The wild fodder Sorghums of the section Eu-Sorghum


Snowden, J.D.

Botanical Journal of the Linnean Society 55(358): 191-260

1955


A key to the 21 spp. within the section Eu-Sorghum is given followed by the history, description, diagnostic characters and affinities, distribution and utility of each sp. and associated vars. The description of the spp. and vars. are based mainly on herbarium specimens, occasionally amplified by information from collectors' notes.

J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
191
The
wild
fodder
Sorghums
of
the
section
E
u
-
S
or
g
hu
m.
By
J.
D.
SNOWDEN,
F.L.S.,
Economic
Botanist
(Retired),
Uganda.
(With
20
Text-figures)
[Read
6
May
1954]
HISTORY
AND
CLASSIFICATION
Sorghums
have
been
cultivated
for
their
grains
or
sweet
stems
since
ancient
times,
but
little
attention
was
given
to
the
small-grained
wild
Sorghums
until
the
beginning
of
the
twentieth
century
when
some
of
them
were
introduced
into
the
United
States
of
America,
where
they
soon
became
popular
and
were
extensively
cultivated
as
fodder
for
stock.
Linnaeus,
in
the
first
edition
of
his
Species
Plantarum,
published
in
1753,
included
the
Sorghums
known
to
him
under
the
genus
Holcus,
and
described
the
wild
race
he
knew
as
H.
halepensis.
The
same
species
was
referred
to
by
Scopoli
(Flora
Carniolica,
ed.
2,
2,
274,
1772)
under
Andropogon
arundinaceus.
Later,
when
the
Holcus
of
Linnaeus
was
divided
into
several
genera,
the
Sorghums
were
transferred
to
the
genus
Sorghum
by
Moench
(in
Methodus
Plantas,
p.
207,1794)
;
and
Persoon
(in
his
Synopsis
Plantarum,
1,
101,
1805),
the
latter
referring
the
wild
Sorghums
to
S.
halepense.
Brotero,
however
(in
Fl.
Lusitanic,a,
1,
89,
1804),
placed
the
Sorghums
in
the
genus
Andropogon
and
made
the
combination
A.
halepensis
for
the
wild
species.
In
his
Monograph
of
the
Andropogoneae
Hackel
(in
De
Candolle's
Monographiae
Phanerogamarum,
6,
499,
502,
1889)
classified
all
the
races
of
Sorghum
under
Andropogon
sorghum,
placing
the
cultivated
species
under
the
sub-
species
sativus
and
the
wild
ones
under
the
subspecies
halepensis.
It
was
suggested,
by
Hackel,
in
'
Die
kultivirten
Sorghum-Formen
and
ihre
Abstam-
mung
'
(in
Engl.,
Bot.
J
ahrb
.7,
115-26,
1885),
and
Koernicke
(in
H
andbuchcles
Getreidebaues
,
1,
300,
1885),
that
the
cultivated
Sorghums
had
been
derived
from
varieties
of
A.
hale-
pensis.
This
led
some
botanists
to
infer
that
all
the
cultivated
varieties
had
been
derived
from
one
wild
ancestor.
Piper,
who
had
studied
many
of
the
Sorghums
under
cultivation
in
the
United
States,
as
well
as
the
herbarium
material
in
that
country,
at
Kew,
and
in
Berlin,
in
his
revision
of
A
.
halepensis
and
A.
sorghum
(in
Proc.
Biol.
Soc.
Wash.
28,
25-44,
1915)
pointed
out
that
Hackel
had
indicated
that
they
had
arisen
from
those
varieties
without
perennial
rootstocks,
such
as
effusus,
virgatus,
and
aethiopicus,
and
not
from
those
having
original
rhizomes.
Piper
then
classified
the
wild
races
of
the
former
under
eleven
subspecies
of
A.
sorghum,
and
placed
those
with
rhizomes
under
six
subspecies
of
A.
halepensis.
In
his
revision
of
the
African
Sorghums
Stapf
(in
PraM,
Fl.
Trop.
Afr.
9,
104-54,
1917)
decided
to
retain
Sorghum
as
a
separate
genus.
He
discussed
the
classifications
of
Hackel,
Koernicke
and
Piper,
pointed
out
the
desirability
of
recognizing
smaller
units,
and
described
all
the
wild
and
cultivated
species
of
Tropical
Africa
known
to
him.
He
divided
them
into
two
sections,
namely
Eu-Sorghum,
containing
twenty-seven
species,
where
among
the
wild
Sorghums
many
of
Piper's
subspecies
were
given
specific
rank,
and
Sorghastrum,
in
which
he
placed
six
species
of
a
distinctive
group
that
Nash
(in
Britton,
Manual
of
the
Flora
of
the
Northern
States
and
Canada,
p.
71,
1901)
had
described
as
a
new
genus
under
the
name
of
Sorghastrum.
In
my
'Classification
of
the
Cultivated
Sorghums'
(in
Kew
Bull.
pp.
222,
223,
1935)
I
decided
to
keep
Sorghastrum
Nash
separate
from
Sorghum,
as
the
plants
of
this
group
differ
from
the
true
Sorghums,
especially
in
the
solitary
primary
branches
of
the
panicle,
which
are
often
much
branched
quite
near
the
base
so
that
they
appear
to
be
falsely
semi-verticillate
;
the
pedicelled
spikelets
usually
reduced
to
the
hairy
pedicels
;
and
the
glabrous
lodicules.
The
lower
glume
is
also
usually
shorter
than
the
upper.
The
separation
JOURN.
LINN.
SOC.—BOTANY,
VOL.
LV
192
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
of
Sorghastrum
left
a
more
uniform
genus
Sorghum
which
was
divided
into
two
sections,
characterized
as
follows:
I.
Eu-Sorghum:
sheath-nodes
glabrous
or
finely
pubescent,
not
bearded;
primary
branches
of
the
panicle
divided
(at
least
the
lower);
racemes
lateral
and
terminal;
number
of
chromosomes
(in
those
species
that
have
been
examined)
2n
=
20,
or
2n=
40.
A.
Subsection
Arundinacea:
annuals,
or
tufted
perennials
without
rhizomes;
number
of
chromosomes
usually
2n
=
20
(rarely
2n,
=
40).
This
subsection
is
divided
into
two
series,
Spontanea,
containing
the
wild
Sorghums,
and
Sativa,
embracing
the
culti-
vated
Sorghums.
B.
Subsection
Halepensia:
perennials
possessing
distinct
elongated
rhizomes;
number
of
chromosomes
usually
2n=
40;
wild
Sorghums.
II.
Para-Sorghum:
sheath-nodes
bearded
(at
least
the
upper);
primary
branches
of
the
panicle
simple;
racemes
terminal;
number
of
chromosomes
2n=
10;
wild
Sorghums.
An
account
of
the
cytological
work
carried
out
by
Garber
(in
Amer.
Nat.,
78,89-94,
1944)
supports
the
separation
of
Para-Sorghum
as
a
distinct
section
and
tends
to
confirm
that
the
Sorghums
of
this
group
are
unlikely
to
have
been
concerned
in
the
evolution
of
the
cultivated
Sorghums.
Most
of
the
species
of
this
group
are
slender
grasses
with
little
bulk
and,
as
they
do
not
appear
to
have
any
special
importance
as
fodder
for
stock,
they
are
not
dealt
with
further
in
this
account
of
the
Arundinacedus
Forage
Sorghums.
The
wild
Sorghums
of
the
two
subsections
of
Eu-Sorghum
are
very
similar
in
most
characters,
but
those
of
the
Halepensia
group,
besides
possessing
rhizomes
and
usually
having
a
higher
number
of
chromosomes,
generally
show
little
tendency
to
hybridize
with
the
cultivated
grain
Sorghums.
They
also
have
a
different
area
of
distribution,
which
extends
from
the
Mediterranean
region
eastwards
to
India
and
south-western
Asia.
On
the
other
hand,
the
wild
Sorghums
of
the
series
Spontanea
and
the
cultivated
races
of
the
series
Sativa,
which
belong
to
the
subsection
Arundinacea,
are
frequently
found
to
hybridize
with
one
another
in
regions
where
Sorghums
are
grown
for
their
grains
and
sweet
stems,
and
the
wild
races
of
this
subsection
are
mainly
distributed
in
tropical
and
subtropical
Africa.
The
relationships
of
the
wild
and
cultivated
Sorghums
were
reviewed
in
my
account
of
the
`Grain
and
Sweet-stemmed
Sorghums'
(The
Cultivated
Races
of
Sorghum,
pp.
234-
40,
1936).
A
few
of
the
more
important
intermediate
races
that
appear
to
have
arisen
naturally,
and
not
as
a
direct
result
of
artificial
cross-fertilization,
are
included
below,
together
with
the
other
wild
races
of
the
section
Eu-Sorghum.
Although
they
generally
agree
with
the
wild
Sorghums
in
possessing
small
grains
enclosed
by
the
glumes,
they
have
somewhat
tougher
racemes
resembling
those
of
the
cultivated
races.
Those
culti-
vated
Sorghums
that
show
signs
of
degeneration
as
a
result
of
intercrossing
with
wild
species,
generally
indicated
by
the
racemes
shattering
and
the
spikelets
readily
falling,
will
be
excluded
from
this
revision.
In
connexion
with
the
cross-pollination
that
occurs
between
the
different
races
of
Sorghum,
it
is
of
interest
to
record
that
in
India,
in
the
course
of
experiments
undertaken
to
procure
early-maturing
races
of
sugar
cane,
by
means
of
artificial
cross-fertilization,
hybrids
between
Sorghum
and
Saccharum
have
been
raised,
thus
indicating
the
possibility
of
a
closer
relationship
between
these
two
genera
than
has
generally
been
accepted.
As
so
many
specimens
of
wild
Sorghums
have
been
received
at
Kew
since
Dr
Stapf's
account
of
the
Tropical
African
species
was
published
in
1917,
a
revision
of
those
be-
longing
to
the
section
Eu-Sorghum
has
now
become
practicable
and
indeed
has
been
found
necessary
in
order
to
facilitate
the
identification
of
the
material,
especially
that
received
from
North-east,
Central,
and
East
Africa.
In
dealing
with
these
Sorghums,
however,
it
should
be
noted
that
the
descriptions
of
the
species
and
varieties
are
based
on
the
dry
herbarium
specimens,
although
amplified
in
some
cases
by
information
from
collectors'
notes.
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
193
GENERAL
CHARACTERS
OF
THE
WILD
SPECIES
OF
SECTION
EU-SORGHUM
Annual
or
perennial
grasses
one
to
several
metres
high.
Cu/ms
slender
to
stout,
usually
terete,
often
woody
and
reed-like,
few
to
many
noded,
simple
or
sometimes
branched
above,
usually
glabrous
and
green
but
often
with
purplish
markings
and
a
grey
waxy
bloom.
Leaves
alternate;
sheaths
usually
clasping
the
culms
when
young,
often
shorter
than
the
mature
internodes,
generally
glabrous
except
for
a
little
pubescence
near
the
base,
along
the
margins
and
near
the
junction
with
the
blade;
ligule
usually
a
membra-
nous
or
scarious
rim
1-3
mm.
long,
ciliate
from
the
back;
blades
convolute
in
bud
but
soon
becoming
flat,
usually
linear
to
linear-lanceolate,
narrowing
upwards
to
an
acumi-
nate
tip,
glabrous
except
for
a
few
hairs
near
the
ligule
and
the
junction
with
the
sheath,
mostly
light
to
dark
green
with
a
paler,
often
broad
and
pronounced,
mid
rib,
sometimes
with
purplish
markings,
generally
spinulously
scabrid
on
the
margins,
especially
upwards.
Panicles
usually
compound
or
decompound,
varying
much
in
shape
and
size,
generally
loose
and
open
on
erect
peduncles,
at
length
more
or
less
exserted
from
the
uppermost
leaf-sheath
;
rhachis
usually
continuous,
often
striate
or
channelled,
with
the
angles
spinulously
ciliate,
at
least
upwards;
branches
1-6
(rarely
more)
from
a
node,
usually
hairy
at
the
base
and
also
in
the
angles
of
the
divisions,
often
scabrid.
Racemes
terminal
and
lateral
on
the
branchlets,
which
may
arise
from
near
the
base,
or
1-5
cm.
or
more
from
the
base,
of
the
branches,
1-
to
7-noded
(rarely
more),
mostly
fragile
and
generally
readily
disarticulating
at
maturity,
but
sometimes
rather
tough;
internodes
(joints)
and
pedicels
very
similar,
but
the
internodes
often
more
dorsally
compressed
and
almost
flat
on
the
face,
slender
to
fdiform,
ciliate
along
the
sides,
the
pedicels
generally
with
almost
discoid
or
sometimes
oval
tips;
spikelets
mostly
in
pairs
composed
of
one
sessile
and
one
pedicelled,
except
at
the
end
of
each
raceme
where
there
is
usually
one
sessile
and
two
pedicelled
spikelets.
Sessile
spikelets
dorsally
compressed,
varying
in
shape,
mostly
awned
but
sometimes
awnless,
with
a
small
blunt,
usually
bearded,
callus
at
the
base
of
the
lower
glume
;
florets
2,
the
lower
floret
usually
reduced
to
its
lemma.
Glumes
equal
or
almost
equal
in
length,
usually
coriaceous
for
the
most
part
but
thinner
towards
the
tips
and
with
a
thin
transverse
basal
strip
0.3-0.5
mm.
wide,
rarely
thinner
and
subcoriaceous
to
papery
or
crustaceous
throughout,
the
margins
papery
to
membranous,
narrowly
incurved,
those
of
the
lower
clasping
the
upper
(except
when
in
flower),
so
that
when
closed
they
more
or
less
conform
to
the
shape
of
the
spikelet
in
outline;
lower
glume
flattened
or
only
slightly
convex
on
the
back,
mostly
10-
to
15-nerved,
but
usually
with
only
3-5
nerves
out-
wardly
evident
near
the
tip,
generally
with
short
cross-nerves
scattered
in
the
middle
third
or
three-fifths,
the
margins
narrowly
inflexed
upwards,
2-keeled
from
about
the
middle
or
in
the
upper
third
or
fourth,
with
the
scarious
spinulously
ciliolate
keels
usually
narrowly
winged
(marginate)
and
terminating
in
minute
teeth
just
below
the
hyaline
apex,
or
sometimes
gradually
merging
imperceptibly
with
it
;
upper
glume
convexly
boat-shaped,
usually
7-nerved
(rarely
more),
the
central
nerve
finely
keeled
and
scaberulous
above
the
middle
or
near
the
tip,
short
cross-nerves
usually
present
in
the
middle
third
or
three-fifths,
the
incurved
margins
hyaline
and
usually
finely
long-ciliate,
at
least
upwards.
Lemmas
hyaline,
thinly
membranous
and
often
very
fragile,
the
mar-
gins
incurved
and
usually
finely
ciliate
;
lower
lemma
(when
flattened
out,
as
in
the
following
descriptions)
almost
similar
in
shape
and
size
to
the
lower
glume
but
a
little
shorter
and
with
the
apex
obtuse
or
slightly
notched,
usually
with
two
fine
nerves,
one
on
each
side,
but
sometimes
with
a
third
central
nerve
;
upper
lemma
usually
1-2
mm.
shorter
than
the
lower,
ovate
to
ovate-elliptic,
finely
1-
to
3-nerved,
the
two
lateral
often
short
or
obscure,
the
apex
2-lobed
and
with
an
awn
from
the
sinus,
or
sometimes
entire,
or
almost
entire,
and
then
muticous
or
mucronate
;
awn
usually
abruptly
bent
near
the
middle,
the
lower
half
tightly
twisted
to
form
a
straight
stem
or
column
and
glabrous
or
finely
pubescent
along
the
spirals,
the
upper
part
or
bristle
being
only
slightly
twisted
N
2
194
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
and
finely
scaberulous
along
the
margins.
