belts. It was shown that where
gradients of several environmental factors exist, those areas
where the greatest number of extremes occur are the ones which
show the highest percentages of endemism, while those areas
along the gradient where the environment is most moderate are
characterized by a relatively low degree of endemism.

This being true, it should hold that over any extensive geo-
graphical region those centers in which the extremes of a number
of climatic or other environmental gradients occur together will
be the centers of endemism as regards their flora. This should
serve as a basis for dividing such a region into vegetation areas,
‘each with its characteristic flora.

The present paper describes the procedure which has been
followed in mapping such areas for the Pacific Northwest, and
shows how the areas thus established are related to the distribu-
tion of a few of the dominant plant species.

The environmental factors upon which this investigation is
based are all climatic features, viz., annual precipitation, average
January temperature, average July temperature, and length of
the period between the last killing frost in the spring and the first
killing frost in the fall. These features in themselves may not al-
ways be what are most important as the limiting factors in the dis-
tribution of plant species, but they undoubtedly reflect those fea-
tures which are the limiting factors. Seasonal distribution of
rainfall may be of greater importance to plant life in some areas
than the annual total, but in the geographical region under consid-
eration the relative distribution of moisture among the seasons is
fairly constant, no matter what may be the actual total precipita-
tion. Again, the limiting factor for plants in winter temperatures
is the actual lowest temperature that occurs, but it is found in com-
paring these lowest temperatures where they are available that they

Maprono, Vol. 9, No. 5, pp. 137-168. March 20, 1948.

JUN 15 1948

170 MADRONO [ Vou. 9

vary from one locality to another directly as the average January
temperature varies, and are thus reflected with sufficient accuracy
in these January averages to make the latter valid for purposes of
our comparative study. |

The data used in this investigation are readily available in the
annual reports of the United States Weather Bureau. The cli-
matic maps are adapted from the 1941 Yearbook of the United
States Department of Agriculture, “Climate and Man.”’ Data on
the climatic features which might have more direct bearing on the
flora, such as those just discussed, are not available from a suf-
ficient number of stations over as large an area as the Pacific
Northwest, and to collect them would have been a task of too
large proportions for one worker to undertake.

Edaphic factors are of great importance in determining the
character of a flora. They frequently serve to explain local
variations, or in some instances even form the basis for subdivid-
ing the major vegetation areas into smaller units, but climate still
remains the chief basis upon which these major floras are deter-
mined, and for this reason edaphic conditions have not been
included in this study.

The first step in the mapping procedure was to determine the
gradients for each individual environmental factor. These are
shown in the climatic maps of Figures 1 to 4. A cursory inspec-
tion of the maps is sufficient to reveal the location of the extremes
of these gradients. Examples of extremes, either high or low,
may be noted in the annual precipitation in the Olympic Moun-
tains, in the Big Bend of the Columbia River, and at Pyramid
Lake in western Nevada, in the average July temperature on the
Oregon and Washington coasts, in the Rogue River Valley, and in
the Big Bend of the Columbia River, in the average January tem-
perature on the southern Oregon coast, the Cascade Mountain sum-
mits, and the lower Clearwater River in Idaho, and in the length
of growing season on the southern Oregon coast and on the pla-
teau of central Oregon.

The location of the more noteworthy of the gradient extremes
for all four climatic factors is indicated in Figure 5. High
extremes of temperatures and precipitation and the longest grow-
ing seasons are marked by solid dots, the opposite extremes by
circles. The letters accompanying these marks indicate the cli-
matic factors to which they appertain, as follows: P, annual pre-
cipitation; S, July (summer) temperature; W, January (winter)
temperature; G, growing season.

When the extremes for all factors are mapped together in this
manner it is noted that there are certain points around which they
tend to be concentrated. According to the findings of the study
previously referred to it is at these points that we should expect
the highest degree of endemism. The next step, therefore, was to
locate the major concentrations of extremes, which would serve as

1948 | DETLING: ENVIRONMENTAL EXTREMES 171

340 50 80 70 30 16 16 20

Fic. 1. Annual precipitation in inches in the Pacific Northwest.

the environmental centers of the vegetation areas for the region
under consideration. The solid triangles in Figure 6 indicate the
location of these centers.

Figure 6 shows, in addition to the environmental centers
already mentioned, the limits of the floral areas of the Pacific
Northwest. In fixing boundaries for the areas it was assumed
that such boundaries should represent as nearly as possible for

172 MADRONO [Vox. 9

Fic. 2. Average July temperature in °F. in the Pacific Northwest.

each climatic factor the isopleth marking the median point be-
tween gradient extremes. Since these isopleths do not coincide
for all factors, it was necessary to establish a line which most
nearly expressed their average. When so constructed, the bound-
aries between adjacent areas lie along the lines which are farthest
removed from what might be termed the “influence” of their

1948 | DETLING: ENVIRONMENTAL EXTREMES 173

Fic. 3. Average January temperature in °F. in the Pacific Northwest.

respective centers of extremes. In other words, the territory
within a vegetation area comprises what is environmentally most
closely related to the point of greatest concentration of extremes
around which the area is built. The sinuous courses frequently
followed by the boundary lines correspond to the courses of some
or all of the isopleths upon which they are based.

174 MADRONO [ Vou. 9

It should be emphasized that the important thing in each area
is its center of environmental extremes, not its boundaries. This
center is the center of endemism as well, and the farther we go
from this point of greatest influence the less pronounced does en-
demism become, until at the boundary of the area the flora par-
takes equally of the characteristics of two adjacent floras.

It will be noted that in some instances more than one area is
featured by the same combination of extremes. For example, most
of the mountainous areas are characterized by high annual pre-
cipitation, low summer and winter temperatures and short growing
season. It would seem that such similarity in climate would tend
to cause a corresponding similarity in floras. Such areas, how-
ever, are always separated from one another by regions in which
at least one, and sometimes more, of the climatic factors are at
its opposite extreme. These intervening belts, because of this
environmental difference, apparently serve as barriers to the
migration and consequent mingling of many plant species.

The concept of vegetation areas based upon centers of environ-
mental extremes offers an explanation for a number of features of
the distribution of plant species in the Pacific Northwest. The
location in western Nevada of the environmental center for the
long arm that extends northward into southern Lake and Klamath
counties in Oregon explains why so many species listed for the
latter locality extend into California and Nevada rather than
northward. Similarly, the location on the Columbia River of a
center of an environmental unit including north central Oregon
and central Washington explains why the flora of north central
Oregon is so closely related to that of central Washington and so
different from that of south central Oregon. The reason for the
sharp break in the coastal flora in the vicinity of Coos Bay, Ore-
gon, becomes evident when one notes the two coastal centers in
Del Norte County, California, and near Grays Harbor, Washing-
ton. Another interesting question is, ““Why is the Coast Range in
Oregon a region with so few endemic species?” It will be noted
on the climatic maps that there is no striking concentration of
extremes in these mountains, a fact which probably accounts for
the low degree of endemism. It is this lack of extremes which
excludes the Coast Range as a vegetation area in the present
scheme, although it has been so considered by other botanical
writers.

THe VEGETATION AREAS OF THE Pactric NORTHWEST

The following summary of the vegetation areas attempts to
present briefly the climatic features of each, along with the main
floral elements which characterize the area. No attempt is made
here to compile an extended list of endemic species, but merely
to show the framework for each area, within which a greater or

1948 | DETLING: ENVIRONMENTAL EXTREMES 175

Fic. 4. Average length in days of the growing season in the Pacific Northwest.

lesser number of endemics do occur, as may be ascertained by con-
sulting any manual of the flora of the region.

The names selected for the areas are for the most part those
of some geographical feature at or near the environmental center
of the area.

The present treatment is limited to those areas occurring either
wholly or in part in the states of Washington and Oregon, except

176 MADRONO [ Vou. 9

that the Sierra Area has been included because of its close floristic
relationship with the areas of southern Oregon.

Det Norte Area. This is a region of high annual precipita-
tion, low summer temperatures, high winter temperatures and
long growing season. The dominant species throughout much of
the area is Pseudotsuga taxifolia, but the two most characteristic
tree species are Sequoia sempervirens and Chamaecy paris Lawsoniana.
Other species marking the area are Umbellularia californica, Litho-
carpus densiflora, Quercus chrysolepis, Ceanothus thyrsiflorus, and
Rhododendron occidentale.

Nortu Coast Area. The climatic extremes here are the same
ones found in the Del Norte Area, the most significant difference
lying in the lower winter temperatures: occurring in the more
northerly area. Pseudotsuga tazifolia is again the dominant spe-
cies, but Sequoia sempervirens and Chamaecyparis Lawsoniana are
replaced by Picea sitchensis and notable stands of Thuja plicata
and T'suga heterophylla.

Otyrmpic Area. This mountain massif, representing the highest
elevation lying between the North Coast and Puget Areas, is
the center of the extremes of high precipitation, low summer and
winter temperatures, and short growing season. ‘The dominant
vegetation consists of a forest of Pseudotsuga taxifolia, Thuja plicata
and T'suga heterophylla. The occurrence of Abies lasiocarpa, A.
amabilis, A. procera, Tsuga Mertensiana and Chamaecy paris nootkaten-
sis, along with those species already named, emphasizes the gen-
eral similarity of the flora of this area to that of the North Cas-
cade. The Olympic Area, however, lacks the Picea Engelmanniz,
Lari« Lyallii and Pinus albicaulis of the Cascades, as well as a
number of herbaceous species as listed by Piper (1906). On the
other hand, a considerable number of herbaceous or shrubby spe-
cies are endemic to the Olympics.

Siskiyou ArEA. The mountain massif composed of the Siski-
vou and Trinity mountains, marked by a high extreme of precipi-
tation and low extremes of summer and winter temperatures and
of growing season, is separated climatically from the similar ex-
tremes of the Cascade-Sierra divide by opposite extremes in the
lower elevations of the Klamath, Pit and upper Sacramento val-
leys. The dominant vegetation in the western part of the area
(its environmental center) is a Pseudotsuga forest mixed with
Lithocarpus densiflora and Quercus chrysolepis. This changes east-
ward to a forest of Pinus ponderosa and P. Jeffreyz. ‘The area is
one of a high degree of floral endemism, marked by such species
as Pinus Balfouriana, Picea Breweriana, Cupressus MacNabiana,
Juniperus californica siskiyouensis, Lithocarpus densiflora echinoides,
Quercus Breweri, Q. Sadleriana and Q. vacciniifolia.

Rocur Area. This area is one of relatively low rainfall, high
summer temperature, moderate winter temperature, and long
growing season. The dominant vegetation is a deciduous wood-

177

DETLING: ENVIRONMENTAL EXTREMES

1948 |

; e:
aD REO CRG OS Ce Cntais J
es yy : "®@ eS er ay, i
Ww Sea m
of

Ma gta, oo

i a

ced a Thee eed
™e a

© Shes at
zt: : atten, A
SO SIDR EG wee Sally OMG

ne HY Oy 6

rey 7

° ot es ne
Qa iS nee: Sik
. Lé
SARs Ae
: “ <€
is “Y raed “y Ae CUP me
ax Ym FE
We tly Ok bes
Ouvi\i te €
PAW?
s* ¢f
2 ebeas
-* &

© re B [
a aay yh Some
Principal gradient extremes in the Pacific Northwest for annual

Fie. 5.
precipitation (P), average July temperature (S), average January temperature

(W), and length of growing season (G); high ®, low ©.

178 MADRONO [ Vou. 9

land consisting of Quercus Garryana, Q. Kelloggit and Arbutus Men-
gies, with a generous mixture of Pinus ponderosa, especially on
slopes just above the preceding. Important shrubs character-
izing the area are Rhus diversiloba, Ceanothus cuneatus and Arctos-
taphylos viscida.

Pucet Area. The great trough between the Coast Range and
the Cascades, extending from Puget Sound southward to the Cala-
pooya Mountains in Oregon, has the same climatic extremes as
does the preceding area. However, at its center, on lower Puget
Sound, the long growing season extreme is more pronounced than
in the Rogue, the high summer temperature extreme is less pro-
nounced, while winter temperatures are about the same. The veg-
etation of the Puget Area is predominantly the Pseudotsuga forest,
mingled in cool valleys and on north slopes with Thuja plicata and
Tsuga heterophylla, and in its drier phase with Arbutus Menziesii
and Quercus Garryana. It lacks the Picea sitchensis of the adjacent
North Coast Area, and the Quercus Kelloggu and Pinus ponderosa
of the Rogue.

Nortu CascapE Area. The high mountains of this area result
in a high extreme of precipitation, and low extremes in both
winter and summer temperatures as well as in length of growing
season. The vegetation at the altitudes where these extremes
occur is a subalpine forest consisting of T’suga Mertensiana, Abies
lasiocarpa and Pinus albicaulis, flanked at lower elevations by
Pinus contorta Murrayana and Picea Engelmannii, and by Pseudo-
tsuga taxifolia mixed with Abies procera and A. amabilis.

SoutH Cascape ArEA. This area, with climatic extremes sim-
ilar to those of the preceding one, is separated from it physio-
graphically and climatically by the break in the Cascade Range
where the Columbia River cuts through. The plant species dom-
inant in the North Cascade Area are present also in the South
Cascade, but the association in the latter is altered by the addition
of Pinus Lambertiana, P. attenuata and Abies magnifica and its vari-
ety shastensis.

SrerRA AREA. Geographically and physiographically this area
occupies a position similar to that of the North Cascade and South
Cascade Areas, and like them is marked by a high extreme of
precipitation and low extremes of summer and winter temper-
ature and of growing season. An environmental break at the
Pit and Klamath rivers marks the transition from the flora of the
South Cascade to that of the Sierra Area. Typical of the vegeta-
tion of the latter area are Pinus Jeffreyi, P. Lambertiana, P. Sabin-
tana, Sequoia gigantea, Carpenteria californica, Jamesia americana
californica, Staphylea Bolanderi, Rhamnus Purshiana anonaefolia, R.
rubra, Ceanothus Lemmonii and Fremontia californica.

Cotumsia ArgA. This comprises the valley of the middle por-
tion of the Columbia River and the lower portions of its tributaries,
chief of which are the Snake, Yakima, John Day and Deschutes.

i CK : :
ED ES\CHU RES Se
: i

are
vo’ é 2:

Fic. 6. Vegetation areas of the Pacific Northwest. Triangles indicate
the centers of environmental extremes.

Its low precipitation, high summer and relatively high winter
temperatures and long growing season are extremes of climatic
gradients whose opposite extremes are in the Cascade Mountains,
Blue Mountains and Okanogan Highlands, which together roughly
surround the area. The region is primarily grassland, with sage-

180 MADRONO [ Von. 9

brush keenly competing with the grasses. Characteristic species
are Agropyron spicatum, Poa secunda and Artemisia tridentata. In
the more alkaline situations Elymus glauca and Sarcobatus vermicu-
latus are common. Along streams and in the canyons some of the
species which mark the vegetation of this area are Rhus glabra
occidentalis, R. Toxicodendron and Prunus demissa.

CLEARWATER AREA. Centered in the lower valley of the Clear-
water River in Idaho, with its extremes of low rainfall, high sum-
mer and winter temperatures and long growing season, this area
extends up the valleys of the Clearwater and its tributaries and
out over the rolling hills of southeastern Washington. Its vege-
tation is typically a bunchgrass association like that of the Colum-
bia Area to the west. The hills and lower mountains within its
eastern limits support a forest of Pinus contorta Murrayana, P.
monticola, P. ponderosa and Pseudotsuga tazifolia. No strictly en-
demic species appears to be sufficiently prominent to serve as a
marker for the area, although a number of endemic species occur
locally.

Descuutes AREA. This area, occupying the extreme northern
portion of the Great Basin and the plateau region immediately
adjacent to it presents, with a few local exceptions, a fairly uni-
form set of environmental conditions extending over a consider-
able territory. The climatic extremes featuring the area are low
‘precipitation, low winter temperature and short growing season,
but with no marked extreme of summer temperature. The vege-
tation is basically a desert shrub formation consisting of Artemisia
tridentata, Chrysothamnus spp., and in the more alkaline localities
Sarcobatus vermiculatus, Grayia spinosa, Eurotia lanata, Kochia vestita
and various species of Atriplez. Arborescent species frequently
occupying slopes and some of the higher elevations are Juniperus
occidentalis, Pinus ponderosa and P. contorta Murrayana.

SNAKE River AreEA. The climatic center of this area, located
along the Snake River in Washington County, Idaho, and Baker
County, Oregon, differs from that of the adjacent Deschutes Area
in its high winter and summer temperatures and long growing
season. The general character of the vegetation, especially as it
extends up the Snake River Valley across southern Idaho, is
strikingly similar to that of the sagebrush plains of the Deschutes
Area, and the species listed as characterizing the latter might be
repeated for the region under discussion. However, especially
in its northern extremity in the vicinity of its climatic center,
endemic species or subspecies of more local distribution occur
with sufficient frequency to emphasize the distinctiveness of its
flora.

Wasnor Area. The Washoe Area occupies that portion of the
Great Basin lying just to the south of the Deschutes Area. Its
climatic extremes, viz., low precipitation, high summer and winter
temperatures and long growing season, are centered in the Pyra-

1948 | DETLING: ENVIRONMENTAL EXTREMES 181

mid Lake Valley of western Nevada. The characteristic vege-
tation of this area is again desert shrub, and the species com-
prising this formation are largely the same as those in the Deschu-
tes Area. However, the Juniperus occidentalis of the latter is here
replaced by J. utahensis and Pinus monophylla, while other species
distinguishing the Washoe Area include Ephedra spp., Celtis
Douglas and Prunus Anderson.

Biue Mountain Area. The high precipitation, low summer
and winter temperatures and short growing season of this elevated
area represent the extremes of climatic gradients extending out
into the valieys of the Columbia River to the northwest and the
Snake River to the east and south. The lower levels of the cli-
matic gradients are occupied by a forest of Pinus ponderosa associ-
ated with an undergrowth of Ceanothus velutinus. This forest is
replaced toward the opposite extremes by Pinus contorta Murray-
ana, Larix occidentalis, Abies grandis and Pseudotsuga taxifolia. The
Blue Mountain Area holds many species in common with the Rocky
Mountains, but at the same time is characterized by the occurrence
of many endemics. |

Oxanocan AreEA. This comprises largely the mountainous
region of northeastern Washington known as the Okanogan High-
lands, with its extremes of high precipitation, low summer and
winter temperatures and short growing season. The southern
margin of the area, representing its lower environmental levels, is
given over to a grassland formation similar in composition to the
grasslands of the Columbia Area. Going toward the environ-
mental center in the higher mountains the vegetation type changes
progressively to forests of Pinus ponderosa, Pseudotsuga tazifolia,
Larix occidentalis, Pinus contorta Murrayana, Thuja plicata, Picea
Engelmannit and Pinus monticola. The area is also marked by the
occurrence of Juniperus scopulorum, Betula papyrifera occidentalis,
Populus tremuloides aurea, Acer Douglasii, Ceanothus sanguineus, C.
velutinus and Rhus glabra occidentalis.

AREAL DISTRIBUTION IN A Few Paciric NorTHWEST GENERA

That the areas delimited in the manner previously shown do
constitute centers of endemism is indicated in the work of recent
writers who have monographed plant genera or sections of genera
which are well represented in the Pacific Northwest. An analysis
of these monographs reveals that a relatively large number of
species or subspecies are confined wholly or largely to a single
area as here treated. A random selection from the monographs is
presented in the following table, in which the first column indi-
cates the number of species or subspecies occurring in the Pacific
Northwest, and the second column the number of these which are
endemic to one area. These endemics are then listed by name and
their distribution indicated. Besides the species reported as

[ Vox. 9

x sny}ueUuoUl

% suo}sisiod

x TaUVILD

x HAT P80) 5 |

x snioyiun

sn}10yoo[eD tL =|) OW

x unipuesdéiyo wniopipueazs

winoprodeosiq[nul

ssusuUNONn}y

susoseindind

x | UnApueoNnes] UNUOTI10

x TTTOMO |

eet wunruoryy AI 9 cI
| | Weer)

x | | | | sisuoyeyn yseurendy |
| lafodye A yseurendy

| eIpsUlia}Ul ysemeny |
Xa | | Wee S|
| | evoinze yseuend
eUTXeUl YSeUeNy

) eIsseules)

Sey BULOSqe BSOUT}NIS
BIP[PYOL

ia

rac

MADRONO
*

ad

ia

~
Lam
=

am
op)

uesouryd
ST 9Q
soynyosaq
IIJEMIVITD
eIquIn[o)
BAILII
apRosey) “Ss
apeosen "N
noALYSIS
onsoy
psng
o1dwtA[O
$8807) "N
9}10N [2G
dIUIEpUy
[#191L

VUANAL) ISAMH LYON OLMIOVG MAJ V NI NOLLAGHLSid IVaANY ~“[ ATAVY,

182

183

ENVIRONMENTAL EXTREMES

DETLING

1948]

uResoury()

“STIN PNT

soynyosaq

1IJVMIVITD

Aw KA

PIquin[o;)

nr K AK

BAILII

apRoseg ’sS

rad

apeosrg "N

NOATYSIS

tad

onsoy
yosndg
o1duA

te)

BOD N
IWON PG

vnisqnd
susqnio1jze eLopisreds
TPpPrsn)
TyTeFNN
TPF MA
sIje}UOZI1OY SUZdSo}NAYNS
ear
SI[RJUSPINIO BYeAAT
laMoIg
epymeuurdqns
[1a[Yo0 y
susoseindind eyeoing
eordui Ajo eyediny

are ty
snye[norde
sapo
snso[noeu sndivd019eUl
snuvise[snog
sna[n1909
snjjAydouour
snqaodns
SNNUITUTUL
pea]

SI

66

(panwjuo,))—sny10yaso[e—

(panwuo))

‘[ 1av

dTULIpUy

TPI9L

[ Vou. 9

MADRONO

urvsoureyaO

“

‘SHA ON

JOYsEe

ayxeug

a

x
xX
x
x
ie
BP .<
xX
xX
>,€
>,¢
xX
a) C24) OR ae a ee
a Seal tae ee ei ran
= Se | & dk
we he ee lea ballet pa cieeny Bsc |S
Se |S) S | Se
+ | @ | =. a a =
nan | o | ® aah B=”
ar} fa)

A

v
S
ag
oO
cr

onsoy
o1du ATO
S80) °N
9}10N [PPC

S9proyyUelpoy
Saprlorusyoy
BIYyOA MA
sn[n}uesoya
snjze[nyyeds
stijsnyted snjzenue}ye
snupneyd
sny}e.iosqns
snuevAase A
sIsuo}seys
sn.oydoreydao snxopoiszoy
snoid 4} snxopo.19zoy
IPpId
SISUJUOSUIYSeM
snoumsue
sn}eao
snitodsoy
snoid A} 1ue47eY
uoUIa4ysSUuag
wino1d 4} winyzesnyqo
9[v9.10q WIn}eSn}qQo
wuinotd A} UNXP]
WINI[OJ1}eT WINXP]
UINpIs

184

(papnpuo)) *T atavy,

ST

pun) ~

JIU

ioe)

[POL

1948 | DETLING: ENVIRONMENTAL EXTREMES 185

endemic to one area only, there are others which occur in two or
three areas only, these areas being similar as to climatic extremes.

The genera and their monographers herein referred to are the
following: Tofieldia, Hitchcock (1944) ; Camassia, Gould (1942) ;
Erythronium, Applegate (1935); Calochortus, Ownbey (1940) ;
Arabis, Rollins (1986) ; Sedum, Clausen (1942); Penstemon, Keck
(1945); Wyethia, Weber (1946).

SUMMARY

In any region extensive enough to show physiographic variability,
the climatic or other environmental factors occur as gradients. Pre-
vious investigation has shown that where several extremes of these
gradients occur together they mark the center of an area rich in en-
demic plant species. In this study the Pacific Northwest has been
divided into sixteen vegetation areas, each built around one of these
centers of environmental extremes. Each area is briefly characterized
as to its dominant plant association. Examples from the recent works
of monographers show how the distribution of most plant species fits
into the pattern of these areas.

Museum of Natural History,
University of Oregon, Eugene

LITERATURE CITED

APPLEGATE, ELMER Ivan. 1935. The genus Erythronium: a taxonomic and
distributional study of the western North American species. Madrono
3: 58-113.

CLAusEN, Ropert T. 1942. Studies in the Crassulaceae—III. Sedum, subgenus
Gormania, section Hugormania. Bull. Torrey Bot. Club 69: 27-40.

Detiinc, LERoy E. 1948. Environmental extremes and endemism. Madrono
9: 137-149. -

Govu tp, Frank W. 1942. A systematic treatment of the genus Camassia Lindl.
Am. Midland Nat. 28: 712-742.

Hircucock, C. Leo. 1944. The Tofieldia glutinosa complex of western North
America. Am. Midland Nat. 31: 487-498.

Keck, Davin D. 1945. Studies in Penstemon—VIII. A Cyto-taxonomic ac-
count of the section Spermunculus. Am. Midland Nat. 33: 128-206.

Ownsry, Marton. 1940. A Monograph of the genus Calochortus. Ann. Mo.
Bot. Gard. 27: 371-560.

Piper, CHartes V. 1906. Flora of the state of Washington. Contr. U. S. Nat.
Herb. Vol. 11.

Roiirs, Reep C. 1936. The genus Arabis in the Pacific Northwest. Res.
Stud. State Coll. Washington 4; 1-51.

CLIMATE AND Man. 1941 Yearbook of Agriculture, United States Department
of Agriculture.

Weser, Witttam A. 1946. A taxonomic and cytologic study of the genus
Wyethia, family Compositae, with notes on the related genus Balsa-
morhiza. Am. Midland Nat. 35: 400-452.

186 MADRONO [ Vou. 9

THE GENUS HELIANTHELLA IN OREGON
Wittiam A. WEBER

Dr. Morton E. Peck’s Manual of the Higher Plants of Oregon
(1941) listed Helianthella Douglasii T. & G. as the sole species
then known to occur in that state. The present paper brings to
the attention of students of the Oregon flora three previously un-
reported species in this genus of Compositae and presents a
simple key to all the species now recorded from Oregon.

The collections upon which these new records are based are
not recent ones; on the contrary, they were made about fifty
years ago and are probably widely distributed in herbaria, but
for want of critical examination these have passed for individ-
uals of the common H. Douglasii.

An annotated list of the Oregon species of Helianthella fol-
lows. All specimens cited are from the herbarium of the United
States National Museum.

HELIANTHELLA CALIFORNICA A. Gray var. NEVADENSIS (Greene)
Jepson, Man. FI. Pl. Calif. 1081. 1925. H. nevadensis Greene in
Bull. Calif. Acad. Sci. 1: 89. 1885. Dry yellow pine woods, east
side of Johnson Prairie, Klamath County, Oregon, 1300 m. alt.,
June 13, 1898, EF. I. Applegate 2439. Grassy slopes, head of Elk
Creek, Umpqua divide, Douglas County, Oregon, 1500 m. alt.,
July 2, 1899, J. B. Leiberg 4191.

HELIANTHELLA QUINQUENERVIS (Hook.) A. Gray in Proc. Amer.
Acad. 19: 10. 1883. Helianthus quinquenervis Hook. in Lond.
Jour. Bot. 6: 247. 1847. Warner Range, Lake County, Oregon,
1650 m. alt., July 26, 1896, F. V. Coville §& J. B. Leiberg 82.
The westernmost station previously recorded for this essentially
Rocky Mountain species'was in Elko County, Nevada.

HELIANTHELLA UNIFLoRA (Nutt.) T. & G., Fl. N. Am. 2: 334.
1842. Helianthus uniflorus Nutt. in Jour. Acad. Sci. Phila. 7: 37.
1834. White Horse Mts., south of old Fort Smith, Harney
County, southeastern Oregon, Aug., 1901, David Griffiths & E. L.
Morris 450.

HELIANTHELLA UNIFLoRA (Nutt.) T. & G. var. Douglasii (T. &
G.) comb. nov. Helianthella Douglasii T. & G., Fl. N. Am. 2: 334.
1842. The common GHelianthella in central and northeastern
Oregon from the Ochoco Mountains to the Wallowa Mountains
and northward into Washington, northern Idaho, Montana, and
British Columbia appears to be a weakly differentiated north-
ward extension of H. uniflora of the Rocky Mountains and Great
Basin ranges. Its larger flower heads, ciliate rather than uni-
formly cinereous-pubescent phyllaries, and its occurrence at
lower altitudes separate the present race from H. uniflora proper.
Although most specimens may be placed without much difficulty

1948 | DAVIDSON: POLEMONIUM 187

in either of these two racial categories, transitional intermediates
do occur.

Key To THe OrecGon Species oF HELIANTHELLA

Cauline leaves alternate, except the lowermost pair. Leaves
chiefly basal, at the summit of a slender caudex...... HT. californica
var. nevadensis
Cauline leaves opposite. Leaves chiefly cauline. Caudex
stout.
Cauline leaves attenuate at both ends, long petiolate.
Phyllaries ovate, conspicuously ciliate, blackish on
CIEPAIAD 1 9 Ed Na? sel NON ee ae en OEE Ss Se eS Soa H. quinquenervis
Cauline leaves merely acute, short-petiolate or sessile.
Phyllaries lanceolate, variously pubescent, drying
green.
Heads large, 2.0-2.5 cm. broad excluding the rays.
IPhyilaries ciliate: #0). 21. s60a' sn tartan Me ede ees H. uniflora
var. Douglasti
Heads small, 1.5-2.0 cm. broad excluding the rays.
Phyllaries uniformly cinereous-pubescent......... H. uniflora

University of Colorado,
Boulder, Colorado.

A NEW POLEMONIUM FROM MEXICO

JoHN F. Davipson

Polemonium glabrum sp. nov. Erecta, humilis, 2-3 dm. alta;
rachis foliis angustate-alatis; foliolis 6—8-junctis, glabris vel cilia-
tis; calycis glabris, angustate-campanulatis, 10 mm. longis, seg-
mentis angustatis acutis, tubis equalibus; corollis ceruliis, in-
fundibuliformibus, 25-30 mm. longis, 15-20 mm. latis, lobis
tubarum brevioribus, spathulatis, apiculatis.

A slender erect perennial 2-3 dm. tall from a rootstock or
horizontal rhizome; leaves 4—8 cm. long with 13-17 leaflets 3-11
mm. long, 2-4 mm. wide, elliptical, acute, glabrous or ciliate on
a slightly winged rachis, the bases of the distal five leaflets com-
monly confluent; calyx glabrous, narrowly campanulate, 10 mm.
long, 4mm. broad, the segments narrow, acute, equalling the tube ;
corolla blue, truly funnelform in limb as well as in tube, 25-30 mm.
long, 15-20 mm. broad, the lobes two-thirds as long as the tube,
spatulate and apiculate; stamens inserted 3 mm. from the base of
the corolla-tube, pubescent at, and slightly above, the point of
insertion, 20 mm. long; style slightly exceeding the stamens,
shorter than the corolla; capsule ovoid, many seeded; seeds not
becoming mucilaginous when wet.

