rchn>-^S

^1

tation Bulletin 461

May 1959

Breeding and Improving
Ornamentals

^

F'^ iHM

P^^l

&-«#"♦» "«^

, ^" "^. ^1^^'-^^'

^#1

••^^4

: ■ > ; n*

Ml^%

r^lPEH

l^'ii^^

1

W^ .

y ' / vlP^

Of^^H

^^

^^^^^■T^^^^^H

""W

^^^^^^H

•:^.^j^^^^^^^^t-->-:'"^- "-"'^- -"• ■■:-:'?'S^''

Three new late blooming lilacs. Left to right: Nellie Bean,
Anna Anihoff, James Maefarlane.

By

A. F. YEAGER, E. B. RISLEY,

E. M. MEADER, R. B. PIKE

Agricultural Experiment Station

University of New Hampshire

Durham, New Hampshire

Foreword

In the breeding work with ornamentals, the following indi-
viduals should be credited as taking the lead in the following fields:
E. B. Risley, roses, callas, gladiolus, and monarda; E. M. Meader,
hardy chrysanthemums; R. B. Pike, conifers, rhododendrons and
azaleas; A. F. Yeager, lilacs, coleus and gloxinia.

The color photo on the cover is made possible by a contribution
from The Lilac Endowment Fund, which is sponsored by the New
Hampshire State Federation of Garden Clubs.

Contents

Azalea 3

Calla 5

Chrysanthemum 5

Coleus 7

Evergreens, Dwarf 8

Geranium 10

Gladiolus 10

Gloxinia 11

Lilac 11

Monarda 13

Rose, Rambler 14

Breeding and Improving Ornamentals

By a. F. Yeager, E. B. Risley, E. M. Meader, and R. B. Pike

Azaleas

The breeding work with the group of rhododendrons which are known as
azaleas has three objectives: 1. The production of superior seed propagated
strains of our three native New Hampshire azaleas, 2. The production of
azaleas roughly equivalent to the Ghent and Mollis hybrids that will be
reliably hardy in northern New England, and 3. The production of azaleas
akin to the semi-evergreen types that should be hardy in at least part of
New Hampshire.

1, Rhodora {Rhododendron canadense Torr.), June Pink iR. roseum
Rehd.) and Swamp Honeysuckle (/?. viscosum Torr.) are our three native
New Hampshire azaleas. All three are excellent garden plants in their su-
perior forms. Since there is no practical means of vegetative propagation,
presently known, it is necessary to try to produce strains of enough genetic
purity to make seed propagation satisfactory. A deep colored, wide petaled
selection of Rhodora upon selfing has produced even more attractive off-
spring and a selection of these has in turn been selfed. The resulting seed-
lings have not yet bloomed. It is hoped that this process of selfing can be
continued until a high percentage of the seedlings are superior. However
as Rhodora is one of the rare tetraploid azaleas this process may be long.
In the meantime the possibility of vegetative propagation is being more
fully explored.

A superior wild selection of June Pink {R. roseum) has been selfed and
200 progeny have been grown. These plants display considerable vegetative
variation and it is thus assumed that there may be a comparable variation
in flower color, form, cluster, size and scent from which further selection
and breeding may be continued. Two plants with the highly desirable
character of stoloniferous growth have already been selected for future
breeding.

The Swamp Honeysuckle {R. v'scosum) is still in the stage of first se-
lection and a few hundred plants are being grown from collected seed to
locate a superior plant or plants. Seed reproduction of superior forms will
be attempted with this azalea as with the two former ones. R. roseum and
R. viscosum are both diploids and if self fertility can be maintained a sat-
isfactory degree of genetic purity may be more easily achieved than with
R. canadense which is tetraploid.

2. In the second area of azalea breeding, i.e., the production of hybrids
comparable to the Ghent and Mollis azaleas, we are using as parent stock
only those species and individuals which are either native to New Hamp-
shire or have demonstrated their hardiness here.

The species presently being used are the natives R. roseam and R. vis-
cosum, two other eastern American species, the stoloniferous R. atlandcum
Rehd. and the red flowered R. cumberlandense E. L. Br., and an Asiatic
species R. japonicum Suringar.