Palea
like
the
lemmas
in
texture,
generally
a
little
shorter
than
the
upper
lemma,
almost
linear
to
linear-lanceolate,
nerveless,
with
the
margins
finely,
and
usually
sparsely,
ciliate.
Lodicules
2,
somewhat
wedge-shaped,
about
1-2
mm.
long,
usually
ciliate,
but
the
cilia
sometimes
reduced
to
a
few
or
one
at
each
corner.
Stamens
3,
anthers
linear,
usually
yellow.
Styles
2,
terminal
or
sub-
terminal,
with
the
plumose
stigmas
sometimes
purplish
and
usually
laterally
exserted
for
pollination.
Ovary
glabrous.
Grains
(caryopses)
elliptic,
obovate-elliptic,
or
obovate,
generally
not
exceeding
half
the
length
of
the
glumes
when
mature,
more
or
less
dorsally
compressed,
with
the
hilum
face
almost
flat
and
the
embryo
face
convex,
the
embryo-
mark
oval
and
usually
about
half,
or
rather
more
than
half,
the
length
of
the
grain.
Pedicelled
spikelets
mostly
linear-lanceolate
to
subulate,
often
much
narrower
than
the
sessile,
but
sometimes
little
different
in
size,
usually
awnless,
the
upper
floret
(
5`
or
some-
times
reduced
to
the
lemma
like
the
lower;
glumes
thinner
than
those
of
the
sessile
spikelets,
herbaceous
to
membranous,
with
the
nerves
usually
plainly
visible
throughout
their
length,
otherwise
much
like
those
of
the
sessile
in
shape
and
nervation;
lemmas
like
those
of
the
sessile
in
texture,
shape,
and
nervation,
but
often
reduced
in
size
and
the
upper
usually
muticous
;
palea
small
or
often
absent;
lodicules
like
those
of
the
sessile
spikelets
when
present,
but
smaller
and
often
almost
glabrous;
anthers
mostly
smaller
than
those
of
the
sessile
spikelets,
but
similar
in
shape
and
colour.
THE
POISONOUS
PROPERTIES
OF
SORGHUM
Cases
of
Sorghum
poisoning
or
cyanogenesis,
resulting
from
the
presence
of
hydrocyanic
or
prussic
acid
in
plants
used
for
fodder,
have
been
reported
from
many
countries
during
the
last
fifty
years.
The
toxicity
varies
not
only
in
different
species
but
also
in
the
same
species,
because
the
amount
of
acid
present
differs
according
to
the
conditions
under
which
the
plants
are
grown
and
the
particular
stage
of
growth.
The
poisonous
tendencies
are
generally
greatest
in
the
young
immature
plants,
especially
in
such
luscious
dark
green
growths
as
are
apt
to
follow
the
application
of
an
excess
of
nitrogenous
manures,
heavy
rainfall,
or
irrigation,
after
a
dry
spell
of
weather,
and
in
the
stunted
growths
due
to
checks
caused
by
drought,
frost,
or
insect
and
fungous
attacks.
Generally
it
is
not
considered
advisable
to
feed
animals
on
Sorghums
until
the
plants
have
reached
the
flowering
stage,
although
sometimes
the
immature
grains
may
also
prove
injurious.
At
maturity,
however,
most
of
the
Sorghums
grown
under
normal
conditions
are
considered
harmless.
The
danger
of
poisoning
tends
to
be
reduced
if
the
Sorghums
are
cut
and
dried
before
being
fed
to
stock,
but
drying
does
not
always
immediately
destroy
the
toxicity;
much
depends
on
the
method
of
drying.
The
poisonous
properties
of
some
of
the
Sorghums
under
review
will
be
referred
to
again
under
the
notes
dealing
with
the
utility
of
the
species.
KEY
TO
THE
GROUPS
AND
SPECIES
OF
SECTION
EU-SORGHUM
Subsection
Halepensia:
perennials
with
well-developed
more
or
less
extensively
creeping
rhizomes;
sessile
spikelets
deciduous
when
mature,
3.8-6-2
mm.
long;
grains
small,
enclosed
by
the
glumes;
number
of
chromosomes
usually
2n
=
40;
wild
Sorghums:
Sessile
spikelets
more
or
less
obtuse,
elliptic
to
subelliptic
or
elliptic-lanceolate,
4-5.5
mm.
long;
glumes
coriaceous;
lower
glume
with
the
keels
ending
in
distinct
minute
teeth,
forming
with
the
subhyaline
apex
a
more
or
less
distinct
3-toothed
tip:
Leaf-blades
narrow,
0.5-2
em.
wide
(rarely
more);
cuims
slender,
0.5-1-5
m.
high,
up
to
5
mm.
wide;
panicles
small,
often
somewhat
contracted
after
flowering,
up
to
25
cm.
long
and
5
cm.
wide,
with
the
lower
branches
about
5-8
cm.
long.
1.
halepensis.
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
195
Leaf-blades
somewhat
broad,
mostly
2-4
cm.
wide
when
mature;
culms
slender
to
robust,
up
to
1
cm.
wide,
usually
2-3
m.
high;
panicles
large,
loose
and
spreading,
generally
25-55
cm.
long,
10-25
cm.
wide,
with
the
lower
branches
10-25
cm.
long.
2.
miliaceum.
Sessile
spikelets
acute
to
acuminate;
glumes
subcoriaceous
with
somewhat
papery
tips;
lower
glume
with
the
keels
ending
without
or
with
only
obscure
teeth
:
Sessile
spikelets
5-6.2
mm.
long,
acuminate
to
acute,
often
awned;
panicle
small,
15-30
cm.
long,
lower
branches
5-15
cm.
long;
culms
usually
0.5-
2
m.
high,
0.3-1
cm.
wide;
leaf-blades
narrow,
0.5-2
cm.
wide.
3.
controversum.
Sessile
spikelets
3.8-5
mm.
long,
abruptly
acute
with
a
short
fine
point,
usually
awnless;
panicle
large,
20-60
cm.
long,
lower
branches
15-20
cm.
long;
culms
2-3
m.
high,
1-3
cm.
wide;
leaf-blades
broad,
often
3-5
cm.
wide.
4.
propinguu
m.
Subsection
Arundinacea:
annuals,
or
tufted
perennials
without
rhizomes,
the
tiller-
eulms
arising
from
basal
buds
near
the
foot
of
the
previous
season's
culms;
number
of
chromosomes
usually
2n
=
20
(rarely
2n
=
40):
Series
Sativa
:
racemes
tough;
mature
sessile
spikelets
persistent;
grains
large,
often
exceeding
the
glumes
in
size,
usually
much
exposed;
Sorghums
cultivated
for
their
grains
or
sweet
stems
(the
species
of
this
series
are
not
dealt
with
in
this
revision).
Series
Spontanea
:
racemes
usually
fragile,
rarely
somewhat
tough
;
mature
sessile
spikelets
usually
deciduous,
sometimes
rather
tardily;
grains
small,
en-
closed
by
the
glumes;
wild
Sorghums:
Racemes
fragile,
readily
disarticulating
at
maturity,
the
sessile
spikelets
falling
to-
gether
with
the
adjoining
internode
of
the
rhachis
and
its
pedicelled
spikelet,
or
at
least
its
pedicel:
Sessile
spikelets
lanceolate,
6-8
mm.
long,
awns
8-18
mm.
long;
leaf-blades
usually
0.5-2
cm.
wide,
rarely
up
to
3
cm.:
Panicle
long
and
narrow,
15-60
cm.
long,
1-5
cm.
wide,
with
the
branches
usually
suberect;
sessile
spikelets
6.5-7
mm.
long,
2-2.5
mm.
wide,
acute,
awn
slender;
grains
obovate-elliptic,
about
2.5-3
mm.
long,
1.5-2
mm.
wide.
5.
virgatum.
Panicle
ovate-oblong
to
lanceolate,
20-40
cm.
long,
up
to
15
(rarely
20)
cm.
wide,
with
the
branches
soon
spreading;
sessile
spikelets
6-8
mm.
long,
2-3
mm.
wide,
shortly
acuminate,
awn
rather
stout
;
grains
obovate-
elliptic
to
obovate-oblong,
3-3.5
mm.
long,
2-2.5
mm.
wide.
6.
lanceolat
um.
Spikelets
not
as
above
:
Sessile
spikelets
6-9
(rarely
5.5)
mm.
long:
Sessile
spikelets
2.5-3-5
mm.
wide,
6-9
mm.
long
:
Sessile
spikelets
8-9
mm.
long,
elliptic-
to
oblong-lanceolate,
acuminate,
awn-
less,
or
more
often
with
an
awn
8-16
mm.
long;
panicle
large,
loose,
15-45
cm.
long,
5-25
cm.
wide.
7.
vogelian
um
.
Sessile
spikelets
6-8
mm.
long,
usually
with
awns
20-30
mm.
long:
Panicle
very
narrow,
10-40
cm.
long,
3-10
cm.
wide;
branches
usually
suberect;
sessile
spikelets
elliptic-
to
ovate-lanceolate,
usually
densely
and
persistently
tomentose
with
mostly
whitish
silky
hairs;
awn
20-30
mm.
long.
8.
aethiopicum.
Panicle
rather
wide
and
loose,
up
to
35
cm.
long,
8-15
cm.
wide,
branches
subflexuous
and
spreading;
sessile
spikelets
elliptic-lanceolate
to
elliptic-oblong,
shortly
and
finely,
often
fulvously,
hairy,
or
at
length
almost
glabrous
on
the
backs
of
the
glumes;
awn
20-25
mm.
long.
9.
macrochaeta.
Sessile
spikelets
2-2.5
mm.
wide,
usually
6-7.5
mm.
long,
but
sometimes
reduced
to
5.5
mm.,
elliptic-
to
oblong-lanceolate;
grains
obovate-
elliptic
to
obovate,
2-3
mm.
long,
1.5-2
mm.
wide:
196
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
Upper
lemma
of
the
sessile
spikelet
awnless
or
with
an
awn
5-10
mm.
long
;
sessile
spikelets
elliptic-
to
oblong-lanceolate,
narrowly
acuminate;
leaf-blades
usually
broad,
mostly
3-6
cm.
wide;
panicle
large,
loose
and
much
branched,
20-60
cm.
long,
10-25
cm.
wide.
10.
arundinaceum.
Upper
lemma
of
the
sessile
spikelets
with
an
awn
10-16
(sometimes
18)
mm.
long;
leaf-blades
narrow,
mostly
1-2.5
(but
sometimes
3)
cm.
wide:
Leaf-blades
mostly
45-60
cm.
long,
narrowing
upwards
to
an
acuminate
tip;
sessile
spikelets
elliptic-lanceolate,
acuminate,
6-7
(or
sometimes
5.5)
mm.
long;
panicle
pyramidal
to
elliptic-oblong,
at
length
loose
and
spreading,
15-50
cm.
long,
5-15
cm.
wide,
lower
branches
up
to
15
cm.
long.
11.
verticilliflorum.
Leaf-blades
up
to
30
cm.
long;
panicle
narrow-oblong,
1-2
cm.
wide:
Leaf-blades
10-30
cm.
long,
with
a
long-tapering
filiform
tip
;
sessile
spikelets
elliptic-lanceolate
to
elliptic,
acute,
5.5-6
mm.
long;
awn
16-18
nun.
long;
panicle
10-25
cm.
long.
12.
somaliense.
Leaf-blades
4-10
cm.
long,
acuminate;
sessile
spikelets
oblong-
lanceolate,
acuminate,
5.5-6-5
mm.
long;
awn
12-14
mm.
long;
panicle
4-12
cm.
long.
13.
pugionifolium.
Sessile
spikelets
min.
long,
1.5-2-5
mm.
wide:
Upper
lemma
of
the
sessile
spikelet
with
an
awn
10-16
mm.
long
:
Leaf-blades
broad,
usually
3-5
(but
sometimes
2)
cm.
wide
when
mature;
panicle
wide
and
loose,
15-50
cm.
long,
10-30
cm.
wide
:
Sessile
spikelets
elliptic-ovate,
acute
to
acuminate,
4.5-5-5
mm.
long,
2-2.5
mm.
wide.
14.
brevicarinatum.
Sessile
spikelets
elliptic-lanceolate
to
elliptic-oblong,
obtuse
or
subacute,
4.5-5-2
mm.
long,
1.5-2
mm.
wide.
15.
usambarense.
Leaf-blades
narrow,
1.5-2-7
cm.
wide;
panicle
narrow
and
loose,
up
to
40
cm.
long
and
10-15
cm.
wide
;
sessile
spikelets
broadly
elliptic
to
elliptic-oblong,
obtuse,
4-4.5
mm.
long,
2-2.5
mm.
wide.
16.
castaneum.
Upper
lemma
of
the
sessile
spikelets
with
a
small
mucro
;
sessile
spikelets
elliptic
to
elliptic-oblong,
subacute,
4-5
mm.
long,
1.5-2
mm.
wide;
panicle
narrow,
loose
or
somewhat
contracted,
20-25
cm.
long,
about
6
cm.
wide;
leaf-blades
(uppermost)
1.5-2
cm.
wide.
17.
panicoide,s.
Racemes
more
or
less
tough
or
tardily
disarticulating;
sessile
spikelets
persistent
or
at
length
deciduous
:
Panicle
dense
and
contracted
with
the
rhachis
often
completely
hidden
by
the
branches,
10-20
cm.
long,
3-8
cm.
wide;
leaf-blades
broad,
4-8
cm.
wide;
sessile
spikelets
ovate
to
broadly
elliptic-ovate,
acute
to
acuminate,
6-7.5
mm.
long,
2.5-3-5
mm.
wide.
18.
hewisonii.
Panicle
loose,
with
the
rhachis
more
or
less
freely
exposed
when
mature;
leaf-
blades
narrow,
1-3
cm.
wide
:
Sessile
spikelets
6-7.5
mm.
long,
2-3
mm.
wide,
elliptic
to
elliptic-
or
oblong-
lanceolate,
subacute;
pedicelled
spikelets
persistent;
grains
elliptic
to
obovate-elliptic,
3.5-4-5
mm.
long,
1.5-2
mm.
wide.
19.
sudanense.
Sessile
spikelets
4.5-6
mm.
long,
ovate,
ovate-lanceolate,
or
elliptic-ovate
to
elliptic-rotund;
pedicelled
spikelets
deciduous:
Sessile
spikelets
ovate
to
ovate-lanceolate,
4.5-5.5
mm.
long,
2-2.5
mm.
wide,
acute
to
acuminate;
grains
broad-elliptic
to
obovate-elliptic,
2-2.5
mm.
long,
1.5-1-8
mm.
wide;
panicle
sublinear,
about
18
cm.
long
and
2
cm.
wide.
20.
elliotii.
Sessile
spikelets
elliptic-ovate
to
elliptic-rotund,
4.5-6
mm.
long,
2-3
min.
wide,
acute
to
acuminate
and
closed,
or
subacute
and
slightly
open
at
the
tip;
grains
obovate-elliptic
to
obovate,
3-4
mm.
long,
1.5-2-5
mm.
wide;
panicle
oblong
to
elliptic-oblong,
15-40
cm.
long,
8-15
cm.
wide.
21.
niloticum.
T.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
197
DESCRIPTION
AND
ENUMERATION
Subsection
Halepensia
Snowden
in
Kew
Bull.
p.
222,
1935
1.
Sorghum
halepense
(Linn.)
Pers.
Holcus
halepensis
Linn.,
Sp.
Pl.
ed.
1,
p.
1047,
1753.