Type. Mt. Mohinora (10 miles west of Guadalupe y Calvo),
southwest Chihuahua, Mexico, September 1, 1898, E. W. Nelson
4865 (United States National Herbarium).

MADRONO

188

——

Sw ZA

WZ
Ze Y
AS

Z 2

Rm

aa,
J
AY H \N ’ XX
5 ip Y

Y, Fig. 2, calyx, X 1%.

Fig. 1, habit; X<

PLatE 23. PoLEMONIUM GLABRUM.
Fig. 3, floral dissection, natural size.

leaf apex, X 144.

Fig. 5,

Fig. 4, capsule dissection, x 1'/.

1948 | REED: EASTERN ASIATIC FERNS 189

Polemonium glabrum may be distinguished readily from any
other naturally occurring Polemonium by the shape and size of the
corolla, and by its glabrous calyx. The only other record of a
similar corolla is found in the report by Ostenfeld (Genetic studies
in Polemonium, Hereditas 12: 31-39. 1929.) of crosses between
P. mexicanum Cerv. ex Lag. and P. pauciflorum S. Wats. The
affinities of the present species may well be with the above, but
neither P. mexicanum nor P. pauciflorum have been reported from
the vicinity of P. glabrum. The probability of P. glabrum being
merely a hybrid (P. meaicanum x pauciflorum) was considered and
discounted because of the absence of the putative parents and
because of its constant pollen size. The pollen of known hybrids
has been found by the author to show irregularities in size,
whereas the pollen of P. glabrum is perfectly normal. Also there
is apparently no reduction in the number of seeds set, as might be
expected in the case of a hybrid plant.

Department of Botany,
University of California, Berkeley

NOTES ON THE TAXONOMY OF SOME EASTERN
ASIATIC FERNS OF THE GENERA PROTO-
WOODSTIA AND PTERETIS

Crype F. Reep

While glancing over a recent paper of Dr. R. C. Ching
(1945, p. 86) and a review of the same paper in Biological
Abstracts (1946), the author noticed the generic name Proto-
woodsia being used for a new genus of ferns. Since he was
familiar with this name as early as 1941, verifying the validity
of the name seemed necessary. In checking over his notes, the
author found the following sequence of circumstances.

In 1940 Dr. Ching (p. 245) used the generic name Proto-
woodsia, listing under it “P. manchuriensis (Hook.) Ching. <A
monotypic genus confined to N. E. Asia.”” The only description
of the genus is to be found in his description of the new family
Woodsiaceae, as “spores bilateral, dark-colored, with perispores,
or tetraedral, smooth and translucent (Protowoodsia).” Con-
cluding his description of the Woodsiaceae, Ching remarks that
this is “‘a very small family of two genera and about forty species;
its affinities with Cyathea and the next two families has generally
been recognized.”

The manner in which the name Protowoodsia appeared in
Ching’s paper led the author to believe that the generic name
had been validly published elsewhere. Therefore, in his doc-
torate thesis in 1942 (p. 73), the author used the generic name
as follows: “Protowoodsia Ching has translucent, smooth, sub-
globose to bilateral spores lacking an exospore (P. manchurien-

190 MADRONO [ Vor. 9

sis).” After studying the spores of specimens of Woodsia man-
churiensis from Japan and Korea (from herbarium sheets at the
Gray Herbarium), I am unable to verify the “tetraedral’” spores
Ching had described Protowoodsia as possessing.

In 1945 Ching (p. 86) published the genus Protowoodsia
properly, accompanied with a Latin description and with cita-
tions of the specific transfers. The following is Dr. Ching’s
description as originally written:

‘“ProTowoopsia, gen. nov. Genus Woodsiae § Euwoodsia,
habitu, configuratione et stipite continuo similis, differt praecipue
folia utraque nuda, pilis articulatis et paleis brunneis linearibus
desunt, indusiis inferioribus globosis griseo-membranaceis, satis
magnis, apice ore rotundata contracto dehiscentibus, spores
tetraedricae, translucentibus, laevibus.

Species unica in Asia boreali-orientalis incola.”’

Again, Dr. Ching describes the spores as being “‘tetraedral”
in shape. However, the apparently invalid use of the com-
bination Protowoodsia manchuriensis (Hook.) Ching, as well as of
the generic name, in the Sunyatsenia publication are totally
ignored in the later publication. Obviously, the earlier publica-
tion was a prepublication.

The spores in the specimens studied by the author are sub-
globose, totally lacking any trace of the triradiating crest so
prevalent in tetrahedral spores. They are definitely monolete,
subglobose to ellipsoidal in shape and light brown in color, lack-
ing an exospore, and therefore appearing smooth. Dr. Ching
called the loose outer covering of the spores the perispore; I
have designated it as the exospore. The perispore, which is the
outer spore-wall, in the case of Woodsia manchuriensis has a few
broad pits, but is otherwise smooth.

While studying the annular cells of the sporangia of various
species of the genus Woodsia, I observed that the range of vari-
ability in the number of cells in the annulus is confined to groups
of species. In Table 1 are given some of the species studied,
the locality of the specimens, and the number of cells in the annuli.

Judging from the range in the number of cells in the annulus
in P. manchuriensis (10-13), it belongs by itself. The spores and
the indusium also bear out isolating this species, probably in the
new genus Protowoodsia, as proposed by Ching, with an emenda-
tion in regard to the spore characters.

Protowoopsia Ching, Sunyatsenia 5(4): 245 (nomen subnud.).
1940; Lingnan Sci. Jour. 21: 86. 1945.

PRoTOWoopDsIA MANCHURIENSIS (Hook.) Ching, Sunyatsenia 5(4):
245 (nomen subnud.). 1940; Reed, Thesis, Harvard Univ., p.
73.1942 (ined.) ; Ching, Lingnan Sci. Jour. 21: 36. 1945.
Syn.: Woodsia (Physematium) manchuriensis Hook., 2nd. Cent.

1948 ] REED: EASTERN ASIATIC FERNS 191

Ferns, pl. 98. 1861; Diacalpe manchuriensis 'Trev., Nuov.
Giorn. Bot. Ital. 7: 160. 1875; Physematium manchuriense
Nakai, Bot. Mag. Tokyo 39: 176. 1925; Woodsia insularis
Hance, Ann. Sci. Nat. 1V. 15: 228. 1861 (non sensu Baker,
in Hook., Syn. Fil. ed. 2, 47. 1874).

Distribution: North China, Korea and northern part of Japan.

Another case of “prepublication” in these papers of Dr. Ching
(1940, p. 224) is the combination Pteretis intermedia in the de-

Taste 1. SuMMARY OF THE CELL NUMBER IN THE ANNULUS.

Species Locality Number of cells in annulus

Woodsia lanosa ...... Kansu 26, 27, 28, 27, 27, 26, 28.
W. glabella .......... Newfoundland 23, 25, 26, 23, 21, 22, 25, 21.
W. glabella .......... Spitzbergen 19, 17, 19, 21, 17.
W. macrospora ....... Kansu 24, 23, 22, 23, 23, 23.
We. CLON GOLA? cei. ss Kumaon 21, 24, 23.
W. oregana .......... Quebec 24, 18, 18, 20, 20, 23, 17, 18, 21.
W. cathcartiana ...... Michigan 20, 21, 23, 21, 18, 17, 18, 23.
WW WOCnStS. 22.0. eke Newfoundland 20, 21, 26, 20, 21, 21, 21, 22, 21.
Wo QUPING 2a hee ae New Brunswick 20, 18, 15, 18, 16, 17.
W. fragilis ..... de Maeciteh Caucasus 18, 18, 20, 18, 19, 18, 18.
W. polystichoides .... Manchuria 20, 17, 19, 18, 19, 17, 18.
W. subcordata ....... Manchuria 1 Ui Be pi fore ay ee rf
W. peruviana ........ Peru 18, 17, 18, 18, 18.
Vie CT ONAL Ua ee. st Bolivia 18, 18, 16, 17, 17, 16.
W. montevidensis .... Brazil P16, VE
W. meawicana ......... New Mexico 15, 17, 16, 17, 16, 18, 17, 18.
W. Plummerae ...... Arizona 14, 15, 14, 15, 16, 15.
WURTVOUAS: Min com tae eas Mexico 15, 14, 14.
W. scopulina ........ South Dakota 15, 15, 16, 16, 15, 16.
Wa ODUUSH. not ion 2 Maryland 14, 14, 15, 15, 14, 14, 17, 16.
Protowoodsia manchu-

PUCTESUS® 9 isbn ens Japan & Korea 12, 12, 10, 12, 10, 12, 13, 12.

scription of the new family Onocleaceae. In describing the
family there appears the following passage: “... sori superficial,
dorsal on the veins, globose, borne on cylindrical or convex re-
ceptacle and provided with white, membranaceous, fugaceous
inferior, shell-shaped or globose indusium (lacking in Pt. inter-
media)... . Ching only recognized two genera in the family,
saying nothing of the genus Pentarhizidium of Hayata (1927, p.
716 and 1928, p. 345). However, at this place Ching does not
cite the author of the specific epithet intermedia.

In his thesis again the present author (1942, p. 70) recognized
three genera as follows, based on spore characters: “In Onoclea L.
there are from thirty to thirty-two cells in the annulus. The spores
(O. sensibilis) are green, large, bilateral, with a large loose punctate
exospore. In Pentarhizidium MWHayata the spores are similarly
green, large, bilateral, with a very large loose punctate exospore (P.
orientale). The annulus has from thirty-three to thirty-four cells.

192 MADRONO [ Vor. 9

In Pteretis Raf. (Matteuccia Todaro) the spores are brownish, with
a smooth exospore, about one-half the size of those in Onoclea and
Pentarhizidium and of entirely different structure. There are only
twenty to twenty-seven cells in the annulus.”

In 1945 Dr. Ching (p. 36) makes the new combinations Pteretis
japonica (Hay.), Pteretis intermedia (C. Chr.) and Pteretis ori-
entalis (Hook.), citing the proper synonyms for the specific epithets.
Ching has taken the three species Hayata had placed in Penta-
rhizidium and transferred them to Pteretis, thus reducing the former
genus to Pteretis.

The sori in Pteretis are medial and biserial, protected by a
marginal lobe; in Pentarhizidium the sori are intramarginal and
uniserial, protected by the continuous dark-brown leaf-margin. The
spores of these two genera differ as stated above. The number of
cells in the annulus is quite significant: Pentarhizidium having
thirty-three to thirty-four cells with four to seven stomium-cells be-
sides; Pteretis having twenty to twenty-seven rarely thirty, cells with
four to six stomium-cells. These are probably distinct genera.

Preretis Raf., Amer. Monthly Mag. Crit. Rev. 2: 268. 1818;
Nieuwland, Amer. Mid]. Nat. 3: 197. 1914; 4: 334. 1915; C. Chr.,
Ind. Fil. Suppl. 2: 30. 1917; Small, Ferns Vicinity of New York,
140-143, fig. 1935; Mattfeld, Repert. Sp. Nov. Fedde 44: 289.
1938; Merrill, Amer. Fern Jour. 33: 55—56. 19438.

Syn.: Onoclea Hoffm., Deutsch. Fl. 2: 11.1795 (partim) ; Michx.,
Fl. Bor. Ame: 2: 272. 1803; Hook., Sp. Fil. 4: 161. 1862;
Hook. et Bak., Syn. Fil. 46. 1866; Struthiopteris Willd., Mag.
Ges. Nat. Fr. Berlin 3: 160. 1809 (non Struthiopteris Weis,
1770 nec Struthopteris Bernh., 1801); Pterilis Raf., Amer.
Monthly Mag. Crit. Rev. 4: 195. 1819; Pterinodes Kuntze,
Rev. Gen. Pl. 2: 819. 1891 (partim); Matteuccia Todaro,
Syn. Pl. Acot. Vasc. Sicilia, 30. 1866.

This genus contains the European species Pteretis sions
(L.) Nieuwl. (Amer. Midl. Nat. 3: 197. 1914) and the American
species Pteretis nodulosa (Michx.) Nieuwl. [Amer. Mid]. Nat. 4:
334.1916 (1915) ].

PenTarRHizipIum Hayata, Bot. Mag. Tokyo 41: 715-716. 1927;
42: 345. 1928.

PENTARHIZIDIUM INTERMEDIUM (C, Chr.) Hayata, Bot. Mag. Tokyo

41: 715. 1927; 42: 346. 1928.

Syn.: Matteuccia intermedia C. Chr., Bot. Gaz. 56: 337. 1913;
Pteretis intermedia (C. Chr.) Ching, Sunyatsenia 5(4): 224
(nomen nudum). 1940; Reed, Thesis, Harvard Univ., p. 70.
1942 (ined.) ; Ching, Lingnan Sci. Jour. 21: 36. 1945.

Distribution: China (Shensi and N.W. Yunnan), India (Sikkim).

PENTARHIZIDIUM JAPONICUM Hayata, Bot. Mag. Tokyo 42: 345. 1928.
Syn.: Matteuccia japonica (Hayata) C. Chr., Ind. Fil. Suppl. 3:

1948 | REVIEW 193

127. 1934; Pteretis japonica (Hayata) Ching, Lingnan Sci.
Jour. 21: 36. 1945.
Distribution: Japan.

PENTARHIZIDIUM ORIENTALE (Hook.) Hayata, Bot. Mag. Tokyo 41:

715-716. 1927; 42: 345. 1928.

Syn.: Struthiopteris orientalis Hook., 2nd. Cent. Ferns, pl. 4.
1860; Onoclea orientalis Hook., Sp. Fil. 4: 161. 1862; Syn.
Fil. 46. 1867; Matteuccia orientale (Hook.) Trev., Atti Ist.
Veneto III. 18: 586. 1869; C. Chr., Ind. Fil. 420. 1906;
Pteretis orientalis (Hook.) Ching, Lingnan Sci. Jour. 21: 36.
1945.

Distribution: Temperate China, East Himalayas and Japan.

Baltimore, Maryland.

LITERATURE CITED

BroLocican Assrracts. 1946. 20(1): No. 1432.

Cuine, R. C. 1940. On natural classification of the family ‘Polypodiaceae.’
Sunyatsenia 5(4): 201-268. (Oct.)

Cuine, R. C. 1945. The studies of Chinese ferns, XXXII. Lingnan Sci.
Jour. 21: 31-37.

Hayara, B. 1927. On the systematic importance of the stelar system in the
Filicales. I. (Japanese). Bot. Mag. Tokyo 41: 697-718, 25 text-fig.
Abstract 11, Jap. Jour. Bot. 4(4): 1928-29.

Hayara, B. 1928. On the systematic importance of the stelar system in the
Filicales. III. (Japanese). Bot. Mag. Tokyo 42: 334-348, 8 fig.
Abstract 163, Jap. Jour. Bot. 4: (57). 1928-29.

Reep, Crype F. 1942. The morphology of fern spores and its relation to
taxonomy. Thesis, Harvard. 87 pages, 33 plates. (Unpublished.)

REVIEW

A study of the genus Paeonia. By F.C. Stern. London, Royal
Horticultural Society. viii+ 155 pp., 15 colored plates, 28 text
figures, 8 maps. 1946. 638s.

There is probably no group of non-professional botanists to
whom plant science owes a greater debt than the botanical and
horticultural enthusiasts of Great Britain. Their energy in gath-
ering together collections of both specimens and living plants
from all corners of the earth, their care in raising a great variety
of rare, exotic, and “difficult”? species in their gardens, and their
generosity in financing the explorations and research studies of
their friends in the professional field of botany has widened
immensely our knowledge of the world’s flora. And their stand-
ards of execution have been consistently high, both as to the
accuracy of the research and the elegance of the publication.
Consequently it is more of a pleasure than a surprise to learn that
during Britain’s “darkest hour” of the last war there was being
prepared a botanical work which is not only a fitting successor to

194 MADRONO [Vor. 9

its great array of predecessors, but which in addition is in many
ways a model for progressive monographic studies of the future.

Mr. F. C. Stern’s work on Paeonia, modestly entitled a “‘study,’
is actually a magnificent folio volume, superbly illustrated with
accurate and highly artistic color plates of most of the species,
supplemented by line drawings of some of the technical details
and a complete set of distributional maps. The technical portions
of the monographic treatment; nomenclatural history, synonymy,
species descriptions, and discussions of diagnostic morphological
characters, are full and accurate. The artificial key to the species
is concise and as easy to follow as one can make it in a genus like
Paeonia, which simply does not have the clear cut diagnostic
characters found in many other groups. Most of the species
descriptions are accompanied by helpful and well written accounts
of the appearance of the living plants and of the proper methods
of culture. Mr. Stern has shown that he is not an amateur in any
sense of the word which refers to his degree of competence, but
that in regard to its literal meaning he is a true lover of the plants
to which he has devoted so much of his life.

But even more interesting than these parts of the work are
the sections on the cytology and distribution of the species. The
chromosome numbers are given of 34 of the 47 species, varieties,
and forms; of the 13 not listed, 8 are poorly differentiated varieties
of species of which the number is known, 2 are members of the
tetraploid mascula and officinalis complexes, and undoubtedly have
the same numbers as their close relatives; and the remaining three
are the rare P. kesrouanensis, native to Syria, and two closely
related endemics of southwestern China, P. Mairei and P. oxypetala.
Paeonia thus ranks with Crepis, Nicotiana, and Gossypium as one of
the best known cytologically of plant genera. The species are
all either diploids with the somatic number 2n = 10, or tetraploids
with 2n = 20.

The patterns of distribution of the various species provide
material for a most interesting discussion. Endemism is common
in the genus; 8 of the 33 recognized species are restricted to a
single island, mountain range, or other small area. The genus
as a whole occurs in five disjunct areas; the Mediterranean region;
central Asia from the Urals to Siberia with an outlier in eastern
Lapland; the western Himalaya; eastern Asia from southwestern
China to Manchuria and Japan; and Pacific North America. Such
a distribution is evidence of the great age of the genus, as is also
the primitive nature of its morphological characteristics. Stern
points out that the cytological condition of the species is charac-
teristic for each separate area of distribution. North America
contains only two diploid species, which have a distinctive type
of chromosome behavior at meiosis. The species of eastern Asia
are diploid with one exception, while those of Central Asia and
the Himalaya are strictly diploid. The Mediterranean species

1948] REVIEW 195

include both diploids and tetraploids, with the latter having by far
the widest distributions. This latter fact brings forth a very
plausible hypothesis as to the origin of these tetraploids. The
diploid species are believed to be preglacial relics, which were
pushed southward by the advancing ice sheet of the Pleistocene
period, and took refuge in the islands of the Mediterranean and
other warm areas. The tetraploids, which are believed to have
arisen from the diploids by autopolyploidy, were supposedly the
only forms which were able to migrate northwards in postglacial
times.

There are probably few genera about which two botanists,
studying the species independently and with different materials
available, would agree completely as to the true relationships and
boundaries of the species. Paeonia is no exception. The writer
has spent some time studying this genus, his work being based
largely on the living plants and interspecific hybrids kindly made
available to him by Dr. A. P. Saunders of Clinton, New York, but
supplemented by inspection of nearly all of the specimens available
in American herbaria. His synopsis (Univ. Calif. Publ. Bot. 19:
245-266. 1939) differs in some respects from the arrangement of
the species as given by Stern, and the evidence presented in Mr.
Stern’s study has not been sufficient to persuade him to change
more than a few of his concepts, except in the case of names which
must be altered for nomenclatural reasons.

In the first place, Mr. Stern’s concept of the species is entirely
morphological, and based chiefly on the ease with which they can
be recognized in herbarium specimens. The present writer con-
cluded that the three sets of characters which most sharply set off
the majority of the species are those of the sepals, the carpels
and stigmas at anthesis, and the mature seeds. Since, as Mr.
Stern states, none of these can be readily studied in herbarium
specimens, they are not included in either his key or the species
descriptions. In the writer’s decisions as to which forms should
be recognized as species and which as subspecies, the ability of
forms to cross in the garden and form fertile hybrids played an
important role, particularly if the types concerned were known to
occur naturally in the same or adjacent regions. Mr. Stern refers
in some instances to the observation of the writer and Dr. Saunders
that certain types hybridize freely in the garden, but fails to men-
tion the significant fact that in those instances where the genetic
evidence caused the writer to group different forms into the same
species, as in P. Delavayi, P. lutea, and P. Potanini; and in P. daurica
(“P. triternata”’) and P. Mlokosewitschii; the hybrids formed were
fully fertile. Some valid species, like P. Veitchi and P. Emodi, as
well as P. daurica and P. tenuifolia, also hybridize easily, but pro-
duce almost completely sterile F, hybrids characterized by very
irregular meiosis. On the other hand Mr. Stern places in the
same species as the yellow flowered P. Wittmaniana the plant from

196 MADRONO [ Vou. 9

the Caucasus first believed by Dr. Saunders and the writer to be
P. macrophylla and later P. tomentosa. This plant has white, not
yellow flowers; its sepals and petals are much broader than those
of P. Wittmaniana, and the shape of the sepals is entirely different;
its carpels are not only tomentose, but both the shape of the
carpels and that of the stigmas is entirely different from those of
the yellow flowered form recognized by Stern as P. Wittmaniana var.
nudicarpa. Furthermore, the hybrid between these two forms is
completely sterile. In every respect they appear to the writer
far more distinct than such species as P. arietina and P. officinalis,
which Stern places in different subsections of the genus, but which
are able to form fully fertile hybrids.

Considerations like these cast considerable doubt on the valid-
ity of the two subsections recognized by Stern in the section (or
subgenus) Paeon, namely Foliolatae and Dissectifoliae. They are
in general distinct and natural groupings, but exceptions to
this situation exist in the species groups of P. officinalis, P. pere-
grina, and P. arietina, all of them tetraploid, and admittedly by far
the most difficult species groups inthe genus. Mr. Stern has done
a great service in describing the characteristics of leaf morphology
by which their “species” may be identified, and in stating clearly
their geographic distributions. But his evidence that they occur
in adjacent areas rather than together in the same region, and
that in at least some instances they intergrade where their ranges
overlap, suggests to the present writer that they represent mem-
bers of a typical ““Rassenkreis” or polytypic species as recognized
by zoologists, and that each of Mr. Stern’s “species,” in these three
groups, with the exception of the more distinct and cytologically
diploid P. Clusi, is a typical geographic subspecies. This point
of view is supported by the hybridization experiments of Dr.
Saunders.

That these differences of opinion in regard to the limits of
species are not purely academic is evidenced by the fact that Mr.
Stern and the writer hold opposite points of view in regard to the
nature and origin of the tetraploids. His hypothesis that they
arose and spread in response to the climatic changes which took
place during the Pleistocene epoch is entirely plausible, but the
belief that each tetraploid species arose separately and inde-
pendently as an autotetraploid from a different diploid species is
not in accord with a number of facts. In the first place, many of
the tetraploids could, on morphological grounds, be just as easily
connected with an entirely different diploid species from the one
chosen by Mr. Stern. For instance, a comparison of the illustra-
tions in Mr. Stern’s study suggests that P. Russi, which he relates
to P. Cambessedesii, is in many respects like P. Broteri, and could be
derived from that species almost as easily, and study of details
of floral structure supports this view. P. mascula, which is con-
sidered to be derived from P. daurica, is in many respects equally

1948 | REVIEW 197

_ similar to P. Cambessedesiti and P. Broteri, and Stern remarks (p.

68) that “the form of P. mascula in Sicily looks very like P. Russi
when examined as dried specimens.” P. banatica, which such
botanists as Kitaibel, and Ascherson and Graebner have con-
sidered to be a variety of P. officinalis, is regarded by Mr. Stern as
derived from P. mascula, and so indirectly by autotetraploidy from
P. daurica. But in the descriptive section he states (p. 72): “It
is dificult to say whether this plant may be a variety of P. mascula
or of P. arietina, since it has some of the characters of both of these
paeonies.” P. arietina is believed to be an autopolyploid from
P. rhodia, a species very different from P. daurica. Finally, Mr.
Stern considers that the two genetically isolated species grouped
by him under P. Wittmaniana are autopolyploids of P. Mlokose-
witschi. But although P. daurica and P. Mlokosewitschii are inter-
fertile and differ in nothing except flower color and leaf shape,
their two supposed autotetraploids, P. mascula and P. Wittmaniana
(including P. tomentosa) are widely divergent in a number of mor-
phological characteristics, and would almost certainly form highly
sterile hybrids if intercrossed.

All of these facts support the writer’s belief that the tetraploid
peonies of the Mediterranean region form a typical polyploid com-
plex, in which autopolyploidy has figured to a certain extent, but
of which the majority of the species are allopolyploids derived
from crossing between either the present day diploids or their
ancestors or autopolyploid forms. Those belonging to the sub-
section Foliolatae are derived from the diploids of this subsection,
P. Cambessedesii, P. Broteri, P. rhodia, and P. daurica. But the
tetraploid Dissectifoliae, namely P. officinalis and its relatives, prob-
ably represent ancient allopolyploids involving on the one hand
Mediterranean diploids, like P. Clusii, P. rhodia, and P. daurica,
and on the other, the central Asiatic P. anomala. The best mor-
phological evidence for this hypothesis lies in the, appearance of
P. peregrina, the most easterly of these tetraploids, which is one of
the two species for which Mr. Stern could not find a diploid an-
cestor. But P. officinalis, which Mr. Stern believes to be an auto-
tetraploid of P. Clusii, differs from the latter species in its rela-
tively narrow leaves, while the most common effect of autopoly-
ploidy on leaf shape in dicotyledons is to make the leaves shorter
and broader. The influence of P. anomala, which has narrow as
well as strongly lobed leaflets, would tend to produce precisely
the divergence in leaf shape which is found in P. officinalis and P.

peregrina as compared to P. Clusii. Furthermore, these tetra-

ploids have one leaf characteristic not observed by Mr. Stern which
is characteristic of P. anomala and its relatives, but is not found in
any of the Mediterranean diploids, including P. Clusii; namely the
presence of short, scabrous pubescence along the veins of the
upper surface of the leaf. Finally, the hybrids produced by Dr.
Saunders between P. Mlokosewitschii and P. anomala as well as P.

198 MADRONO [ Vor. 9

Veitch resemble closely members of the P. officinalis complex in
all of their features of external morphology.

The cytological evidence, also, supports the present writer’s
hypothesis. In the species of Paeonia, with their large chromo-
somes and random distribution of chiasmata, a high proportion of
multivalents is to be expected if the component genomes of a
tetraploid are completely homologous, or even if they are not quite
so. But in most of the tetraploids investigated by the writer and
Mr. S. O. S. Dark (Jour. Genetics 32: 353-3872) including P. off-
cinalis, P. peregrina, and P. Wittmaniana, the number of quadrivalent
configurations per nucleus is only one or two, with most of the
chromosomes paired as bivalents. This would suggest that the
four component genomes of these tetraploids are not completely
homologous, and that they belong to the category recently charac-
terized by the writer (Advances in Genetics 1: pp. 417-421) as
segmental allopolyploids, or polyploids of which the component
genomes bear the majority of their chromosomal segments in com-
mon, but in which these genomes differ from each other by a large
enough number of such segments so that free interchange between
them is barred by complete sterility on the diploid level. The
fact that most diploid inter-specific hybrids of Paeonia may form
as many as four or five bivalents suggests that polyploids derived
from them would be of this nature.

The hypothesis that the members of the P. officinalis complex
arose as allotetraploids from hybrids between the Mediterranean
diploids and P. anomala presupposes that at the time when these
hybridizations took place the distributions of the diploid species
were very different from what they are now. But both Mr. Stern
and the writer are agreed that the present Mediterranean diploids
are relics which had a considerably wider distribution before the
beginning of the Pleistocene ice age. And fossil remains of late
Tertiary age from western Europe, particularly the abundant
seeds collected by Reid and Reid in the Pliocene deposits of the
lower Rhine basin, contain a large proportion of species of flower-
ing plants now confined to Asia, indicating the presence of a
strong Asiatic element in the European flora at that time, which
might easily have included Paeonia anomala or a relative of that
species. The writer, therefore, would like to modify Mr. Stern's
hypothesis in regard to the origin and evolution of the tetraploid
species of Paeonia, and believes that they originated through a
series of hybridizations between diploid species or their auto-
tetraploid derivatives. The first of these hybridizations took
place not later than the middle or end of the Pliocene epoch, but
the process very likely was continued during the interglacial
periods of the Pleistocene. The tetraploids have persisted and
spread not only because of such beneficial qualities as might have
been given them by their increased chromosome number, but also,
and perhaps chiefly because they possess favorable combinations

1948 ] NOTES AND NEWS 199

of genes derived from ecologically as well as morphologically
different ancestral species, which gives them a relatively wide
range of tolerance of diverse ecological conditions.

The final decision as to the correctness of one or the other of
these hypotheses, as well as to the validity of the writer's species
concepts insofar as they differ from those of Mr. Stern, cannot be
made through any attempt to improve on Mr. Stern’s fine mono-
graphic study by means of examining further the herbarium speci-
mens and garden plants now available to us. Careful studies are
needed of the critical species as they grow in nature, and the
splendid series of interspecific hybrids produced by Dr. A. P.
Saunders needs to be increased and studied more carefully. Un-
fortunately the present state of the world makes both of these
types of studies seem like remote ideals rather than actualities for
the immediate future. Such parts of the globe as Dalmatia,
Greece, Syria, and the Caucasus are considered by most people at
present to be critical areas for very different reasons from the
fact that those regions will yield important information about the
relationship of Paeonia species. And the years of labor and de-
votion expended by Dr. Saunders on his beautiful creations are a
scarce commodity in this age of fear, hurry, and utilitarianism.
But peonies have existed on this earth for many millions of years,
and they will still be with us when the world settles down to a
more normal way of living. And when that time comes, Mr.
Stern’s “‘study” may be looked upon as one of the outstanding
achievements of the present period in the history of plant science.
—G. L. Stessins, Jr., University of California, Berkeley, Cali-
fornia.

NOTES AND NEWS

RANGE EXTENSIONS oF GrassEs INTO CoLorapo.’ In connection
with the preparation of a flora of Colorado many plants not listed
for the state in the various manuals and monographs have come
to light. Among these unrecorded plants are 32 grass species.

Because of the great economic importance of the grass family
in this region, and because, as far as can be ascertained, the
majority of these grasses are a part of the actual flora of the state,
it was considered worth while to put them on record, together
with the herbaria wherein the specimens are deposited. The fol-
lowing abbreviations are used: University of Colorado (CU);
Colorado Agricultural and Mechanical College (CA); Colorado
College (CC); United States Forest Service, Regional Office,
Denver (FS); Rocky Mountain Forest and Range Experiment
Station, Fort Collins (FES); private herbarium of Paul Ginter,
Fort Collins (G); Soil Conservation Service, maintained by the
Department of Range and Pasture Management, Colorado Agri-
cultural and Mechanical College (SS); United States National

1 Scientific Series Paper 216, Colorado Agricultural Experiment Station.

200 MADRONO [ Vox. 9

Museum (US); Rocky Mountain Herbarium, University of Wyo-
ming (W).