When
he
described
this
species
Linnaeus
did
not
cite
any
specimens
in
support
of
it,
but
there
are
two
specimens
in
his
herbarium
at
The
Linnean
Society
of
London,
and
one
of
them
which
I
have
examined
is
written
up
halepensis
in
his
own
handwriting.
Linnaeus
gives
the
habitat
of
his
species
as
Syria
and
Mauritania.
Andropogon
arundinaceus
Scop.,
Fl.
Cam.
ed.
2,
2,
274,
1772,
was
the
name
used
for
a
specimen
from
Carniola,
but
Scopoli
cites
earlier
synonyms
of
Linnaeus
and
his
species
is
evidently
based
on
Holcus
halepensis
Linn.
10
0
10
2
0
30
4
0
50
60
70
80
90
100
110
120
130
40
1
.za•
''
30
_
I
l
i
\
.....—______
.
2
0
20
1
I
L
10
0
'
to
10
11
14
,-.,
0-.,
o
20
if
--___
20
At
30
10
0
l0
20
30
40
50
60
70
80 90 too
no
120
t30
Distribution
of
the
chief
species:
1,
Sorghum
halepense;
2,
S.
miliaceum;
4,
S.
propinquum;
5,
S.
virga-
tum;
6,
S.
lanceolatum;
8,
S.
aethiopicum;
10,
S.
arundinaceum;
11,
S.
verticilliflorum;
14,
S.
brevi-
carinatum.
Milium
halepensis
(Linn.)
Cay.,
Descr.
Pl.
p.
306,
1802,
is
founded
on
Holcus
halepensis
Linn.
Blumen,bachia
halepensis
(Linn.)
Koel.,
Descr.
Gram.
p.
29,
1802,
is
based
on
Holcus
halepensis
Linn.
His
new
genus
was
not
generally
accepted
by
other
botanists.
Andropogon
halepensis
(Linn.)
Brot.,
Fl.
Lusit.
1,
89,
1804,
rests
on
Holcus
halepensis
Linn.
Sorghum
halepense
(Linn.)
Pers.,
Synops.
1,101,
1805,
is
based
on
Holcus
halepensis
Linn.
Andropogon
sorghum
(Linn.)
Brot.
subsp.
halepensis
(Linn.)
Hack.
var.
halepensis
Hack.
in
DC.
Monogr.
Phan.
6,
502,
1889,
is
founded
on
Holcus
halepensis
Linn.,
but
besides
reducing
halepensis
to
a
subspecies
Hackel
also
includes
some
elements
in
his
var.
halepensis,
which
in
my
opinion
do
not
belong
to
Holcus
halepensis
Linn.
Some
of
the
species
cited
by
Hackel
in
synonomy
are
therefore
excluded,
particularly
Andropogon
198
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
laxus
Roxb.,
Fl.
Ind.
1,
271,
1820,
A.
amtroversus
Steud.,
Synops.
1,
391,
1854,
and
Sorghum
giganteum,
Edgew.,
J.
As.
Soc.
Bengal,
21,
181,
1853,
as
well
as
all
the
varieties
of
Hackers
subsp.
h,alepensis
other
than
var.
halepensis
(sensu
stricto).
DESCRIPTION
A
perennial
with
abundant
creeping
rhizomes
25-75
cm.
long,
often
descending
to
a
depth
of
25-50
cm.
Cu/ms
rather
slender,
mostly
simple
but
sometimes
branched,
0.5-1-5
m.
high,
up
to
5
mm.
wide
near
the
base.
Leaves
usually
7-10;
blades
linear,
narrowed
to
a
point
at
the
apex,
20-60
cm.
long,
0.5-2
(rarely
2.5)
cm.
wide,
the
stoutish
I
16'
I
B
C
el
;,,
I
E
I
1
A
D
F
Sorghum
halepense
(Linn.)
Pers.
(1)
A,
Sessile
and
pedieelled
spikelets;
B,
lower
glume;
C,
upper
glume,
exterior
view
(awnless
form,
from
Snowden,
Kew,
1933).
(2)
D,
sessile
and
pedicelled
spikelets;
E,
lower
glume;
F,
upper
glume
(awned
form;
from
Haradjean
3470).
A
and
D,
x
5;
B,
C,
E
and
F,
x10.
midrib
pale
green
to
almost
white
and
contrasting
with
the
darker
green
of
the
remainder.
Panicle
loose
when
in
flower,
somewhat
contracted
after
flowering,
lanceolate
to
lanceo-
late-oblong,
mostly
10-25
cm.
long,
and
3-8
cm.
wide;
branches
slender,
suberect,
one
to
several
from
each
node,
the
lower
usually
about
half
the
length
of
the
panicle,
bare
of
spikelets
for
1-5
cm.
from
the
base.
Racemes
fragile,
1-
to
5-noded,
up
to
2
cm.
long;
internodes
and
pedicels
slender,
2.5-4
mm.
long,
rather
densely
ciliate
with
whitish
to
tawny
or
purplish
hairs
0.7-1
mm.
long.
Sessile
spikelets
elliptic
to
subelliptic,
mostly
J.
D.
SNOWDEN
:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
199
4.5-5
mm.
long,
but
sometimes
up
to
5.5
mm.,
1-8-2.3
mm.
wide,
obtuse
or
subacute,
at
first
more
or
less
hairy
and
cream
to
buff-yellow
or
tawny,
with
3-4
nerves
evident
near
the
tip
of
the
lower
glume,
at
length
often
purplish
to
blackish
brown
or
black
and
sometimes
almost
glabrous
;
awned
or
awnless
;
callus-beard
copious
to
somewhat
scanty.
Glumes
firmly
coriaceous
with
slightly
thinner
tips;
lower
usually
7-
to
9-nerved,
with
a
few
cross-nerves
above
the
middle,
keeled
in
the
upper
fourth,
the
wings
of
the
keels
widening
upwards
to
end
in
minute
teeth,
forming
with
the
short
apex
a
distinctly
3-toothed
tip,
spinulously
scabrid
on
the
keels;
upper
7-nerved,
with
a
few
cross-nerves
from
the
middle
upwards,
rough
on
the
keel
in
the
upper
third.
Lemmas
moderately
and
finely
ciliate
;
lower
elliptic
to
elliptic-oblong,
4-5
mm.
long
;
upper
ovate
to
ovate-
elliptic,
3-3.5
mm.
long,
1-
to
3-nerved,
acute
and
minutely
mucronate,
or
2-lobed
to
about
0.5
mm.
deep
and
with
an
awn
10-16
mm.
long.
Anthers
2.5-3
mm.
long.
Grains
light
to
dark
brown,
oblong-obovate,
2.5-3
mm.
long.
Pedicelled
spikelets
at
length
deciduous,
often
purplish,
c
3`
or
sometimes
neuter,
4.5-6
mm.
long
;
lower
glume
7-
to
9-
nerved,
upper
5-
to
7-nerved;
anthers
2.5-3
mm.
long.
DIAGNOSTIC
CHARACTERS
AND
AFFINITIES
The
distinguishing
features
of
Sorghum
halepense
are
the
abundance
of
the
slender
wide-spreading
rhizomes,
the
rather
slender
dwarf
culms,
narrow
leaf-blades
20-60
cm.
long,
0.5-2
(rarely
2.5)
cm.
wide;
small
panicles
that
are
usually
somewhat
contracted
after
flowering;
and
the
elliptic
to
subelliptic
sessile
spikelets,
4.5-5
or
sometimes
5.5
mm.
long,
1.8-2.3
mm.
wide,
with
the
tip
of
the
lower
glume
obtuse
or
subacute
and
distinctly
3-toothed.
It
is
closely
related
to
S.
miliaceum
(Roxb.)
Snowden,
but
the
latter
has
taller
culms,
larger
wide-spreading
panicles,
and
somewhat
more
elliptic-
lanceolate
sessile
spikelets,
as
well
as
shorter
thicker
rhizomes
closely
knitted
to
form
dense
mats.
DISTRIBUTION
AND
UTILITY
S.
halepense
is
distributed
throughout
the
Mediterranean
Region,
where
it
is
most
fre-
quent
in
the
islands
and
coastal
lands
of
the
countries
bordering
the
Mediterranean
Sea.
It
also
occurs
in
many
inland
districts
where
the
soil
and
climatic
conditions
are
favourable.
It
is
usually
found
in
or
near
cultivated
lands
and
pastures,
but
it
is
also
met
with
on
unfilled
lands,
in
waste
places
and
by
roadsides.
Many
specimens
are
recorded
as
having
been
collected
in,
or
near
the
borders
of,
vineyards,
and
it
would
appear
that
S.
halepense
thrives
under
similar
conditions
to
those
which
have
been
found
to
be
most
suitable
for
Grape
Vine
cultivation.
The
occurrence
of
S.
halepense
in
some
areas
adjacent
to
the
Mediterranean
Sea,
such
as
Portugal,
the
Canary
Islands
and
Madeira
Islands,
may
be
part
of
its
natural
distri-
bution
or
due
to
some
early
introduction
from
Spain
and
North-west
Africa.
Eastwards
of
the
Mediterranean
its
distribution
extends
across
Asia
Minor
and
the
Levant
to
Iran,
Turkmen,
Afghanistan,
Western
Pakistan
and
Kashmir,
although
in
the
drier
districts
of
some
of
these
territories
it
is
generally
confined
to
irrigated
or
cultivated
lands.
After
examining
all
the
Sorghum
material
available
at
Kew
I
am
of
the
opinion
that
S.
hale-
pense
is
not
indigenous
to
any
part
of
India
as
at
present
constituted,
but
is
replaced
somewhere
in
the
neighbourhood
of
the
North-West
Frontier
Province
of
Pakistan
and
Kashmir
by
the
closely
related
S.
miliaceum,
which
has
a
more
south-easterly
distribution
and
succeeds
better
in
a
warmer climate.
S.
halepense,
however,
must
have
been
introduced
into
the
Punjab
and
United
Provinces
of
India
at
some
early
period,
and
is
now
found
in
most
provinces
of
India,
in
some
cases
having
been
used
for
the
purpose
of
experimental
cultivation
and
trial
as
a
fodder
crop
for
domestic
animals.
During
the
last
hundred
and
twenty
years.
S.
halepense
has
been
introduced
into
most
countries
throughout
the
world
where
suitable
warm-temperate
or
subtropical
conditions
prevail.
In
some
of
them,
such
as
those
of
Central
and
Western
Europe,
including
200
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
ETJ-SORGHUM
southern
England
(where
it
is
not
usually
grown
for
feeding
to
stock),
it
is
not
common,
but
is
sometimes
found
as
an
alien
on
rubbish
tips,
and
waste
lands,
or
occasionally
it
is
cultivated
in
Botanic
Gardens.
In
countries
where
the
winters
are
severe,
it
is
rarely
able
to
establish
itself
under
natural
conditions,
although
it
may
persist
in
sheltered
gardens
for
many
years.
On
the
other
hand,
in
many
territories
where
it
has
been
introduced
and
cultivated
as
fodder
for
stock,
it
has
found
conditions
so
favourable
that
it
has
become
a
noxious
weed,
difficult
to
eradicate
because
of
the
vigour
and
abundance
of
its
wide-
spreading
rhizomes.
It
is
of
interest
to
note,
however,
that
S.
halepense
has
not
extended
from
North
Africa
southwards
into
Tropical
Africa,
nor
has
it
established
itself
to
any
great
extent
in
other
tropical
countries.
This
may
be
partly
due
to
the
fact
that
fodder
crops
are
not
generally
grown
for
feeding
livestock
in
the
tropics
and
little
attention
has
been
paid
to
its
cultiva-
tion,
but
it
seems
more
probable
that
this
species
requires
a
longer
growing
period
to
mature
than
most
tropical
African
Sorghums,
as
well
as
a
distinct
resting-period
brought
about
by
cold
weather
during
a
comparatively
wet
season
which
keeps
the
rhizomes
reasonably
moist,
in
contrast
to
the
conditions
prevailing
in
most
tropical
countries,
where
the
resting
stage
usually
occurs
during
the
hot
dry
season
when
the
land
is
usually
baked
hard
with
the
sun
and
the
old
vegetation
often
subject
to
destructive
fires.
Attention
to
the
importance
of
S.
halepense
as
a
fodder
plant
was
largely
due
to
its
successful
cultivation
in
many
parts
of
the
United
States
in
the
first
half
of
the
nineteenth
century.
According
to
Piper
(in
Proc.
Biol.
Soc.
Wash.
28,
27,
1915)
it
was
first
intro-
duced
from
Turkey
in
1830.
Since
then
it
has
been
widely
cultivated
(under
the
name
of
Johnson
Grass)
in
many
States,
especially
those
south
of
latitude
36°,
where
in
some
districts
it
has
become
naturalized.
From
the
United
States
it
has
spread
to
Mexico
and
many
countries
of
Central
and
South
America,
as
well
as
the
islands
of
the
Carribean
Sea.
It
has
also
been
grown
on
a
large
scale
in
South
Africa
and
in
some
States
of
Australia,
especially
Queensland,
where
it
has
also
become
naturalized
in
some
areas.
During
later
years
its
cultivation
as
fodder
has
declined
to
some
extent,
partly
because
of
the
difficulty
of
eradicating
its
rhizomes
from
cultivated
lands,
and
also
because
of
the
more
recent
introduction
of
S.
sudanense
(Piper)
Stapf,
commonly
known
as
Sudan
Grass,
which
has
been
found
to
provide
equally
good
feed
for
stock
as
well
as
having
the
advantage
of
being
more
easily
controlled
because
of
the
absence
of
rhizomes,
which
is
characteristic
of
the
Sorghums
of
the
subsection
Arundinac,ea,
to
which
it
belongs.
Hybrid
strains
between
Sorghums
of
this
latter
subsection
and
S.
halepense
have
now
been
successfully
obtained
under
cultivation
in
South
America.
As
these
are
said
to
have
less
troublesome
rootstocks
than
the
true
species
it
is
possible
that
some
of
these
new
varieties
may
help
to
restore
S.
halepense
to
its
former
good
favour.
Specimens
of
these
hybrids
that
I
have
seen
at
Kew
show
close
relationship
with
cultivated
Sorghums
of
the
subseries
Drummondii.
Since
the
beginning
of
the
present
century
the
poisonous
properties
of
Sorghum
hale-
pense
have
received
much
attention
from
scientists,
and
the
results
of
their
investigations
have
been
published
in
various
scientific
and
agricultural
journals,
including
those
of
the
United
States
of
America,
South
Africa
and
Australia.
Useful
data
concerning
some
of
these
experiments
have
been
brought
together
in
The
Poison
Plants
of
New
South
Wales,
by
Evelyn
Hurst,
published
in
1942.
In
Montana,
United
States,
the
plants
were
found
to
be
dangerous
when
growth
was
stunted,
and
especially
noxious
when
rank
growth
was
produced
in
irrigated
soils.
In
Queensland,
Australia
(according
to
White,
in
Qd
Agric.
J.
1934),
the
plants
are
least
harmful
when
in
full
seed.
The
danger
is
minimized
if
the
plants
are
cut
and
allowed
to
wilt
before
being
fed
to
stock.
Dealing
with
Sorghums
in
the
United
States
in
1935,
Ball
considered
that
Johnson
Grass
contained
a
greater
percentage
of
HCN
than
any
other
cultivated
Sorghum.
According
to
the
New
South
Wales
Department
of
Agriculture,
the
underground
parts,
which
are
white
and
succulent,
are
said
to
have
caused
losses
in
both
cattle
and
pigs.
The
results
of
other
experiments
show
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
201
that
positive
results
of
hydrocyanic
acid
have
been
obtained
both
from
the
older
leafy
portions
and
from
young
leafy
shoots.
On
the
other
hand,
in
an
experiment
where
3
lb.
of
freshly
cut
plants,
consisting
of
old
shoots,
mainly
with
heads
well
matured,
were
fed
to
sheep
of
103
lb.
weight,
negative
results
were
obtained.