I. Casual, perhaps cultivated, introductions:

1. Anthoranthum odoratum L., CC, SS.

2. Brachypodium distachyon (L.) Beauv., CU.
3. Bromus catharticus Vahl., CA, W.

4. Cynodon Dactylon (L.) Pers. CA.

5. Cynosurus cristatus L., CC, SS.

6. Eleusine indica (L.) Gaertn., CA, FS.

8. Muhlenbergia Schreberi Gmel., CA.

9. Paspalum racemosum L., CU.

0. Poa trvialts Li. CU; |

1. Zizania aquatica var. angustifolia Hitche., CC.

II. Native or well established grasses:

Agropyron latiglume (S. & S.) Rydb., FS.
2. Aristida Curtiss (A. Gray) Nash, CA.

3. Aristida purpurea Nutt., W.

4. Bouteloua barbata Lag., CA.
5
6

—s

Calamagrostis montanensis (Scribn.) Seribn., CA.
Calamovilfa gigantea (Nutt.) Scribn. & Merr., CA, CC,
FES, SS.
7. Danthonia unispicata (Thurb.) Munro, CA.
8. Elymus salina Jones, US.
9. Eragrostis Barrelieri Daveau., CA,SS.
10. Eragrostis trichodes (Nutt.) Wood, CA.
11. Eriochloa contracta Hitche., CA, CU.
12. Muhlenbergia arenacea (Buckl.) Hitche., CA.
13. Panicum huachucae Ashe, CA.
14. Panicum Wilcoxrianum Vasey, FS, G.
15. Paspalum stramineum Nash, CA, W.
16. Poa bulbosa L., CA, FS, W.
17. Poa scabrella (Thurb.) Benth., CU.
18. Sorghum halepense (L.) Pers., CA.
19. Sporobolus heterolepis (Gray) Gray, CU, CA, SS, CC.
20. Sporobolus neglectus Nash, CA.
21. Stipa spartea Trin., CA, CC, W.

None of the above 32 grasses has been previously directly
recorded for Colorado in the knowledge of the writer. All speci-
mens considered doubtful were sent to the United States National
Museum for determination.—H. D. Harrineron, Department of
Botany and Plant Pathology, Colorado Agricultural and Mechani-
cal College, Fort Collins, Colorado.

MADRONO

A West American Journal of Botany

—_

A quarterly journal devoted to important
and stimulating articles dealing with
plant morphology, physiology, taxonomy,
and botanical history. These volumes
should be a part of every botanist’s li-
brary and should be made accessible to
students of all universities and colleges.

Volume I, 1916-1929. . . $5.00
Volume II, 1930-1934 .. 5.00
Volume III, 1935-1936 . 5.00
Volume IV, 1937-1938 . 5.00
Volume V, 1939-1940 . . 5.00
Volume VI, 1941-1942 . 5.00.
Volume VII, 1943-1944 . 5.00
Volume VIII, 1945-1946 . 5.00
Single numbers....... 1.00

The subscription price of MADRONO
is $3.50 per year. We solicit your pat-
ronage,

Address all orders to:

Rimo Bacigalupi, Bus. Mgr.

Natural History Museum,
Stanford University, California .

VOLUME IX | | NUMBER 7

a i
‘

ih
hy
Ge
ay
a
ipa
Ty
i
are
*h'

<<

Cl

fy

. MADRONO

a A WEST AMERICAN JOURNAL OF

E BOTANY
i Contents

i Somer Prostems in THE GeNvus Gita, Herbert L. Mason and Alva D. Grant. 201
‘ POTAMOGETON LATIFOLIUS IN Texas, W. C. Muenscher .................... 220
f THe Prace or Wittis Linn Jepson 1x Catrrornia Botany, David D. Keck. 223
ee :

if Reviews: J. B. Hutchinson; R. A. Silow and S. G. Stephens, The Evolution

ay of Gossypium and the Differentiation of the Cultivated Cottons (Thomas
ie H. Kearney); Carl B. Wolf and ee W. Wagener, The New World

Xf. Cuypresses (David) Deck yt). 5 NRO ae I eS gale 228
‘ Notes ann News: Elymus aristatus in California, Frank W. Gould ....... 232

rh Sh A
i ve D oy “a
i v bend fed © bea Shs ,
f

4, * ff
4 uF {i ' , whl a
T AS fF ity: a RES on >* “et
1 —~ es deo .
: | eat
1
| Published at North Queen Street and McGovern Avenue,
| Lancaster, Pennsylvania
;|

| July, 1948

MADRONO

A WEST AMERICAN JOURNAL OF BOTANY

Board of Editors

Herpert L. Mason, University of California, Berkeley, Chairman.

LeRoy Asrams, Stanford University, California.

Epcar ANDbDERSON, Missouri Botanical Garden, St. Louis.

Lyman Benson, Pomona College, Claremont, California.

Hersert F, CopeLtanp, Sacramento College, Sacramento, California.

Ivan M. Jounsron, Arnold Arboretum, Jamaica Plain, Massachusetts.

Mitprep EE. Marutas, Dept. of Botany, University of California, Los Angeles 24.
Bassett Macuire, New York Botanical Garden, N. Y. C.

Marion Ownsey, State College of Washington, Pullman.

Secretary, Editorial Board—ANwnetra Carter
Department of Botany, University of California, Berkeley

Business Manager—Rimo Bacica.upi
North Queen Street and McGovern Avenue, Lancaster, Pennsylvania
or
Natural History Museum,
Stanford University, California

Entered as second-class matter October 1, 1935, at the post office at
Lancaster, Pa., under the act of March 3, 1879.

Established 1916. Published quarterly. Subscription price $3.50 per year.
Completed volumes I to VIII $5.00 each; single numbers $1.00.

Papers up to 15 or 20 pages are acceptable. Longer contributions may
be accepted if the excess costs of printing and illustration are borne by the
contributor. Range extensions and similar notes will be published in con-
densed form with a suitable title under the general heading “Notes and News.”
Articles may be submitted to any member of the editorial board. Manuscripts
may be included in the forthcoming issue provided that the contributor pay the
cost of the pages added to the issue to accommodate his article. Reprints of
any article are furnished at .a cost of 4 pages, 50 copies $5.90; 100 copies $6.48;
additional 100’s $1.23; 8 pages, 50 copies $8.55; 100 copies $9.49; additional 100’s
$1.88;°16 pages, 50 copies $12.00; 100 copies $13.44; additional 100’s $2.88.
Covers, 50 for $3.95; additional covers at $2.38 per hundred: Reprints should
be ordered when proofs are returned.

Published at North Queen Street and McGovern Avenue, Lancaster,
Pennsylvania, for the

CALIFORNIA BOTANICAL SOCIETY, INC.

President: Adriance S. Foster, University of California, Berkeley. First
Vice-President: G. Ledyard Stebbins, Jr., University of California, Berkeley.
Second Vice-President: Herbert F. Copeland, Sacramento College, Sacramento,

California. Secretary: John Whitehead, University of California Botanical

Garden, Berkeley. ‘Treasurer: Rimo Bacigalupi, Natural History Museum,
Stanford University, California.

Annual membership dues of the California Botanical Society are $3.50,
which includes a year’s subscription to Madrofio. For two members of the same
family the dues are $4.00, which includes one copy of Madrofio and all other

privileges for both. Dues should be remitted to the Treasurer. General |

correspondence and applications for membership should be addressed to the
Secretary.

1948 | MASON: GILIA 201

SOME PROBLEMS IN THE GENUS GILIA
Hersert L. Mason anno Atva D. Grant

Of all of the genera of Polemoniaceae, the genus Gilia, when
viewed in terms of the treatments accorded it in the literature of
botany, presents the most confused picture. In the first place
there has been confusion as to generic limits. With the exception
of Phlox and Polemonium all of the herbaceous genera have at one
time or another been included in Gilia. The nature of this con-
fusion has been discussed previously by the senior author (1945)
and will not be discussed further here. Another aspect of the
confusion in Gilia results from the fact that the genus is replete
with polymorphic species and intergrading populations that seem
to defy rational treatment along traditional taxonomic lines.
This, it seems to us, relates itself to the nature of the environment
and its influence on the genetic elaboration of the populations of
species. Gilia has its distributional center in the arid southwest-
ern United States. Here, the moisture factor approaches the
minimum in several of its aspects and soils display great local and
geographic diversity as to origin, maturity, hydrogen ion concen-
tration and degree of leaching of the mineral content. The
moisture factor, approaching the minimum as it does, according
to Liebig’s (1843) law results in striking habitat differences owing
to moisture differences of small amount. These local differences
result from differences in the annual rainfall or from seasonal
fluctuations from year to year in the same area. This latter
aspect is of very great significance in the floristic expression in
any given desert or semi-arid area from season to season, and
appears to manifest itself in a selective way on the threshold of
germination of the stored seed that may be present in the soil.
One sequence of moisture-temperature variables will cause a
given set of seed to germinate while another sequence appears to
favor another set. When this fluctuation and geographic varia-
tion of the moisture factor is superimposed over the geographic
variation in other edaphic conditions there result an enormous
number of significant habitats such as one does not encounter in
more humid areas. Striking differences in floristics from one
habitat to another and from one season to another in the same
habitat result. Through their genetic and physiological re-
sponses to these varied habitats many genera in many of the
families of the arid southwest have become very complex. It is
not surprising that the taxonomist working on such genera be-
comes frustrated in his interpretations unless he has a full appre-
ciation of the potentialities of genetic processes as they function
to elaborate the species populations over such habitats. Even
then he can utilize this information only in organizing his problem

Manrono, Vol. 9, No. 6, pp. 169-200. June 9, 1948.

SEP 2 4 Me ak 2

202 MADRONO [Vox. 9

and in producing a taxonomic arrangement that will provide the
necessary background for further field studies, breeding, and
cytological investigation where these are needed for a more
thorough analysis of relationships. The present treatment at-
tempts only the preliminary organization aimed to point the way
to such studies and attempts to discuss the problems that have
arisen during the preparation of the manuscript for the treatment
of the genus Gilia in Abrams, Illustrated Flora of the Pacific
Coast States. Those species that have presented no particular
problems have been omitted from this discussion as are also the
general key and the formal descriptions. All of these are in-
cluded in the treatment in the above mentioned flora.

Tue Taxonomic ENTITIES

It is our firm conviction that the main objective of taxonomy
is to give expression to the interrelationships that the taxonomist
construes to exist in the group of plants under investigation. It
would indeed be wonderful if the taxonomic categories with which
he had to deal were each discrete with values fixed by definition
or by legislation and utilized characters that offered no great
problems in their interpretation. The interrelationships that are
apparent in Gilia are exceedingly complex and display much inter-
gradation. In some of the subgenera, races of almost every con-
ceivable taxonomic magnitude exist ranging from the small pop-
ulation with one or two distinctive characters to what we choose
to regard as species, and groups of species within the subgenera.
Whether we look upon this complex in terms of morphological
characters, evident crossability or any of the many aspects of
ecological differentiation, the same complex situation exists. It is
not a gradient of variation but rather a mosaic of interlocking cen-
ters of variation wherein groups of greater or less distinctness are
evident and are distinctive because of any of several variants which
may involve either morphology, behavior pattern, or ecology. To
sort out these entities and express them in terms of values con-
strued to designate or to delimit such categories as species or sub-
species would serve only to fit into these categories groups of pop-
ulations that are obviously of a very heterogeneous nature. Species
and subspecies can only be applied to such groups in the relative
sense in which they are outlined in the International Rules of Bo-
tanical Nomenclature (1935, Arts. 10, 12). What we may desig-
nate as species another may regard as subspecies and what we may
designate as subspecies another will construe to be species. Tax-
onomic evaluation is only a tool for the expression of relationship
and it is not too important that two workers agree precisely upon
it if by their diverse concepts they arrive at the same pattern of
relationship. We hope that in the entities that we have desig-
nated as species we have included groups of populations that
possess a reasonably high degree of morphological uniformity cor-

1948 | MASON: GILIA 203

related with a range of physiological capacity that expresses itself
in a particular pattern of ecological and geographical distribu-
tion. Judging from the presence of intermediates between these
groups of populations, they produce hybrids which become estab-
lished in nature. On the other hand, these same groups either
do not hybridize with other populations, or if they do, the progeny
fail to establish themselves in the dispersal range of the seed pro-
ducing parent. Where species would be so large as to become
unwieldy for practical taxonomic use we have not followed a
strict interpretation of this philosophy. Our chief objective has
been to orient the entities so as to depict and characterize constel-
lations of relationship as we see them. .

Since little is to be gained by postulating relationships below
the rank of subspecies without the aid of genetic manipulation of
representative material, the present treatment is not carried below
the level of subspecies. We do not imply by this that we regard
the category “subspecies” and “variety” as being synonymous but
rather that we accept both in the sequence as outlined in the Inter-
national Rules of Botanical Nomenclature (1985, Art. 12).
Every individual plant is potentially a member of every category
in the taxonomic structure and although we list varieties under
our subspecies and make new combinations involving subspecies
we do not imply by this that the variety. is raised to the level of
subspecies or submerged in synonomy with the subspecies or the
species. Technically the variety is at least part of the subspecies
or the species. We include it in our literature citations only for
bibliographic completeness.

Gilia ranges from southern British Columbia southward
through the mountains and valleys into Mexico, thence eastward
across Texas to the south Atlantic Coast. It is adventive as far
north as Massachusetts. It recurs along the west coast of South
America from Peru to Patagonia. The great preponderance of
species, however, occurs in the arid regions from southern Cali-
fornia to western Texas and northward into the Great Basin,
with the Colorado and the Mohave deserts being especially rich
in species. Although Gilia is predominantly a North American
genus, it was first described by Ruiz and Pavon and was based
upon the Peruvian species G. laciniata. It is named in honor of
Felipe Luis Gil, a Spanish botanist. The following description
outlines our concept of the constitution of the genus.

Git1a Ruiz and Pavon, Prodr. Fl. Peru, 25: t. 4. 1794.

Annual, biennial, or perennial herbs, rarely subshrubby.
Leaves alternate, herbaceous rarely slightly rigid, entire or vari-
ously pinnately lobed, toothed, or dissected, often disposed in a
basal rosette. Flowers solitary on slender pedicels in the leaf
axils, or in paniculately branched or thyrsoid inflorescences, or
congested in glomerules or sessile in capitate heads. Calyx lobes

204 MADRONO ['Vou. 9

usually equal, cleft nearly to the base and often flanked on the
margins by a membrane, that of adjoining sepals often uniting to
form a pseudotube which becomes distended or ruptured by the
growing capsule. Corolla funnelform or salverform or less often
campanulate, usually regular, rarely slightly irregular, blue, pink,
red, yellow or white. Stamens equally inserted on the corolla
tube or the throat, most often in or just below the sinuses of the
corolla lobes, sometimes unequally inserted, usually equal in
length, rarely unequal. Capsule 38-celled, the valves remaining
joined at the base and campanulately spreading on dehiscence.
Seeds usually several to many in a locule, rarely 1 or 2, rapidly
taking up water when wetted and becoming mucilaginous on the
surface, rarely not so affected.

For the most part relationships within the genus aggregate
the species into fairly distinctive groups which have been vari-
ously treated as sections or as subgenera at the hands of several
botanists. Most of these subgenera as here utilized are natural,
although to treat them so it has been necessary to confine the
application of some of them to a few or even single species. To
do otherwise would frustrate our announced objectives. Ina few
of the subgenera we were strongly tempted to break them en-
tirely from Gilia as separate genera but it soon developed that it
would be more difficult to give expression to the interrelationships
between such groups if they were so separated. The following
key will serve to differentiate the subgenera.

KEY TO THE SUBGENERA

Seeds very many to a locule, ellipsoidal, reddish
brown, not mucilaginous when wetted; leaf
blades chiefly broadly elliptic sometimes shal-
lowly lobed, dentate, the teeth often aristate,
plants annual or perennial .................. Subgenus Gilmania
Seeds several to a locule, rarely one or two,
usually mucilaginous when wetted; leaves vari-
ously dissected or lobed or entire, rarely with
a broad elliptic blade.
Plants biennial or perennial, if annual, the in-
florescence leafy-bracted.
Corolla 20-30 mm. long, red, pink, yellow, or
white; inflorescence a thyrsoid panicle ... Subgenus Ipomopsis
Corolla 4-10 mm. long, white; inflorescence
capitate-congested or glomerate, usually
leaf y—bracted? icc Pte Uae eee ee ees Subgenus Elaphocera
Plants annual.
Ovules 1 or sometimes 2 to a locule; leaves
irregularly toothed or lobed or lanceolate-
entire; stamens unequally inserted on the
long, narrow throat 27) ea hieee ee Subgenus Greenianthus
Ovules several to a locule, stamens usually
equally inserted on the throat or tube or
in the sinuses of the corolla lobes; leaves
various.

1948] MASON: GILIA 205

Leaves variously toothed, lobed, dissected,
or divided, rarely entire; flowers in pa-
niculate, thyrsoid, glomerate or capitate
inflorescences, rarely solitary in the
upper leaf axils.

Stems conspicuously leafy, leaves becoming
reduced only high in inflorescence;
throat usually full campanulate, often
equal or longer than tube; _ basal
rosette rarely well differentiated at
maturity of plant; inflorescence often
Cab abere arte Mite Shs a ahmed eich aise a Subgenus Capitata

Stems not conspicuously leafy, cauline
leaves much smaller than basal, basal
rosette prominent; throat usually
ample and short, less commonly as
long or longer than the elongate tube;
inflorescence never capitate .........

Plants with one to several erect stems

usually branching above; corollas

usually with elongate tubes; basal

leaves strap-shaped or dissected,

usually 2-10 cm. long ............. Subgenus Hugilia
Plants low and divaricately spreading;

corollas with short tubes, sometimes

the throat elongate; basal leaves

ovate to ovate-lanceolate, rarely

Over slCmin LOM Paseo eel Subgenus Campanulastrum

Leaves, or most of them, linear to linear-
filiform, rarely a few pinnately dissected
into few filiform lobes, never broad and
toothed; flowers solitary in the leaf

axils.
Corolla tubular to narrow funnelform,
pink, white or pale blue ............ Subgenus Kelloggia

Corolla open campanulate ............. Subgenus Tintinabulum

Subgenus Gilmania subgen. nov.

Annua aut perennis, laminis foliorum latis, simplicis, dentatis
vel lyrati-lobatis aut partitis, dentes subulatis vel aristatis; corol-
lis parvis splendidis-rosaceis; multispermatis subrubris-fulvis.

Annual or perennial, leaf blades broad, simple, toothed, or
lyrately lobed or parted; the teeth subulate to aristate; corollas
small, bright pink ; seeds many, reddish brown. Type. Gilia lati-
folia.

The subgenus Gilmania is composed of two well-marked species
differing in several details though obviously closely related to one
another. Giulia latifolia Gray is an annual and G. Ripley: Barneby
(G. Gilmani Jepson) is a perennial. They have in common the
very many ellipsoid seeds to a capsule, each of which is pigmented
with a red-brown color, the broad leaf blades with their aristi-
form teeth, and the numerous, small, pink corollas. It is one of
the most distinctive groups within the genus yet its inclusion in
Gilia seems beyond question, since it ties in closely with G. lepto-
meria of the subgenus Eugilia.

206 MADRONO [Vor. 9
Subgenus Ipomopsis (Michaux) Milliken, Univ. Calif. Publ. Bot.
2: 24. 1904

The species included in this group were regarded by Michaux
as constituting a distinct genus based upon the eastern Gilia rubra.

Bentham included it under Gilia as a subgenus, a position that -

clearly expresses the relationships of its species. The biennial
character of the members of this subgenus is outstanding. The
chief problem in the group centers in the G. aggregata complex
wherein specific and subspecific segregation in the southwest is
very complicated. In the Pacific Coast states, however, only
typical G. aggregata occurs.

Subgenus Evapnuocera (Nuttall) Milliken, Univ. Calif. Publ. Bot.
2:24. 1904

The members of this subgenus have been treated in
detail under the heading “the Gilia congesta complex’”’ by Con-
stance and Rollins and will not be further elaborated here except
to point out that the concept of the group is here expanded to
include the annual species G. polycladon. The outstanding char-
acters of the group include a short tubular or salverform white to
pale blue corolla with short stamens in or just below the sinuses
of the corolla lobes, capitate or leafy-bracted, glomerate inflor-
escences and 1- to 2-seeded capsule locules. The species may be
annual or perennial, herbaceous or shrubby.

Subgenus Greenianthus subgen. nov.

Annua, foliis integeris vel irregulariter aut regulariter pinnati-
sectis aut furcatis; corollis tubiformibus infundibuliformibus,
jugulus angustissimatis tubis multo longiore; staminis inaequalis-
insertatis, longitudine inaequalis; loculis 1- raro 2-ovulatis.

Annuals, leaves entire to irregularly or regularly pinnately
cleft or forked. Corolla tubular, funnelform, throat very narrow,
much longer than tube. Stamens unequally inserted on throat,

unequal in length, locules 1-seeded rarely 2-seeded. Type. Gulia

gilioides.

This subgenus is characterized by its broad, cleft or entire
leaves with lanceolate teeth or lobes of very diverse size but never
dissected into linear filiform segments. The usually deep violet
to purple or sometimes white corolla is likewise distinctive with
its very long, almost tubular throat and very unequally inserted
stamens. The subgenus includes the Gila gilioides complex and the
desert species G. depressa. This latter species presents no prob-
lem so will not be further dealt with here.

Although there is great morphological, genetical, and ecologi-
cal diversity within Gilia gilioides, it stands as one of the most
distinctive units within the genus Gilia. Its unequal and un-
equally inserted stamens together with the usual condition of

1948 | MASON: GILIA 207

uniovulate locules serve to set it apart from the rest of the genus.
In fact, on the basis of these characters, it was once placed in
Microsteris, and because of its general leafiness which extends well
up into the inflorescence, in addition to the stamen and ovule
characters, Bentham at one time included it in Collomia. On the
other hand, its calyx and corolla and the nature of the capsular
dehiscence as well as the lobing and alternate insertion of the
leaves clearly indicate its close relationship with the other species
of Gilia. Erection of a separate genus for G. gilioides would only
serve to defeat the objectives of taxonomy by separating it from
its obvious relatives. Despite the wide geographic range of the
species and the variation that exists within it, a synonomy of only
fourteen names is recorded in our treatment.

Giza GILIoIDEs subsp. volcanica (Brand) comb. nov. G. divari-
cata var. volcanica Brand in Engler, Pflanzenreich 47°: 94. 1907.

This subspecies, with its pink corolla lobes, purple throat, and
portion of the stamens exserted from the corolla tube is here
regarded as having sufficient supplementary characters to warrant
nomenclatural status. The geographic ranges of this and other
color races usually do not overlap. Exceptions are a violet race
and a white race in the middle altitudes of the Sierra Nevada
which usually occur alone but sometimes are found intermixed.
When occurring together, they seem to retain their distinctness.

Leaf variation, although extreme, does not manifest itself
along lines that could be expressed in terms of taxonomic diverg-
ence. Variation occurs in both form and size of leaves. They
may be simple, lanceolate, and entire or they may be irregularly
cleft into 2 to several divisions or they may be toothed. Some
may be regularly pinnately cleft. A population rarely may ex-
hibit relative uniformity as to leaf character but it is not uncom-
mon for a single large plant to display the entire range of leaf type
variation found in the species as a whole.

Another point of variation pertains to stamen insertion and
exsertion. Throughout most of the populations the stamens are
usually all included although they may be unequally inserted and
equal or slightly unequal in length. In G. gilioides subsp. volcanica
one or two of the stamens are exserted and the remainder included.

GILIA GILIOIDEs subsp. glutinosa (Bentham) comb. nov. Collomia
glutinosa Bentham, Bot. Reg. 19: sub t. 1622. 1833.

Throughout southern and insular California as well as in
northern Baja California a population with all of the stamens
exserted has been variously treated in the literature. It has been
described as Collomia glutinosa Benth. and as Gilia Traskeae East-
wood. Its morphological distinctness and geographic unity war-
rant its inclusion in a subspecific status.

Geographically Gilia gilioides ranges from northern Baja Cali-
fornia to southern Oregon and eastward into Nevada. It occurs

208 MADRONO [Vou. 9

from sea level to near timberline throughout a great range of
habitats.

The remainder of the synonymy of the group is regarded as
reflecting chiefly the changes in nomenclatural status of members
of the group hence will not be further discussed here.

Subgenus Capitata Milliken, Univ. Calif. Publ. Bot. 2: 37. 1904.

This is a very natural subgenus characterized by leafy stems
with the leaves little reduced upward and with highly dissected
blades and by flowers with usually full campanulate throats and
short tubes. The calyces are often woolly. The group can be
divided into two main types on the basis of the fact that in some
plants the inflorescence is a compact head while in others it is
made up of open glomerules or of solitary peduncled flowers.
Under certain habitat conditions however, some species normally
producing heads develop instead open paniculate inflorescences.
The members of this subgenus lend themselves to manipulation
genetically since they are adaptable to garden culture and since
the seeds display a high percentage of germination. Cytogenetic
work on this problem is at present being carried on by students;
so we shall confine our remarks only to those points that demand
our immediate attention and await a fuller report on the problem.
It will suffice here to point out that five different specific names
have been applied within the group of plants that we include
under Gilia achilleaefolia. These are G. achilleaefolia Benth., G.
stricta Scheele, G. abrotanifolia Nuttall ex Greene, G. staminea
Greene, and G. chamissonis Greene. Of these, G. stricta is of

horticultural origin, probably derived directly from seed of G. -

achilleaefolia sent to Europe by Douglas. The remainder vary
geographically to such an extent that it is impossible to clearly
differentiate them. We therefore for the present accept the
following:

GILIA ACHILLEAEFOLIA subsp. CHAMISSONIS (Greene) Brand.

GILIA ACHILLEAEFOLIA subsp. staminea (Greene) comb. nov.
G. staminea Greene, Erythea 3: 105. 1895.

Gitta caPiTaTa Douglas.

GILIA MULTICAULIS Bentham, Bot. Reg. 19: sub t. 1622. 18338.

The taxonomy of this species presents many complications
resulting largely from its great diversity. Several variants have
been described within the complex which we believe are best
treated as subspecies since we are unable to, clearly differentiate
between them. Giulia multicaulis subsp. eu-multicaulis Brand is the
common species of the central coast ranges. It produces flowers
on short peduncles in few flowered glomerules, is exceedingly
variable as to pubescence and its ascending or erect stems are
quite leafy well up into the inflorescence.

1948 | MASON: GILIA 209

GILIA MULTICAULIS subsp. peduncularis (Eastwood) comb. nov.
G. peduncularis Eastwood ex Milliken, Univ. Calif. Publ. Bot. 2:
34. 1904.

Often growing with typical G. multicaulis but occurring inde-
pendently also is the form with the flowers on elongate slender
peduncles. It intergrades with the typical form in this character,
but differs in being much less leafy.

GILIA MULTICAULIs subsp. millifoliata (Fischer & Meyer) comb.
nov. G. millifoliata Fischer and Meyer, Ind. Sem. Hort. Petrop. 5:
35. 1838. ;

This is a stout, glandular, divaricately branched type with an
accrescent calyx which occurs along the coastal sand dunes from
central California to southern Oregon.

GILIA MULTICAULIs subsp. Nevinii (Gray) comb. nov. G. Nevinii
Gray, Syn. Fl. N. Am. ed. 2, 2 (suppl.): 411. 1886.

On San Clemente and Guadalupe Islands occur populations
striking because of their finely dissected leaves and corollas much
longer than in the type. They were first described by Gray as
Gilia multicaulis var. millifolia, and later raised to specific rank by
Gray under the name G. Nevinii. Since the name “‘millifolia’” is so
close in orthography and pronounciation to the preceding sub-
species and since we apply it to another rank we believe it ex-
pedient to accept Gray’s name in the role of a trinomial.

GILIA TRICOLOR Bentham.

GIL1a TRICOLOR subsp. diffusa (Congdon) comb. nov. G. diffusa
Congdon, Erythea 7: 186. 1900. G. tricolor var. longipedicellata
Greenm., Rhodora 6: 154. 1904. G. inconspicua subsp. sinuata
var. oreophila subvar. diffusa Brand, Pflanzenreich 4°°°: 105. 1907.

Occurring occasionally in the range of the species but extend-
ing farther south in the Sierra Nevada foothills and in the hills
bordering the southern San Joaquin Valley, this subspecies is rec-
ognized for its diffuse branching and its smaller flowers which are
borne on longer, slender pedicels.

Subgenus Evaeitira (Bentham) Milliken, Univ. Calif. Publ. Bot.
2: 23. 1904

The subgenus Fugilia is the major problem in the genus Gilia.
It occurs chiefly in the deserts and semi-arid basins and valleys of
the west and southwest with some races extending well up into
the intervening mountains, and reaches the Pacific Coast in sand
dune areas from Santa Cruz County, California, southward.

To one beginning a study of this group certain features stand
out. Most obvious is the strong tendency for parallel variation

among the entities of this complex. Many of the entities com-

prise small-flowered subspecies with short corolla tubes and large-
flowered subspecies with long corolla tubes. Examples of such

210 MADRONO [Vou. 9

parallel variation are: G. splendens and its long-corolla-tubed form,
G. splendens subsp. Grinnelli; G. latiflora and its long-corolla-tubed
form G. latiflora subsp. speciosa. Nevertheless intergradation
between these species and subspecies occurs.

The subgenus breaks clearly into two well-marked subdi-
visions between which we have seen no evidence of intergradation.
The most obvious differentiating character has been overlooked
in the past largely because it involves a character that in many
groups is often unstable, namely the character of the pubescence.
Gilia Abramsi, G. ochroleuca, G. tenuiflora, G. latiflora and G. sinuata
have in common a pubescence consisting of long, tangled hairs
so fine that an individual hair is not readily seen with the naked
eye. This pubescence is found mainly on the lowermost leaves
and stems and may be thick and woolly or tufted or very sparse.
In a few cases where relationships to one or another ‘of these five
species are clear through other characters, the plants may be
entirely glabrous. In contrast to this situation the other members
of the subgenus, G. splendens, G. caruifolia, G. stellata, G. scopulorum
and G. leptomeria, have a pubescence of coarse hairs of various
types but never long and tangled. The individual hairs may be
readily seen with the naked eye. A completely glabrous con-
dition is unknown to us in this section of the subgenus.

One of the most baffling problems in taxonomic treatments of
the subgenus Fugilia has been the variation in the nature of the
leaves. The numerous distinct leaf forms which occur in various
combinations with the characters of leaf size and degree of
pubescence suggests that there are numerous races within a
species. Considerable variation, moreover, can be seen in a
single population. It has been our observation that in any given
population the larger the individual the more complex the dis-
section of the leaf; and in two populations, related, but of dis-
tinctly different leaf form, the smallest individuals may appear
quite similar. The situation in Giulia latiflora will provide an
example. Specimens in one mass collection displayed leaf types
1, 4 and 5 (text fig. 1) in the order of their development from
simple to complex whereas another such collection displayed
types 1, 2 and 8 (text fig. 1). In the first case the most highly
developed leaves were bi- to tri-pinnate with narrow rachis and in
the second case the most highly developed ones were bipinnate
with broad rachis. In both cases plants bearing only leaf type 1
were indistinguishable. This suggests that the degree of dissection
of the leaves may be related to the rate of growth as influenced
by local ecologic differences or seasonal differences. In a poor
flowering year on the desert such potential leaf variation is
masked and to make certain of the type of plant being dealt with
in any season, one must look for the better-developed individuals.

In G. latiflora an attempt was made to correlate leaf form with
other characters and with geographic distribution, and this was

1948 | MASON: GILIA

211
1 2
3
4 E
|

6

Fic. 1. Leaf types in Gilia, subgenus Eugilia. 1, G. latiflora, 1 mile east of
Lancaster, Los Angeles County, California (Mason 6869); 2, G. latiflora,
Barstow, San Bernardino County, California (W. W. Jones, April 15, 1921);
3, G. latiflora, 5 miles west of Barstow, San Bernardino County, California
(Minthorn 80); 4, G. latifiora, Kramer Station, San Bernardino County, Cali-
fornia (Constance & Mason 2110); 5, G. latiflora subsp. speciosa, summit be-
tween Nine Mile Canyon and Kennedy Meadows, Tulare County, California
(Alexander & Kellogg 2962) ; 6, G. latiflora, 2.5 miles east of Coso Hot Springs,
Coso Range, Inyo County, California (Alexander & Kellogg 2771); 7, G. lati-
flora subsp. cana, Carroll Creek, southwest of Lone Pine, Inyo County, Cali-
fornia (Alexander & Kellogg 2814); 8, G. tenuiflora subsp. interior, Red Rock
Canyon, Kern County, California (Mason 9278) ; 9, G. ochroleuca, Carrizo Plain,
San Luis Obispo County, California (Hsau, April 13, 1935). All drawings x 1.

possible in the subspecies; but the situation of typical G. latiflora
in the Mohave Desert. presents a confused picture. In the area
between Victorville, Barstow, and Lancaster, California, collec-
tions display specimens similar in most of their features except
leaf form; yet here every one of the first six types depicted in

212 MADRONO [Vou. 9

text fig. 1 is represented, and these vary widely in dimension as
well as dissection so that in their vegetative characters the plants
appear very different. To one who is familiar with the G. latiflora
complex in the herbarium and who has had only a limited experi-
ence with it in the field the impression is carried that there are more
species in the herbarium than there are in the field. The liter-
ature reflects this situation. Although we have corrected some of
the fallacies inherited from previous treatments, only detailed
population studies will furnish us with a true picture of variability
in the G. latiflora complex.

In the past the small flowered types in the subgenus Eugilia
have generally been regarded as centering around what has been
termed Giulia inconspicua, an epithet which we reject because of the
difficulty of establishing just what is the type upon which the
name rests. The case for this decision has been presented by
the senior author (1945) and will not be further elaborated here.
We construe most of these small flowered plants to group them-
selves around three well-marked species, namely G. ochroleuca
Jones, G. sinuata Douglas, and G. leptomeria Gray, centering geo-
graphically in the Great Basin and desert areas south to northern
Baja California. The remainder of the group in the past has
been variously treated as many distinct species or has been aggre-
gated under G. tenuiflora or G. latiflora as synonyms or as varieties.

The outstanding characters of the subgenus Eugilia are the
annual habit, the usual varied leaf dissection, the basal rosette
(except in depauperate individuals), the much-reduced cauline
leaves, the stems appearing almost naked, the subglomerate to
open-paniculate, stipitate-glandular inflorescences, and the stamen
insertion in the sinuses of the corolla lobes (one exception). Fol-
lowing is a discussion of the individual species.

GILIA SsPLENDENS Douglas ex Paxton, Mag. Bot. 3: 260. 18387.

This is the species that has been regarded in the literature as
G. tenuiflora var. altissima Parish. The name G. splendens as
applying to this seems to have been overlooked in spite of the
excellent illustration in Lindley’s Botanical Register (1836, t.
1888) under the caption “G. tenuiflora Benth.” This illustration
was made from living plants grown in England from seed collected
by David Douglas and labelled by him “Gilia splendens” and is a
faithful reproduction of the species. The range of G. splendens
barely overlaps into the range of G. tenuiflora but we have seen no
evidence of intergradation between them. Giulia splendens is
readily distinguished from that species by its pubescence of
coarse translucent hairs on the basal leaves, the bi- and tri-pinnate
basal leaves with finely toothed lobes and the rose or bright pink
color of the corolla as opposed to the purple and yellow of G.
tenuiflora. It is typically a montane species occurring from the
mountains of southern Monterey County to those of Santa Bar-

1948 | MASON: GILIA 213

bara and Ventura counties and in the San Gabriel, San Bernar-
dino, and San Jacinto mountains, California.

GILIA SPLENDENS subsp. Grinnellii (Brand) comb. nov. Gilia
Grinnellii Brand in Engler, Pflanzenreich 4°°°: 101. 1907.

This is a long-tubed form of the species that seems to be re-
stricted to the San Gabriel and San Bernardino mountains, Cali-
fornia.

GILIA SPLENDENS subsp. australis subsp. nov. A G. splendens
differt capsulis magnioris (5-7 mm. longis) et corollis multis
brevioris quo limbis est longioris proportionalis tubis et jugulis.

Differs from G. splendens in the larger capsules (5—7 mm. long)
and in the much shorter corolla in which the limb is proportion-
ately longer.

San Bernardino and Riverside counties, California, to Baja
California, Mexico.

Type. Temecula Valley, Riverside County, California, Ma-
son 38195 (Herb. Univ. Calif. 748763).

GiLiA caRuIFOLIA Abrams, Bull. Torrey Bot. Club 32: 540. 1905.

Gilia caruifolia resembles G. splendens closely in vegetative
aspect, but differs in the smaller blue, violet, pink or white corolla
with a short throat and long stamens inserted midway on the
throat. All other species of the subgenus Eugilia have the sta-
mens inserted in the sinuses of the corolla lobes. Their geo-
graphic distributions are completely distinct, G. caruifolia occur-
ring farther to the south. The gap between their ranges is filled
by G. splendens subsp. australis which occurs also farther south in
the region of G. caruifolia. Throughout its range this subspecies
maintains the corolla tube and throat proportions and stamen char-
acter of G. splendens but the smaller flowers give a suggestion of its
intermediate nature between G. splendens and G. caruifolia.

Ginia sTeELLATA Heller, Muhlenbergia 2: 117. 1906. G. tenui-
flora var. Newloniana Jepson, FI. Calif. 3: 179. 1943.

The form of the lower leaves as well as the flower size and
color clearly mark this species as related to the interior form of
G. splendens. It differs in its peculiar pubescence of several-
celled, translucent, geniculate hairs and much smaller corollas.
It is primarily a desert species rather than a montane plant.

GILIA scopuLorum Jones, Bull. Torrey Bot. Club 8: 70. 1881.
Gilia scopulorum and G. stellata are unusual in the subgenus Eugilia
in that the calyx tends to be accrescent rather than to be ruptured
by the capsule. Also they are unique in that even the highest
cauline leaves tend to be toothed rather than reduced to an entire
bract. They also have spherical rather than the cylindric cap-
sules, like Gilia splendens and G. caruifolia. 'They may be readily
distinguished from one another by the broader leaf segments and

214 MADRONO [Vou. 9

long corolla tube of G. scopulorum and by the geniculate hairs of
G. stellata. Giulia scopulorum occurs chiefly in washes in the
canyons of desert mountains.

GILIA LEPTOMERIA Gray, Proc. Am. Acad. 8: 278. 1870.

This species is outstanding among the small flowered mem-
bers of the subgenus Eugilia because of its broad leaf blades and
its pubescence. The leaf blades when lobed or dissected have the
lobes opposite or sub-opposite, a condition not typical of the other
members of the subgenus. The cauline leaves are simple and
entire. In pubescence, G. leptomeria possesses glandular hairs,
not only in the inflorescence as in the other species but on the
basal leaves as well. There have been frequent references to
trident-lobed flowers in this group of Gilia. All such specimens
known to the writers are referable to G. leptomeria. They are G.
leptomeria var. tridentata Jones, G. inconspicua dentiflora Davidson,
G. leptomeria var. myriacantha Jones, G. triodon Eastwood, and
Aliciella triodon (Eastwood) Brand. This latter impressed Brand
sufficiently to cause him to segregate it as a distinct genus. It has
been the experience of the senior author in the field that this form
of the petals occurs on soils high in gypsum. The difference
though no doubt genetic scarcely warrants subspecific status.

In addition to typical Gilia leptomeria the following subspecies
seem to warrant recognition:

GILIA LEPTOMERIA subsp. micromeria (Gray) comb. nov. G.

micromeria Gray, Proc. Am. Acad. Sci. 8: 271. 1870.

The pedicels are more slender than in the species and are often
reflexed, the corolla is often minute and the petals sometimes are
3-toothed but usually entire. The opposite leaf lobes and the
entire, upper cauline leaves clearly place this with G. leptomeria.
It ranges from eastern Oregon to the Rocky Mountains, the type
having come from the hills above Bear River near Evanston, Utah.

GILIA LEPTOMERIA subsp. rubella (Brand) comb. nov. G.
arenaria var. rubella Brand in Engler, Pflanzenreich 47°°: 103. 1907.

The basal leaves of this subspecies are more deeply cut than
in the species and are often bipinnate. The literature displays
some confusion as to the red pigment in the stems which resulted
in the name applied by Brand. Rydberg, under G. Hutchinsifolia
(Bull. Torrey Bot. Club 40: 472. 1913) maintains Brand confused
red sand with plant pigment. We noted a red coloration on the
base of the stems in all collections seen, including the specimen
upon which G. leptomeria subsp. rubella rests (Jones 1651). Here
again the entire cauline leaves and the opposite lobes of the basal

leaves clearly relate this to G. leptomeria. It is known from Red

Rock Canyon in Kern County, California, east to southern Nevada,
Utah and northern Arizona.

GILIA OCHROLEUCA Jones, Contrib. West. Bot. 8: 35. 1898.
There are three outstanding features whereby this species may

1948 | MASON: GILIA 215

be readily distinguished from other small flowered species namely,
(1) the linear segments or dissections of the basal leaves, (2) the
linear, finger-like lobed cauline leaves, and (3) the yellow or
cream colored flowers whose lobes sometimes are tipped with
violet. Although we designate only two subspecies for the Pacific
Coast States we are aware of some interesting developments in
this species in southwestern Nevada and Arizona which deserve
further study.

GILIA OCHROLEUCA subsp. typica stat. nov.

Leaf rachis and lobes almost filiform, not exceeding 1 mm. in
width; inflorescence full, divaricately much branched, branches
filiform. This subspecies is based upon the type of the species
and is the least widespread of the two subspecies, being restricted
to the Mohave and Colorado deserts and the hills of Inyo County,
California, and the southwest border of Nevada.

GILIA OCHROLEUCA subsp. transmontana subsp. nov. Lobis
foliorum 1-2 mm. latis, primibus ramis inflorescentium virgatis,
inflorescentibus angustatis.

Leaf lobes 1-2 mm. wide, main branches of the inflorescence
virgate, the inflorescence narrow.

Eastern Washington and Oregon southeast of the Sierra
Nevada to the mountains of southern California and northern
Baja California, Mexico; east to Wyoming, Utah and New
Mexico.

Type. Beaver Dam River, Arizona Strip, Arizona, Maguire
et al. 4923 (Herb. Univ. Calif. 553752).

Gitia stinuaTA Douglas ex Bentham, in DC. Prodr. 9: 313. 1845.

Gilia sinuata is an, exceedingly variable species particularly in
leaf dissection and degree of pubescence. It differs from G.
ochroleuca chiefly in the bract-like cauline leaves, the short-toothed
lobes of the basal leaves, and the proportionately longer corolla
tube. Its stems are usually stout.

Although G. ochroleuca and G. leptomeria differ widely from
each other in leaf form and pubescence and offer no difficulties in
identification, there are two lines of evidence making the possibil-
ity worth considering that chance interbreeding of these two spe-
cies has given rise to at least the ancestors of G. sinuata. (1) A
study of the geographical distributions of these hypothetical par-
ents shows that they occur sympatrically in the central and eastern
Great Basin area, but G. ochroleuca extends a little farther west
than does G. leptomeria. On the western margins of the range of
G. leptomeria can be found also G. sinuata but to our knowledge it
does not occur east of here. In eastern Washington and Oregon,
where there are no other species of the section to confuse the issue,
it may be significant that G. sinuata, G. ochroleuca, and G. leptomeria,
have been collected from the same localities and in some cases, at
least, G. ochroleuca and G. sinuata have been mixed in the same col-

216 MADRONO [Vor. 9

lections. (2) In respect to its leaf characters, G. sinuata is inter-
mediate between G. leptomeria and G. ochroleuca. The characters
in question can be most easily compared if presented in tabular
form.

Taste 1. Comparison oF LEAF CHARACTERS

Characters G. leptomeria G. sinuata G. ochroleuca
Basal broadly strap- pinnate or bipin- pinnate or bipin-
leaves - shaped, sinuately nate with rachis nate with narrow

toothed or shallow broad and shal- rachis and slender,

lobed. lowly to deeply cut | linear lobes, which
into lobes which are longer than
are not linear. 2 times the width

of the rachis.

simple pinnate with

Cauline entire or shallow- similar to basal narrow rachis, the

leaves toothed; abruptly leaves but gradu- lobes linear and
smaller than basal ally or abruptly longer than 2 times
leaves, the stems shorter; the stems width of rachis;
appearing almost somewhat leafy, or | the leaves shorter
naked above the appearing almost than basal ones,
prominent basal naked. but not abruptly
rosette. so, stems appear-

ing somewhat leafy.

In southwestern Nevada and southern California G. sinuata
has reached a more complex state of development than in the
north, due in part, at least, to the comparative abundance of related
species as a source of new characters, and to the more varied
geography and ecology of the region, providing a wide choice of
habitats for new forms. In this region of overlap, G. sinuata
grows side by side with its close relative, G. latiflora, and the two
forms seem to hybridize freely.

Gilia Abramsii (Brand) comb. nov. G. arenaria var. Abramsii
Brand in Briquet, Ann. Conserv. et Jard. Bot. 15-16: 330. 1913.

Gilia Abramsii appears to differ from G. ochroleuca principally
in the nature of the corollas which are larger with abruptly ex-
panding, conspicuous throat. Although it is known to have
occasional intermediates with G. ochroleuca, its distribution is
distinct, being at higher elevations to the west and south of G.
ochroleuca.

Gitia Asramsi subsp. integrifolia subsp. nov. Foliis inferiori-
bus plerumque simplicis et integeris, linearis, per occasionem ali-
quot cum 1-2-lineari-lobis. |

Basal leaves mostly simple and entire, linear, occasionally a
few with one or two linear lobes.

Type. Temecula Canyon one mile south of Temecula, River-
side County, California, Mason 3112 (Herb. Univ. Calif. 748762).

1948 | MASON: GILIA 217

GILIA TENUIFLORA Bentham, Bot. Reg. 19: sub t. 1622. 1833.

We include in synonymy with G. tenuiflora what has been
described as G. arenaria of Douglas collected presumably at
Monterey, California. We have seen specimens from the Del
Monte sand dunes and from sandy hills in the Santa Cruz Moun-
tains. They differ only in having the lobes of basal leaves reduced.
Recent mass collections have shown both leaf types to occur in
the same populations, the reduced-lobe type occurring particularly
in depauperate specimens. The type of Gulia arenaria is evi-
dently just such a specimen; thus we consider the two names
synonymous.

GILIA TENUIFLORA subsp. interior subsp. nov. Caulis erectis,
e basi ramosissimis divaricatis, foliis: inferiore vix longiore; foliis
levi-vel moderate-lanatis; corolla 6-14 mm. longis, calycis 2—4-plo
longiore, tubis 3-5 mm. longis purpureo, jugulo flaveo 5-purpure-
maculoso infra lobus palide-violescens.

Stem erect, much branched and spreading from the base,
barely exceeding the basal rosette; leaves lightly to moderately
woolly pubescent; corolla 6-14 mm. long, 2—4 times the calyx,
tube 3-5 mm. long, purple, throat yellow with 5 purple spots sub-
tending the light violet lobes.

Inner coast ranges from Mount Hamilton to Santa Barbara
County, southern San Joaquin Valley to the mountains of SEU
County and the western Mohave Desert, California.

Type. Walker Pass, Kern County, California, Mason 5340
(Herb. Univ. Calif. 748761).

Gilia tenuiflora as here interpreted is a variable entity both in
corolla size and proportions and in leaf characters. In an inland
direction corollas tend to become smaller, and towards the south
the outstanding tendency is toward a proportionate shortening of
the corolla-tube. Thus, in Monterey County, California, in the
northern part of its range, where the type of the species origi-
nated, the corolla tube may attain a length of three times greater
than the throat; but in the southern part of its range, in the
Cholame Valley, Kern County, and in the northwestern reaches
of Antelope Valley, Los Angeles County, where G. tenuiflora inter-
grades with G. latiflora, the corolla proportions gradually approach
those of G. latiflora, namely the tube is less than two times the
throat, and the throat is more broadly expanding.

The leaf form is typically similar to that of G. ochroleuca,
being pinnately or bipinnately lobed with slender linear lobes and
a linear rachis, but frequently the lobes of the basal leaves are
reduced to teeth as previously discussed. The most constant
leaf feature is the fingerlike, linear lobes of the cauline leaves.

The inland transition of G. tenuiflora toward smaller corolla
size and also more diminutive habit reaches its ultimate in G.
tenuiflora subsp. interior. Were it not for the purple and yellow
coloration of the corolla, as in typical G. tenuiflora one would

@¢

218 MADRONO [Vou. 9

confuse the subspecies with G. ochroleuca; since its flowers are
small enough to.be within the upper limits of the size range for
G. ochroleuca flowers, and like the typical G. tenuiflora, it possesses
the type of cauline leaf so characteristic of G. ochroleuca.

As one observes the aspect of this subspecies in its range from
north to south he finds, as in typical G. tenuiflora, an intergradation
with desert forms. In this case, however, not only G. latiflora but
also G. ochroleuca influences the complex. It is significant that
in mapping this group in the Mohave Desert we found G. tenui-
flora subsp. interior to occur only in localities where G. ochroleuca
and G. latiflora also occurred. Furthermore, several collections
are obviously heterozygous, as evidenced by the wide range in
corolla length and leaf form. Detailed population studies are
much to be desired for the whole desert complex.

GILIA LATIFLORA Gray, Syn. Fl. 2(1): 147. 1878.

We have discussed some of the outstanding problems of leaf
variation as they pertain to G. latiflora. Certainly without more
detailed genetic studies and without more detailed field work,
and taking into consideration the erratic seasonal conditions of the
area, one can only express the range of variability of leaves
and flowers and treat this highly polymorphic group as a single
entity having in common corollas with short tubes and ample,
broadly-expanding throats. There are, however, some outstand-
ing variations that seem to be correlated with one another and
seem to have distinctive patterns of geographic distribution.

GILIA LATIFLORA subsp. speciosa (Jepson) comb. nov. G.
tenuiflora var. speciosa Jepson, FI. Calif. 3: 181. 1943.

This subspecies varies from the typical form in having an
elongated corolla tube which may be as much as 4 em. long, but
which varies from 2—8.5 times the length of the throat. Its leaves
may be of types 1, 4,5, or 7 (text fig. 1). It occurs in the northern
Mohave Desert where it integrades with typical G. latiflora. It
likewise intergrades with G. latiflora subsp. Purpusii.

GILIA LATIFLoRA subsp. Purpusii (Milliken) comb. nov. G.
tenuiflora var. Purpusi Milliken, Univ. Calif. Publ. Bot. 2: 29. 1904.

In many respects this subspecies gives the impression of being
a small form of G. latiflora subsp. speciosa. The corolla tube, how-
ever, is more slender and the lobes narrower. Leaf types 1 and
4 (text fig. 1) are, characteristic of it although the lobes tend to
be somewhat shorter and more crowded than in type 4. It occurs
in the southern Sierra Nevada in Tulare County, California.

GILIA LATIFLORA subsp. cana (Jones) comb. nov. G. latiflora
var. cana Jones, Contr. West. Bot. 8: 35. 1898.

The corolla tube of this subspecies varies from 2—3 times the
throat. Its leaves are covered with a dense layer of white wool.
This white wool and the broader leaf lobes (type 7, text fig. 1)
distinguish it from G. latiflora subsp. Purpusii from which it is

1948 | MASON: GILIA 219

separated geographically by the crest of the southern Sierra
Nevada. It intergrades with G. latiflora subsp. triceps to the east
and with typical G. latiflora to the south. It occurs on the east
slope of the Sierra Nevada in Mono and Inyo counties, California.

GILIA LATIFLORA subsp. triceps (Brand) comb. nov. G. tenuiflora
var. triceps Brand in Engler, Pflanzenreich 4°°°: 102. 1907.

This subspecies is outstanding for its full, many-flowered
inflorescences and filiform corolla tubes. The leaves may be of
types 5, 6 or 7 (text fig. 1). It occurs in the valleys and moun-
tains east of the southern Sierra Nevada to southern Nevada and
south to the San Bernardino Mountains, California.

GILIA LATIFLORA subsp. leptantha (Parish) comb. nov. G. lep-
tantha Parish, Zoe 5: 74. 1900.

This subspecies resembles G. latiflora subsp. Purpusiw in many
characters but differs from it chiefly in the shorter corolla tube
and in its long-exserted stamens. The leaves are of types 4 and 5
(text fig. 1). It occurs in the Mount Pinos region of Ventura
County and in the San Bernardino Mountains, California.

GILIA LATIFLORA subsp. exilis (Gray) comb. nov. G. latiflora
var. exilis Gray Syn. Fl. ed. 2, 2(suppl.): 411. 1886.

The corolla proportions of this subspecies are similar to
typical G. latiflora, the tube being shorter than the long full throat.
The flowers, however, are smaller and the whole plant is diminu-
tive with numerous slender branches from the base. It inter-
grades with the type and occurs in the San Gabriel and San
Bernardino Mountains, California.

Subgenus Campanulastrum (Brand) comb. nov.

Gilia subgenus Greeneophila Brand, section Campanulastrum
Brand in Engler, Pflanzenreich 4°°°: 144. 1907.

This subgenus, based upon G. campanulata, but construed by Brand
as a section involving several species belonging to the genus Linanthus,
is here restricted to include Gilia campanulata Gray and G. inyoensis
Johnston. It is closely related to the subgenus Eugilia from which it
differs in the short broad corollas, the low spreading form of the plant,
and the broad short leaves.

Subgenus Kelloggia subgen. nov.

Folis plerumque linearibus vel lineari-filiformibus raro aliquot
pinnatisectis in paucis filiformibus lobis; floribus solitariis axillis,
corollis tubiformibus vel angustati-infundibuliformibus vel turbinatibus
hic calycis vix longiore, alio modo calycis multo longiore.

Leaves, or most of them, linear or linear-filiform, rarely a few
pinnately dissected into few filiform lobes, flowers solitary in leaf axils,
corolla tubular to narrow funnelform, or turbinate, then barely exceed-
ing the calyx, otherwise much exceeding the calyx. Based upon G.
capillaris Kellogg.

This subgenus includes G. leptalea, G. capillaris, G. minutiflora and

220 MADRONO [Vou. 9

G. tenerrima. Except for a subspecies of G. leptalea the leaves are all
linear-filiform and entire. Of these species only G. leptalea needs
special consideration here.

GILIA LEPTALEA subsp. pinnatisecta subsp. nov. Speciei simili
autem foliis pinnati- vel laciniati-lobatis aut dissectis, planta totus
saepe glandulosus-viscidus.

Similar to the species but the leaves pinnately to laciniately lobed
or dissected, and the whole plant often glandular-viscid. North Coast
Ranges, Lake County to Humboldt County, California; San Marcos,
Brandegee (Santa Barbara County?).

Type. Open ground about Whispering Pines resort, Lake County,
California, Baker 2299a (Herb. Univ. Calif. 353868).

GILIA LEPTALEA subsp. bicolor subsp. nov. Speciei simili autem
jugulus flavus tubo subaequantibus.

Similar to the species, but the throat subequal the tube and yellow.
Canadian zone; central Sierra Nevada, California.

Type. Dardanelle, Tuolumne County, California, Alexander &
Kellogg 3736 (Herb. Univ. Calif. 702227).

Subgenus Tintinabulum (Rydberg) comb. nov.

Lintinabulum Rydberg, Fl. Rocky Mountains, pp. 698 and 1065.
LOZ

In view of the close relationship between the single species of this
subgenus with the entire linear-leaved members of the subgenus Kel-
loggia it seems scarcely necessary to recognize Tintinabulum of Rydberg
as a genus. It would stand only on the open campanulate yellow
corollas of Gilia filiformis. There are occasional colonies with cream
colored flowers.

Department of Botany
University of California, Berkeley
LireraTurE CITED

International Rules of Botanical Nomenclature. 1935.

Lirzic, J. 1843. Chemistry in its relation to agriculture and physiology.
Third edition.

Linptey, J. 1836. Botanical Register. London.

Mason, H. L. 1945. The genus Eriastrum and the influence of Bentham and
Gray upon the problem of generic confusion in Polemoniaceae.
Madrono 8: 65-91. 19465.

POTAMOGETON LATIFOLIUS IN TEXAS
W. C. MvuenscHer

In June, 1945, my attention was attracted by an abundant
growth of an unfamiliar Potamogeton in the outlets of springs
about Fort Stockton, in western Texas. When Dr. William T.
Winne and I began to collect some of these specimens for press-
ing, it became apparent that we had a robust species belonging to

1948 | MUENSCHER: POTAMOGETON 221

the subgenus Coleogeton. <A series of specimens brought back
to the herbarium at Cornell University was determined as
Potamogeton latifolius (Robbins) Morong, only after considerable
study.

Potamogeton pectinatus var. latifolius Robbins (in S. Watson,
Botany Kings Expl. p. 388. 1871) was described from material
collected in running brackish waters of Humboldt River below
Humboldt Lake, Nevada (W. W. Bailey 1142).

Potamogeton latifolius (Robbins) Morong (Mem. Torrey Bot.
Club 32: p. 52, 1893) is based on the description of Robbins
as amplified by Morong from fruiting material collected by Mrs.
R. M. Austin in Goose Lake in northeastern California in 1884.

The only records of Potamogeton latifolius that could be found
are from northwestern Nevada, northeastern California and
southern Oregon.

Through the kindness of its officials, I have had an opportun-
ity to examine the following specimens of Potamogeton latifolius
in the Gray Herbarium: Nevapa. Humboldt Lake, 1867, Bailey
1142; Hot Springs, Nevada desert, 1874, Lemmon 1164. Catt-
FORNIA. Mono County, 1898, Congdon 9915; Kings River, 1876,
Lemmon 1521. Orecon. Goose Lake?, 1897, Everman.

The Fort Stockton material, in all its essential features, com- |
pares with the above specimens. Thus, the Texas station repre-
sents a considerable extension of the recorded range of Potamoge-
ton latifolius.

Good material of this species appears to be rare in herbaria.
The only illustration that could be located, Morong (18938), is
hardly adequate to permit it to be recognized as Potamogeton
latifolius. The following description and illustration were both
based upon the Texas collection (in running water, Comanche
Springs, Fort Stockton, Pecos County, Texas, June 380, 1945,
Muenscher and Winne 16516). Specimens have been deposited in
the herbarium of Cornell University.

PoTAMOGETON LATIFOLIUS (Robbins) Morong. Rootstock slen-
der, long-creeping, about 2 mm. in diameter, rooting freely at the
nodes. Stems simple below, repeatedly branched above, 380-
100 cm. long, the ultimate branches mostly less than 10 cm. long
and bearing 1 or more spikes. Leaves linear, somewhat crisp
and fleshy, the blades 2-7 mm. wide, mostly from 3 to 7 cm. long,
those of the basal leaves somewhat longer and those on the
ultimate and spike-bearing branches often less than 3 em. long,
green to bronze, rather opaque; leaf margin entire; apex obtuse,
rounded or mucronate to acute on the upper leaves; blade with
3 nearly-equal main nerves and often with 1 or 2 smaller nerves
near each margin, reticulated by prominent cross-nerves between
the main parallel nerves. Stipules 8 to 12 mm. long, scarious,
adnate to the leaf-base to form a sheath; ligule about 1 to 4 mm.
long, hyaline along the margin, entire or becoming erose. Ped-

222 MADRONO [Vou. 9

LZ

ZA

<A

as
es +a we
<e

=<

SS
SS

SS

\
I

=>
=

i

|
|

——sa

\

)
i

ya
~
Sa
—s Z
S Se = oH
LSS —_ =
~ =
=
NN

1948 | KECK: WILLIS LINN JEPSON 223

uncles ‘2-5 cm. long, 1-1.5 mm. in diameter, tapering above,
green to bronze-red. Spike 2 to 3 (4) em. long, dense, ultimately
of 4 to 6 distant nodes with 2 or more flowers at each; basal
internodes 5-12 mm. long, the upper shorter. Sepaloid con-
nectives 1-1.5 mm. high, 2 mm. wide. Pollen grains spherical
to elliptical in outline. Fruits olive-green to fulvous, 4 mm.
long, 3 to 4 mm. broad, somewhat compressed but the sides con-
vex; stigma somewhat capitate but oblique, on a short stout
style in fruit; beak about 1 mm. long, slightly curved. No
winter buds observed although some of the ultimate branches
are abbreviated and fleshy and may function as such.

Department of Botany,
Cornell University.

THE PLACE OF WILLIS LINN JEPSON IN CALIFORNIA
BOTANY

Davw D. KecK«

For three-score years Willis Linn Jepson, 1867-1946, was actively
connected with the Department of Botany of the University of Cali-
fornia as student, professor, and professor emeritus. Throughout this
long period he was thoroughly devoted to the study of the flora of his
native state and to furthering its interpretation and appreciation. ‘T'o
this end he founded the California Botanical Society in 1913, which he
served as president, with the exception of three years, until 1929. In
1916 he launched the organ of the Society, Maprono, which he edited
continuously through 1934. Much earlier, with the aid of E. L. Greene,
he had founded and edited the journal Erythea.