It
was
also
found
that
drying
does
not
immediately
destroy
the
toxicity
of
Johnson
Grass,
as
lucerne
hay
containing
75
%
of
Sorghum
caused
the
loss
of
twenty-eight
head
of
cattle
in
an
hour,
although
the
hay
had
been
cut
2
days
previously
and
the
sun
had
been
hot
during
that
period.
In
this
latter
experiment
the
cattle
were
very
hungry
when
fed.
Strong
positive
reactions
for
HON
were
obtained
both
from
the
hay
and
the
ingesta
of
the
dead
animals.
The
above
examples
will
be
sufficient
to
indicate
some
of
the
dangers
that
may
result
from
using
this
Sorghum
as
fodder
for
stock,
especially
in
the
case
of
freshly
cut
and
immature
plants,
although
in
countries
where
stock
have
been
accustomed
to
feed
on
it
no
harm
results
under
normal
conditions,
and
it
provides
the
animals
with
good
nourishing
food.
FORMS
AND
VARIETIES
Since
S.
halepense
was
first
recognized
as
a
distinct
species,
a
number
of
forms
or
varieties
have
been
described
by
various
authors.
A
frequent
variation
of
the
spikelets
is
the
presence
or
absence
of
an
awn
to
the
upper
lemma.
The
majority
of
the
plants
have
awned
spikelets,
but
sometimes
in
specimens
of
the
same
gathering
and
number
the
spikelets
of
one
panicle
are
awned
and
those
of
another
awnless,
and
it
is
not
uncommon
to
find
most
of
the
spikelets
in
a
panicle
awnless
but
accompanied
by
a
few
that
are
distinctly
awned.
The
awned
form
has
generally
been
accepted
as
typical
of
the
species,
and
Hackel
(in
DC.
Monogr.
Phan.
6,
502,
1889)
describes
it
under
Andropogon
sorghum
(Linn.)
Brot.,
subsp.
halepensis
(Linn.)
var.
halepensis
subvar.
genuinus
Hack.
It
is
referred
to
by
Stapf
(in
Hook.
Fl.
Brit.
Ind.
7,
182,
1896)
under
A.
halepensis
(Linn.)
Brot.,
var.
genuinus
Hack
(under
which
he
also
includes
Andropogon
miliaceus
Roxb.)
;
and
by
Acherson
and
Graebner
(in
Synops.
Mitteleur.
Fl.
2,
Pt.
1,
47,
1898)
under
A.
halepensis
(Linn.)
Brot.
var.
typicus.
The
awnless
state
was
described
by
Hackel,
loc.
cit.,
as
A.
sorghum
(Linn.)
Brot.,
subsp.
halepensis
(Linn.)
var.
halepensis
subvar.
muticus
Hack.,
and
it
has
also
received
the
following
names:
A.
halepensis
(Linn.)
Brot.
var.
muticus
(Hack.)
Aschers.
&
Graebn.
(loc.
cit.);
A.
halepensis
(Linn.)
Brot.
subsp.
anatherus
Piper
(in
Proc.
Biol.
Soc.
Wash.
28,
28,
1915);
and
Sorghum
halepense
(Linn.)
Pers.
f.
muticum
(Hack.)
C.
E.
Hubbard
(in
Hook.,
Ic.
Pl.
34,
1938,
under
t.
3364,
p.
4).
A
glabrous
form
of
the
awned
state
was
described
by
Hackel,
loc.
cit.,
as
Andropogon
sorghum
(Linn.)
Brot.
subsp.
halepensis
(Linn.)
var.
halepensis
subvar.
leiostachys.
In
addition,
Hackel,
loc.
cit.,
refers
to
a
broad-leaved
form,
Sorghum
halepense
(Linn.)
var.
latifolium
Willk.
&
Lange
(Prodr.
Fl.
Hisp.
1,
48,
1861),
which
is
described
as
5
ft.
high,
with
leaves
-1
in.
wide,
and
a
panicle
a
foot
or
more
long.
Although
I
have
seen
two
of
Willkomm's
specimens
from
Spain,
neither
of
them
is
from
the
type
locality,
nor
do
they
agree
with
the
above
description.
On
the
other
hand,
this
name
and
description
might
well
be
applied
to
some
of
the
robust
forms,
found
in
Europe
and
other
regions,
that
have
evidently
been
grown
under
most
favourable
conditions
and
are
probably
either
cultivated
plants
or
grown
in
or
near
to
cultivated
lands.
In
contrast
to
these
robust
forms
there
are
dwarf
specimens
with
slender
culms
about
50
cm.
high,
leaves
0.5-1
cm.
wide,
and
small
panicles,
that
evidently
represent
im-
poverished
plants
grown
under
adverse
conditions,
such
as
infertile
soils
and
low
rainfall,
or
may
have
resulted
from
close
grazing
by
stock
or
frequent
cutting
for
hay.
Often
such
specimens
have
well-developed
rhizomes,
and
under
favourable
conditions
would
have
grown
into
average-sized
plants.
There
are
also
slender-culmed
forms
that
appear
to
have
matured
under
shady
conditions
that
have
led
to
elongation
and
weakness
of
the
culms
with
looser
longer-branched
panicles
than
usual.
When
grown
in
the
open
the
202
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
lower
branches
of
the
panicle
are
generally
comparatively
short
except,
as
sometimes
happens,
where
only
one
or
two
of
the
lowermost
whorl
develop,
because
the
panicle
fails
to
emerge
properly
from,
or
is
too
tightly
held
by,
the
uppermost
leaf-sheath.
In
such
cases
the
remaining
branch
or
two
tend
to
be
longer
than
usual
and
should
be
disregarded
when
reckoning
the
average
length
of
the
panicle
branches.
The
size
of
the
spikelets,
also,
may
differ
to
some
extent
according
to
the
robustness
of
the
plants,
but
they
also
vary
somewhat
in
shape
and
size
with
the
stage
of
development.
They
are
generally
longest
at
the
flowering
period
or
immediately
after
flowering,
tending
to
become
a
little
shorter
and
narrower
as
they
become
more
plump
and
rounded
with
the
increasing
size
of
the
grains.
As
most
of
the
above
forms
and
variations
appear
to
be
mainly
due
to
the
influence
of
climate
and
habitat,
and
careful
examination
of
the
material
at
Kew
has
failed
to
reveal
reliable
characters
suitable
for
dividing
the
species
into
distinct
varieties,
I
have
thought
it
best
to
leave
these
variable
forms
to
be
investigated
in
the
field,
by
growing
them
under
similar
conditions,
in
order
to
determine
if
any
of
them
are
worthy
of
varietal
status.
In
the
enumeration
of
the
specimens
given
below,
however,
the
awned
state
is
denoted
by
an
*
immediately
after
the
name
or
number
of
the
collector.
A
few
more
notable
variations
are
mentioned
in
the
text.
ENGLAND.
Southern
counties:
Fishponds,
Bristol,
Sandwith
*
!
Brighton,
Sussex,
Smart
*!
Kew,
Surrey,
Snowden!
Middlesex,
Welch
*!
Dagenham,
Essex,
Lousley*!
Melville
*!
Wanstead,
Nash
*!
FRANCE.
Mostly
in
southern
districts:
Banks
of
the
Loire
River,
Perret
*
!
Avignon,
Hort.
Bot.
Divionensis
*!
Delacour
*!
Requiem!
Tarascon,
Gay
!
CORSICA.
Etang
de
Biguglia,
Mabille
66
!
66
*
!
Corte
district,
Kralik,
Pl.
Gorses
*!
SPAIN.
Motril,
Granada
province,
Willkomm,
iter
Hispanicum
286
!
Chiva,
near
Valencia,
Willkomm,
iter
Hispanicum
528
*
!
Near
Murcia,
Bourgeau,
Pl.
d'Espagne
1533
*
!
BALEARIC
ISLANDS.
From
sea-level
up
to
1220
m.
Minorca:
Fornells
(Sept.
1935),
Kennedy
86
*
!
Majorca
:
Pollensa
and
Aleudia
(July,
1936),
Kennedy
86
!
253
*
!
PORTUGAL.
Near
Lumiar,
Welwitsch,
Pl.
Lusitanicae,
Inlio
*
!
SWITZERLAND.
At
the
Rhone
dam,
near
Sion,
Kneucher,
Gram.
Exsicc.
182
!
HUNGARY.
Aleppoi
Czirok,
Degen
in
Gram.
Hungarica
101
*
ITALY.
Lombardia
district,
Petri
Portae
(July
1865)
*!
(Oct.
1867)
*
!
Near
Riva,
on
Lake
Garda,
Kneucher,
Gram.
Exsicc.
63
*
!
Parma,
Gay
(Aug.
1826)
*
!
Near
Naples,
Alexander
(1845)
!
(July
1845)
*
!
SICILY.
Near
Pelizzi,
Nebrodi
mountains,
Flora
Nebrodensis,
Strobly
(Aug.
1873)
*
!
Palermo,
Todaro,
Fl.
Sieuta
Exsicc.
496
!
496
*
!
TRIESTE.
Without
precise
locality:
Prihoda
(Oct.
1864)
*
!
Tidoll
*!
YUGOSLAVIA.
Istria,
Zaule,
Schultz
herb.
normale
Englehardt
2599
*!
Near
Fiume
(Rijeka),
Smith
(July
1869)
*
!
Constantza
near
Mangalia,
Cretzoiu
(July
1827)
*
!
Isaccea,
in
Tulcea
district,
Cretzoiu
(July
1837)
*
!
Lesina
Island,
Dalmatia,
Bottori
*!
The
specimen
from
Istria,
Schultz
2599,
is
a
robust
form
that
may
have
been
grown
on
cultivated
lands.
ALBANIA.
Between
Durazzo
(Durres)
and
Berat,
Alston
&
Sandwith
2516
!
This
is
a
robust
specimen
with
the
panicle
in
fruit.
RUMANIA.
Without
precise
locality,
Trivaldski
(1837)
*
!
BULGARIA.
Hills
north
of
Varna,
Gilliat-Smith
984
*
!
GREECE.
Macedonia,
at
Kopriva
and
south
of
Karamudli,
Turrill
368
*
!
490
!
490
*
!
Thrace,
at
Xanthi,
Todd
444
*
!
Attica,
Heldreich,
Pl.
exsicc.
de
Graecia
*!
Drakofto,
Pelloponnesus,
Davis
1012
!
Without
precise
locality,
in
Graecia
(Berger)
Mus.
Month
(1837)
*!
CEPITALONIA.
Without
precise
locality,
in
vineyards,
Schimper
&
Wiest
(1834)
*
!
CRETE.
Canea
province:
Kissamos
and
Platania,
Elie&
Reverchon,
Pl.
de
Crete
(1883
&
1884)
183
*
!
Gandoger,
Herb.
Creticum
7359
*
!
J.
D.
SNOWDEN
:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
203
TURKEY
(in
EUROPE).
Thrace
:
Gallipoli,
Suala,
and
Anzac,
Ingoldby
595
!
595
*
!
Kett
90
*
!
TURKEY
(in
ASIA).
Antalya,
30
m.
Tengwall
31
*!
Ain-Halakim,
Mount
Nusairy,
760-900
m.,
Haradjian
3470
*!
Near
Ermenck,
Plains
of
Cilicia,
Cosson
213
*!
Near
Mardin,
Kurdistania,
Sintensis
1341
*
!
Schiherb
sur,
Kurdistania,
Hausknecht
(1867)
!
Karadagh,
Fischer!
CYPRUS.
Kythraeae,
Sintensis
679
*!
Timben
and
Limassol,
15
m.,
Syngrassides
389
!
717
*!
Kyrenia,
at
sea-level,
and
Platres,
1130
m.,
Kennedy
82
!
83
*!
84
*!
Near
Perapedhi,
1067
m.,
Davis
1851
*!
Larnaca,
Ussher
124
*!
Nicosia,
Lindberg
*
!
With-
out
precise
locality,
often
causing
poisoning
to
animals,
Rae
*
!
SYRIA.
Northern
Syria,
Aintab,
610
m.,
Haussknecht,
Syriaco-Armeniacum
(1865)
*!
Southern
Syria,
El
Hutch,
Lowe
(1863-4)
*!
Without
precise
locality,
Gandoger
(1920)
*
!
Kneucker
Gram.
Exsiec.
63a
*
!
Blanche
56
*!
PALESTINE.
Amaltha,
Hayne
(Oct.
1872)
*
!
Ghor-es-Safieh,
Hart
(1883-4)
*!
Wad
el
Kelt,
Khan
Yubb
Yusif
to
AM
el
Tabigha,
sea-level
to
100
m.,
ex
American
Colony,
Jerusalem
B
6066
*!
B
66
*!
Northern
Circle,
poisonous
to
grazing
animals
(except
horses,
mules
and
donkeys),
Agric.
&
Forests
Dep.
(Mason)
4
*
!
Jerusalem,
Forest
Dep.
(Gabrielith)
48
*!
48
!
TUNISIA.
Near
Zaghouan,
Kralik
310
*
!
ALGERIA.
Oran
province,
Sidi-bel-Abbes,
Warion
2286
*
!
Alger
province,
Blidah,
Lefebre
186
!
Stora,
Bove
(Aug.
1889)
!
MOROCCO.
Southern
Morocco,
Mzuda,
Hooker
(May
1871)
*
!
CANARY
ISLANDS.
Gran
Canaria,
Beo
de
Angustura,
ex
Herb.
Monig.
Murray
*
!
Without
precise
locality,
presented
Hooker,
Feb.
1868,
Gay!
MADEIRA
ISLANDS.
Near
Gorgulho,
ex
Herb.
Monig.
Murray
*
!
Mandon
262
!
Near
Funchal,
Clarke
94
*!
Grabham
31
!
Lowe
26
!
Dos
Saltos,
Hooper
(Jan.
1842)
*
!
Without
precise
locality,
Lawrence
(July
1837)
*
!
Leman!
Lowe!
Masson
*!
BELGIAN
CONGO.
Introduced
as
a
fodder
plant,
Eala,
Lebrun
604
!
SOUTHERN
RHODESIA.
Introduced
for
trial
as
a
fodder
plant,
Agricultural
Experiment
Station,
Salisbury,
Mundy
50
b,
Trapnell
754
*!
Insiza,
1370
m.,
Capstick
2328
*!
Lomagundi
district,
1070
m.,
Eyles
3149
*!
PORTUGUESE
EAST
AFRICA.
Zambesi
Delta,
Boruma,
Meyrtharth,
1046
*!
SOUTH
AFRICA.
Transvaal:
Pretoria,
Froenkloof
Experiment
Farm,
and
without
pre-
cise
locality,
Nat.
Herb.
Pretoria
H.
20939
*
!
8498
*
!
(Waterval
No.
7)
1544
*
!
Natal
:
in
Botanic
Gardens,
not
indigenous,
Medley
Wood
6675
!
Medley
Wood's
specimen
has
a
loose
panicle
and
looks
as
though
the
plant
had
been
grown
in
a
shady
habitat.
IRAQ.
Northern
district,
Imperial
Institute
I!
Remount
Farm,
Karredah,
Graham
*
!
West
Baghdad,
Rogers
0303
!
Rustam
Experimental
Farm,
Baghdad,
Lozer
3416
*!
Guest
184
!
275
*!
2459
*!
Aziziyah,
Guest
3567
!
Zauritch,
915
m.,
Guest
3750
!
Basra,
Haussknecht,
Syriaco-Armeniacum
(1865)
!
IRAN.
Near
Teheran,
Trott
863
*!
TURKMEN
(TURCOMAN).
Kizil
Arvat,
Karakala,
Sintensis
1890
!
AFGHANISTAN.
Without
precise
locality,
Griffith
356
*!
661
*!
6825
!
PAKISTAN.
Baluchistan:
Bolan
Pass
(ex
Herb.
Leman,
1852)
Griffith
*
!
Chitral,
1500
m.,
Harris
*
!
Drosh,
1370
m.,
Toppin
307
*!
Without
precise
locality,
Wingate
*
!
Gilgit:
Hunza
Valley,
1675
m.,
Duthie
12318
*!
KASHMIR.
Shadipur—Jhelum
Valley,
Ludlow
&
Sherriff
7803
!
Without
precise
locality,
near
towns,
Jacquemont
(388)
656
*
!
Herb.
Falconer
*
!
INDIA.
Punjab:
West
of
Chenab
River,
Hooker
&
Thomson
(Nov.
1846)
*!
Plains
below
Mumior,
Hooker
&
Thomson
*
!
Bhyrowal,
45
miles
east
of
Lahore,
Hooker
&
Thomson
(Sept.
1846)
1542
!
Timoo
Ghat,
Hooker
&
Thomson
(Oct.
1846)
!
Without
precise
locality,
Hooker
&
Thomson
*
!
Jellundur,
Wingate
(Sept.
1888)
*!
In
valley
204
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
below
Nahan
Sirmur,
Jacquemont
(144)
2518
!
United
Provinces:
near
Mustafabad,
Hooker
&
Thomson
(Oct.
1846)
!
Bengal:
Botanic
Gardens,
Calcutta,
Narayanaswami
(April,
1918)
*!
Serampore,
Griffith
*
!
Madras:
Samalkota,
Godaveri
district;
Betta
Mogilam,
Hosur,
and
Metigiri,
Salem
district,
Madras
Herb.
S.
India
Flora
(June
1917)
*
!
(Dec.
1913)
!
(Dec.
1916)
!
15451
*!
Shencottah
to
Aryankavu,
Travancore
State,
Calder
&
Ramaswami
669
*!
Province
and
locality
uncertain,
Wight
Herbarium
1673
*!
Robust
specimens
from
plants
used
for
hybridization
experiments
with
Saccharum,
Janaki
Ammal
1236,
which
have
persistent
spikelets,
are
probably
hybrid
forms
of
Sorghum
halepense.
A
specimen
from
the
Botanic
Gardens,
Calcutta,
Nashe
1317,
which
also
has
persistent
spikelets,
may
have
resulted
from
intercrossing
between
S.
halepense
and
a
cultivated
Sorghum.
CEYLON.
Peradeniya
Herbarium
(seeds
from
Drieberg)
Galston
2240
!
JAVA.
Without
precise
locality,
Posthumus
89
!
A
poor
specimen
from
Java,
raised
from
seeds
introduced
from
Valparaiso,
Brigsman,
(Mar.
1914),
may
belong
to
S.
halepense,
but
the
material
is
too
scanty
for
exact
determination.
FORMOSA.
Kontay
Experiment
Station,
near
Koshun,
from
a
sample
of
Johnson
Grass
hay,
Price
(June
1912)
*I
AUSTRALIA.
Western
Australia:
Bassendean,
Palmer
*
!
Bayswater,
near
Perth,
Stoward
234
!
Williambury,
Minilya
River,
Bunbury
*
!
Queensland:
near
Brisbane,
Bailey
(Dec.
1913)
*
!
Blake
318
*!
Burpengary,
Blake
136
*!
Cloncurry,
Blake
201
!
Near
Brisbane;
Moreton
district;
between
Grandchester
and
Laidley;
near
Mt.
Gravat
;
between
Laidley
and
Forest
Hill;
and
between
Teirotville
and
Anthony,
Hubbard
3026
!
4316
*!
4732
*!
5285
*1
5352
*1
53881
8075
*!
8906
*1
Forest
Hill,
White
5
*!
Johnson
River,
Harding
(Nov.
1907)
*!
Kigaroy,
Jensen
*
!
Atate
Farm,
Hermitage,
Warwick,
Liverseed
8
*!
Riversleigh
Eidsvold
Line,
Turner!
Christmas
Creek,
McIntyre
Bros.
7
!
New
South
Wales:
Ashfield,
Cheel!
Without
precise
locality,
Brisbane
Botanic
Gardens
Herbarium
9
*
!
Victoria
:
Without
precise
locality,
National
Herbarium,
Melbourne!
NEW
ZEALAND.
Napier,
naturalized,
Kirk
1387
*1
FIJI
ISLANDS.
Introduced
experimentally
to
feed
stock
but
now
a
troublesome
weed,
without
precise
locality,
Greenwood
392
!
Greenwood's
specimen
has
rhizomes
and
leaves
typical
of
Sorghum
halepense
but
it
has
no
inflorescence.
HAWAII.
Nuwanu
Valley,
Oahu,
Forbes!
UNITED
STATES.
California:
Redding,
Shasta
county,
Smith
744
*!
Santa
Cruz
county,
Woolley-Dod
453
!
Mesa
Grande,
760
m.,
San
Diego
county,
Spencer
1393
*!
Oklahoma:
Harmon
county,
Stevens
1087
!
Texas
:
San
Antonio,
Bexar
county,
185
m.,
Heller
1706
*!
Without
precise
locality,
ex
United
States
National
Herbarium,
Walley
(1886)
*!
Arkansas:
Van
Buren
Bluffs,
Crawford
county,
Demaree
3167!
Louisiana:
Without
precise
locality,
Ball
80
!
Tennessee:
Near
Nashville,
Curtiss
3644A
!
Virginia:
Norfolk,
Hitchcock
277
!
South
Carolina:
near
Columbia,
Curtiss
3644
*
!
Florida:
Indian
River,
Merritt
Island,
Curtiss
5724
!
Hillsborough
county,
Fredholm,
6346
!
Miami
and
Buena
Vista,
Dade
county,
Moldenke
522
!
704
*!
Only
a
few
spikelets
of
Moldenke's
704
are
awned.
MEXICO.
Sonora:
Guamas,
Palmer
64!
Coahuila:
Saltillo,
1600
m.,
Arsgne
3464
*!
Nuevo
Leon:
near
Monterrey,
540
m.,
Abbon
*
!
Arsene
498
*!
Michoacan:
Zitacuara
to
Las
Anonas
and
to
San
Jose
Purua,
1575-1775
m.,
Hinton
13140
*!
13270
!
13270
*!
Guerrero
:
Pungarabato,
Coyuca,
Hinton
5457
!
Mexico
:
Temaxcaltepec
district,
1750
m.,
Hinton
4031
!
1427
!
1427
*!
Yucatan:
Izamal,
Gaumer
720
!
The
specimen
Hinton
4031
has
the
appearance
of
having
been
grown
in
a
shady
habitat.
J.
D.
SNOWDEN
:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
205
BERMUDA
ISLANDS.
Paget,
and
Trott's
Pond,
Brown
&
Britton
130
!
135
!
CUBA.
Near
Havana,
Curtiss
561
*
!
PORTO
Rico.
Finca,
Bueno
Causejo,
Rio
Piedras,
Hioraw
*
!
SANTO
DOMINGO.
Cibas
Valley,
Santiago
province,
Ekman!
JAMAICA
Blue
Mountains
district
:
St
Andrew
and
Resource,
Perkins
1479
*
!
1139
*
!
Resource,
1100
m.,
Harris
11521
!
1000
m.,
Harris
11582
*
!
Richmond
Park,
St
Andrew,
Harris
12706
!
Hope
Grounds,
215
m.,
Harris
*
!
Without
precise
locality,
Botanic
Department,
Jamaica
2004
*
!
The
unnumbered
specimen
of
Harris,
from
Hope
Grounds,
has
only
a
few
spikelets
with
awns.
ST
LUCIA.
Near
Government
Experiment
Station,
Union,
Box
323
!
COSTA
RICA.
Las
Concavas,
Lankester
210
*
!
VENEZUELA.
Cetu,
Caracas
and
La
Guayra,
610-915
m.,
Fendler,
Plantae
Venezue-
lanae,
1854-55,
1653
!
Caracas,
915
m.,
Munro,
Plantae
Venezuelanae,
July
1858,
2629
!
Cumana,
Flincke
245
!
BRITISH
GUIANA.
Botanic
Gardens,
Georgetown,
Actson
115
*
!
This
specimen
is
a
shady
habitat
form.
PARAGUAY.
Cordillera
de
Altos,
Hassler,
Plantae
Paraguarienses,
3346
*
!
ARGENTINA.
Villa
Sujan,
460
m.,
Tucuman
province,
Venturi
313
*
!
Near
Cnel.
Roca,
250-360
m.,
Fischer
216
!
Near
Santa
Fe,
Santa
Fe
province,
Job
1067
*
!
URUGUAY.
Soriano,
Mercedes
district,
Berro
6387
*
!
2.
Sorghum
miliaceum
(Roxb.)
Snowden
Andropogon,
miliaceus
Roxb.,
Fl.
Indica,
1,
276,
1820,
is
based
on
specimens
grown
at
Calcutta
from
seeds
collected
in
the
mountains
north
of
Oude,
which
were
sent
from
Lucknow
under
the
name
of
Hill
Grass.
There
are
no
authentic
specimens
of
Roxburgh's
plant
at
present
known
to
me,
but
the
drawing
of
it,
no.
1717
in
Icones
Roxburghiana;e
1
,
illustrates
the
large
size
and
loose
branching
of
the
panicle,
and
the
dissections
of
the
sessile
spikelets
show
them
to
be
short
and
wide
with
the
apex
obtuse,
as
stated
in
Roxburgh's
description.
Two
enlarged
drawings
of
the
sessile
spikelets
are
given.
In
one
of
them
the
upper
lemma
has
a
short
awn
less
than
twice
the
length
of
the
lemma,
but
in
the
other
it
is
awnless.
The
glumes
are
shown
as
shortly
hairy
and
the
lemmas
as
shortly
ciliate.
Roxburgh
states
that
the
leaves
are
ensiform-lanceolate,
2-3
ft.
long
and
1-2
in.
wide
near
the
base
where
they
are
broadest.
Andropogon
miliformis
Schultes,
Mantissa,
2,
448,
1824,
is
a
change
of
name
for
A.
miliaceus
Roxb.
This
change
of
name
was
made
by
Schultes
because
the
name
A.
miliaceum
had
appeared
(without
description
or
author)
in
Brouss.,
Cat.
Monsp.
p.
5,
1804,
among
several
other
species
of
Andropogon,
also
without
descriptions
or
authors
except
for
A.
aculeatum
Forsk.,
which
appeared
near
the
bottom
of
page
4.
Roemer
et
Schultes,
Syst.
Veg.
2,
828,
1817,
refer
in
a
note
to
four
species
of
Andro-
pogon,
which
are
insufficiently
known
to
them,
thus
:
Quid:
Andropogon
foveolatus
Descript
de
l'Egypt,
t.
8?
A.
aculeatus
Forsk.
apud
Brouss.,
Cat.
Monsp.
1804.
A.
miliaceus
Forsk.
apud
Brouss.,
Cat.
Monsp.
1804.
A.
radiatus
P.
de
Beauv.
There
is
nothing
in
Brouss.,
Cat.
Monsp.
1804
to
connect
Forskal
with
A.
miliaceum,
and
it
seems
possible
that
Schultes
may
have
been
mistaken
in
assuming
that
because
it
appeared
immediately
below
A.
aculeatus
Forsk.
in
Roemer
et
Schultes,
loc.
cit.,
A.
miliaceus
was
also
one
of
Forskal's
species.
Kunth,
Enum.
Pl.
1,
507,
1833,
and
Dietrich,
Synops.
Pl.
1,
410,
1839,
refer
to
A.
miliformis
Schultes
as
a
synonym
of
1
These
Icones
consist
of
a
few
sets
of
unpublished
drawings
of
Roxburgh's
plants,
one
set
of
which
I
have
seen
at
Kew.
206
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
E17-SORGHUM
A.
miliaceus
Roxb.,
and
Steudel,
Nonzencl.
Bot.
ed.
2,
1,
92,
1840,
refers
to
the
latter
as
A.
miliaceus,
Brouss.,
Forsk.,
Aegyptus
?
Hackel,
in
DC.,11
Monogr.
Phan.
6,1889,
makes
no
reference
to
A.
miliaceus
Forsk.,
but
(loc.
cit.,
p.
541)
places
A.
miliformis
Schultes
in
synonomy
under
A.
miliaceus
Roxb.
In
Roxburgh's
Hortus
Bengalensis,
p.
1814,
A.
miliaceus
is
stated
as
having
been
received
from
General
Martin,
from
Hurdwar
(probably
north
of
Delhi).
As
General
Claude
Martin
died
on
13
September
1800,
the
seeds
must
have
been
sent
to
Roxburgh,
f
,
I
;
I
1
/,
A
C
r
B
;
/,
;
411
if
D
/
0
E
ii
I.
'
F
(1)
Sorghum
propinquum
(Kunth)
Hitchc.
A,
sessile
and
pedicelled
spikelets;
B,
lower
glume;
C,
upper
glume,
exterior
view
(from
Loher
1807).
(2)
S.
miliaceum
(Roxb.)
Snowden
var.
miliaceum.
D,
sessile
and
pedicelled
spikelets;
E,
lower
glurne;
F,
upper
glume,
exterior
view
(from
Snowden,
Kew,
Sept.
1933).
A
and
D,
x
5;
B,
C,
E
and
F,
x
10.
who
came
to
the
Calcutta
Gardens
in
1793,
between
the
latter
date
and
1800.
As
Rox-
burgh
was
exchanging
seeds,
etc.,
with
Sir
Joseph
Banks,
and
Broussonet
mentions
receiving
seeds
from
Sir
Joseph,
it
is
quite
possible
that
the
A.
miliaceum
mentioned
by
Broussonet
was
that
of
Roxburgh,
and
had
been
raised
from
seeds
sent
to
Sir
Joseph
Banks
by
Roxburgh
and
passed
on
to
Broussonet.
After
considering
all
the
available
evidence,
I
have
come
to
the
conclusion
that
there
was
only
one
A.
miliaceus,
namely
the
one
described
by
Roxburgh
in
Fl.
Indica,
1,
276,
1820.
Sorghum
giganteum
Edgew.,
J.
As.
Soc.
Bengal,
21,
181,
1853,
was
described
in
an
account
of
the
plants
growing
in
the
Banda
district,
Allahabad
division,
of
the
United
Provinces,
India.
Both
from
the
description
and
from
the
locality,
it
is
evident
that
this
species
is
the
same
as
the
one
Roxburgh
had
previously
described
as
Andropogon
milia-
J.
D.
SNOWDEN
:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
207
ceus.
The
type
specimen
of
Edgeworth's
species
has
not
been
located,
but
in
his
herbarium
at
Oxford
there
is
a
specimen,
no.
361,
Keirat
Rangal',
which
agrees
with
the
above
description
and
may
be
part
of
the
same
gathering.
Sorghum
miliaceum
(Roxb.)
Snowden,
comb.nov.,
based
on
Andropogon
miliaceus
Roxb.,
Fl.
Indica,
1,
276,
1820,
for
the
reasons
already
mentioned
above,
is
here
proposed
as
the
correct
name
for
this
group.
There
are
three
specimens
in
the
Wallich
Herbarium
at
Kew,
nos.
8778
A,
B
and
C,
that
belong
to
this
species
and
agree
well
with
the
illustration
no.
1717
in
hones
Roxburghianae,
especially
if
the
panicles
had
been
spread
out
in
a
more
natural
manner
when
mounted.
Of
these
I
select
no.
8778
B
to
be
linked
with
the
above
drawing
of
Roxburgh's
plant,
as
a
typical
specimen,
to
be
used
until
a
more
authentic
type
has
been
located.
DESCRIPTION
A
tall
loosely
tufted
perennial
with
short
thick
somewhat
densely
matted
rhizomes.