The botanical writings of Jepson are both extensive and profound,
and they have exerted a lasting influence upon our knowledge of the
botany of California. The present account attempts to evaluate
Jepson’s lifework, as made known by these contributions, on the his-
torical background.’ A bibliography of authors who have named
flowering plants occurring in the wild in California now includes well
over 900 names! Where does Jepson stand among these?

Three stages can be recognized in the study of the California flora:
(1) its study by Europeans; (2) by Americans along the eastern sea-
board; and (3) by Californians. The first stage dates back to the late
eighteenth century, when European explorers began to collect the
objects of natural history that they found on these shores. By the
early nineteenth century people in England had become greatly in-
terested in horticulture, and expeditions were sent out to the four

1 For sketches of Jepson, the man, giving more details of his active life,
refer to (1 )Herbert L. Mason in Madrono 9: 61-64, 1947; (2) Lincoln Con-
stance in Science 105: 614, 1947; (3 & 4) Emanuel Fritz in California Forester
14: 6-8, 1947, and in Jour. Calif. Hort. Soc. 9: 23-26, 1948; (5) Marion R.
Parsons in Sierra Club Bull. 32: 104-107, 1947; and (6) Joseph A. Ewan in
Jour. Wash. Acad. Sci. 37: 414-416, 1947.

224 MADRONO [Vou. 9

corners of the earth in search of plants to enrich British gardens. Bent
on this purpose and showing amazing activity, David Douglas alone,
Scotch collector for the Horticultural Society of London, now the
Royal Horticultural Society, in his single season in California pro-
vided the material from which some 300 species were to be described.

The second stage, led by Thomas Nuttall, began around 1830, when
the botanical exploration of the West by American botanists was under
way. Soon John Torrey and Asa Gray were vying with the British
botanists, W. J. Hooker and George Bentham, in the volume of West
American species that they were bringing to light. During much of
the latter half of the century, collectors by the dozen were sending
West American plants to Dr. Gray, the highest authority of the period
on the flora of this region. As a culmination of this stage there ap-
peared the monumental two volume Botany of California by W. H.
Brewer and Sereno Watson, with a large section contributed by Gray
(1876, 1880). This invaluable work, based principally on the large
accumulations of western material that had gravitated to Harvard and
also the collections of Brewer and others made in connection with the
Geological Survey of California, has been the starting point for all
subsequent floras that have been produced in the state.

Gradually, as the third stage in the elucidation of the California
flora, the West developed its own botanical authors. The first to pub-
lish a number of native species new to science was Albert Kellogg, a San
Francisco physician. His contributions appeared particularly from the
1850’s to the 1870’s in the Proceedings of the California Academy of
Sciences, the institution of which he was a founder.

By 1880 the botanical activities of the Reverend Dr. E. L. Greene
had begun. His contributions through the years were very large, but
were so rarely of a monographic nature that the proportion of his
specific proposals that were to be widely accepted is not to be compared
with that of Gray, Watson, or Jepson. Yet, as a pioneer worker in a
region outstanding for the richness of its flora, and having a keen eye
for small variations, which he named, it was inevitable that Greene's
name should be associated with a goodly percentage of our California
species. He contributed two local floras of value: Manual of the Botany
of the Region of San Francisco Bay, 1894, and Flora Franciscana, 1891—
SHEE

The first of the major botanical works produced by Jepson, who
was a student of Greene, was A Flora of Western Middle California,
1901, second edition, 1911. He usually had several manuscripts in
preparation simultaneously. His work on one yielded information or
suggested ideas applicable to another. About the time he finished work
on this book he projected The Silva of California (1910), The Trees of
California (1909, second edition, 1923), and A Flora of California,
which from the first he looked upon as his greatest life’s work. Jepson
himself says the Flora was planned in 1894. The first two parts ap-

2 Flora of California 2: 7, 1936.

1948 | KECK: WILLIS LINN JEPSON 225

peared in 1909, the twelfth part in 1943. Work was actively pro-
gressing on the thirteenth part until illness interrupted, and the
author’s death a year later in 1946 found the Flora about three-fourths
completed and published. Volume I starts on page 33, the first 32
pages being reserved for an introduction that was to have appeared
upon the completion of the whole work. Its seven parts, otherwise
complete, are not indexed. Volume II is complete, but the index is to
families and genera only. The two completed parts of volume III are
not indexed.

Jepson was thoroughly aware of these deficiencies and was almost
reticent in advertising the parts of the Flora that were available. As
in the case of A Manual of the Flowering Plants of California, which
was also issued in parts, from 1923 to 1925, he preferred to withhold
advertising of the parts as they appeared individually, because for
the general user the completed work would prove more useful, and it
was desirable not to deplete the stock of any one part before the en-
tire volume could be bound. Because of these shortcomings, Jepson’s
Flora is definitely less convenient than his Manual, particularly for
use in the field, and therefore has not received the general recognition
and use that it deserves, but the quality of workmanship in the later
parts is unsurpassed in any similar American work.

Jepson’s major projects built progressively upon one another.
As the Flora of Western Middle California built upon Brewer and
Watson’s Botany of California, Gray’s Synoptical Flora of North
America, and the works of Greene, appreciably advancing our knowl-
edge of the plants of its area, so the Manual drew upon this work and
the portions of A Flora of California then completed to become one of
the finest botanical handbooks extant. Similarly, succeeding parts of
the Flora mark a distinct advance over the Manual. As would be an-
ticipated in a work that was to appear in parts over more than a third
of a century, A Flora of California is uneven in treatment. The pro-
gressive improvement noted in volumes II and III as compared with
volume I reflect not only the scientific growth of the author, but also
the growth of botany in the West.

At a very early time Jepson had to decide whether to use the
system of measurement based on the foot, inch, and line, used by the
English botanists and the Harvard school, or to adopt the metric sys-
tem coming into vogue on the Continent. He chose to follow the former,
and, having committed his Flora to this system, was forced to continue,
even though it was soon evident that the English system had been be-
coming obsolete from the turn of the century. By the time his Manual
appeared in 1925, Jepson was originating the only major flora in
America that did not follow the metric system.

This relatively minor fault, if fault it be, is nevertheless one of the
few mechanical details to which exception can be taken in the works of
one who put mechanical perfection very high indeed among the obliga-
tions of an author. Jepson’s works are freer from typographical error
than those of almost any other American botanist due to the fact that he

226 MADRONO [Vou. 9

meticulously read proof himself and left no mechanical detail to the
discretion of his printer.

Jepson strove for a uniform treatment and avoided introducing
chromosome numbers, genetical data, and other experimental results
that, by the time the later parts of the Flora were appearing, were be-
coming a determinative influence in taxonomy. It is well that he stayed
on wholly familiar ground, continuing to rely on those tools which he
handled as an adept—accurate descriptive morphology and analysis
and a keen perception for the place of the plant in its natural environ-
ment. ‘The story that he wrote he was perhaps better prepared to write
than any other person.

The inclusion in later parts of the Flora of excerpts from Jepson’s
very extensive field notebooks on the ecology, physiology, and mor-
phology of many species is of great value. The reader finds much in-
teresting and original information under such a variety of titles as:
geographical note, field note, leaf variation, taxonomic note, note on rela-
tionship, biological note, etc. Jepson was not only an astute observer ;
he was a facile writer whose written word was forceful, clear, and often
of great beauty.

His appreciation of the historical precedent and the classical style
stemmed not only from his teacher, E. L. Greene, who valued these
especially highly, but also from his study of the works of the greatest
systematists and from a reading of the classics. He urged upon his
students the desirability of becoming familiar with great works on
travel and biography as a proper foundation for work in taxonomy.

Students of west coast botany are fortunate that the principal task
of organizing their flora has been done by one with the sound botanical
judgment of Jepson. This he did not learn from Greene, nor from
other contemporaries in California, but from a devoted study of the
artistry of the great British systematists of the nineteenth century.
That he profited much from this study is evident from the quality of
his work, which has made an impress on the writings of others.

Jepson, with an intuitive grasp of what are good species and genera,
organized the scattered knowledge of the complex California flora in a
remarkable way. He introduced the Englerian system of phylogeny to
California, but here and there made his own appraisals of the proper
positions for the families. His species concept was grounded on so
sound a morphological basis that, on the whole, it has been widely ac-
cepted, and the present-day methods of the experimental gardens and
the cytological laboratories usually substantiate rather than displace
Jepson’s judgments. Relatively few of his contemporary authors
have found their work so generally acceptable.

In gauging Jepson’s place in California botany, the writer was
prompted by curiosity to tabulate the number of species in the state
named by each author, using unchanged the data as given in Jepson’s
Manual, our last complete list. Despite the shortcomings of the Man-
ual data, such as the incomplete synonymy, the resulting list is of
some interest. Here are the top 15 names, including all those who

1948 | KECK: WILLIS LINN JEPSON 227

have named 50 or more species in the Manual, together with the num-
ber of names contributed by each.

1. Asa Gray 717 9. Willis Linn Jepson 154
2. Carolus Linnaeus 431 10. David Douglas 86
3. Edward Lee Greene 365 11. George Engelmann 74
4. Sereno Watson 283 12. A. P. de Candolle 68
5. Thomas Nuttall 266 13. Frederick T. Pursh 64
6. John Torrey 245 14. George A. W. Arnott 58
7. George Bentham 188 15. Joseph Nelson Rose 51
8. William J. Hooker 160

The only Californians among the first 15 are Greene and Jepson, and
these are grouped among the classical students of the California flora.
Albert Kellogg, however, with 48 species, is in sixteenth place. Other
Californians among the first 50 are A. A. Heller, T. S. Brandegee, Alice
Eastwood, H. M. Hall, Katherine Brandegee, and S. B. Parish, in that
order.

Jepson worked in that transitional period between the time of
Greene, when new species were yet to be found on almost every moun-
tain range and valley floor, and the present, when even monographic
researches uncover relatively few acceptable new species. Consider-
ing the conservative stand that he took on the matter of describing new
species, it is interesting how high in the list his name is found. Jepson
preferred to evaluate critically his own proposals before offering them
to the world. This is one reason that his work has attained a lasting
character.

The influence of Jepson does not rest wholly upon his writings.
The relatively small number of graduate students that he found time
to encourage came impressionably under the influence of his strong
character. Their training would doubtless be considered unorthodox
and irregular, but certain fundamentals about meticulous detail in ob-
servation of the plant, whether in the field or in the laboratory, and a
broad appreciation for the contributions from related fields were drilled
into the memory. His graduate student seminars were often his sole
contact with the student. These were broadening and often dramatic
experiences that challenged the imagination to reach out; they served
to turn the student’s attention from the local flora, with which Jepson’s
life would seem to be engrossed, to the far corners of the earth and
to many fields untouched by Jepson’s writings. The beneficial in-
fluence of this training is apparent from the sound taxonomic practices
of those trained by him and, in turn, of their students.

Jepson succeeded in imparting to his public, which consisted in
good part of laymen as well as of students, his deep feeling for nature.
He looked upon the plant not only with the discriminating eye of the
master systematist, but also with the enthusiasm and reverence of the
naturalist and woodsman. Perhaps most beautifully expressed was
his love for trees, so obvious in the Silva. One’s love of nature is apt

228 MADRONO ['Vor. 9

to govern in direct proportion one’s concern for conservation, and so
it was that Jepson was a founder and prominent spokesman for the
Save-the-Redwoods League and a staunch advocate of forest conserva-
tion measures and such other endeavors as the Point Lobos Reserve.
All in all, it has been through many channels that the works of Jepson
the botanist have become known, not only to his California audience,
but to the world at large.

; Carnegie Institution of Washington

Division of Plant Biology
Stanford, California

REVIEWS

The Evolution of Gossypium and the Differentiation of the Culti-
vated Cottons. By J. B. Hutcuinson, R. A. Sitow and S. G. STE-
PHENS. Oxford University Press. 160 pp. 10 figs. 1947. 15s.

The genus Gossypium is of outstanding interest not merely
because of the economic importance of cotton, but in respect to the
taxonomy, genetic relationships, morphology, and physiology of
the plants. To botanists of the Pacific Coast the interest is en-
hanced by the fact that of approximately 16 truly wild species of
this genus, 4 are indigenous to the shores and islands of the Gulf
of California.

The scope of the admirable little book under review is indi-
cated by its title and by the titles of the parts: Part 1, The classi-
fication of the Genus Gossypium, by Hutchinson; Part 2, The
Evolution of the Species of Gossypium, by Hutchinson and Ste-
phens; Part 8, The Differentiation of the True Cottons, by Hutch-
inson and Silow; and Part 4, The Significance of Gossypium in
Evolutionary Studies, by Hutchinson and Stephens. It would be
hard to find authors more competent to deal with these matters,
all three of them having carried on original research that has
advanced, substantially, our knowledge of the subject.

In treating ‘““The Relationships of the Genus,” the authors fol-
low Edlin in transferring the tribe Hibisceae, to which Gossypium
belongs, from the Malvaceae to the Bombacaceae, on the ground
that the fruits are capsular (loculicidally dehiscent), not septi-
cidally dehiscent into schizocarps, as in the other tribes of Mal-
vaceae. But this distinction is not absolute because normally
in Bastardia and occasionally in Sphaeralcea and other genera
which no one would think of removing from the Malvaceae, the
septicidal dehiscence is very imperfect. There seems to be no
sharp line of demarkation between the two families and perhaps
we should return to the classification of Bentham and Hooker, and
regard the Bombacaceae as merely a tribe or subfamily of the
Malvaceae.

In the classification of the species, the authors follow, in the
main, that which was first outlined by Zaitzev, and later amplified
by S. C. Harland, on the basis of cytogenetic studies. Three

19481 REVIEWS 229

main groups are recognized: (1) The approximately 16 wild spe-
cies, of which all that have been studied cytologically have 13 as
the haploid chromosome number. None of them has true lint
hairs on the seeds. These species are widely dispersed in tropical
and subtropical regions of both the Eastern and Western Hemi-
spheres, with scarcely any overlapping of their ranges. (2) The
Asiatic cultivated cottons (G. herbaceum and G. arboreum), likewise
with 13 haploid chromosomes, but producing lint hairs of spin-
nable length. (3) The American cultivated cottons (G. hirsutum
and G. barbadense), with 26 haploid chromosomes and long lint
hairs on the seeds. To the last group belongs G. tomentosum, of
the Hawaiian Islands, where it is supposed to be endemic, although
its affinity to the cultivated American cottons is unquestionable.

The remarkable discovery was made by Skovsted, that the
American cultivated cottons possess two sets of chromosomes, 13
larger ones, similar to those of the Old World cultivated cottons,
and 13 smaller ones, similar to those of American wild diploid spe-
cies. This suggested that the American cultivated cottons origina-
ted as allopolyploids. Subsequent genetic investigations by Har-
land and Atteck, Silow, Beasley, and Stephens have strengthened
the evidence of such origin, pointing to the Peruvian wild cotton,
G. Raimondii, or a near relative, as one of the probable ancestors.

The difficulty has been to explain how contact could have been
brought about between an Old World and a New World species,
since the origin of the American cultivated cottons was, assuredly,
pre-Columbian. Harland assumed a southern trans-Pacific land
bridge in Cretaceous or early Tertiary times, but apparently
insuperable objections to this hypothesis are given in the book
under review (pp. 75, 76). The authors (pp. 77-80) prefer the
assumption that seeds of an Old World cotton were brought to
America and planted there by prehistoric voyagers across the
Pacific. Such an explanation cannot be dismissed as impossible,
but a third, and in the reviewer’s opinion a more plausible hy-
pothesis, has been advanced recently by Stebbins (Ecol. Mon. 17:
155), who points out the significance, in relation to this problem,
of the mingling of Asiatic and New World floral elements in
Eocene deposits of North America. Although no traces of Gos-
sypium have been found, as yet, in such deposits, it is an attractive
possibility that a cotton, related to the Old World cultivated
diploid species, might have existed in the Western Hemisphere
during the Tertiary, in contact with an American diploid species,
and that hybridization between them might have produced the
ancestors of the American tetraploid cultivated cottons.—THomas
H. Kearney, California Academy of Sciences, San Francisco.

The New World Cypresses. Part I. Taxonomic and Distribu-
tional Studies of the New World Cypresses, by Cart B. Woir. Part
IT. Diseases of Cypresses, by Wituis W. Wagener. Part III.

230 MADRONO [Von. 9

Horticultural Studies and Experiments on the New World Cypresses,
by Cart B. Worr. Pp. xvi+ 444. Appearing as volume I of El
Aliso, A Series of Papers on the Native Plants of California,
published by the Rancho Santa Ana Botanic Garden, Anaheim,
California. 1948. $4.00 buckram, $3.55 paper bound.

The new publication El Aliso gets off to an auspicious be-
ginning with this excellent treatise on the New World cypresses
as its first volume. As explained in the forword written by the
editor, Philip.A. Munz, this journal is founded to care for the
botanical and horticultural papers issuing from the Rancho Santa
Ana Botanic Garden and will appear at irregular intervals. The
name, E] Aliso, was that used by the Spanish Californians for the
native sycamore. ;

The genus Cupressus has received excellent attention at the
hands of Wolf and Wagener. These scientists were well quali-
fied for their task, each with nearly 20 years’ experience with
cypresses behind them before writing their contributions. Dr.
Wolf, until recently Botanist at the Garden, has contributed Part
I, a 250-page detailed taxonomic and distributional account of the
species, and Part III, 115 pages on their horticulture. Dr. Wage-
ner, of the Division of Forest Pathology of the United States
Department of Agriculture, has written Part II, of 68 pages, on
the diseases of the American species. Their work was mutually
cooperative from the year 1934. The first cypresses were planted
at the Rancho Santa Ana Botanic Garden in the year of its found-
ing, 1927, and a thorough investigation of this genus of trees soon
became the most ambitious single project undertaken by the
Garden.

Eleven of the sixteen entities recognized are native only in
California, and Dr. Wolf has collected and grown material from
three-fourths of the known groves in the state. In Part I fifteen
species are recognized as native in the New World, one of which,
C. Bakeri, is composed of two subspecies. ‘Two species are new:
C. Abramsiana, known definitely but from two stations on Ben
Lomond, Santa Cruz Mountains, and C. Stephensonu, known from
a scattered stand about one mile in extent along the upper limits
of King Creek, Cuyamaca Peak, San Diego County, both of which
must be counted among the rarest tree species in California.
Also new is C. Bakeri subsp. Matthewsit of the Siskiyou Mountains
of Josephine County, Oregon, and Siskiyou County, California.

In the systematic section the order of treatment appears to
be rather arbitrary. One feels that Wolf uncovered very few
clear patterns of relationship within the genus, but that had the
scattered discussions been drawn together in one strong sub-
chapter on relationship the picture would have been clearer to
the reader, and one or two improvements in the order of treat-
ment would have been obvious.

The species concept employed is easily defended from the

19481 REVIEWS 231

point of view of the horticulturist, to whom differences in habit,
herbage color, etc., are of importance, although Wolf grants that
some of the separations are difficult to make from herbarium
material. Similar difficulties confront us in the classification of
the pines, oaks, and many other arboreal genera. Perhaps in
woody plants of horticultural or economic importance we are
being most expedient in recognizing differences of a sort that are
passed over more lightly in the herbaceous species. Wolf is
aware of the problem and more than once discusses what his more
conservative treatment of the New World cypresses could have
been—7 species instead of 15. From the relatively minor nature
of the morphological differences between the species, one specu-
lates as to whethter there might prove to be even fewer biologi-
cally distinct entities, but no crossing experiments have been
attempted. Wolf appearently has seen no tree in the wild that
he considers to be an interspecific hybrid.

In Part II Dr. Wagener gives a clear account of the diseases
found in American cypresses, the most serious of which is the
Coryneum canker, known in California only since 1915 but now
spread through and having almost eliminated most of the planted
Monterey cypresses in the state. Fortunately, protective mea-
sures have been effective in keeping it out of the native groves of
the native groves of the species. Susceptibility to the disease
was tested in two plots established at Stanford University, and the
coastal species were found to be most susceptible, and the inland
the most resistant.

In Part III Wolf describes the growing of cypresses and gives
the experience obtained in numerous test plots. An evaluation
of the horticultural possibilities of the species is then presented.

Good stock, presswork, and binding have combined to make an
attractive book. A candid review, however, must touch upon the
weak organization and lay-out of the material. For example, the
book is liberally illustrated, a number of the 80 halftones being
composed of 2 or 3 photographs, but it is unfortunate that 66 of
the 80 have their legends on the opposing, or some other, page.
Judicious trimming of the photographs would have improved
many of them and allowed this number to be reduced to an approx-
imate 10. In part I the narrative style results in long descriptions
and citations of localities and collections. One wades through
much extraneous matter to get at any one class of information,
such as the ecology of cypress. The synonomy is given twice, in
an effort to draw like materials together, but it is disturbing to
find that the two references do not always agree, as in the author-
ity for the combination C. Goveniana var. Sargentii. Headings
are not graded in descending order of importance, nor do those
in the text always parallel those in the table of contents, and the
dual legends employed in the last eleven tables should preferably
have been avoided. Finally, in so comprehensive a treatment of

232 MADRONO [Vou. 9

a genus like Cupressus, with species distributed in groves so
admirably suited for the purpose, the lack of distribution maps is
keenly felt. All of these detractions are minor, however, as
weighed against the wealth of material to be found in this volume,
and the thanks of all persons interested in cypress are due the
authors. |

We can now set this cooperative treatment of a native tree
important to horticulture beside a similarly conceived book on
another horticultural subject, Ceanothus, put out in 1942 by the
Santa Barbara Botanic Garden. Our California botanic gardens
appear to be producing just the type of work for which they are
better fitted than anyone else. May the list of their contribu-
tions become a long one !—Davwp D. Keck, Carnegie Institution of
Washington, Stanford, California.

NOTES AND NEWS

ELYMUS ARISTATUS IN CALIFoRNIA. What is apparently the first
detinite California station for Elymus aristatus Merr., a species
occurring mainly from Washington and Montana south to Nevada
and California, was established by the writer in 1941 when this
species was collected in Mono County, twenty miles northwest
of Bridgeport (Gould 1325). According to Mrs. Agnes Chase of
the United States National Museum this grass was known previ-
ously in the state only from two Bolander collections whose loca-
tions were stated merely as “California.” At the site of the
Bridgeport collection E. aristatus was observed growing with E.
cinereus Scribn. & Merr. on a dry, open hillside in an Artemisia
tridentata association. Both species of Elymus had a strongly
developed bunch-grass habit with culms in large clumps. £lymus
aristatus, however, differed conspicuously from the more familiar
E. cinereus in its long awns and non-glaucous culms and leaves.
The paucity of California records of E. aristatus is probably not
a true index to the abundance or range of the species in the state.
Further collections from the foothills of the central and southern
Sierra Nevada are certainly to be looked for. The writer is in-
debted to Mrs. Chase for information concerning the Bolander
collections and for confirming the identity of his own specimens.
—Franx W. Goutp, Department of Botany, University of Arizona,
Tucson, Arizona.

MADRONO

A West American Journal of Botany

A quarterly journal devoted to important
and stimulating articles dealing with
plant morphology, physiology, taxonomy,
and botanical history. These volumes
should be a part of every botanist’s li-
brary and should be made accessible to
students of all universities and colleges.

Volume I, 1916-1929. . . $5.00
Volume II, 1930-1934 .. 5.00
Volume III, 1935-1936 . 5.00
Volume IV, 1937-1938 . 5.00
Volume V, 1939-1940 . . 5.00
Volume VI, 1941-1942 . 5.00
Volume VII, 1943-1944 . 5.00
Volume VIII, 1945-1946 . 5.00
Single numbers....... 1.00

The subscription price of MADRONO
is $3.50 per year. We solicit your pat-
ronage,

Address all orders to:

Rimo Bacigalupi, Bus. Mgr.
Natural History Museum,
Stanford University, California

ra Y
Avy)
my ‘
Lig bs A
’ ;
‘ ‘ phsdeathl sin Mnwths ada biy oo bse wae ‘ F
; ,
v3
‘
5
Z 2 A t¢ id é
‘ i
a
4
.
/
; ’
{ '
, ‘
i
1
S
;
¥ ’
y
\
‘

VOLUME Ix NUMBER 8

Axes wit SHITE
as nes CEN

MADR

A WEST AMERICAN JOURNAL OF
BOTANY

we

Contents
Tae Iventrry anp Detmrration oF Attiom Totmir1 Baker, Marion
CET) SEMIN LIE OIE T athe oe) ls BRI oC CA MC Ss TO 233
Nores ON THE GENUs ToOwNSENDIA IN WeEsTEBN NortH Amenica, Charles
PEPER ERO E TT ol CMY ane elie ace cual eteve eRe sss Ch KE ee uOs Mn hah RISE AORN, 238
Some PaRaLLELs BETWEEN Desere ann ALPINE F'Lora 1n Catirornnia, F. W.
UA a es A DORE ROP TUT SIE OR lei aie aie RO LY aD SS she a ak ed Ve 241
Some AvorrionaL Notes on Potemontiacear, Herbert L. Mason ........... 249

A New Speecies or Puacetia from Sonora, Mexico, Lincoln Constance ... 255

Curomosome Number Poustication, J. A. Rattenbury ............... ee 257

PR ENMES TONOWUME EX jk i te ade dies SR ANS ae 250 #
a

Published at North Queen Street and McGovern Avenue,
Lancaster, Pennsylvania

October, 1948

MADRONO

A WEST AMERICAN JOURNAL OF BOTANY

Board of Editors

Hersert L. Mason, University of California, Berkeley, Chairman.

LeRoy Anrams, Stanford University, California.

Epcar Anperson, Missouri Botanical Garden, St. Louis.

Lyman Benson, Pomona College, Claremont, California.

Herserr F. Coprtann, Sacramento College, Sacramento, California.

Ivan M. Jonnston, Arnold Arboretum, Jamaica Plain, Massachusetts.

Mixprep E. Matuias, Dept. of Botany, University of California, Los Angeles 24.
Bassett Macume, New York Botanical Garden, N. Y. C.

Marion Ownesey, State College of Washington, Pullman.

Secretary, Editorial Board—Anwnerra Carter
Department of Botany, University of California, Berkeley

Business Manager—Rimo Bacica.Lupi
North Queen Street and McGovern Avenue, Lancaster, Pennsylvania
or
Natural History Museum,
Stanford University, California

Entered as second-class matter October 1, 1935, at the post office at
Lancaster, Pa., under the act of March 3, 1879.

Established 1916. Published quarterly. Subscription price $3.50 per year.
Completed volumes I to VIII $5.00 each; single numbers $1.00.

Papers up to 15 or 20 pages are acceptable. Longer contributions may
be accepted if the excess costs of printing and illustration are borne by the
- contributor. Range extensions and similar notes will be published in con-
densed form with a suitable title under the general heading “Notes and News.”
Articles may be submitted to any member of the editorial board. Manuscripts
may be included in the forthcoming issue provided that the contributor pay the
cost of the pages added to the issue to accommodate his article. Reprints of
any article are furnished at a cost of 4 pages, 50 copies $5.90; 100 copies $6.48;
additional 100’s $1.23; 8 pages, 50 copies $8.55; 100 copies $9.49; additional 100’s
$1.88; 16 pages, 50 copies $12.00; 100 copies $13.44; additional 100’s $2.88.
Covers, 50 for $3.95; additional covers at $2.38 per hundred. Reprints should
be ordered when proofs are returned.

Published at North Queen Street and McGovern Avenue, Lancaster,
Pennsylvania, for the

CALIFORNIA BOTANICAL SOCIETY, INC.

President: Adriance S. Foster, University of California, Berkeley. First
Vice-President: G. Ledyard Stebbins, Jr.. University of California, Berkeley.

Second Vice-President: Herbert F. Copeland, Sacramento College, Sacramento,

California. Secretary: John Whitehead, University of California Botanical
Garden, Berkeley. Treasurer: Rimo Bacigalupi, Natural History Museum,
Stanford University, California.

Annual membership dues of the California Botanical Society are $3.50,
which includes a year’s subscription to Madrofio. For two members of the same
family the dues are $4.00, which includes one copy of Madrofio and all other
privileges for both. Dues should be remitted to the Treasurer. General
correspondence and applications for membership should be addressed to the
Secretary.

1948] OWNBEY: ALLIUM TOLMIEI 233

THE IDENTITY AND DELIMITATION OF ALLIUM
TOLMIEI BAKER

Marion OwnsBEY

Biosystematic understanding of any group of species is a long-
time undertaking during the course of which many minor studies
must be made. These are of little consequence in themselves, but
in toto they are indispensable to the accuracy and acceptability of
the final conclusions. One series of such minor studies is con-
cerned with the establishment of the taxonomic identity of each
of the proposed names within the group. Others deal with the
genetic, cytological, ecological, distributional, and phylogenetic
relationships of. the biological entities themselves. The results of
these minor studies are usually incorporated into the final treat-
ment where each takes up at most only a few lines, and are not
scattered through the literature in separate papers. Occasion-
ally, it may be advisable to present the conclusions from such a
study in advance of the appearance of the final treatment. The
identity of Allium Tolmiei seems to be such a case. The recent
proposal of two superfluous names for this species (Traub, 1947)
calls for a clarification of its identity in accord with the estab-
lished principles of plant classification.

Allium Tolmiei generally has been accepted for the last seventy
years as a valid species of the northwestern states (Watson, 1879 ;
Coulter, 1885; Howell, 1902; Piper, 1906; Rydberg, 1917;
Abrams, 1923; Peck, 1941), but there has not been similar
unanimity as to the characteristics of the species to which the
name should be applied. In the writer’s opinion, all of these
descriptions were drawn for the most part from specimens which
he would refer to A. Tolmiei. In other words there has been una-
nimity in the acceptance of the name, but not in the delimitation of
the species which must bear it.

The history of Allium Tolmiei in the literature precedes by
many years the first appearance of the binomial. In his “Flora
Boreali-Americana,’ Hooker (1839) mentions an unnamed
variety 8 of A. Douglasii, with leaves longer than the scape, col-
lected in the “Snake Country” allegedly by Tolmie. Whether
or not these specimens were actually collected by Tolmie or by a
friend, as he expressly states according to Piper (1906), is not
relevant to the problem. In the following discussion, they will
be referred to as Tolmie’s specimens. The “Snake Country”’
would undoubtedly be in southwestern Idaho or adjacent Oregon
whence come more recent collections closely resembling Tolmie’s.

The binomial, Allium Tolmiei, was first proposed in 1876 by
Baker, who undoubtedly had examined critically Tolmie’s speci-
mens preserved at Kew and arrived at the conclusion that they

Manprono, Vol. 9, No. 7, pp. 201-232. September 2, 1948.