Cu/ms
slender
to
rather
stout,
2-3
m.
or
more
high,
0.5-1
cm.
wide
near
the
base,
simple
or
branched
above.
Leaves
about
10-12
;
blades
ensiform-lanceolate
to
linear,
from
a
rather
broad
base,
gradually
attenuated
upwards
to
the
pointed
apex,
30-90
cm.
long,
1-4
cm.
wide,
more
or
less
smooth
except
for
the
serrate-ciliate
margins,
the
midrib
paler
than
the
remainder
and
somewhat
prominent
below.
Panicle
large,
loose
and
spreading,
ovate-elliptic,
ovate-lanceolate,
or
pyramidal,
mostly
25-55
cm.
long,
and
10-25
cm.
wide
;
branches
subverticillate,
slender
and
flexuous
with
the
tips
often
drooping,
lower
often
4-
to
5-nate,
10-20
cm.
long,
often
naked
for
2-5
cm.
from
the
junction
with
the
main
axis.
Racemes
fragile,
2-
to
5-noded,
10-22
mm.
long;
internodes
and
pedicels
slender,
2-4
mm.
long,
densely
ciliate
with
whitish
to
fulvous
or
tawny
hairs
0.5-1
mm.
long.
Sessile
spikelets
elliptic-lanceolate
to
subelliptic,
more
or
less
obtuse,
4-5.5
mm.
long,
1.5-2-3
mm.
wide,
straw
yellow
to
deep
cream
with
3-5
greenish
nerves
showing
near
the
tip
of
the
lower
glume
when
young,
at
length
tawny
to
mahogany
red
or
almost
black,
sparingly
strigose,
especially
near
the
base,
sides
and
above
the
middle;
awned
or
awnless;
callus
densely
to
moderately
bearded.
Glumes
coriaceous,
but
some-
what
thinner
towards
the
tips;
lower
8-
to
12-nerved,
with
a
few
cross
nerves
above
the
middle,
keeled
in
the
upper
third,
with
the
keel-wings
spinulously
ciliolate
and
ending
in
minute
teeth
below
the
hyaline
apex
to
form
a
3-toothed
tip;
upper
7-nerved,
the
two
adjoining
the
median
nerve
often
rather
short,
with
a
few
cross
nerves
in
the
upper
half,
scabrid
on
the
keel
in
the
upper
third.
Lemmas
moderately
to
sparingly
ciliate
;
lower
elliptic-oblong
to
elliptic-lanceolate,
3.5-5
mm.
long;
upper
ovate
to
ovate-elliptic,
2.5-4
mm.
long,
2-lobed
to
0.5-1
mm.
deep
and
with
an
awn
10-16
mm.
long,
or
some-
times
subentire
and
mucronate.
Anthers
2.5-3.5
mm.
long.
Grains
obovate-elliptic
to
obovate-oblong
2-2.5
mm.
long,
1-1.5
mm.
wide,
yellowish
brown
to
dark
purplish
brown;
embryo-mark
obscure.
Pedicelled
spikelets
usually
6,
5-6.5
mm.
long,
at
length
deciduous,
often
purplish;
glumes
herbaceous,
lower
7-
to
8-nerved,
upper
5-
to
7-nerved;
lemmas
well
developed;
anthers
2.5-3-5
mm.
long.
DIAGNOSTIC
CHARACTERS
AND
AFFINITIES
The
distinguishing
features
of
Sorghum
miliaceum
are
the
thick
rhizomes
;
tall
culms
2-3
m.
or
more
high
;
leaf-blades
1-4
cm.
wide
;
large
loose
panicles,
mostly
25-55
cm.
long,
10-25
cm.
wide;
sessile
spikelets
elliptic-lanceolate
to
subelliptic,
more
or
less
obtuse,
4-5.5
mm.
long,
1.5-2-3
mm.
wide,
awned
or
awnless;
and
the
coriaceous
glumes,
with
the
lower
more
or
less
distinctly
3-toothed
at
the
tip.
It
is
closely
related
to
S.
halepense
but
differs
from
it
in
the
thicker
rhizomes,
taller
culms,
wider
leaf-blades,
larger
panicles,
and
the
sessile
spikelets
usually
tending
more
towards
elliptic-lanceolate
JOURN.
LINN.
SOC.—BOTANY,
VOL.
LV
0
208
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
in
outline.
From
S.
controversum
it
differs
in
the
comparatively
broader
and
more
obtuse
sessile
spikelets,
the
more
coriaceous
and
less
hairy
glumes,
with
the
lower
more
distinctly
3-toothed
at
the
tip,
and
the
broader
leaf-blades
up
to
4
cm.
wide.
DISTRIBUTION
AND
UTILITY
The
natural
distribution
of
S.
miliaceum
extends
from
Kashmir
and
the
North-West
Frontier
Province
of
Pakistan,
across
northern
India,
including
the
Punjab,
United
Provinces,
Central
Provinces,
and
Bengal,
southwards
to
Madras
Province.
Its
taller
and
more
robust
culms
indicate
that
it
is
more
of
a
jungle
grass
than
S.
halepense,
and
this
is
confirmed
by
it
having
been
found
growing
in
forests,
probably
in
clearings
or
near
the
outskirts.
Little
information
is
available
as
to
its
cultivation
and
its
introduction
into
other
countries,
but
it
has
been
grown
experimentally
in
the
United
States,
and
also
at
Kew.
It
may
also
have
reached
Mexico,
either
from
the
United
States
or
direct
from
India
as
some
Mexican
specimens,
referred
to
below,
are
remarkably
like
the
typical
S.
miliaceum
of
India.
According
to
Piper
(who
refers
to
it
under
Andropogon
halepensis
miliformis
Schultes,
in
Proc.
Biol.
Soc.
Wash.
28,
29,
1915)
seed
collected
by
Hartless
at
Saharanpur,
in
the
United
Provinces
of
India,
was
introduced
into
the
United
States
and
cultivated
at
Arlington
Farm,
Virginia,
where
the
plants
grew
to
a
much
larger
size
than
those
of
typical
Sorghum
halepense,
and
differed
conspicuously
in
the
larger
looser
panicles,
besides
having
very
abundant
short
thick
rootstocks
in
a
dense
tangled
mass.
In
the
Royal
Botanic
Gardens,
Kew,
where
in
1933
I
collected
specimens
of
a
Sorghum
which
proved
to
be
S.
miliaceum,
from
plants
that
were
growing
near
to
some
clumps
of
S.
halepense,
similar
differences
were
also
noticeable,
as
well
as
the
broader
leaves,
3-5
cm.
wide.
Unfortunately
all
the
plants
of
S.
miliaceum
at
Kew
have
since
died,
apparently
being
unable
to
survive
the
English
winters,
although
those
of
S.
halepense
have
persisted
for
more
than
twenty
years.
As
S.
miliaceum
has
not
generally
been
recognized
as
distinct
from
S.
halepense,
there
is
little
information
available
regarding
either
its
utility
or
its
toxic
properties.
Although
its
stouter
culms
and
wider
leaves
would
provide
plenty
of
good
fodder,
it
might
not
be
so
easy
to
handle
with
machinery.
The
plants
would
also
seem
to
require
a
warmer
climate
than
those
of
S.
halepense,
and
their
thick
rootstocks
might
be
just
as
difficult
to
eradicate
as
those
of
the
latter
species.
On
the
other
hand,
S.
miliaceum
is
likely
to
prove
more
suitable
than
S.
halepense
for
cultivation
under
tropical
conditions
in
countries
where
there
is
no
cold
winter
season.
(i)
Var.
miliaceum
Var.
miliaceum
is
the
typical
variety
founded
on
Andropogon
miliaceus
Roxb.
(Fl.
Indica,
1,
276,
1820)
and
Sorghum
giganteum
Edgew.
(J.
As.
Soc.
Bengal,
21,
181,
1853).
As
the
type
specimens
of
the
above
species
have
not
yet
been
located,
I
have
selected
Wallich
8778B,
in
the
Wallich
Herbarium
at
Kew,
linked
with
Icones
Roxburghianae
Fig.
no.
1717,
to
be
used
in
their
place
until
they
are
found.
Sessile
spikelets
4.8-5.5
mm.
long,
1.8-2.3
mm.
wide,
usually
awned;
panicle
with
the
lower
branches
usually
1-
to
4-nate,
stoutish
and
stiff
below
after
flowering,
slender
and
flexuous
upwards.
ENGLAND.
Growing
in
herbaceous
grounds,
Royal
Botanic
Gardens,
Kew,
Sept.
1933,
Snowden!
PAKISTAN.
North-West
Frontier
Province
:
Hazara,
Hilkoti,
Saran
Range,
Inayat
(Sept.
1898)
presented
Duthie!
Hazara,
Kagan
Valley,
Inayat
(Aug.
1896;
Herb.
Bot.
Dep.
N.
India
20294a)
presented
Duthie!
Gilgit,
Winterbottom
976
!
KASHMIR.
Sunajar,
1615
m.,
Gammie!
In
Kashmir
valley,
1676
m.,
?Clarke
14805
A
!
T.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
209
INDIA.
Punjab
:
Karnal—Ghogripur,
Drummond
21,089
!
21,090
!
Umritsur,
Clarke
22,226A
!
22,226B
!
Chumba,
1524
m.,
and
914
m.,
Clarke
23,519
!
24,308
A
!
Himalayas,
Barogh,
1524
m.,
Rich
908
!
United
Provinces
:
Mahannddee
below
Muldah,
Hooker
&
Thomson
(May,
1850)
!
Saharanpur,
Gov.
Botanic
Gardens
(Sept.
1884)
!
Kampti
Falls
near
Mussoorie,
1067
m.,
Stewart!
Province
and
locality
uncertain,
Wallich
8778B
!
Central
Provinces
:
Chanda
forests,
along
streams,
Haines
5318
!
Allapalli,
Haines
3647
!
Bihar
and
Orissa
:
Baudh
State,
near
Tikerapara,
Mooney
634
!
Madras
:
Rampa
county,
beyond
Erragonda,
Godaveri
district,
Narangaswami
542
!
Nilgiris,
S.E.
Wignaad,
Gudalur,
914
m.,
Gamble
15,422
!
Khoondas,
near
water,
Hohenacker
1284
!
Hill
Grove,
Bourne
Herbarium!
Begur,
Malabar,
Wyanaad,
762
m.,
Fischer
4511
!
Nungambakam,
Herb.
Madras.
Myson!
ASSAIVI.
Without
precise
locality,
Fielding!
Several
of
the
above
specimens,
including
those
of
Clarke,
Mooney,
and
Fielding,
are
weak
or
shady-habitat
specimens.
Hohenacker
1284,
from
Khoondas
in
the
Nilgiris,
matches
the
specimen
from
plants
grown
in
the
Royal
Botanic
Gardens,
Kew.
Andropogon
halepensis
subsp.
miliformis
(Schult.)
Piper
(in
Proc.
Biol.
Soc.
Wash.
28,
28,
1915),
in
so
far
as
it
is
based
on
A.
miliaceus
Roxb.,
comes
under
this
variety.
The
species
A.
laxus
Roxb.
and
A.
controversus
Steud.,
cited
in
synonomy
by
Piper,
are
here
excluded,
however,
and
will
be
described
below
under
Sorghum
controversum.
In
his
de-
scription
of
Andropogon
halepensis
subsp.
miliformis
Piper
gives
the
width
of
the
leaves
as
5-15
mm.,
whereas
in
the
material
I
have
examined
at
Kew
the
leaves
of
Sorghum
miliaceum
are
often
3-4
cm.
wide.
I
have
been
able
to
accept
only
four
of
the
specimens
in
Kew
Herbarium
cited
by
Piper
as
belonging
to
S.
miliaceum,
the
others
I
think
are
better
placed
under
S.
halepense.
I
have
not
seen
any
of
the
specimens
of
the
plants
referred
to
by
Piper
as
having
been
grown
in
the
United
States
and
none
of
the
specimens
from
that
country
that
I
have
examined
at
Kew
belongs
to
S.
miliaceum.
It
seems
possible
that
the
plants
mentioned
by
Piper
failed
to
establish
themselves
in
the
United
States,
but
died
out
like
those
grown
in
the
Royal
Botanic
Gardens,
Kew.
The
following
specimens
from
Mexico,
however,
which
have
tall
culms,
broad
leaves,
and
large
panicles,
agree
well
with
Sorghum
miliaceum.
It
is
possible
that
they
may
be
the
progeny
of
some
of
the
plants
grown
in
the
United
States,
or
from
seed
introduced
direct
from
India,
but
it
is
difficult
to
account
for
their
appearance
in
Mexico
and,
until
further
information
is
available,
they
are
somewhat
doubtfully
placed
here
:
MEXICO.
Sonora:
north-eastern
Sonora,
Colonia
Morelos,
800
m.,
Santos
1879!
Chihuahua
:
Guasaremos,
Rio
Mayo,
Scott
Gentry
2375
!
Nuevo
Leon
:
Fundicion,
near
Monterrey,
Arsene
708
!
Guerrero
:
Coyuca,
Hinton
5552
!
(ii)
Var.
parvispicula
Snowden
Var.
parvispicula
Snowden,
var.nov.
Spiculae
sessiles
4-4.5
mm.
longae,
1.5-1.8
mm.
latae,
plerumque
muticae;
panicula
nutans
rami
inferiores
5-7-nati,
in
totum
graciles
et
flexibiles.
Sessile
spikelets
4-4.5
mm.
long,
1.5-1-8
mm.
wide,
usually
awnless.
Panicle
nodding,
lower
branches
5-
to
7-nate,
slender
and
flexuous
throughout.
The
type
of
this
variety
is
a
specimen
in
Kew
Herbarium,
Gamble
27,139,
collected
at
Debra
Dun,
United
Provinces,
India.
PAKISTAN.
Western
Himalaya
:
Baru,
Duthie!
West
Punjab
:
Rawalpindi,
Aitchison
116!
INDIA.
East
Punjab:
without
precise
locality
(near
Karnal
?)
Drummond
21,088!
United
Provinces:
Dehra
Dun,
670
m.,
Gamble
27,139
!
Dehra
Dun,
in
new
forest,
670
m.,
Umrao
Singh
383
!
Province
and
locality
uncertain,
Wallich
8778
A
!
8778
C
!
Madras:
Syadapet
(Saidapet),
Government
Farm
(Nov.
1885)
Gamble
Herbarium!
Nilgiris,
Burliar,
914
m.,
Gamble
14,380
!
Coimbatore,
1310
m.,
Raju
&
Nagathan
4927
!
Salem,
el
9
210
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
Hosur
Cattle
Farm,
914
m.,
Narangaswami
3061D
!
Kudirai
Madai,
990
m.,
Naranga-
swami
3944D
!
Biligirirangano,
N.
Coimbalai-Mysore,
Alagranathan!
Dindigul,
914
m.,
Wallich
8778D
(2)
!
The
panicle
of
Wallich
8778D
(2)
is
rather
weak
and
depauperate.
3.
Sorghum
controversum
(Steud.)
Snowden
Andropogon
laxus
Roxb.,
Fl.
Indica,l,
275,
1820,
was
the
name
originally
given
to
this
species,
but
it
was
changed
by
Steudel
because
of
the
earlier
A.
laxus
Willd.
It
was
described
from
plants
growing
in
north-east
India
that
were
locally
known
by
the
names
Kalamoocha
(Beng.)
and
Gaddi-jan,00
(Teling.).
The
figure
no.
894
in
Icones
Roxburgh-
ianae
shows
its
loose
spreading
panicle,
and
gives
dissections
of
the
sessile
spikelets.
A.
controversus
Steud.,
Syn.
Pl.
Glum.
1,
391,
1854,
is
based
on
A.
laxus
Roxb.,
loc.
cit.,
the
change
of
name
being
necessary
on
account
of
the
older
A.
laxus
Willd.,
Sp.
Pl.
4,
907,
1806.
Sorghum
controversum
(Steud.)
Snowden,
comb.nov.,
is
a
change
that
is
necessary
on
transferring
this
species
to
the
genus
Sorghum.
No
authentic
type
specimen
of
Andropogon
laxus
Roxb.
has
been
located,
but
in
the
Wallich
Herbarium
at
Kew
there
is
one
sheet
written
up
A.
laxum
and
marked
H.B.,
which
suggests
that
the
two
specimens
mounted
on
the
sheet
may
have
been
grown
in
Hort.
Bot.
Calcutta,
from
seed
of
the
plants
from
which
the
illustration
in
Icones
Rox-
burghianae
was
drawn.
The
right-hand
panicle
of
this
sheet
certainly
agrees
well
with
the
illustration,
but
it
differs
in
having
the
sessile
spikelets
awnless.
Awned
and
awnless
states
belonging
to
the
same
species,
however,
are
common
in
certain
species
of
the
sub-
section
Halepensia,
and
these
Wallich
specimens
agree
with
awnless
states
of
Sorghum
controversum
that
have
been
received
at
Kew
from
south-east
India.
The
right-hand
panicle
of
Wallich
8778
H
has
therefore
been
selected
as
typical
of
the
awnless
state
of
this
species,
to
be
linked
with
figure
no.
894
in
Icones
Roxburghianae.
Until
Roxburgh's
type
specimen
has
been
located,
a
specimen
in
Kew
Herbarium,
Bourne
3270,
from
Tape-
swarani,
Godavery
district,
Madras
province,
dated
December
1907,
has
been
selected
as
characteristic
of
the
awned
state
of
this
species.
DESCRIPTION
A
more
or
less
loosely
tufted
perennial,
with
the
rhizomes
often
somewhat
woody.
Cu/ms
erect
or
sometimes
decumbent
near
the
base
and
rooting
at
the
lower
nodes,
slender
to
stoutish
and
tending
to
become
woody,
0.3-1
cm.
wide,
0.5-2
m.
high,
or
sometimes
(according
to
Roxburgh)
under
favourable
moist
shady
conditions,
such
as
river
banks,
etc.,
attaining
a
height
of
3-4.5
m.
Leaves
few
to
many;
blades
linear,
from
a
rounded
and
slightly
contracted
base,
gradually
narrowing
upwards
to
the
acuminate
apex,
12-35
or
sometimes
60
cm.
long,
mostly
0.5-2
cm.
wide,
more
or
less
smooth
except
for
the
scabrous
margins.
Panicle
ovate
to
pyramidal,
loose
and
spreading,
often
nodding,
15-30
cm.
long,
5-15
cm.
wide
;
branches
subverticillate,
widely
spaced,
slender
and
flexuous,
except
near
the
base,
woolly
near
the
junction
with
the
main
axis
and
in
the
axils
of
the
divisions,
the
lower
usually
3-
to
4-nate,
5-16
cm.
long,
often
naked
for
1-5
cm.
from
the
base.
Racemes
fragile,
2-
to
5-noded,
8-22
mm.
long;
internodes
and
pedicels
slender,
2-4
mm.
long,
densely
ciliate
with
whitish
to
tawny
or
pale
purplish
hairs
0.7-1
mm.
long.
Sessile
spikelets
elliptic-lanceolate
to
subelliptic,
acute
to
acumi-
nate,
5-6.2
mm.
long,
1.5-2-2
mm.
wide,
pale
straw
to
cream
colour
with
3-5
nerves
evident
near
the
tip
of
the
lower
glume
when
young,
at
length
tawny
to
purplish,
usually
more
or
less
densely
softly
hairy
throughout;
awned
or
sometimes
awnless;
callus
short
and
narrow,
densely
bearded.
Glumes
rather
thin
and
subcoriaceous
with
somewhat
papery
tips;
lower
usually
9-
to
11-nerved,
finely
keeled
from
about
the
middle
or
in
the
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
211
upper
third,
with
the
narrow
keel-margins
ending
in
minute
teeth
or
sometimes
merging
almost
imperceptibly
into
the
hyaline
apex,
finely
spinulously
ciliolate
on
the
keels
;
upper
7-nerved,
with
two
or
sometimes
four
of
the
lateral
nerves
often
rather
short,
slightly
scabrid
on
the
keel
in
the
upper
third
or
fourth.
Lemmas
densely
but
finely
long-
ciliate;
lower
elliptic-lanceolate
to
elliptic-oblong,
4
4.5
mm.
long;
upper
ovate
to
elliptic-ovate,
3-4
mm.
long,
2-lobed
to
0.7
or
1
mm.
deep,
with
an
awn
10-14
mm.
long
or
sometimes
with
a
mucro
1-3
mm.
long.
Anthers
3-3.5
mm.
long.
Grains
(immature
?)
obovate-elliptic,
2-2.5
mm.
long,
1-1.3
mm.
wide,
tan
to
reddish
brown.
Pedicelled
spikelets
linear-lanceolate
to
subulate,
usually
S,
4.5-7
mm.
long,
1-1.5
mm.
wide,
straw
yellow
or
more
often
purplish,
sparsely
shortly
hairy;
glumes
thin
and
papery,
lower
7-
to
9-nerved,
upper
5-
to
7-nerved;
lemmas
large,
well
developed;
anthers
2-3.5
mm.
long.
/
/
//
V•r•
V
\`
1
1
A
B
/
/
/
//
//
/
C
%/
1
)1/
9,
/1
Sorghum
controversum
(Stead.)
Snowden.
A,
sessile
and
pedicelled
spikelets;
B,
ower
glume;
C,
upper
glume
of
sessile
spikelet,
exterior
view.
A,
x
5;
B
and
C,
x
10.
(From
Bourne
3270.)
DIAGNOSTIC
CHARACTERS
AND
AFFINITIES
The
most
distinctive
features
of
this
species
are
the
somewhat
woody
rhizomes
;
slender
to
rather
stout
and
woody
culms,
sometimes
rooting
at
the
lower
nodes,
short
in
dry
but
tall
in
moist
situations;
narrow
leaf-blades,
mostly
0.5-2
cm.
wide;
loose
rather
thin
panicles
15-30
cm.
long,
5-15
cm.
wide;
the
elliptic-lanceolate
to
subelliptic,
acute
to
acuminate
sessile
spikelets
5-6.2
mm.
long,
1.5-2-2
mm.
wide,
more
or
less
densely
hairy,
often
with
awns
10-14
mm.
long;
and
the
rather
thin
subcoriaceous
glumes,
the
lower
one
very
finely
keeled
and
narrowly
margined,
the
keels
without
or
with
only
obscure
minute
teeth.
Sorghum
controversum
differs
from
S.
halepense
and
S.
miliaceum
in
the
more
acuminate
tip
of
the
usually
densely
hairy
sessile
spikelets,
the
thinner
subcoria-
ceous
glumes,
the
lower
one
entire
or
obscurely
and
very
minutely
toothed
near
the
apex,
and
from
S.
propinquum
in
the
more
slender
culms,
narrower
leaves
and
larger,
often
awned,
sessile
spikelets.
DISTRIBUTION
AND
UTILITY
This
species
is
apparently
confined
to
eastern
and
southern
India,
extending
from
Bengal
province
southwards
through
Bihar
and
Orissa
to
Madras
province.
Its
utility
is
probably
comparable
with
that
of
S.
halepense
as
regards
size
and
foliage,
but
little
information
is
available
at
present
as
to
its
usefulness
under
cultivation,
as
hitherto
it
212
J.
D.
SNOWDEN
:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
has
probably
been
treated
as
a
form
or
variety
of
that
species.
S.
controversum
seems
to
succeed
best
under
moist
shady
conditions,
and
when
growing
on
moist
river
banks
it
often
sends
out
adventitious
roots
at
the
lower
nodes.
In
dry
soils
or
when
subject
to
drought,
it
tends
to
become
dwarf
and
stunted.
It
would
appear
to
require
a
tropical
climate
in
order
to
succeed.
In
the
specimens
enumerated
below
those
with
awned
sessile
spikelets
are
distinguished
by
an
*
immediately
after
the
name
or
number
of
the
collector.
INDIA.
Bengal:
Botanic
Gardens,
Calcutta,
Wallich
Herbarium
877811
!
Monghyr,
Mokim
1417
!
Bihar
and
Orissa:
without
precise
locality,
Haines
706
*!
S
706
*!
Karlapat
and
Lanigarh,
Kalahandi
State,
Mooney
1189
*!
1736
*
!
Madras:
Tapeswarani,
Godaveri
district,
Bourne
3270
*!
Gamble
15,780
*!
at
1067
m.,
Shevaroy!
Bellary
district,
Brougham
*
!
Travancore,
Countalbum,
Ramon
Rao
2111
!
Coimbatore,
Imperial
Dep.
Agric.
India,
Janaki
Ammal
1205
*!
Ayinigira
Delta,
Geddesol,
1371
m.,
Cherian
Jacob
356
*!
Ramandrug
and
South
Arcot
district,
Madras
Herb.
S.
India
14,383
*!
17,860
*
!
Several
of
the
above
specimens
are
damaged
by
`Smut
Disease'.
In
such
cases
the
sessile
spikelets
are
usually
awnless.
They
include
Mokim,
1417
and
Shevaroy.
A
specimen
from
Bombay
Reversing
Station,
Hull
Ghat,
460
m.,
Wroughton,
34,
damaged
by
the
same
disease,
may
also
belong
to
this
species.
Hackel
(in
DC.,
Monogr.
.
Phan.,
6,
502,
1889)
cited
Andropogon
laxus
Roxb.
(Fl.
Indica,
1,
275,
1820)
and
A.
controversus
Steud.
(Synops
.
Pl.
Glum.
1,
391,
1854)
in
synonomy
under
A.
sorghum
(Linn.)
var.
halepensis
(Linn.);
and
Piper
(in
Proc.
Biol.
Soc.
Wash.
28,
28,
1915)
cited
them
in
synonomy
under
A.
halepensis
(Linn.)
miliformis.
I
think,
however,
that
although
this
group
belongs
to
the
subsection
Halepensia
it
is
best
kept
as
a
separate
species
as
originally
described
by
Roxburgh.
4.
Sorghum
propinquum
(Kunth)
Hitchc.
Andropogon
affinis
J.
S.
Presl,
in
C.
B.
Presl,
Rel.
Haenk.
1,
343,
1830,
is
the
earliest
name
known
for
this
species,
which
was
described
from
specimens
collected
by
Haenke
in
Luzon,
one
of
the
Philippine
Islands.
A.
propinquus
Kunth,
Rev.
Gram.
1,
Suppl.
40,
1830,
and
Enum.
1,
502,
1833,
is
based
on
A.
affinis
J.
S.
Presl,
but
the
specific
name
was
changed
because
it
was
invalidated
by
the
older
A.
affinis
R.
Br.,
Prod.
Nov.
Holl.
1,
201,
1810.
Sorghum
propinquum
(Kunth)
Hitchc.
in
Lingnan,
Sci.
J.
7,
249,
June
1929,
is
based
on
Andropogon
affinis
J.
S.
Presl.
Although
I
have
not
been
able
to
see
any
of
the
specimens
collected
by
Haenke,
I
have
examined
much
material
from
the
Philippine
Islands,
including
Loher
1806
and
1807
from
Luzon
Island,
as
well
as
other
specimens
in
Kew
Herbarium
identified
by
Piper
as
belonging
to
this
species.
DESCRIPTION
A
tall
perennial,
loosely
tufted,
with
a
few
stoutish
rhizomes
often
15-30
cm.
long
and
3-4
mm.
wide.
Cu/m8
mostly
2-3
(rarely
5)
m.
high,
1-3
cm.
wide
near
the
base,
some-
times
branched
above.
Leaves
15-26;
blades
linear,
from
a
broad
somewhat
clasping
base,
tapering
upwards
to
the
acuminate
tip,
0.3-1
m.
long,
mostly
3-5
cm.
wide,
with
the
stoutish
mid
rib
prominent
below.
Panicles
large,
loose,
oblong
to
elliptic-oblong,
5-10
(rarely
15)
cm.
wide,
20-60
cm.
long
;
branches
slender,
suberect
or
slightly
spreading
above,
1-4
from
each
node,
the
lower
often
15-20
cm.
long
and
exceeding
half
the
length
of
the
panicle,
usually
bare
of
spikelets
for
1-5
cm.
from
the
base.
Racemes
slender,
fragile,
crowded,
1-
to
5-noded,
mostly
1-2
cm.
long;
internodes
and
pedicels
slender,
2.5-4
mm.
long,
sparsely
to
moderately
ciliate
with
whitish
to
fulvous
hairs
about
0.5
mm.
long.
Sessile
spikelets
elliptic-lanceolate
or
narrowly
elliptic,
abruptly
acute
J.
D.
SNOWDEN
:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
213
with
a
short
fine
point,
3.8-5
mm.
long,
1.2-2
mm.
wide,
narrowed
below
to
about
0.5-
0.8
mm.
near
the
callus;
sparsely
strigose
with
whitish
to
fulvous
hairs
at
the
base,
sides,
and
above
the
middle,
straw-yellow
to
cream-colour
with
3-4
delicate
nerves
visible
be-
tween
the
keels
of
the
lower
glume
when
young,
at
length
becoming
tawny
to
mahogany
red,
blackish
brown,
or
black
;
usually
awnless;
callus
moderately
to
sparsely
bearded.
Glumes
subcoriaceous
to
thinly
crustaceous
in
the
lower
half,
somewhat
membranous
above;
lower
usually
9-
to
11-nerved,
finely
keeled
in
the
upper
third,
not
or
only
obscurely
winged,
with
the
keels
tapering
almost
imperceptibly
into
the
hyaline
tip
or
rarely
ending
in
very
minute
teeth,
very
finely
and
shortly
rigidly
ciliolate
on
the
keels
;
upper
7-nerved,
with
2
on
each
side
of
the
median
nerve
usually
somewhat
short,
some-
what
scabrid
on
the
keel
in
the
upper
fourth.
Lemmas
thinly
ciliate
about
the
middle
or
in
the
upper
two-thirds;
lower
elliptic-oblong,
3.5-4.5
mm.
long;
upper
ovate
to
elliptic-ovate,
about
3
mm.
long,
entire
and
acuminate
or
minutely
notched
to
0.5
mm.
deep
and
with
a
very
small
mucro
(rarely
a
short
awn),
usually
1-nerved.
Anthers
2-2.5
mm.
long.
Grains
obovate
to
obovate-elliptic,
1-5-1.8
mm.
long,
1-1.2
mm.
wide,
reddish
brown
to
blackish
purple,
embryo-mark
often
obscure.
Pedicelled
spikelets
usually
d,
4-5.5
mm.
long,
straw-yellow
to
pale
purplish;
glumes
almost
membranous,
lower
8-
to
9-nerved,
upper
5-
to
7-nerved;
lemmas
well
developed;
anthers
1.5-2.5
mm.
long.
DIAGNOSTIC
CHARACTERS
AND
AFFINITIES
Sorghum
propinquum
is
distinguished
by
its
few
thick
rhizomes;
tall
stout
culms,
1-3
cm.
wide
near
the
base
;
numerous
leaves
with
broad
blades
mostly
3-5
cm.
wide;
large
loose
panicles
;
small
narrow
acute
sessile
spikelets,
almost
elliptic
in
outline,
3.8-5
mm.
long,
1.5-2
mm.
wide,
usually
awnless;
the
thin
glumes,
subcoriaceous
below
and
almost
papery
above
the
middle
;
and
rather
large
anthers
in
comparison
with
the
size
of
the
spikelets.