DEc 27 IQ49

234 MADRONO [Vol. 9

represented a species distinct from A. Douglas. Unfortunately,
he did not describe his new species at that time, but merely cited
Hooker’s earlier reference to the collection. Thus the name
might have remained a nomen subnudum had not it been taken up
three years later by Watson (1879) and provided with a clear
and unmistakable description. With this first adequate descrip-
tion, Watson cites: (1) Tolmie’s specimens; (2) Hooker’s figure
of A. Douglasi (in part), about which he undoubtedly was con-
fused; and (8) his own collection from Parley’s Park in the
Wasatch Mountains, Utah, which he had identified earlier as
A. tribracteatum, but which now proves to be A. Brandegei. As far
as can be ascertained, Watson’s description was drawn entirely
from Tolmie’s specimens which are still preserved in an identifi-
able condition in the Gray Herbarium. He does not mention,
for instance, the characteristic cellular reticulations on the bulb
coats of A. Brandegei which he clearly illustrates in fig. 7 of Plate
XXXVIII of the Botany of the King Expedition. It is necessary,
then, in the typification of A. Tolmiei, to exclude both elements
2 and 8, above. This leaves A. Tolmiei Baker ex Watson (1879)
exactly equivalent to A. Tolmiet Baker (1876) and A. Douglas
var. 8 Hooker (1839), these all being based on the same collec-
tion. Itis probable that the first adequate description of the spe-
cies which he attributes to Baker was actually drawn by Watson
himself from the specimens preserved in the Gray Herbarium.
It seems proper, therefore, to designate this sheet as the type
rather. than that presumably preserved at Kew.

Six years following the publication of Watson’s description of
Allium Tolmiei, Coulter (1885) accepted this species in his ““Manual
of the Botany of the Rocky Mountain Region.” His description
was compiled directly, word for word, from Watson, with certain
rearrangements and deletions. Certainly, there is no basis for
the assumption that this author had any first-hand knowledge of
the species whatsoever, or that this description applies to any
species other than that represented by Tolmie’s specimens. This
would be unimportant had Traub (1947) not made Allium Tolmiei
Baker the basis of a new varietal combination under A. Douglasii,
and A. Tolmiei “Coulter . . . non Baker’ the basis of a new name.
From the foregoing, it is clear that A. Douglasii var. B = A. Tolmiei
Baker = A. Tolmiet Baker ex Watson = A. Tolmiei Baker ex Coulter,
these being based on one and the same collection and that A.
Douglas var. Tolmiet (Baker) Traub and A. idahoense Traub,
being equal to the same thing are equal to each other, and are
accordingly superfluous synonyms of A. Tolmiez.

Once the taxonomic identity of a proposed name is established,
the next step is the association of this name with a natural bio-
logical population. The methods developed by modern sys-
tematists for the association of a name with the proper biological
entity differ materially from those used by their predecessors.

1948] OWNBEY: ALLIUM TOLMIEI 235

It is a well-known truism that no two individuals, in the ultimate
analysis, are exactly alike. One cannot, therefore, restrict the
application of a name to individuals which are exactly like the
original ones and arrive at anything which could be called a useful
classification. Classical systematy was essentially a mechanical
sorting process whereby individuals were associated with extreme
morphological forms deemed to represent species—usually on
the basis of very slender evidence. That such species frequently
coincided with natural units can be attributed to the nature of the
material rather than to the reliability of the method. The mod-
ern emphasis is on the species as a natural biological phenomenon,
whereas, the stress formerly was on actual or supposed specific
differences. Modern species are bounded by discontinuities ;
classic species were marked by distinguishing morphological char-
acters. The two are not the same. The aim of the old syste-
matics was to provide each species with a name and a description.
That of the new systematics is to understand the species and to
name it only after the need for a name has been clearly estab-
lished. Indeed, the modern systematist would prefer to arrive
at his conclusions apart from and uninfluenced by pre-existing
concepts. This is the only way in which he may be sure of avoid-
ing the pitfalls inherent in the older method. Allium Tolmiei is
a case in point.

Sporadically scattered over much of eastern Oregon and over-
lapping into adjacent states, there is a series of closely related
local populations of the genus Allium. In a given locality, the
plants are usually very much alike, although they may show some
evident variation. In another locality, perhaps close at hand,
they may be slightly different or even very different. The plants
from locality to locality vary greatly in size, in the relative and
absolute length, breadth, curvature, and glaucescence of the
leaves, in the relative and absolute length, breadth, and thickness
of the scape, whether this structure is slightly, moderately, or
strongly flattened, wingless, narrowly or broadly winged, in the
number of flowers in the umbel, and the relative and absolute
lengths and thicknesses of the flowering pedicels, in the color of
the perianth, and even in the intimate details of the floral struc-
ture, such as the presence or absence of crests on the ovary, and
their development from obsolete to obscure to prominent. The
total variation is enormous. Confronted by a half dozen speci-
mens representing as many extremes, no person unacquainted with
the complexities of intraspecific variation would question for a mo-
ment that each represented a distinct and definite species. If he
had fifty of them at once, he might become suspicious, and con-
sider them that anathema of the systematist, the polymorphic
species. In all probability, however, he would pick out some
three or four of the most conspicuously distinct and aberrant

236 MADRONO [Vol. 9

types, and group the others around these, disregarding the fact
that most of them could go into one pile as well as another. These
departures from his type concept, he would attribute vaguely to
ecological factors.

Modern experimental taxonomy provides a method of investi-
gating such perplexing variation between natural populations.
Some thirty collections from as many localities representing the
above series have been assembled and grown side-by-side at
Pullman under essentially uniform conditions. The distinctive
characteristics which marked the parental populations are main-
tained in the garden. From herbarium studies, it can be inferred
that only a small percentage of the local races within the series
are represented in this living collection. There is no reason to
suppose that within the series there is a single morphological
hiatus which cannot be bridged or detoured through intermediate
biological populations. Within the same area, however, and
extending beyond it to the north and the south,.is another series
of similar populations, the A. parvum series, apparently distin-
guished at all times by a constant hiatus, the magnitude of which
is much less than that of the difference separating any two of a
number of extreme populations of either series. This is illus-
trative of the fact that the magnitude of a difference does not in
itself make a species.

The thirty collections of the first series have been studied mor-
phologically and cytologically, and the results of these studies
form the basis of a paper in preparation (Ownbey and Aase,
unpublished). It is sufficient to say here that Dr. Aase has found
most of the local races of the series to be diploid, but that in one
limited area, there exist, sometimes side-by-side, morphologically
distinguishable diploid and tetraploid races, and that in another,
the plants apparently are uniformly hexaploid. The tetraploid
race is morphologically nearly indistinguishable from a diploid
race growing in a nearby area. No diploid exactly corresponding
to the hexaploid race is known, but the attenuated morphological
characteristics by means of which it may be recognized are of
exactly the same nature and are much less conspicuous than those
which distinguish many of the diploid races. Thus it may be
concluded that the entire series of intergrading populations rep-
resents only a single biological species for which a name must
now be selected.

With specimens representing three different local populations
of the above species at hand, Watson (1879) proposed, in the
same paper, three species, Allium Cusicki, A. pleianthum, and A.
Tolmiei, into which pigeon holes subsequent botanists have been
vainly struggling to make their specimens fit. Later, Jones
(1902) added A. anceps var. aberrans, and Tidestrom (1916)
described the tetraploid as A. platyphyllum. Both of these were

1948 | OWNBEY: ALLIUM TOLMIEI 237

promptly reduced to synonymy. Other later proposals probably
belonging here have been made, but their identity has not been
unequivocally established. The International Rules do not rec-
ognize priority of position, but give the subsequent author the
privilege of choosing between them should two or more simul-
taneous proposals prove synonymous. Ordinarily, he picks the
one in the prior position (A. Cusickii), but in this instance another
choice seems imperative. There will always be those who con-
sider Hooker’s three Latin words an adequate botanical descrip-
tion, and date the effective publication of A. Tolmiei from 1876
instead of 1879, in spite of the fact that these three words describe
equally well any one of at least half of the western American
species of the genus. To avoid this confusion, the name selected
must be the last in the series, Allium Tolmiei.
The established synonymy follows:

I. Tyreponyms!

Allium Tolmiei Baker ex Watson in Proc. Am. Acad. Arts and
Sci. 14: 234. 1879, excluding references to illustration of A.
Douglasii and Watson’s collection from Parley’s Park; Coulter,
Man. Bot. Rocky Mountain Reg., p. 349. 1885; Howell, Fl. N. W.
America, p. 642. 1902; Piper in Contr. U. S. Nat. Herb. 11 [FI.
Wash.]: 188. 1906, excluding specimens cited; Rydberg, FI.
Rocky Mountains and Adj. Plains, p. 161. 1917; Abrams, I[llust.
Fl. Pacific States 1: 887. 1923; Peck, Man. Higher Plants Oregon,
p. 195. 1941.

Allium Douglasit var. 8. Hooker, Fl. Bor.-Am. 2: 185. 1839.

Allium Tolmiet Baker in Bot. Mag. Ser. III. 32: under t. 6227.
1876, nomen subnudum.

Allium Douglasii var. Tolmiet (Baker) Traub in Herbertia 12
(1945): 68. 1947.

Allium idahoense Traub, Ibid., p. 69.

II. MretonymMs

Allium Cusickii Watson in Proc. Am. Acad. Arts and Sci. 14:
228. 1879; Howell, Fl. N. W. America, p. 642. 1902; Rydberg,
Fl. Rocky Mts. and Adj. Plains, p. 161. 1917; Abrams, Illus. FI.
Pacific States 1: 887. 1923; St. John, Fl. S. E. Wash. and Adj.
Idaho, p. 85. 1937; Peck, Man. Higher Plants Oregon, p. 195.
1941.

Allium pleianthum Watson in Proc. Am. Acad. Arts and
Sci. 14: 2338. 1879; Howell, Fl. N. W. America, p. 642. 1902;
Rydberg, Fl. Rocky Mts. and Adj. Plains, p. 161. 1917; Abrams,

1 The terms “typonym” (a name based on the same type) and “metonym”
(a name based on another member of the same group), defined in the “Code
of Botanical Nomenclature,” proposed in Bull. Torrey Bot. Club 31: 249-290.
1904, have not been generally adopted. They represent exceedingly useful
concepts.

238 MADRONO | [Vol.9

Illus. Fl. Pacifie States 1: 386. 1923; Peck, Man. Higher Plants
Oregon, p. 195. 1941.

Allium anceps var. aberrans Jones, Contr. West. Bot. No. 10,
p. 10, fig. 9. 1902.

Allium platyphyllum Tidestrom in Torreya 16: 242. 1916.

State College of Washington
Pullman, Washington

NOTES ON THE GENUS TOWNSENDIA IN WESTERN
NORTH AMERICA

Cuartes B. Hetser, Jr.

While identifying the Compositae collected by Miss Annie
M. Alexander and Miss Louise Kellogg in the Sweetwater Moun-
tains of California and Nevada during the summer of 1945, I en-
countered a number of specimens of Townsendia. An investiga-
tion of herbarium material of the genus was undertaken, and
since field studies seem out of the question at present, I am record-
ing here some of my results.

In the revision of the genus by Larsen (1927), nineteen spe-
cies are recognized for the genus. Larsen lists only two species,
Townsendia scapigera and T. Watsoni, as occurring in the states of
California and Nevada. The last few years have witnessed in-
creasing collecting activity in the Great Basin area and additional
material has been obtained so that seven species are now known
to occur in these states.

Since the publication of Larsen’s paper, two new species have
been described, T. minima Eastwood (1986) from Utah and T.
diversa Osterhout (1928) from Colorado. The description of an-
other new species in the present paper brings the total number of
species recognized to twenty-two, some of which doubtfully de-
serve specific rank.

All of the specimens cited are deposited in the Herbarium of
the University of California, Berkeley, unless otherwise indicated.
During the course of this study, material has been examined from
the California Academy of Science (CA), the Missouri Botanical
Garden (MBG), the Dudley Herbarium of Stanford University,
the United States National Herbarium, the Intermountain Herba-
rium of Utah State Agricultural College (IH), and the Rocky
Mountain Herbarium of Wyoming University. I would like to
thank the curators of these herbaria for the privilege of examin-
ing their specimens.

Townsendia sericea has been collected in both California (Mono
County: Maguire & Holmgren 26109; Duran 1661) and Nevada
(Nye County: Maguire & Holmgren 25818, 25944). Townsendia
incana is known from Nevada from a specimen collected by

1948 ] HEISER: TOWNSENDIA 239

Shockley (103) in Nye County, and has recently been reported
from Lincoln County by Barneby (1947). Townsendia arizonica
has been collected several times in Clark County, Nevada, chiefly
in the Charleston Mountains (Alexander 774, 780; Alexander &
Kellogg 1558, 1568, 1676; Clokey 7772, 7773; Ripley & Barneby
2910).

Townsendia spathulata is known from high altitudes at three
localities in Mono County, California (Duran 1662, Maguire &
Holmgren 26109a, Alexander & Kellogg 4061). Alexander and
Kellogg have noted that the single plant of this species which
they collected was growing with T. scapigera. The Maguire and
Holmgren specimen is only a fragment mounted on a sheet with
T’. sericea.

Townsendia florifer has been collected several times in Nevada:
[Elko County: May 10, 1942, Cantelow s. n. (CA); Ripley &
Barneby 4613 (CA); Maguire & Holmgren 2828; Holmgren & Lund
3 (IH). White Pine County: Ripley & Barneby 3596 (CA) ].
Townsendia Watsonii was reported from Nevada by Larsen on the
basis of a specimen collected in 1891 by A. J. Jones, without
definite locality. This entity, however, is scarcely specifically
distinct from T’. florifer.

One of the most interesting species in the genus, T'. scapigera,
was known in California from only one collection at the time of
Larsen’s revision, and was not known from Nevada at that time.
It is now known for several stations in both states. Only one
collector is cited for each county. [Cattrornia. Inyo County:
Alexander & Kellogg 2492, 2998, 3020, 30386. Mono County: Alex-
ander & Kellogg 3959, 4053, 4556, 4556A, 4561. Modoe County:
May 1879, Lemmon s. n. Nevapa. Elko County: Holmgren 1034,
00245 (IH). Esmeralda County: Maguire & Holmgren 25640.
Mineral County: Alexander & Kellogg 4440. Nye County: Train
27388. White Pine County: May 1918, King s.n. (CA). Eureka-
Lander counties: Eastwood & Howell 168,175 (CA) ]}.

Townsendia scapigera is an extremely variable species, and the
forms found in Inyo and Mono counties deserve special mention.
In Inyo County, dwarf forms occur which have rays 7-11 mm.
long, heads 11-13 mm. high, and 12—22 involucral bracts (Alea-
ander & Kellogg 3036, 3020, 2993). The plants from the Sweet-
water Mountains (Alerander & Kellogg 4556, 4556A, 4053, 3959),
on the other hand, are large, with rays 14-16 mm. long, heads
15-20 mm. high, and 30-37 involueral bracts. Examination of
the pollen of the latter specimens revealed a high percentage of
empty grains, as well as the presence of both 8- and 4-pored
grains similar to those found in many apomictic species. It would
not appear wise to give formal taxonomic recognition to these
entities until further studies can be undertaken.

During the summers of 1946 and 1947 I had the privilege of

240 MADRONO [Vol.9

examining the specimens of Townsendia in the herbarium of the
Missouri Botanical Garden. Of particular interest were the speci-
mens collected by von Schrenk in Wyoming which Larsen in her
revision interpreted as T. scapigera. These specimens do not fall
within the range of any known species of the genus and accord-
ingly are described as new.

Townsendia anomala sp. nov. Herba perennis ad 4 cm. alta,
foliis spathulatis usque 1 cm. longis 8 mm. latis dense strigillosis,
capitulis in ramis foliaceis brevi-pedunculatis, involucri bracteis
lanceolatis acuminatis, marginibus membranaceis latis, pappis
plurisetosis, setis disciflorum ca. 5 mm. longis, ligulis ca. 4 mm.
longis, achaeniis 83-4 mm. longis leviter pubescentibus, pilis brevi-
bus crassis, plerumque simplicibus aliquando emarginatis vel
brevibidentatis.

Perennial up to 4 cm. in height; leaves spathulate, 1 cm. or
less long, 3 mm. or less wide, densely strigillose; heads on leafy
branches, short pedunculate; involucre 2—3 seriate, 6—8 mm. wide;
bracts of the involucre lanceolate, acuminate with wide mem-
braceous margins; rays about 15, 5-7 mm. long, about 1 mm.
wide; pappus plurisetose, the setae of the disk-flowers slightly
longer than that of the ray-flowers; achenes 3—4 mm. long, lightly
pubescent with short, thick, mostly simple or sometimes emar-
ginate or short-bidentate hairs.

Specimens examined. Wyoming. Park County: dry ridge,
Howell Ranch, August 26, 1922, H. von Schrenk s. n. (type, her-
barium of the Missouri Botanical Garden, no. 901271); Holm
Lodge, about 40 miles west of Cody, August 27, 1922, H. von
Schrenk s. n. (MBG).

The Howell Ranch, on which Holm Lodge is located, is ten
miles east of the east entrance of Yellowstone Park at an altitude
of approximately 7000 feet.

The new species appears to be more closely related to T.
spathulata than to T. scapigera, and is found in the range of the
former species. Townsendia spathulata occurs at altitudes of from
8000 to 12000 feet. Townsendia anomala is rather readily dis-
tinguished from it by the pubescence of the leaves and the much
smaller heads which are borne on short leafy branches rather than
being sessile. The hairs of the achenes of both species are more
or less similar. Larsen describes the hairs of the achenes of T.
spathulata as bidentate, but as has been pointed out by Hitchcock
and Thompson (1945) the hairs may be simple in this species.

A number of problems in the genus Townsendia call for ex-
tensive field work and experimental studies. The majority of the
species recognized at present are rather clearcut over most of
their range, but this distinction is frequently blurred at the bound-
aries. For example, T. strigosa is a rather well defined entity
throughout the southern part of its range, but in Wyoming this
species appears to approach T. florifer. Whether hybridization or

1948] WENT: DESERT AND ALPINE FLORA 241

some other factor is responsible can only be revealed by future
studies.

Townsendia, for the most part, is confined to rather high alti-
tudes in the Rocky Mountains. From my preliminary survey of
the genus it is apparent that certain mountain ranges possess
distinctive races or species. A critical correlation of the geo-
graphical distribution and morphological variation in the genus
should reveal the effects of isolation and the origin of new forms

or species.
Department of Botany, Indiana University,
Bloomington, Indiana

LITERATURE CITED

Barnesy, R. C. Distributional notes and minor novelties. lLeafl. West. Bot.
5: 66. 1947.

Eastwoop, A. A new Townsendia from Utah. Leafl. West. Bot. 1: 206. 1936.

Hrircucock, C. L., and J. W. THompson. Noteworthy plants of Idaho. Leafl.
West. Bot. 4: 204. 1945.

Larsen, E. L. A revision of the genus Townsendia. Ann. Mo. Bot. Gard.

14: 1-46. 1927.
OsterHouT, G. E. New plants from Colorado. Bull. Torrey Bot. Club. 55:
(Oeel928:

SOME PARALLELS BETWEEN DESERT AND ALPINE
FLORA IN CALIFORNIA

F. W. Went

At first sight it may seem that a desert flora is the opposite of
an alpine flora, just as the climatic conditions seem so different.
The alpine flora is usually largely influenced by the long cold
winters, whereas the desert flora derives its specific character
from the hot summers and lack of water.

A comparison will be made between the flora of the central
and southern Sierra Nevada (Yosemite and Sequoia National
parks) and the Mohave and Colorado deserts of California. In
and around these deserts several mountain ranges reach into the
alpine zone so that a continuous range of climatic conditions links
the two chosen areas; for comparison, however, the extremes will
be discussed: montane and alpine conditions at 2000 meters and
higher, the desert conditions below 1000 meters.

The alpine climate is one of a very short growing season
of about two months duration (July and August) at altitudes of
3000 meters (Clausen, Keck, and Hiesey, 1940) and a little longer
at 2500 meters. Due to the relatively small precipitation, snow
cover is in most localities not the limiting factor determining the
beginning and end of the growing season. Only towards the end
of June do the mean minimal daily temperatures reach values near
0° C.; before that the freezing point is reached every night, which

242 MADRONO [Vol.9

makes growth for most plants impossible. Melting snowbanks
indicate how few plants can develop at all temperatures around
the freezing point. LErythronium, Caltha biflora and some Carex
species are examples of plants which can grow to a very limited
extent under melting snow, which means at 0° C., but most other
plants covered by snow (such as Saliz) do not start to show visible
signs of growth until the snow has disappeared. Since in most
plants actual growth occurs during night, no appreciable growth
is possible until the night temperatures remain above the freezing
point. And in the beginning of September the nights become too
cold again for growth. Day temperatures during the growing
season become quite high (20 to 25° C.). Precipitation is very
limited during the growing season, usually not exceeding 40 mm.
during the growing period, and is irregularly distributed as
thunderstorms. Therefore the plants have to depend on soil
moisture, which restricts their distribution. Also it increases the
percentage of xerophytic plant types compared with the moister
alpine regions of the mountains farther north. Yet the main limi-
tation of growth is due to low temperatures.

In the lower deserts of southern California the rainfall occurs
almost exclusively during the winter, when temperatures are
fairly low; at sea level freezes occur only seldom, but, at altitudes
around 1000 meters, growth during the winter months is sus-
pended due to low temperatures. Then a short growing season
(March through May) follows before the soil is too dry for fur-
ther plant development.

Therefore the desert and alpine climates have in common a
very short growing season: in the desert it is limited by cold in
the beginning and by moisture in the end, and under alpine con-
ditions it is limited by cold both as far as beginning and end are
concerned, dryness also entering in as a factor. During the grow-
ing season in both localities a high rate of insolation and extremes
in daily range of temperature are common, especially low night
temperatures. Considering all this it is not amazing that marked
parallels in vegetation occur.

Whereas at least one-half of the California alpine plants be-
long to genera migrating from the north (Achillea, Aquilegia, An-
drosace, Antennaria, Carex, various conifers, Draba, Epilobium,
Pedicularis, Potentilla, Primula, Ranunculus, Saxifraga, Silene, To-
fieldia and Viola), a considerable number developed from typically
Western North American genera. In general the latter genera
have a much wider distribution over California than those coming
from the north. In Table 1 the approximate altitudinal distribu-
tion of the circumboreal and Western North American genera is
shown. Most of the data are taken from Jepson’s Manual (1925)
with occasional additions and changes based on personal obser-
vations. In general the altitudinal range for most species is
higher in the southern Sierra Nevada than indicated by Jepson.

1948 | WENT: DESERT AND ALPINE FLORA 243

TABLE 1. APpprROxIMATE ALTITUDINAL DIsTRIBUTION OF CIRCUMBOREAL AND
Western NortH AMERICAN GENERA

No. of Main Genera with No.of Main

Genera with Cali- distri- pa main distri- Cali-_ distri- = i
oT oan an fornia bution alpine bution in fornia bution Ane
istribution species above “*'. Western species above ;
total 2000m. *P&cles America total 2000 m. SP&*les

Anemone .... 5 3 60 Brodiaea ..... 21 3 12
Antennaria .. 8 8 100 Calochortus ... 24 A 17
Arenaria .... 14 if 50 Ceanothus .... 29 5 17
Arnica ...... 10 7 70 Collinsia ...... 17 7 41
CORED wes is, 127 58 46 Delphinium.... 16 5 31
CREDIS eee. 8 4 50 Erigeron ...... 32 15 AT
DYCOG 2s. ; 12 9 75 EHriogonum .... 66 18 27
Epilobium ... 138 10 77 Eriophyllum .. 13 4 31
Hieracium ... 7 3 43 Gilia ......... 19 6 ol
Pedicularis .. 6 5 83 Mimulus ...... 39 12 31
Pirola ....... 6 4A 67 Monardella .... 19 5 26
Potentilla .... 44 30 68 Penstemon..... 37 20 54
GliCn en 22 12 54 Phacelia ...... 55 11 20
Savifraga .... i0 8 80 MEU UDE Siem acacen owas 26 10 38
Silene ....... 21 10 48 Solidago ...... 8 2 25
Stellaria ..... 8 4A 50 Streptanthus .. 21 7 oo

64% 30%

Mean ........ 20.1 11.4 57% 27.6 8.4 30%

This table shows what was to be expected: the genera migrat-
ing from the north have remained in the cooler regions, and
relatively few species have adapted themselves to the lower and
warmer regions of California. Many of those occurring at lower
altitudes are directly derived from or are identical with forms
occurring elsewhere at lower altitudes (Arenaria, Carex, Crepis,
Hieracium, Saliz, Silene, Stellaria). Others occur at lower alti-
tudes only in the cool, moist northwestern part of California.

The endemic genera on the other hand have developed every-
where, and have representatives not only in the mountains, but
also in chaparral, deserts and valleys. Therefore the percentage
of their occurrence in the mountains is lower. However, another
factor enters into the problem, and this is moisture. Most of the
northern genera require a fairly high amount of moisture, at least
in the soil, during the growing season. The species of these
genera which have invaded the lower regions usually occur in
moist places, thus having acquired only the ability to grow at
higher temperatures without having altered their water require-
ments.

In contrast with the origin in the north of about 50 per cent
of the entire alpine flora, when we consider the annuals alone
occurring at an altitude of 2700 meters and higher, 75 per cent
of them belong to endemic genera. Only the species belonging
to circumboreal genera have a distribution reaching beyond Cali-

244 MADRONO [Vol.9

fornia as far north as Washington and Alaska, as the following
list shows. There are more annual species known which reach up
into the alpine zone, but most of these belong to the genera listed
below (Table 2). Some others, like Gnaphalium purpureum and G.
palustre, have a much lower distribution, and under exceptional
conditions are found in the alpine zone.

Taste 2. DIstTRIBUTION OF SPECIES CONSIDERED

Species Distribution
1. Collinsia parviflora Mount Shasta to Mount San Jacinto
2. Collinsia Torreyi Mount Shasta to southern California
3. Cryptanthe Torreyana Mount Shasta to Sequoia National Park
4. Eriogonum spergulinum Sierra Nevada
5. Gayophytum humile Washington to southern California
6. Gayophytum ramosissimum Mount Shasta to southern California
7. Gilia leptalea Northern California to Sequoia National Park
8. Linanthus ciliatus
var. neglectus Southern Sierra Nevada
9. Linanthus Harknessi Idaho to Yosemite National Park
10. Mimulus leptaleus Mount Lassen to Sequoia National Park
11. Mimulus montioides Northwestern Nevada to Sequoia National
Park
12. Mimulus rubellus British Columbia to southern California
13. Nemophila spatulata Western Nevada to southern California
14. Streptanthus tortuosus Mount Shasta to Sequoia National Park
15. Draba stenoloba Alaska to Sequoia National Park
16. Galium bifolium Washington to Yosemite National Park
17. Juncus triformis ~ Washington to southern California
18. Polygonum Kelloggii British Columbia to southern California
19. Polygonum minimum Alaska to Yosemite National Park

In the Swiss Alps six to thirteen annual plants occur above
2500 meters. Since 2500 meters in the Alps corresponds cli-
matically with 8500 meters in the Sierra Nevada,’ where almost
no annuals are found at 3500 meters (only occasionally some
Mimulus species, Sharsmith communication), the population of
alpine annuals in the Sierra Nevada is relatively poor. This is
obviously connected with the limited and unreliable precipitation
during summer, which does not favor the development of annuals.
It is significant, however, that the annuals occurring are predomi-
nantly representatives of endemic Western American species. In
the Olympic Mountains of Washington only one single annual
(Polygonum minimum) is found above timberline. Table 3 gives a
comparison of the number of annuals found at corresponding
altitudes in Europe (taken from Raunkiaer, 1908) and the Sierra

Nevada.

1 Timberline in the Alps is 2200 meters at the highest, but in general it is
around 2000 meters, whereas timberline in the southern Sierra Nevada lies as
high as 3100-3300 meters.

1948 ] WENT: DESERT AND ALPINE FLORA 245

TasLteE 3. COMPARISON OF THE NuMBER OF ANNUALS FounpD AT DIFFERENT
ALTITUDES IN E\UROPE AND THE SIERRA NEVADA

Altitude in d
respect to Poschiavo Tatra San Aosta Sierra
cebecling ps Valley Nevada
+700 m.-— higher 1 3 5
+350 m.—+ 700 m. 8 1 13 6
timberline—+350 m. 22 9
—350m.—timberline 30 28 19

—700 m.—350 m. . 39

From these considerations we can draw an interesting con-
clusion. The climatic response of a genus or even a family is a
physiological character which is extremely tenacious, and can
hardly be changed by evolution. Temperature tolerance, drought
resistance, water requirements all seem to be physiological char-
acters, which are as constant and as unalterable as generic or
family characters, and are not of the type which are usually en-
countered in genetic variability. Thus most alpine annuals in the
southern Sierra Nevada are really desert annuals with a higher
altitudinal distribution. The following list shows how many of
the alpine annuals of endemic genera have close relatives in
desert regions (Table 4). Nemophila is a genus of moist places,
but closely related genera (Ellisia, Phacelia) have many desert
annual species. Gayophytum is an exclusively montane genus, but
with many desert species in related genera (Oenothera, Gaura).

Apart from the relations between desert and alpine thero-
phytes, there are many parallels between perennial plants, shrubs
and trees in desert and alpine habitat. These are all basically
conditioned by the short growing season with limited moisture,
high insolation and large temperature fluctuations. <A partial
list of plants occurring in both habitats follows:

A. Geophytes (bulbous plants): Calochortus, Brodiaea.

B. Phanerophytes (trees). At the extreme range it is in both
conditions conifers which dominate. Pinus monophylla and Juni-
perus are the first trees to appear at the upper range of the desert
steppe, where rainfall becomes slightly greater. Pinus albicaulis,
P. flexilis, Juniperus occidentalis, Tsuga Mertensiana and some other
conifers are the last trees found above 3000 meters.