It
is
separated
from
S.
controversum
by
its
stouter
culms,
wider
leaf-blades,
and
smaller,
usually
awnless,
sessile
spikelets.
It
differs
from
S.
miliaceum
in
the
more
acute
tip
of
the
sessile
spikelets
and
the
thinner
glumes,
with
the
lower
one
very
finely
keeled
and
not
distinctly
3-toothed
at
the
tip.
DISTRIBUTION
AND
UTILITY
This
species
has
the
most
easterly
distribution
of
the
wild
Sorghums
belonging
to
the
section
Eu-Sorghum.
It
is
common
in
many
parts
of
the
Philippine
Islands,
where
the
type
specimens
were
collected,
and
it
is
also
found
on
many
of
the
islands
of
south-east
Asia,
including
Palau
Islands,
the
Moluccas,
Borneo,
Sarawak,
and
the
Malay
Islands,
as
well
as
on
the
mainland
of
southern
China,
the
Malay
Peninsula,
and
Siam.
It
is
also
found
in
southern
India
and
Ceylon,
where
it
may
have
been
introduced.
With
its
tall
stout
culms,
broad
leaves,
and
large
panicles,
it
has
the
appearance
of
being
much
more
of
a
jungle
grass
than
Sorghum
halepense,
resembling
S.
miliaceum
in
its
robustness.
According
to
Piper
(in
Proc.
Biol.
Soc.
Wash.
28,
30,
1915)
seeds
from
Manila,
Luzon
Island,
were
introduced
into
the
United
States,
where
under
cultivation
S.
propinquum
was
found
to
resemble
the
cultivated
grain
Sorghums
in
its
coarse
stems
and
wide
leaves.
It
also
required
a
long
time
to
reach
maturity,
for
at
Biloxi,
Missouri,
2-year-old
plants
did
not
bloom
until
the
end
of
October,
and
at
Washington,
D.C.,
the
plants
had
not
bloomed
before
they
were
cut
down
by
frost,
but
the
rootstocks
survived
the
winter.
It
is
probably
on
account
of
its
lateness
in
maturing,
coupled
with
its
thick
culms,
which
would
be
difficult
to
cut
with
machinery,
that
this
species
does
not
appear
to
have
received
much
attention
as
a
fodder-crop
for
feeding
to
stock,
in
spite
of
the
abundance
of
its
broad
leaves.
There
would
also
be
some
difficulty
in
harvesting
the
seed,
as
the
racemes
are
very
fragile
when
mature
and
a
large
proportion
of
the
spikelets
would
probably
fall
before
being
harvested.
214
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
(i)
Var.
propinquum
Var.
propinquum
is
based
on
Andropogon
propinquum
Kunth,
Rev.
Gram.
1,
Suppl.
40,
1830
and
Enum.
Pl.
1,
502,
1833,
which
is
based
on
A.
affinis
J.
S.
Presl,
in
C.
B.
Presl,
Rel.
Haenk.
1,
343,
1830.
A.
sorghum
subsp.
halepensis
var.
propinquus
(Kunth)
Hack.
in
DC.,
Monogr.
Phan.
6,
503,
1889,
excluding
Trimens
specimen
from
Ceylon,
has
the
same
basis
and
is
included
in
this
variety.
A.
halepensis
propinquus
(Kunth)
Piper
in
Proc.
Biol.
Soc.
Wash.
28,
29,
1915,
rests
on
the
same
foundation
and
is
also
included
here.
This
variety
has
small
sessile
spikelets
3.8-4-5
mm.
long,
and
1.2-1-8
mm.
wide.
Its
distribution
is
mainly
east
of
Siam.
CHINA.
Canton:
Hance
4879
!
McClure
9138
!
Kwantung
:
McClure
9985
!
Taipu
(Taifu):
Ford
484
!
Hainan
Island:
Henry
8295
!
Paak
Po
Shan,
Toam-chau
district
(Lignan
University
16,236)
Tsang
Wai-Tak
737
!
PHILIPPINE
ISLANDS.
Luzon
Island:
Elmer
8287
!
Cuming
569
!
Loher
1806
!
1807
!
Merrill
1468
!
Ramos
13,609
!
Mindanao
Island:
Copeland
466
!
Weber
1015
!
Balabac
Island:
Vidal
3996
!
Without
precise
locality,
Loher
7169
!
7209
!
BORNEO.
Bandjarmassin,
Motley
414
!
Sarawak
State:
Beccari
3924
!
Clemens
21,308
!
MOLUCCAS.
Ceram
Island:
Kornassi
448
!
Buru
Island:
Riedel!
MALAY
PENINSULA.
Pahang
:
Sungei
Jelei,
Machado
11,529
!
PALAU
ISLANDS.
Without
precise
locality,
Lederman
14,284
!
(ii)
Var.
siamense
(Piper)
Snowden
Var.
siamense
(Piper)
Snowden,
stat.nov.,
is
based
on
Andropogon
halepensis
siamensis
Piper.
A.
halepensis
subsp.
siamensis
Piper
in
Proc.
Biol.
Soc.
Wash.
28,
30,
1915,
is
based
on
specimens
collected
in
Siam,
the
type
being
Kerr
2156
in
Kew
Herbarium.
This
variety
has
sessile
spikelets
4.5-5
mm.
long,
and
1.8-2
mm.
wide.
It
is
found
from
Siam
westwards
to
southern
India
and
Ceylon.
SIAM.
Near
Kampang,
106
m.,
growing
in
pampas
along
the
banks
of
Mei
Ping
River,
Kerr
2156
!
Pak
Bawag,
Kerr
2006
!
Bangkok,
Kerr
4358
!
Maneron,
Chainat,
50
m.,
Kerr
19,671
!
CAMBODIA.
Phnom
Penk
(Oct.
1878)
Godefroy-Lebeuf
83
!
INDIA.
Madras:
Kodaikanal
Ghat
and
Poombari,
Bourne
1042
!
1896
!
Tadagam
Hills,
Coimbatore
district,
Fischer
2693
!
Without
precise
locality:
Anamallays!
Jacob
17,273!
Wallich
Herbarium
8778D
(i).
Jacob's
specimen
is
a
very
small
piece
of
young
material.
CEYLON.
Without
precise
locality,
Thwaites
C.P.
2484
!
Regarding
Sorghum
affine
A.
Camus,
in
Lecomte,
Fl.
Gen.
Ind.
China,
7,
pt
2,
321,
1922,
it
is
to
be
noted
that
this
is
not
the
same
as
Sorghum
affine
Kuntze,
Rev.
Gen.
Pl.
p.
791,
1891,
which
is
based
on
Andropogon
affine
R.Br.,
Prod.
1,
201,
1810,
and
has
since
been
named
Dichanthium
affine
(R.Br.)
A.
Camus
in
Bull.
Mus.
Hist.
Nat.
Paris,
27,
548, 549,
1921.
SUBSECTION
ARUNDINACEA,
SERIES
SPONTANEA
Snowden
(In
Kew
Bull.
p.
223,
1935)
5.
Sorghum
virgatum
(Hack.)
Stapf
Andropogon
sorghum
var.
virgatus
Hack.
in
DC.,
Monogr.
Phan.
6,
504,
1889,
is
the
basis
of
this
species,
which
is
founded
on
specimens
from
Egypt
and
the
Anglo-Egyptian
Sudan.
Of
those
cited
by
Hackel
I
have
seen
Kotschy
173
and
Schweinfurth,
529
and
538.
Sorghum
virgatum
(Hack.)
Stapf
in
PraM,
Fl.
Trop.
Afr.
9,
111,
1917,
is
based
on
Andropogon
sorghum
var.
virgatus
Hack.
J.
D.
SNOWDEN:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
215
DESCRIPTION
A
loosely
tufted
annual
30-150
cm.
high.
Cu/ms
slender,
mostly
3-6
mm.
wide,
often
with
suberect
branches
from
the
lower
and
middle
nodes.
Leaves
few;
blades
narrow,
usually
5-15
mm.
but
sometimes
20
mm.
wide,
the
longer
ones
attenuated
towards
the
base,
with
the
mid
rib
stout
in
robust
specimens.
Panicle
very
narrow,
oblong
or
elliptic-oblong,
15-60
cm.
long,
mostly
1-5
cm.
wide
;
branches
very
slender,
erect
or
suberect,
the
lower
5-15
cm.
long,
bare
for
1-5
cm.
from
the
base.
Racemes
fragile,
few-
to
7-noded,
up
to
4
cm.
long;
internodes
4-5
mm.
long,
ciliate
with
whitish
hairs
1-2
mm.
long,
pedicels
similar
or
a
little
shorter.
Sessile
spikelets
narrowly
lanceolate,
acute,
6.5-7
mm.
long,
2-2.5
mm.
wide,
pale
straw-colour
or
at
length
cream
to
buff
yellow,
with
a
few
greenish
nerves
visible
above
the
middle
of
the
lower
glume,
rarely
with
purplish
colouring,
becoming
somewhat
glossy
when
mature
;
callus
moderately
to
densely
bearded.
Glumes
coriaceous
from
near
the
base
to
the
middle
or
somewhat
above
the
middle,
then
subcoriaceous
to
papery;
lower
with
11-13
primary
nerves,
the
keels
evident
from
about
the
middle,
narrowly
marginate
and
spinulously
ciliolate
in
the
upper
third,
terminating
in
minute
teeth
below
the
short
hyaline
tip,
generally
sparsely
strigillose
or
at
length
almost
glabrous
;
upper
with
7
primary
nerves,
keeled
above
the
middle,
scabrid
in
the
upper
third
otherwise
almost
glabrous
except
for
the
long-ciliate
hyaline
margins.
Lemmas
thinly
long-ciliate;
lower
elliptic-oblong,
about
5
mm.
long;
upper
ovate-elliptic,
3.5-4
mm.
long,
with
the
lobes
0.5-1
mm.
deep
and
a
slender
awn
from
the
sinus
8-16
mm.
long.
Anthers
2.5-3
mm.
long.
Grains
narrowly
obovate-elliptic,
2.5-3
mm.
long,
1.5-2
mm.
wide,
reddish
brown,
embryo-mark
distinct.
Pedicelled
spikelets
usually
d,
but
sometimes
reduced,
6-7
mm.
long,
readily
deciduous;
lower
glume
9-nerved,
upper
7-
to
8-nerved;
anthers
about
2
mm.
long.
DIAGNOSTIC
CHARACTERS
AND
AFFINITIES
The
long
very
narrow
panicle,
with
its
branches
suberect
and
keeping
near
to
the
rhachis,
combined
with
the
narrow
lanceolate
spikelets
6.5-7
mm.
long,
furnished
with
long
slender
awns,
generally
suffice
to
enable
this
species
to
be
distinguished
from
its
nearest
relative,
S.
lanceolatum.
The
latter
has
a
broader
panicle
as
well
as
wider
lanceo-
late
sessile
spikelets
with
stouter
awns.
When
the
panicles
of
S.
virgatum
have
been
disarranged
and
the
branches
spread
out,
as
sometimes
happens
in
pressing
and
mounting
the
specimens,
they
lose
their
characteristic
shape.
Both
the
panicles
and
the
spikelets
are
usually
of
a
pale
straw
colour
to
cream
or
buff,
and
it
is
only
rarely
that
they
become
tinged
with
mahogany
or
brownish
purple
colour.
DISTRIBUTION
AND
UTILITY
From
the
specimens
at
present
available
it
would
appear
that
the
natural
distribution
of
S.
virgatum
is
mainly
confined
to
the
hotter
and
drier
districts
of
Egypt
and
the
Anglo-Egyptian
Sudan.
Chiovenda
(in
Fl.
Somali,
2,
438,
1932)
also
records
it
from
the
Juba
province
of
Italian
Somaliland.
A
poor
specimen
in
the
Kew
Herbarium,
however,
Cieferi
159,
collected
in
the
same
province,
and
from
which
all
the
racemes
have
fallen,
has
hairy
spikelets
with
stoutish
awns
like
those
of
S.
lanceolatum,
and
it
is
possible
that
some
of
the
other
specimens
from
Juba
province
may
belong
to
this
latter
species.
S.
virgatum
was
introduced
from
Egypt
into
Algeria,
where
it
was
grown
for
fodder
under
the
name
of
Sorgho
Menu.
It
was
recommended
by
Trabut
(in
C.R.
Acad.
Agric.
Fr.
1916)
as
a
small
Sorghum
suitable
for
cultivation
in
dry
regions.
Piper
(in
Proc.
Biol.
Soc.
Wash.
28,
33,
1915)
states
that
seed
of
it
obtained
from
Trabut
was
introduced
into
the
United
States,
where
it
was
grown
for
some
years
under
the
name
of
Tunis
Grass.
Under
cultivation
in
the
United
States
it
behaved
purely
as
an
annual
and
it
was
found
to
cross
naturally
with
S.
sudanense
(Piper)
Stapf,
and
the
cultivated
sweet-stemmed
216
J.
D.
SNOWDEN
:
WILD
FODDER
SORGHUMS
OF
THE
SECTION
EU-SORGHUM
variety
known
as
Amber
[a
variety
of
S.
dochna
(Forsk.)
Snowden].
Among
early
re-
ferences
to
its
cultivation
in
the
United
States
under
the
name
of
Tunis
Grass
are
those
of
Oakley
in
1912
(U.S.
Dep.
Agric.
Year
Book,
pp.
499-501)
and
Vinall
in
1914
(Farmers'
Bull.
U.S.
Dep.
Agric.
no.
605,
p.
4).
The
slender
culms,
narrow
leaves,
and
fragile
racemes,
which
break
up
readily
and
allow
the
fruiting
spikelets
to
fall
so
easily
at
maturity,
tend
to
detract
from
the
usefulness
of
this
species
under
cultivation
and
it
has
been
superseded
in
most
districts
by
taller
species
with
more
abundant
foliage.
Most
of
the
specimens
enumerated
below
have
been
found
in
or
near
cultivated
lands,
often
in
sandy
soils,
but
sometimes
in
irrigated
lands.
EGYPT.
Near
Cairo
:
Talbia,
and
the
Pyramids,
Acherson
316
!
Bove
268
!
269
!
365
!
Bolland
(July
1912)
!
Keller
428
!
Meinertzhagen!
Shabetai
7
!
Ismailia,
Zagazig,
and
Muschler
(Dec.—Mar.
1904-5)
!
ANGLO-EGYPTIAN
SUDAN.
1
Between
Dongola
and
Meru,
Bromfield
32
!
Between
Khartoum
and
Berber,
Schweinfurth
529
!
538
!
Kordofan,
Kotschy
173
!
Khartoum,
Johnson!
Aylmer
322
!
408
!
Without
precise
locality,
Punter
!
Vern.
name.
Garawi.
SOUTH
AFRICA.
Bechuanaland:
Mafeking,
Nat.
Herb.
Pretoria
H.
11,074!
Vern.
name:
Tunis
Grass.
The
specimen
from
Mafeking
is
mixed
with
a
Sorghum
sp.
near
S.
mellitum
var.
australe
Snowden.
WEST
AUSTRALIA.
Swan
River,
Drummond!
The
specimen
from
West
Australia
was
cited
by
Bentham
(in
Fl.
Aust.
7,
540,
1878)
under
S.
halepense
Pers.
It
agrees
well
with
a
specimen
from
near
Cairo,
Egypt
(Bove
268,
collected
in
1830),
and
it
seems
possible
that
Drummond's
label
may
have
been
wrongly
attached
to
the
specimen.
Hackel
(in
DC.,
Monogr.
Phan.
6,
504,
1889)
cites
specimens
from
Upper
Egypt:
at
Damietta,
Ehrenberg;
Alexandria,
Wichura;
and
Suez,
Kotschy
It.
Syr.
882,
as
well
as
from
the
Oasis
of
Thebes,
Unger;
in
addition
to