The Joshua tree (Yucca brevifolia) has no counterpart in the
Sierra Nevada and a large number of moisture-loving shrubs of
the alpine habitat have no relatives or analogs in the desert. But
in any case the number of deciduous trees and shrubs is small in
both habitats. |

C. Hemicryptophytes and Chamaephytes. In both habitats
‘we commonly encounter the rosette habit, which has developed
into the cushion habit under alpine conditions, but is common as

246 MADRONO [Vol. 9

TasBLE 4. ALPINE ANNUALS OF ENDEMIC GENERA HaAvinG CLosE RELATIVES
In Desert REGIONS

Alpine Altitudinal Desert Altitudinal
species range species range

aed Collinsia David-

flora 1500-2500 meters PERC 1000 t
Collinsia Torreyi 1000-3000. “ noes eres
Cryptantha Cryptantha an-

glomeriflora 2000-3000 . gustifolia 0-1500

Eriogonum gracil-
Erigonum sper- limum 0-1500 ss
gulinum 2000-3000 = Eriogonum in-
flatum 0—1500 s
Gilia capillaris 2000-2500 “ Gilia latifolia 0-1500 “

re % Gilia filiformis 500-1500 *
Gaha leptalea Ss Gilia ochroleuca 500-1500“
Linanthus ciliatus 2000-3000 § Linanthus macu-

latus 0—1000 -
Linanthus oblance- Linanthus Par-

olatus 2500-3000 a ryae 500-2000 -
Linanthus Hark- Linanthus Bige-

nessti 1500-3000 lovtt 500-1500 \
Mimulus leptaleus 2000-2500 Mimulus Bigelovit 0—2000 ts
Mimulus monti- :

oides 2000-3500 3 sie Que 500-1000 ‘“
Mimulus rubellus 2000-3000 Ps
Streptanthus Streptanthus in-

tortuosus 2000-3000 =“ flatus 500-1000 “

regular rosettes in the desert. Probably both the cushion and
rosette habit have the same basic evolutionary significance: 1)
low night temperatures counteract stem elongation without in-
terfering too much with organ initiation, 2) during the short
growing season no material is squandered on synthesis of stem
material, 3) it gives snow and grazing protection.

D. Typical of both alpine meadows and deserts is the mass
flowering, in deserts in April and in alpine meadows towards the
end of July. This is probably also associated with the short
growing season, which forces all plants to flower at approximately
the same time, to have a chance to ripen their seed before cold or
drought cuts short further development.

E. Table 5 lists closely related perennial species which occur
in deserts and under alpine conditions. In this list occur a few
plants which are represented by annual species in the desert (e.g.
Calyptridium monandrum) but by perennial species in the alpine
region (C. umbellatum).

If we analyze Table 5 together with the list of alpine annuals
(Table 2) the following general distribution of these genera over

1948 | WENT: DESERT AND ALPINE FLORA 247

the world is found:
Cosmopolitan (occurring on more than two continents) 13

North and South America 8
Northern Hemisphere 4
General North America 2
Western North America only 13

This shows clearly how important the evolutionary pressure
favors the endemic genera, in developing both desert and moun-
tain forms. Actually the percentage of genera with strong
endemism is much greater, since many plants (e.g. Lupinus, Astra-
galus) listed under cosmopolitan or other headings, have a strong
endemic development in Western North America.

In general it seems more common for a desert plant to develop
an alpine relative than for alpine plants to develop forms which

TaslLE 5. GENERA WITH REPRESENTATIVES BOTH IN DeEseRTS AND MOUNTAINS,
ARRANGED ACCORDING TO FAMILIES
Family Alpine species Desert species
Filices Pellaea Bridgesi P. ornithopus
Cheilanthes gracillima C. Covillet
Gymnospermae Pinus contorta P. monophylla
Juniperus occidentalis J. californica
Gramineae Muhlenbergia andina M. Porteri
Sporobolus confusus S. airoides
Oryzopsis Kingit O. hymenoides
Stipa minor S. speciosa
Liliaceae Brodiaea gracilis B. capitata
Calochortus Leichtlinii CO. Kennedyi
Polygonaceae Eriogonum incanum KE. Heermanni
Portulacaceae Calyptridium umbellatum C. monandrum
Cruciferae Streptanthus tortuosus S. inflatus
Leguminosae Lupinus superbus L. rubens
Lotus oblongifolius DL. scoparwus
Astragalus Hookerianus A. tricarinatus
Euphorbiaceae Euphorbia Palmeri E. polycarpa
Loasaceae Mentzelia congesta M. tricuspis
Onagraceae Oenothera subacaulis O. scapoidea
Umbelliferae Cymopterus terebinthinus C. panamintensis
Gentianaceae Swertia albomarginata S. perennis
Hydrophyllaceae Phacelia heterophylla P. calthifolia
Nama Rothrockii N. demissum
Boraginaceae Cryptantha glomerifiora C. racemosa
Scrophulariaceae Penstemon Menziesii P. ambiguus
Castilleia minor C. angustifolia
Compositae Haplopappus eximius H. gracilis
Aster integrifolius A. Orcuttii
Erigeron compositus EH. Parishit
Hemizonia Wheeleri HH. Wrightu
Eriophylium lanatum E. Wallacei
Chaenactis nevadensis C. Fremontii
Artemisia Rothrockii A. spinescens

248 MADRONO | [Vol. 9

survive in the desert. Since in both cases the plants are adapted
to short growing periods, violent temperature fluctuations and low
night temperatures, we might conclude either that genera in
strong evolutionary development (with many endemic species)
have greater inherent adaptability, or that adaptation to frost re-
sistance is more common than adaptation to drought resistance.
Yet the Cactaceae have no alpine representatives; neither have
Agave, Fouquieria, Zygophyllaceae nor other typical desert plants.

It might seem that the plants used by the Carnegie Labora-
tory group at Stanford University disprove the conclusions
reached above. A number of plants with a very wide altitudinal
distribution has been investigated (Clausen, Keck and Hiesey,
1940), some of them belonging to boreal genera (Achillea lanulosa,
Viola purpurea, Aster adscendens, Potentilla glandulosa, P. gracilis).
These plants, however, are exceptions to the rule and were de-
liberately selected among the whole vegetation of California
because of their exceptionally wide altitudinal and latitudinal
distribution. With such exceptions we can say that the great
majority of species has a fairly limited range of distribution.
This is surprising in view of the fact that the same temperatures
occurring in July and August at 3000 meters altitude, occur in
May and October at 1500 meters and in February and November
at sea level. Therefore a summer plant at high altitudes might
grow as a spring or autumn plant at lower altitudes and be sub-
jected to exactly the same temperatures. There are a few plants
which behave in this way, like Erysimum asperum, which flowers
in February at sea level and in July at timberline, but the majority
of plants do not shift their growing season, thus enabling them to
occur at different altitudes. This is due to other climatic factors
such as photoperiod, chilling requirement, and seasonal succession
of temperatures. And this brings us back to the original thesis,
that the physiological responses to climatic factors are only very
little changed in the course of evolution.

SuMMARY

The close relationships of many desert and alpine plants in
California is pointed out. This resemblance is greatest for the
alpine therophytes (annuals) of which 75 per cent have close
relatives in the desert. This is due to several reasons:

1) Climatically the desert and alpine habitats are alike during
the actual growing season.

2) Whereas part of the alpine flora is of circumboreal origin,
at least one-half belongs to genera endemic in Western North
America, which are exactly those which have also developed rep-
resentatives in the desert.

When comparing the altitudinal distribution among the alpine
plants of the cireumboreal genera with those endemic in Western
North America, it was concluded that in general the climatic re-

1948 | MASON: POLEMONIACEAE 249

sponse of a genus or a family is only very little affected by evolu-
tion. This climatic response is due to physiological characters,
such as temperature requirement, frost and drought resistance,
and water requirement.
William G. Kerckhoff Laboratories
of the Biological Sciences,
California Institute of Technology,
Pasadena 4, California.

LITERATURE CITED

Cuavusen, J.. D. D. Keck and W. M. Hirsry. 1940. Experimental studies on
the nature of species. I. Effect of varied environments on Western
North American plants. Carn. Inst. Washington Publ. No. 520: 1-452.

Jepson, W.L. 1925. A manual of the flowering plants of California. Berkeley.
1925.

RaunxisER, C. 1908. The statistics of life-forms as a basis for biological
plant geography. From: Collected papers. Oxford, Clarendon Press,
1934: 111-147,

SOME ADDITIONAL NOTES ON POLEMONIACEAE
Hersert L. Mason

The preparation of the manuscript for the treatment of
Polemoniaceae in Abrams, Illustrated Flora of the Pacific States,
demands that certain points not suitable to develop in that pub-
lication be made clear elsewhere. These are the publications of
certain new species and subspecies and the discussions of reasons
for some of the decisions made where problems seem contro-
versial. Most of the present notes concern the genus Linanthus.
A new species of Collomia is also included.

LINANTHUS ANDROSACEOUS BENTH.

The treatment afforded this species calls for the aggregation
of several entities traditionally regarded as distinct, as subspecies
under L. androsaceous Benth. This move appears imperative
because there seems to be no way to differentiate these subspecies
clearly from one another because they show intergradation of a
type that suggests wholesale introgression. What appears to have
happened is that there developed under the sanction of insular
isolation of late Tertiary time a large number of distinct types
which, when the continent assumed its present form, were per-
mitted to mingle, apparently without effective genetic barriers
between them. The result is the present morphological confusion
in the coast ranges of California and Oregon. Where colonies
have persisted under conditions of isolation they have retained a
certain local uniformity. One such colony occurs in the Sierra
Nevada foothills and necessitates formal description as a new
subspecies.

250 MADRONO [Vol.9

LINANTHUS ANDROSACEUs subsp. laetus subsp. nov. A typico
LL. androsaceo differt corollis albis, papillis testarum in descriptioni-
bus annularibus corrugatis dispositis.

Differs from typical L. androsaceus in the white corolla and
in having the papillae of the seed coat in a definite, corrugated
pattern around the seed.

Sierra Nevada foothills from Butte County south to Amador
County, California. Butte County: Cherokee, altitude 1000 feet,
Mason 12414 (type, Herbarium of the University of California, no.
754224). Eldorado County: near Latrobe, Mason 7007; Greene
Valley, 4 miles north of Shingle Springs, Mason 4516; Pilot Hill,
Mason 7014; Salmon Falls, Crum 1027; Smith Flat, Robbins 995.
Amador County: Cosumnes River opposite Clark Creek, Mason
4482; near Forest Home, Ione to Latrobe highway, Crum 1690.

Linanthus Bakeri sp. nov. Herba annua erecta, 6—25 cm. alta;
internodia tenues et rigida, foliis 8—-7-plo longiore, infra nodis
glanduloso-puberulis, pedicellis glanduloso-puberulis; rami
cymosi, non profusi; folia segmentis linearibus 3—7-partita; in-
florescentia paniculata irregulare, pedicellis tenuibus longis;
calyx profunde in segmentis linearibus fissus, ad apicem puberu-
lus, sinus cum membrana hyalina circa semicompleta, per cap-
sulam accrescentem distentu; corolla anguste hypocrateriformis,
6-10 mm. longa, alba, rosacea vel violacea, quandoque zonae de-
finite, tubus exsertus vel raro inclusus, faucium 1—4-plo longiore
intus cum linea puberuli vel raro glabri, tubus et fauces vulgo
extus puberulo, faucibus angustis, lobis 2-3 mm. longis; stamina
in sinubus corollae affixa vel leviter infra, quam lobi 14 longa,
flamenta glabra, antherae 1-2-plo longiore; stigma exserta,
lobis circa 1 mm. longis; capsula oblongo-cylindrica, semina in
quoque locula plures.

Erect slender annual, 6-25 cm. high; internodes wiry, 3-7
times the leaves; glandular puberulent below the nodes and on
pedicels; branching cymose, not profuse; leaves 3—7 parted into
linear lobes; flowers on long slender pedicels in an irregular
cymose panicle; calyx deeply cleft into linear lobes, these puberu-
lent above toward the tips, sinuses about half filled with a narrow
hyaline membrane which becomes distended by the growing cap-
sule; corolla slender funnelform, 6-10 mm. long, white, pink,
lilac or violet, sometimes with a definite zoning, tube usually
exserted, rarely included, 1—4 times the throat, with a narrow
hairy band within, rarely glabrous, tube and throat usually
puberulent exteriorly, throat narrow, lobe 2-3 mm. long; stamens
inserted in the sinuses of corolla lobes or just below, 14 as long
as corolla lobes, filaments glabrous, 1-2 times the anthers; stigma
exserted from orifice of throat, lobes about 1 mm. long; capsule
oblong cylindric, locules several seeded. (Gilia Bolanderi of Brand
in Engler Pflanzenreich 47°°: 1384. 1907, as to specimens cited.

Non Gray.)

1948 | MASON: POLEMONIACEAE 251

Fresno County and Mount Diablo, north in Coast Ranges and
Sierra Nevada, California, to Klickitat County, Washington.
Cauirornia. Pilot Hill, Eldorado County, H. L. Mason 7015 (type,
Herbarium of the University of California, no. 754226) ; serpen-
tine, Stonyford, Colusa County, Mason 12404, 12405; serpentine
Stonyford, Colusa County, Mason 12404, 1205; serptine Stonyford,
hills, Mount Bullion, Mariposa County, Mason 11759; serpentine
outcrops, Scott Valley, Siskiyou County, Horn 45; Los Molinos,
Tehama County, Wohletz 15; Watts Valley, Fresno County,
Hoover; Mendocino Pass, Glenn County, Howell 19777; Moon-
springs, Bald Mountain, Lassen National Forest, June 8, 1928,
Swift. Oregon. Sterile ground, Tumolo, Deschutes County, Peck
19750; 3 miles above Cave City, May 15, 1935, Adams. ..

LiInaNTHUS BIcoLoR subsp. minimus subsp. nov. A typico L.
bicolor differt corollis minutis albis vel sordidis.

Corolla minute, about 1 cm. long, white to sordid.

Coastal area from Bodega Head, California, north to Puget
Sound, Washington. Gages Point, Skagit County, Washington,
May 8, 1927, Roush (type, Herbarium of the University of Cali-
fornia, no. 709722). Catirornia. Gasquet, Del Norte County,
Tracy 12361; Hoopa Mountain, Humboldt County, Tracy 12576;
Bodega Bay, Sonoma County, Baker.

Linantuus Harxnessiu subsp. condensatus subsp. nov. Caulis
pumilis ramosissimus; corolla calycis longiore; stamina _ sub-
sessila, faucium ad medium inserta.

Low, densely branched; corolla exceeding calyx; stamens
subsessile, inserted midway on throat.

Known only from the type locality. Plaskett Meadows, Glenn
County, California, Baker 10593 (type, Herbarium of the Uni-
versity of California, no. 754225), Howell 19827.

Linanthus Killipii sp. nov. Herba annua erecta, internodiis
infra vulgo congestis supra rigidus, foliis 1-8-plo longiore, ad
nodem puberulis vel aliquando floccosis; cotyledones sessiles
ovatae anguste perfoliatae; folia palmatim in 5—7 segmentis
linearibus 8-10 mm. longis incisa, supra cum capilli albi subtus
puberula vel glabra; inflorescentia cymis congestis quisque 3-—7-
flores; floribus sessilibus; sepalis ad margo conspicue membra-
naceis praeter ad apicem, membranis infra connatis post anthesin
accrescens; corolla 10-15 mm. longa, anguste hypocrateriformis,
tubus 4—5 mm. longus, robustus, faucium subaequans vel vix longi-
ore a calyce inclusus vel vix exsertus, fauces angustae subcy-
lindricae, lobis rhomboideis denticulatis vel ad apicem integeribus,
basim cum macula linearis; stamina faucium, afliixa, filamenta
glabra faucium aequans; stylus antherae longiore, lobis circa 1
mm. longis; semina in quoque locula plures ellipsoidea, sub aqua
inmutata.

Erect annual, branching usually well above the base, the basal

252 MADRONO [Vol. 9

internodes often congested, stems of the upper internodes wiry,
1-8 times the leaves, puberulent to somewhat floccose at the
nodes; cotyledons sessile, ovate, narrowly perfoliate; leaves pal-
mately cleft into 5—7 linear segments, 3-10 mm. long, puberulent
to glabrate below, hairy above with weak white hairs, the lower
somewhat perfoliate ; inflorescence of congested cymes at the ends
of the branches, each 3—7 flowered; flowers sessile; calyx lobes
conspicuously membrane-margined except at the tips, the mem-
branes united below to form the calyx tube and expanding with
the growing capsule; corolla 10-15 mm. long narrowly funnel-
form, the tube 4-5 mm. long, stout, from subequal to slightly
longer than the throat and included or barely exserted from the
calyx, throat narrow, subcylindric, that is forming a narrow
angle; lobes somewhat rhombic, denticulate or entire at apex
and with a linear spot near the base; stamens inserted on the
throat near the junction with the tube, filaments glabrous, equal-
ling the throat and with the anthers disposed in its orifice; style
slightly exceeding the anthers, the stigma lobes about 1 mm. long;
capsule locules several seeded, the valves adhering at the base;
seeds ellipsoid, reddish brown, unaffected by wetting.

Upper desert slopes of the San Bernardino Mountains, Cali-
fornia. Cactus Flat, altitude 5900 feet. San Bernardino Moun-
tains, June 13, 1941, Killip 36343 (type, United States National
Herbarium 1,828,544); Baldwin Lake, San Bernardino Moun-
tains, Peirson 6748.

LINANTHUS NUDATUS GREENE VERSUS L. NASHIANUS JEPSON

In treating this species I have accepted the name L. nudatus
given it by Greene and rejected L. Nashianus of Jepson. The
bracts subtending the inflorescence of this species are unlike
anything found elsewhere in the genus. The lobes are joined to
one another by a scarious membrane. In view of the very dis-
tinctive nature of the bracts and in view of the fact that Brand
illustrated it and saw and cited a specimen labeled “Lake Co.”
seems conclusive proof that such a specimen existed even though
we are unable to locate it at present nor has it been since collected
in Lake County. Greene’s description fits this species except that
he makes no mention of the scarious membrane of the bract.
However, the combination of hispidulose-ciliate lower leaves with
villous-ciliate bracts and hirsute-ciliate calyx lobes occurs no-
where else in the genus so faras amaware. The nearest in this
respect are L. ciliatus and L. montanus, both of which were well
known to Greene. However, these are hispid-ciliate throughout
and neither is either hirsute- or villous-ciliate. Judging from its
occurrence in the southern Sierra Nevada it seems improbable but
not impossible that it occurs in Lake County. There seems no
good reason, however, for rejecting the name L. nudatus simply

1948] MASON: POLEMONIACEAE 253

because of an apparent error in locality on the label and especially
since Brand’s illustration makes it amply clear what Greene had
in mind when he named L. nudatus.

Linantuus DactyLtopHytituM (Torr.) Ryps. versus
L. pemissus (Gray) GREENE

Torrey, when calling attention to a plant collected by the
Ives Expedition referred to it as being scarcely sufficient for
description but he gave a very brief description and listed it as
follows: “GILIA DACTYLOPHYLLUM, (n. sp.?).” Since the word
“Dactylophyllum” is the name of a section of the genus, Gray
interpreted Torrey as merely indicating the section to which
it belonged, a current practice of the time. He then proceeded to
name and describe the species based upon more adequate ma-
terial. Gray’s position is further supported by the question mark
following the letters “n. sp.”’ which to him meant that Torrey
was uncertain about it being new. This led Gray in citing Tor-
rey s name to put a comma between the generic name and the
word Dactylophyllum. The comma points clearly to the reasons
for Gray’s action but unfortunately Torrey did not use the
comma. The only clues to Torrey’s action are the question mark,
and the statement as to the inadequacy of the specimen. I am
inclined to agree with Gray in his interpretation of the matter. I
think that Torrey was meaning only to call attention to the section
of the genus to which the plant belonged. I raise the following
further arguments to support my contention. 1) The question
mark placed where it is, indicates Torrey’s uncertainty about the
plant being a new species and it is my personal opinion that a
question mark associated in any way with a new name should by
international agreement be construed to indicate a tentative name.
2) A species named ‘dactylophyllum” would naturally be sup-
posed to be the type species of the section Dactylophyllum which
this is not. This constitutes a source of confusion. 8) Torrey
and Gray were co-authors of the Ives report and it is reasonable
to suppose that each was aware of the work and motives of the
other. I therefore accept the name Linanthus demissus (Gray)
Greene.

GILIA TULARENSIS BRAND VERSUS LINANTHUS
OBLANCEOLATUS EASTWOOD EX BRAND

In the year 1904 Culbertson collected specimens of a Linan-
thus in Hockett Meadows, Tulare County, California, which he
sent to C, F. Baker. Baker, in turn sent them to Alice Eastwood
for identification. She named them Linanthus oblanceolatus n. sp.
and this name was appended to the specimens which were dis-
tributed by Baker. The name was used as well in Baker’s
published list, but without description. Linanthus oblanceolatus
Eastwood then is a nomen nudum. I wish to emphasize the point

254 MADRONO [Vol. 9

that the name L. oblanceolatus Eastwood was based upon the Cul-
bertson specimen. ,

Brand, in 1907, borrowed the Polemoniaceae material from
the California Academy of Sciences and found there, in addition
to the Culbertson collection, two collections by Hall and Babcock.
He gave to one Hall and Babcock collection (no. 5554) the name
Gilia oblanceolatus, credited the specific name to Eastwood and
cited the name Linanthus oblanceolatus in synonymy. The Hall and
Babcock specimen was used as the type. Then Brand erected
the variety Culbertsoni, based upon the Culbertson specimen.
To the other Hall and Babcock specimen (no. 5211) which had a
slightly longer corolla he gave the name Gilia tularensis.

In handling this problem we can disregard at the outset the
name Linanthus oblanceolatus Eastwood as applying to the Culbert-
son specimen because it was never validly published. The name
Linanthus oblanceolatus Eastwood ex Brand as applying to the first
Hall and Babcock specimen was first published in synonymy with
a literature citation that involved the Culbertson specimens. This
I construe as invalidating the appellation oblanceolatus for use in
the genus Linanthus and we must turn to the name Gilia tularensis
for a name for these three collections which I regard to represent
the same entity. The new combination and synonymy for
Linanthus tularensis follow:

Linanthus tularensis (Brand) comb. nov. Giulia tularensis
Brand in Engler, Pflanzenreich 4°°°: 186. 1907. Giulia oblanceolata
(Eastwood) Brand, l.c. Linanthus oblanceolatus Eastwood ex
Brand in synonymy, l.c. Gilia oblanceolata var. Culbertsoni Brand

(spelled Cubbertsonz) 1.c.

Collomia Tracyi sp. nov. C. tinctoria subvar. luxuriosa Brand,
in Fedde, Rep. Spec. Nov. 17: 817. 1921. C. tinctoria f. lururiosa
(Brand) Wherry, Am. Midland Nat. 81: 227. 1944.

Herba annua erecta vel expansa, 5—20 em. alta; caules bifurcati
glandulosi; folia linearia vel lanceolata acuminata petiolata vel
subsessilia 2-6 cm. longa; flores 2—5, terminali vel in axillis foli-
orum vel in furcis ramorum; lobi calycis lanceolati attenuati minute
glandulosi; corolla 15-25 mm. longa, calycem 3-plo longiore;
stamina valde inaequaliter inserta, infimo in tubo saepe subsessili,
summis in faucibus, filamentis longis glabris; stigma inclusa,
semina solitaria in loculis.

Erect or spreading annual, 5-20 em. high; stems forked, glan-
dular; leaves linear to lanceolate, tapered at both ends, petioled
or subsessile, 2-6 cm. long, those in the inflorescence barely ex-
ceeding the flowers, flowers in clusters of 2—5, terminal on the
branches or in the axils of the leaves and forks of the branches,
clusters subtended by few leafy bracts; calyx lobes lanceolate
attenuate, minutely glandular; corolla 15-25 mm. long, 3 times
the calyx, subequal to slightly exceeding the leaves of inflores-

1948 | CONSTANCE: PHACELIA 255

cence, limb about 1 cm. broad, white to pink, tube sometimes
purple; stamens very unequally inserted, lowermost well down the
corolla tube and often subsessile, the upper on the throat and with
long glabrous filaments; stigma included, capsule obovoid, seeds
solitary in the locules.

Mountains in the drainage basin of the Van Duzen, Mad, and
Klamath rivers of Humboldt and Trinity counties, California,
1000 to 6800 feet. Trinity County: Three Forks of Mad River,
Tracy 10220 (type, Herbarium of the University of California,
no. 754223); head of White’s Creek, Devil’s Canyon Mountains,
Tracy 14606; Mary Blaine Mountain, Tracy 14466; Upper Mad
River, June 26, 18938, Blankinship. Humboldt County: Grouse
Mountain, Tracy 16420, 16670; South Fork Mountain, Tracy 9046;
Horse Mountain, Tracy 8161; Van Duzen River at Dinsmore’s,
Tracy 16373; Trinity Summit, T'racy 10468; Van Duzen River near
Carlotta, Baker 102; Van Duzen River Valley opposite Buck
Mountain, Tracy 2719, 2720; northwest slope of Buck Mountain,
Tracy 2837; Klamath River, Chandler 1475; Hoopa Mountain, Davy
and Blasdale 56765.

Department of Botany
University of California, Berkeley

A NEW SPECIES OF PHACELLA FROM SONORA, MEXICO

Lincotn CONSTANCE

The only species of Phacelia listed by Gentry in his admirable
study (1942, p. 219) of the flora and vegetation of the Rio Mayo
area of southern Sonora was given as “Phacelia cf. congesta Hook.”’
He characterized the plant as a spring-blooming winter annual,
scattered and infrequent on wooded slopes in the Short-tree
Forest at elevations of 800 to 2000 feet. Since Gentry’s collec-
tions are widely distributed, I have frequently met with specimens
of this entity, which I have been guilty of casually annotating as
“P. aff. distans Benth.” Now that several sheets of this plant
have recently been sent me for verification, I have found it neces-
sary to make a more serious study of it, and have concluded that
it is undescribed.

Phacelia Gentryi, sp. nov. Planta annua, basi ramosa, ramis
diffusis, 3-6 dm. longis hirsutis hirsutulisque vel hirtellis, inflores-
centia stipitato-glandulosa; folia oblongo-ovala ovalave, 3—7 cm.
longa, 2.5 cm. lata, pinnata vel pinnatifida, foliolis crenulatis
breviter dentatisve; inflorescentia scorpioidea, cymis solitariis vel
geminatis, 10—20-floribus; pedicelli maturi adscendentes, 0.5—1.5
mm. longi; calycis lobae lineari-oblanceolatae, 3-5 mm. longae,
0.3-0.8 mm. latae, obtusae, plerumque subaequales, dense hirsu-
tae; corolla pallide coerulea, lati-campanulata, 5-7 mm. longa

256 MADRONO [Vol. 9

lataque, pilosa, lobis obovatis, obscure crenulatis; appendiculae
semiovatae, lamellis ca. 1 mm. longis, partibus transversis appen-
diculorum prominentibus; stamina corollae subaequalia, ca. 4 mm.
longa, antheris ovalibus 0.5-0.7 mm. longis, filamentis glabris;
stylus anthesi corollam subequans, maturitate ca. 5 mm. longus,
ad 34 longitudinis partitus, sub medio hirsutulus, ovario hirsuto;
ovula 2 ad quamque placentam; capsula matura globosa, 1.5—2
mm. longa; semina plerumque 4, oblonga, ca. 1.5 mm. longa,
brunnea, alveolata.

Low spreading annual, the branches diffuse, 3-6 dm. long,
hirsute with stiff scattered spreading or reflexed hairs and hirsu-
tulous or hirtellous, the inflorescence glandular with stipitate capi-
tate glands; leaves thin, oblong-oval to oval, 3—7 cm. long, 2.5—5 em.
broad, pinnate or pinnatifid, the divisions oblong to oval, 1-3 cm.
long, 0.5—1 em. board, crenulate to shallowly dentate, sparsely stri-
gose or strigulose; inflorescence scorpioid, of simple or geminate
terminal and axillary 10-20 flowered cymes, the mature pedicels
ascending, 0.5—-1.5 mm. long; calyx lobes linear-oblanceolate, 3—5
mm. long, 0.8—-0.8 mm. broad, obtuse, usually subequal, densely
spreading-hirsute and often glandular at base; corolla pale blue,
broadly campanulate, 5—7 mm. long and broad, the lobes obovate,
obscurely crenulate, pilose on the back, the appendages broad,
semi-ovate, wholly attached on the side away from the filament,
forming a V-shaped pocket at the base of each filament, the
lamella about 1 mm. high, the transverse part prominent; stamens
about equalling the corolla, ca. 5 mm. long, the anthers oval,
0.5—-0.7 mm. long, the filaments glabrous, or nearly so; style
included in flower to slightly exserted, when mature ca. 5 mm.
long, parted 34 of its length, hirsutulous below the middle, the
ovary densely hirsute; ovules 2 to each placenta; mature capsule
globose, 1.5—2 mm. long; seeds usually 4, oblong, ca. 1.5 mm.
long, brown, alveolate.

Type. San Bernardo, Rio Mayo, Sonora, Mexico, 26 Feb-
ruary 1935, H. S. Gentry 1364 (GH, type; MO). [The symbols
used for herbaria are those listed by Lanjouw (1939)]. Other
specimens examined. Sonora... Alamos, 28 January 1899, E. A.
Goldman 806 (US), 17 March 1910, Rose, Standley 5 Russell 13,014
(US), San Bernardo, 12 Feburary 1935, Gentry 1304 (MEXU,
MO), 26 February 1935, 1364.

This species is a member of that portion of section Euphacelia
revised by Voss (1935) under the title of “the Phacelia hispida
group. Voss did not delimit the group in any way, but he in-
cluded in it P. cicutaria Greene [P. hispida A. Gray, non Buckl.],
P. cryptantha Greene, P. umbrosa Greene, and P. vallis-mortae Voss ;
these species all occur substantially to the north of the range of
P. Gentryi, according to his map. In Voss’s key, the new entity
would lead (with difficulty) to P. eryptantha, from which it differs
in its markedly glandular inflorescence, much shorter pedicels

1948 | RATTENBURY: CHROMOSOME NUMBERS © 257

and calyx lobes, broader and differently colored corolla, and
differently shaped appendages. It may be distinguished with
equal ease from P. umbrosa, the third small-flowered member of
this alliance, by its glandular inflorescence, shorter and obtuse
calyx lobes, broader and differently colored corolla, longer sta-
mens, and much smaller seeds.

It is a pleasure to name this species for Dr. Howard Scott
Gentry, now of the Alan Hancock Foundation of the University
of Southern California, whose extensive explorations and pub-
lished accounts have added so much to our knowledge of the

fascinating flora of northern Mexico.
Gray Herbarium,
Harvard University
LirerRATURE CITED

Voss, J. 1935. A revisional study of the Phacelia hispida group. Bull. South.
Calif. Acad. Sci. 33: 169-178.

Lansgouw. 1939. On the standardization of herbarium abbreviations. Chron.
Bot. 5: 142-150.

Gentry, H. S. 1942. Rio Mayo plants, a study of the flora and vegetation
of the valley of the Rio Mayo. Carnegie Inst. Wash. publ. 527.

CHROMOSOME NUMBER PUBLICATION

J. A. RATTENBURY

It is planned to publish periodically in MaproNo lists of chromo-
some numbers of plants comprising species whose chromosome numbers
have not appeared to date in other publications, or are at variance with
previously given figures. An effort is being made to present these
data in a manner conformable both to usefulness and economy of space.

The conscientious taxonomist feels that the identification of the
species listed in tabulations such as this is in part the interpretation of
the author, and with changing nomenclature is subject to revision.
The proposal is, therefore, to restrict publication of chromosome num-
bers to those collections which are documented by reliable vouchers in
the form of herbarium sheets filed in one or preferably more per-
manent herbaria. It is further recommended that permanent cytolog-
ical preparations be preserved, either attached to the herbarium
sheet or in some other easily accessible form, so that critical counts may
be confirmed by interested researchers. Camera lucida drawings from
cells in marked regions of the permanent slides may also be attached
to the sheets. The desirability of making permanent documentation
of the results of research cannot be too strongly stressed.

It is hoped that botanists and geneticists will contribute chromosome
counts from time to time. The data should include as much as possible
of the information shown in the accompanying table. If the response
is sufficiently great, an attempt will be made to group related species,
genera and families into the same issue. Undocumented counts will

not be published.

258 MADRONO [Vol. 9

The chromosome number listed will be that typical for the tissue
from which the count was taken. For example: 2n = 9; (metaphase
I); 2n=18 (somatic cell); n=9 (meiosis II and pollen grain divi-
sion). The method of making slides permanent described by Bradley
[ Stain Tech. 23(1): 41-44. 1948] is recommended for the preservation
of cytological evidence.

Contributions to this section of MaproNo may be addressed to the
editor.

DOCUMENTED CHROMOSOME NUMBERS OF PLANTS

SPECIES NUMBER CouNTED BY CoLLECTION LocaLity
CRUCIFERAE
Cakile 2n = 91 A. R. Krucke- | Kruckeberg Jenner,
edentula berg, Univ. 1656 UC} Sonoma Co.,
(Bigel.) Hook. Calif. Berkeley Calif.
LILIACEAE
Disporum 2n = 91 J. A. Ratten- | Rattenbury Contra Costa
* Hookeri bury, Univ. 96, 104, 111 UC) Co., ‘Calif.
(Torr.) Britt. Calif. Berkeley
Smilacina
*sesstlifolia 2n=181 | M. S. Cave, Rattenbury Contra Costa
Nutt. Univ. Calif. 94 UC Co., Calif.
Berkeley
*amplexicaulis 2n=18n | J. A. Ratten- | Rattenbury Contra Costa
Nutt. bury, Univ. 99,1172 WC Co., Calif.
Calif. Berkeley

* Prepared slide available. :
1 Symbols used for herbaria are those listed by Lanjouw, Chronica Botanica
5: 142-150. 1939.

Department of Botany
University of California, Berkeley.

1948]

INDEX

259

INDEX TO VOLUME Ix

For classified items see: Biographical articles, Chromosome numbers, Re-

views.

Actinocyclus, 65

Aegilops Hystrix, 125

Agropyron, 122: arenicola, 127; ari-
zonicum, 125; dasystachum, 127;
Gmelini var. Pringlei, 125; jun-
ceum, 126; laeve, 126; latiglume,
200; Parishii, 126, var. laeve, 126;
pauciflorum, 126; Pringlei, 125;
repens, 127; riparium, 127; Saun-
dersii, 125; saxicola, 125; Smithii,
127, var. riparium, 127; spicatum,
125, var. arizonicum, 125; subse-
cundum, 126; tenerum, 126

Algae: Generic names of, proposed
for conservation. I, 8; Notes on
Pacific Coast marine, 155

Aliciella triodon, 214

Allium: anceps var. aberrans, 238;
Brandegei, 234; Cusickii, 237;
Douglasii, 233, var. Tolmiei, 237;
idahoense, 237; parvum, 236; pla-
typhyllum, 238; pleianthum, 237;
Tolmiei, 237; Tolmiei Baker, The
identity and delimitation of, 233;
tribracteatum, 234

Alpine flora in California, Some paral-
lels between desert and, 241

Alpine tundra, 143

Amelia, 65

Amplisiphonia pacifica, 155

Anemone: globosa, 6; parviflora, 5

Anisocladella pacifica, 158

Annulina, 8

Antennaria pulvinata subsp. albescens,
8

Aphelandra: Goodspeedii, 154, pl.
153; ornata, 154; Seibertii, 154

Aquilegia formosa var. flavescens, 6

Arabis hirsuta var. pycnocarpa, 6

Aragalus viscidula, 7

Arenaria: propinqua, 5; Rossii, 5

Asperella: 122; californica, 127

Astragalus: alpinus, 6; impensus, 6;
Purshii, 6

Atelophragma alpiniformis, 6

Auricularia tremelloides, 9

Baker, M. S., A new violet from
Mexico, 131

Barkley, F. A. and P. C. Standley:
Noteworthy South American
plants, I and IT, 149

Batrachospermum, 8

Beloperone: cochabambense, 154; Sou-
kupii, 152, pl. 151

Betula glandulosa, 5

New scientific names are printed in bold-face type.

Binghamiella Forkii, 158
Biographical articles: Jepson, W. L.,
61, 223; Komarov, V. L., 57
Blennosperma: A new species of, from
California, 103; Bakeri, 103, fig.,
103; californicum, 104, fig., 103;
chilense, 104; nanum, 104
Botrychium virginianum, 3

Cakile edentula, 258

Calamagrostis neglecta, 4

California: Elymus aristatus in, 232;
A new species of Blennosperma
from, 103

Callithamnion Pikeanum var. pacifi-
cum, 158

Carex: capillaris, 4, var. elongata, 4;
concinna, 4; disperma, 4; gyno-
crates, 4; leptalea, 4; pseudo-
scirpoidea, 4; Vahlii, 4

Castilleja wallowensis, 7

Catenella, 12: Opuntia, 12; repens, 12

Ceramium: gracillimum, 158; procum-
bens, 158; sinicola var. inter-
ruptum, 158; transversale, 158

Ceramiaceae, 14

Champia, 10

Champiaceae, 13, 14

Chantransia, 8: nigricans, 9; nodosa,
9; vesicata, 9

Cheirinia desertorum, 64

Chimaphila, 65, 67, 97: japonica, 67,
98; maculata, 67, 70, 98; Men-
ziesii, 67, 70, 98, figs. 69, 73, 75, 79,
83, 85, 93; umbellata, 67, 70, 98,
figs. 83, 85, 93

Chlorophyceae, 155

Chlorophycophyta, 8

Chondria, 10, 11, 15: tenuissima, 15

Chordaria: divaricata, 10; flagellifor-
mis, 10

Choreocolax Polysiphoniae, 159

Choreonema Thureti, 157

Chromosome numbers: Blennosperma
Bakeri, 104, californicum, 104;
Cakile edentula, 258; Disporum
Hookeri, 258; Smilacina amplexi-
caulis, 258, sessilifolia, 258

Chromosome number publication, 257

Chylocladia, 10, 11: kaliformis, 11;
verticillata, 11

Cladophora, 8: glomerata, 8

Cladophoraceae, 8

Clavaria, 12

Clavatula, 12

260

Cleome: Eyerdamii,
psoraleaefolia, 150

Clinelymus, 122: glaucus, 126

Coahuila: Mexico, Vegetation and cli-
mate of, 33; Vegetation of, pls.,
45, 47, 49, 51, 53

Codium fragile, 156

Collinsia californica, 13

Collomia, 207: glutinosa, 207; tinctoria
f. luxuriosa, 254; Tracyi, 254

Colorado, Range extensions of grasses
into, 199

Condalia: Two new varieties of, from
Texas, 128; lycioides, 130; obovata
var. edwardsiana, 128; obtusi-
folia, 130; viridis var. Reedii, 129

Conferva, 10; albida, 8; compacta, 8;
glomerata, 8; nigricans, 8; nodosa,
9; rivularis, 8; rupestris, 8; sor-
dida, 8

Constance, Lincoln, A new species of
Phacelia from Sonora, 255

Copeland, H. F.: Observations on the
structure and classification of the
Pyroleae, 65

Corallorrhiza trifida, 5

Cornea, 12: capillacea, 12; deformis,
12; filicina, 12; pusilla, 12; sericea,
12; spinosa, 12

Cory, V. L.: Old world plants appar-
ently recently introduced into
Texas, 64; Two new varieties of
Condalia from Texas, 128

Cruoria pacifica, 157

Cryptantha celosioides, 7

Cystoseira osmundacea, 157

150, pl, 151;

Dasyopsis densa, 159

Dasyphila, 11, 14: Preisii, 14

Dasyphylla, 13, 14, 15: articulata, 13;
ovalis, 14; sedoides, 14; tenuis-
sima, 15; Woodwardii, 15

Davidson, J F.: A new Polemonium
from Mexico, 187; The polygonal
graph for simultaneous portrayal
of several variables in population
analysis, 105; The present status
of the genus Polemoniella Heller,
58

Desert and alpine flora in California,
Some parallels between, 241

Desmarestia, 13

Detling, L. E.: Concentration of en-
vironmental extremes as the basis
for vegetation areas, 169; En-
vironmental extremes and endem-
ism, 137

Diacalpe manchuriensis, 191

Dictyopteris Johnstonei, 156

Dictyotaceae, 9

Dilsea, 13

MADRONO

[Vol. 9

Disporum Hookeri, 258

Dodecatheon alpinum, 7

Dryas Drummondii, 6

Draba: nivalis var. elongata, 6; prae-
alta, 6

Ectocarpus: chantransioides, 156; cy-
lindricus, 156, f. codiophilus, 156
Elymus 122: Nomenclatorial changes
in, with a key to the Californian
species, 120; arenicolus, 127; ari-
status, 124, 126, in California, 232;
arizonicus, 123, 125; californicus,
122, 127; caput-medusae, 122, 125;
cinereus, 124, 127; condensatus,
124, 127, f. pubens, 127; dasy-
stachys, 127; elymoides, 123, 125;
glaueus, 123, 126; aristatus, 126, f.
Jepsonii, 126, subsp. Jepsonil, 122,
126, var. Jepsonii, 126, subsp. vi-
rescens, 124, 126; Hansenii, 123,
124, 125; Hystrix, 125; jubatus,
125; junceus, 126; Macounii, 123,
126; mollis, 124, 126; multinodus,
124, 126; multisetus, 123, 125;
Orcuttianus, 127; pacificus, 125,
127; Parishii, 126; pauciflorus,
124, 126, subsp. laeve, 123,
126, subsp. subsecundus, 123,
126; Pringlei, 125; repens, 124,
127; riparius, 124, 127; Saun-
_ dersii, 123, 125, var. californicus,
125; saxicolus, 123, 125; sibiricus,
122; sierrus, 123, 125; simplex,
127; Sitanion, 125; Smithii, 125,
127; spicatus, 123, 125; Steb-
binsii, 123, 124, 126; subvillosus,
124, 127; triticoides, 124, 125, 127,
subsp. multiflorus, 124, 127; van-

couverensis, 124, 126; virescens,
126
Endemism, Environmental extremes

and, 137

Environmental extremes: and endem-
ism, 1387; Concentration of, as the
basis for vegetation areas, 169

Equisetum variegatum, 3

Eriogonum Kingii, 5

Erxlebenia, 65, 68, 94, 96

Erysimum repandum, 64

Erythrocladia: irregularis, 156; sub-
integra, 156

Erythrotrichia pulvinata, 156

Fauchea media, 158

Fertilization and maturation of the
gametes in Nicotiana, 110

Festuca spicata, 125

Fucus, 10: Asplenioides, 14; cordatus,
13; edulis, 13; kaliformis, 11;
Nemalion, 12; pectinatus, 14;
plumosus, 14; verticillatus, 11

1948]

\Cametes, Maturation of the, and fer-
tilization in Nicotiana, 110
Gardneriella tubifera, 157
Gastroclonium, 10, 11, 14; ovale, 11;
ovatum, 11, 14
Gelidiaceae, 12
Gelidium, 11, 12: corneum, 12
Generic names of algae proposed for
conservation. I, 8
Gentiana interrupta, 7
Gentry, H S.: The genus Mimulus in
or adjacent to Sinaloa, Mexico, 21
Germination of Phacelia seeds, 17
Gigartina papillata, 157
Gigartinaceae, 12
Gilia, 203: Some problems in the
genus, 201; subgen. Campanula-
strum, 219; subgenus Eugilia, leaf
types, figs. 211; subgen. Gil-
mania, 205; subgen. Greenian-
thus, 206; subgen. Kelloggia, 219;
subgen. Tintinabulum, 220;
Abramsii, 216, subsp. integri-
folia, 216; abrotanifolia, 208;
achilleaefolia, 208, subsp. chamiss-
onis, 208; subsp. staminea, 208;
aggregata, 206; arenaria, 217, var.
Abramsii, 216; var. rubella, 214;
Bolanderi, 250; campanulata, 219;
capillaris, 219; capitata, 208;
caruifolia, 213; chamissonis, 208;
congesta, 206; depressa, 206; dif-
fusa, 209; divaricata var. vol-
canica, 207; filiformis, 220, 246;
gilioides, 206, subsp. glutinosa,
207, subsp. Volcanica, 207; Gil-
mani, 205; Grinnellii, 213; Hutch-
insifolia, 214; inconspicua, 212;
dentiflora, 214, subsp, sinuata var.
oreophila subvar. diffusa, 209; in-
yoensis, 219; latiflora, 218, subsp.
cana, 218; var. cana, 218, subsp.
exilis, 219, var. exilis, 219, subsp.
leptantha, 219, subsp. Purpusii,
218, subsp. speciosa, 218, subsp.
triceps, 219; latifolia, 205, 246;
leptalea, 219, subsp. bicolor, 290,
subsp. pinnatisecta, 220; leptan-
tha, 219; leptomeria, 214, 216,
subsp. micromeria, 214, var. myri-
acantha, 214, subsp. rubella, 214,
' var. tridentata, 214; micromeria,
214; millefoliata, 209; minutiflora,
219; multicaulis, 208; subsp. eu-
multicaulis, 208, var. millefolia,
209, subsp. millefoliata, 209, subsp.
Nevinii, 209, subsp. peduncularis,
209; Nevinii, 209; oblanceolatus,
254, var. Culbertsoni, 254; ochro-
leuca, 214, 216, 246, subsp. trans-
montana, 215, subsp. typica, 215;

INDEX / 261

peduncularis, 209; polycladon,
206; Ripleyi, 205; rubra, 206;
scopulorum, 213; sinuata, 215, 216;
splendens, 210, subsp. australis,
213, subsp. Grinnellii, 213; stam-
inea, 208; stellata, 213; stricta,
208; tenerrima, 220; tenuiflora,
217, var. altissima, 212, subsp. in-
terior, 217, var. Newloniana, 213,
var. Purpusii, 218, var. speciosa,
218, var. triceps, 219; Traskeae,
207; tricolor, 209, subsp. diffusa,
209, var. longipedicellata, 209;
triodon, 214; tularensis, 253, 254

Glyceria Otisii, 3

Goniotrichum cornu-cervi, 156

Goodspeed, T. H.: Maturation of the
gametes and fertilization in Nico-
tiana, 110

Gould, F. W.: Elymus aristatus in
California, 232; Nomenclatorial
changes in Elymus with a key to
the Californian species, 120

Grant, A. D. and H. L. Mason: Some
problems in the genus Gilia, 201

Grasses, Range extensions of, into
Colorado, 199

Grateloupia abbreviata, 157

Gymnosorus, 10

Gymnostichum californicum, 127

Gymnothamnion elegans, 158

Habenaria obtusata, 5

Harrington, H. D.: Range extensions
of grasses into Colorado, 199

Hedysarum boreale, 7

Heiser, C. B., Jr.: A new species of
Blennosperma from _ California,
103; Notes on the genus Town-
sendia in Western North America,
238

Helianthella: The genus, in Oregon,
186: californica var. nevadensis,
186, 187; Douglasii, 186; nevaden-
sis, 186; quinquenervis, 186, 187;
uniflora, 186, 187, var. Douglasii,
186, 187

Helianthus: quinquenervis, 186; uni-
florus, 186

Helminthocladiaceae, 11

Helminthora, 11, 12: divaricata, 11

Herposiphonia: secunda, 160; tenella,
159

Hollenberg, G. J.: Notes on Pacific
Coast marine algae, 155

Hypnea californica, 157

Hypopitys, 95

Hystrix, 122: californica, 127

Iridaea, 11, 12: cordata, 13

262

Tridea, 12

Iridophycus, 13

Isenberg, I. H., Location of extran-
eous materials in redwood, 25

Jepson, Willis Linn: biographical
sketch of, 61, pl., 63; The place
of, in California botany, 223

Juncus Regelii, 4

Kaliformia, 11, 12: diaphana, 11;

Opuntia, 12; pusilla, 12; verti-
cillata, 11
Kaliformis, 12, 13, 15: articulatus,

13; dasyphyllus, 15; obtusus, 15;
Opuntia, 12; verticillatus, 11
Kearney, T. H., Review: The evolu-
tion of Gossypium and the dif-
ferentiation of the cultivated cot-
tons, 228

Keck, D. D.: The place of Willis Linn
Jepson in California botany, 223

Keck, D. D., Review: The New World
Cypresses, 229

Kobresia simpliciuscula, 4

Komarov, Vladimir L., biographical
sketch, 57

Laurencia, 10, 11, 15: diegoensis, 160;
obtusa, 15

Ledum, 95

Lemanea, 8

Lepidium: Davisii,
nanum, 162

Lepidiums, two perennial caespitose,
of western North America, 162

Lesquerella Sherwoodii, 6

Linanthus: androsaceous, 249, subsp.
laetus, 250; Bakeri, 250; bicolor
subsp. minimus, 251; ciliatus, 252;
dactylophyllum, 253; demissus,
253; Harknessii subsp. condensa-
tus, 251; maculatus, 246; mon-
tanus, 252; Nashianus, 252; nuda-
tus, 252; Killipii, 251; Parryae,
246; oblanceolatus, 253; tularen-
sis, 254

Lomentaria, 13: articulata, 11, 13

Lophosiphonia villum, 159

164, pl., 163;

Mason, H. L., Some additional notes
on Polemoniaceae, 249; Willis
Linn Jepson, 61

Mason, H. L. and A. D. Grant: Some
problems in the genus Gilia, 201

Matteuccia, 192: intermedia, 192;
japonica, 192; orientale, 193

Matthews, O. V.: A possible record of
Quercus Morehus in Oregon, 168

Mazzaella, 13

McMinn, H. E. Review: The Pacific
Coast Ranges, 30

MADRONO

aH

/

Mexico: The genus Mimulus in w ad-
jacent to Sinaloa, 21; A new Pole-

_ monium from, 187; A new violet

from, 131; Vegetation and climate —
of Coahuila, 33 j

Microsteris, 207

Mimulus: in or adjacent to Sinaloa,
Mexico, 21; calciphilus, 21; den-
tilobus, 21, 22; floribundus, 21, 23;
glabratus, 21, 23; guttatus, 21,
23; mohavensis, 246; Nelsoni, 21,
23; pallens, 21, 23; Pennellii, 21,
24; verbenaceus, 21, 25

Mirov, N. T.: Vladimir L. Komarov,
57

Moneses, 65, 67, 100: reticulata, 100;
uniflora, 67, 70, 100, figs., 79, 83, 85,
91, 93

Muenscher, W. C.: Potamogeton lati-
folius in Texas, 220

Muhlenbergia andina, 4

Muller, C. H.: Vegetation and climate
of Coahuila, Mexico, 33

Nemalion, 11, 12: helminthoides, 12;
lubricum, 12

Nicotiana: Maturation of the gametes
and fertilization in, 110, pl., opp.
114; Megasporogenesis and em-
bryo sac formation in, pl., opp.
112; alata, fig., opp. 112; longi-
flora, figs., opp. 114; otophora,
figs., opp. 114; rotundifolia, fig.
opp. 112; tabacum, figs., opp. 112,
opp. 114

Nomenclatorial changes in Elymus
with a key to the Californian
species, 120

Notes and news, 64, 168, 199, 232

Oedogonium, 8: vesicatum, 9

Onoclea: orientalis, 193; sensibilis, 191

Oregon: Certain plant species of the
canyon of Hurricane Creek, Wal-
lowa County, 1; The genus Heli-
anthella in, 186; A possible record
of Quercus Morehus in, 168

Osmundea, 15

Ownbey, M. The identity and delimita-
tion of Allium Tolmiei Baker,
233; Review: The Genus Crepis,
165

Oxytropis: A new species of, from the
Central Rocky Mountains, 133;
Cusickii, 7; Lambertii, 134; ob-
napiformis, 133, fig., 134; vis-
cida, 7

Papenfuss, G. F.: Generic names of
algae proposed for conservation,

Peck, M. E.: Certain plant species of

1948]

the canyon of Hurricane Creek,
Wallowa County, Oregon, 1
Penstemon Wilcoxii, 7

Pentarhizidium, 191: intermedium,
192; japonicum, 192; orientale,
191, 193

Petrocelis franciscana, 157

Petunia, 115, 116, 117

Phacelia: alpina, 18; cicutaria, 256;
compacta, 18; cryptantha, 256;
dasyphylla, 18; distans, 18; Gen-
tryi, 255; hispida, 256; Lemmonii,
18; leucophylla, 18; mutabilis, 18,
19; A new species of, from
Sonora, 255; Quickii, 18; seeds,
germination of, 17; umbrosa, 256;
vallicola, 18; vallis-mortae, 256

Phaeophyceae, 156

Phaeophycophyta, 9

Phorandendron: Mathewsii, 150; semi-
teres, 150; Storkii, 149, pl., 153

Physematium manchuriense, 191

Pinguicula vulgaris, 7

Pinnatifida, 15

Pinus flexilis, 3

Pirola, 66: alba, 67; alpina, 66; ameri-
cana, 67; aphylla, 66; asarifolia,
67; atropurpurea, 66; blanda, 67;
bracteata, 67; chlorantha, 66; Con-
ardiana, 67; Corbieri, 67; coreana,
68; decorata, 67; elliptica, 66;
Faurieana, 67; Forrestiana, 67;
gracilis, 66; japonica, 67; media,
67; minor, 66; morrisoniana, 66;
nephrophylla, 67; oxypetala, 67;
paradoxa, 67; renifolia, 66; rotun-
difolia, 67; Sartorii, 67; septen-
trionalis, 67; soldanellifolia, 66;
sororia, 67; spathulata, 66; suba-
phylla, 67; sumatrana, 68; uli-
ginosa, 67

Pirolaceae, 66

Plumaria, 11, 14: elegans, 14; pecti-
nata, 14

Poa alpina, 3

Pocockiella variegata, 10

Pogonophora californica, 159

Polemoniaceae, Some additional notes
on, 249

Polemoniella Heller, The present sta-
tus of the genus, 58

Polemoniella: antarcticum, 58; Gay-
anum, 58; micrantha, 58

Polemonium, 58: A new, from Mexico,
187; caeruleum, 60; carneum, 59;
confertum, 59; foliosissimum, 60;
glabrum, 187, 189, pl., 188; Lem-
monii, 59; mexicanum, 189; occi-
dentale, 59; pauciflorum, 59, 60,
189; pulcherrimum, 59, 60

INDEX

263

Polygonum: argyrocoleon, 64; vivi-
parum, 5

Polysiphonia: Eastwoodae, 160; Sny-
derae, 160, var. intricata, 160

Population analysis, The polygonal
graph for a_ simultaneous por-
trayal of several variables in, 105

Porter, C. L.: A new species of Oxy-
tropis from the central Rocky
Mountains, 133

Potamogeton: latifolius, 221, pl., 222,
in Texas, 220; pectinatus var.
latifolius, 221

Protowoodsia, 190: Notes on the tax-
onomy of some eastern Asiatic
ferns of the genera Pteretis and,
189; manchuriensis, 190

Psittacanthus: cuneifolius, 152; Hor-
tonii, 150, pl., 151

Pteretis, 192: Notes on the taxonomy
of some eastern Asiatic ferns of
the genera Protowoodsia and, 189;
intermedia, 192; japonica, 192,
193; nodulosa, 192; orientalis, 192,
193; Struthiopteris, 192

Pterilis, 192

Pterinodes, 192

Pterocladia, 11, 12: lucida, 12

Pterosiphonia: Baileyi, 159; califor-
nica, 160; dendroidea, 159

Ptilota, 14: Asplenioides, 14; filicina,
158; plumosa, 14, var. Aspleni-
oides, 14

Pyrola, 65, 98: americana, 99, figs., 79;
aphylla, 99; asarifolia, 99; brac-
teata, 70, 99; chlorantha, 7, 99;
dentata, 70, 99; var. apophylla, 70,
var. integra, 70; elata, 99; ellip-
tica, 99; incarnata, 99; maculata,
65; media, 99; minor, 7, 65, 70, 99,
figs. 79, 85, 87, 93; picta, 70, 99,
figs., 69, 75, 83, 85; renifolia, 99;
rotundifolia, 65, 99; secunda, 65;
uliginosa, 70, 99; umbellata, 65;
uniflora, 65; virens, 70, 99

Pyroleae, 97: Observations on the
structure and classification of the,
65; Structure of, pls., 69, 73, 75,
19, 83, 85, 87, 91, 93

Quercus Morehus, A possible record
of, in Oregon, 168

Quick, C. R.: Germination of Phacelia
seeds, 17

Ramischia, 65, 66, 97: secunda, 66, 97,
figs., 75, 79, 83, 85, 93; secundi-
flora, 97; truncata, 66, 97

Rattenbury, J. A.: Chromosome num-
ber publication, 257

264

Redwood, Location of extraneous ma-
terials in, 25

Reed, C. F.: Notes on the taxonomy
of some eastern Asiatic ferns of
the genera Protowoodsia and
Pteretis, 189

Reviews: Babcock, The genus Crepis,
165; Hutchinson, Silow, and Ste-
phens, The evolution of Gos-
sypium and the differentiation of
the cultivated cottons, 228; Mar-
tinez, Los Juniperus Mexicanos,
135; Peattie, The Pacific Coast
Ranges, 30; Stern, A study of the
genus Paeonia, 193; Wolf and
Wagener, The New World Cy-
presses, 229

Rhabdoniaceae, 12

Rhamnus obtusifolia, 130

Rhizoclonium riparium, 155

Rhododendron, 95: occidentale, 176

Rhododermis elegans, 157

Rhodoglossum parvum, 157

Rhodomelaceae, 15

Rhodophyceae, 156

Rhodophycophyta, 10

Rivularia multifida, 11

Rollins, R. C.: On two perennial
caespitose lepidiums of western
North America, 162

Salix: brachycarpa var. Sansoni, 5;
vestita, 5; Wolfii var. idahoensis,
5

Saxifraga oppositifolia, 6

Sedoidea, 11, 14: olivacea, 11; pur-
purea, 11

Senecio vulgaris, 64

Sequoia: sempervirens, 25, pls., 28, 29

Silene acaulis var. subacaulescens, 5

Sisyrinchium idahoense, 4

Sitanion, 122: elymoides, 125; flexuo-
sum, 125; Hansenii, 125; Hystrix,
125; jubatum, 125; lanceolatum,
125; multisetum, 125

Smilacina: amplexicaulis, 258; sesili-
folia, 258

Solidago ciliosa, 7

Sonora, A new species of Phacelia
from, 255

South American plants, Noteworthy,
149

Sphaerotrichia, 10

Standley, P. C. and F. A. Barkley:
Noteworthy South American
plants, I and II, 149

Stebbins, G. L., Jr., Review: A study
of the genus Paeonia, 193

Stromatocarpus Gardneri, 159

Struthiopteris, 192: orientalis, 193

Swainsona salsula, 64

MADRONO

[Vol. 9

Swertia perennis, 7

Terellia Macounii, 126

Texas: Old world plants apparently
recently introduced into, 64; Po-
tamogeton latifolius in, 220; Two
new varieties of Condalia from,
128

Thalictrum alpinum, 5

Thelaia, 65: spathulata, 99

Townsendia: diversa, 238; anomala,
240; arizonica, 239; florifer, 239,
240; incana, 238; minima, 238;
Notes on the genus, in Western
North America, 238; scapigera,
239; sericea, 238; spathulata, 239;
strigosa, 240; Watsoni, 238

Triticum, 120: dasystachum, 127; jun-
ceum, 126; pauciflorum, 126; rep-
ens, 127, var. dasystachum, 127,
var. subvillosum, 127; Richard-
soni, 126; subsecundus, 126; tra-
chycaulum, 126

Ulva, 10
Ulvella Setchellii, 155
Urospora, 8

Vegetation and climate of Coahuila,
Mexico, 33

Vegetation areas, Concentration of
environmental extremes as_ the
basis for, 169

Veronica Cusickii, 7

Villania, 9, 10

Viola: flagelliformis, 182; galeanaen-
sis, 131, 132, pedunculata, 132

Violet, A new, from Mexico, 131

Weber, W. A.: The genus Helianthella
in Oregon, 186

Went, F. W.: Some parallels between
desert and alpine flora in Cali-
fornia, 241

Wiggins, I. L., Review: Los Juniperus
Mexicanos, 135

Woodsia: alpina, 191; cathcartiana,
191; crenata, 191; elongata, 191;
fragilis, 191; glabella, 191; ilven-
sis, 191; insularis, 191; lanosa,
191; macrospora, 191; manchuri-
ensis, 190; mexicana, 191; mollis,
191; montevidensis, 191; obtusa,
191; oregana, 191; peruviana, 191;

Plummerae, 191; polystichoides,
191; scopulina, 191; subcordata,
191

Zizyphus: lycioides, 130; obtusifolia,
130

Zonaria, 9: Pavonia, 9; squamaria, 9;
Tournefortii, 9; variegata, 10

MADRONO

A West American Journal of Botany

A quarterly journal devoted to important
and stimulating articles dealing with
plant morphology, physiology, taxonomy,
and botanical history. These volumes
should be a part of every botanist’s li-
brary and should be made accessible to
students of all universities and colleges.

Volume I, 1916-1929. . . $5.00
Volume II, 1930-1934... 5.00
Volume III, 1935-1936 . 5.00
Volume IV, 1937-1938 . 5.00
Volume V, 1939-1940 . . 5.00
Volume VI, 1941-1942 . 5.00
Volume VII, 1943-1944 . 5.00
Volume VIII, 1945-1946 . 5.00
Volume IX, 1947-1948 . 5.00
Single numbers....... 1.00

The subscription price of MADRONO
is $3.50 per year. We solicit your pat-
ronage,

Address all orders to:

Rimo Bacigalupi, Bus. Mgr.

Natural History Museum,
Stanford University, California

ale

EME ts ang

=

*

- i SS - wo oS _
sae eee FS oe ee LS le et fe
_ = a 7 a a = _ a Ps 7 a
> > mest © = .. S 6.2 od te, te S - 7
me ’ a
4
: Se
~~ 7
»
: 7
¥ a
]
: -_
o
S _ :
2S .
- ,
= 7. : a
_ 7 = ar
~~ - - 7
" a? : ; -
-