Present progress includes the selection of parents for further breeding
from an unusual progeny of R. japonicum which displayed extreme forms
of color and stature. This progeny resulted from an attempt in 1950 to
cross R. japonicum with R. canadense. No true hybrids appeared nor were
the offspring identical with the female parent but they did show extreme
expression of the female characters of flower color, foliage type and gen-
eral growth habit. These may be homozygous individuals. Only future breed-
ing results can answer this question. Successful crosses have been made be-
tween R. roseum and six selections of these R. japonicum. It will be an-
other year at least before these bloom. Successful crosses have also been
made with R. japonicum as either male or female parent and R. viscosum
and R. altanticum but these are still in either the seed or early seedling

stages.

A progeny of R. altanticum x R. roseum gave scattered bloom this year
with a promising show of some of the better flower and plant characters
of both.

Advantage was taken of this first bloom to make crosses with R, viscosum
and selected R. japonicum, from which plump seeds have been harvested.

Another approach in this same group of azaleas is being made which
may be productive of good results. This is backcrossing to the native species
and R. japonicum of some of the Exbury azaleas which are the result of
Lionel de Rothchilds' lifelong efforts in breeding azaleas. These English
hybrids are the finest things yet seen in this section of azaleas but are of
doubtful hardiness for New Hampshire. Crosses have been made with R.
roseum, R. atlanticum, {R. atlanticum v R. roseum), R. viscosum, R. cum-
berlandense and several selections of R. japonicum. Although considerable
sterility was evident some plump seeds resulted from each combination.
Even if it is necessary to carry these progenies to the F2 generation in
order to carry them through our winters this approach deserves full ex-
ploration.

3. The third area of azalea breeding is based on two eastern Asiatic
species of the R. obtusum group.

This R. obtusum group of azaleas is largely evergreen and includes the
spectacular species and varieties used in gardens as far north as Washing-
ton, D. C. and Long Island. New York. R. poukhanensis Levi, from Korea
and R. kaempjeri Planch, from Japan have little evergreen tendency
but are much hardier than any other species of the group. Specimens of
both species have been growing in New Hampshire and Maine for years
including R. poukhanensis collected in Korea by Elwyn Meader.

Our approach to this breeding project is to backcross the highly de-
veloped garden varieties to these species and follow through to the F^
generation, hoping to combine hardiness with some of the finer qualities
of the tender garden varieties.

Crosses have been made on R. poukhanensis and to a lesser extent on R.
kaempjeri with some of the U.S.D.A. Glendale hybrids and some of the
Joseph Gable's hybrids. Plump seeds have been produced from the follow-
ing combinations:

4

R. poukhanensis x Dazzler (U.S.D.A.)

X Zulu (U.S.D.A.)
X Rose Greeley (Gable)

R. kaempferi x Rose Greeley (Gable)

By the use of supplemental light in the greenhouse during the winter
we hope to be able to accomplish the equivalent of several years of growth
in one year and thus produce an F2 generation for the field sooner than
could be done out-of-doors.

Calla

In 1950, the Yellow Calla. Zantedeschia elliottiana Engler, was crossed with
the Pink Calla, Z. rehmannii Engler, in an attempt to produce a dwarf,
dark red calla that would make an attractive pot plant. An Fj population
of 20 cream-colored and 6 pinkish-cream colored flowers resulted. The pink-
flowered plants had a few white leaf spots and pale green leaves. Seeds
from the pinkish-flowered plants produced only ¥> seedlings of the same
flower color but with darker green leaves than their parents. The cream
colored flowers occurred on vigorous plants having dark green leaves and
many white leaf spots. Seeds from these yielded an F^ of 360 plants, all
having cream colored flowers. When backcrossed to the yellow parent, the
progeny had only cream colored flowers in various shades.

In 1953, another attempt was made to produce a red flowered calla by
crossing the white flowered species, Z. albo-maculata Baill., with the Pink
Calla. Pale pinkish flowered hybrids appeared in the first generation. From
these a second generation may be grown when viable seeds are obtained
from the hybrids.

To further test the breeding behavior of these callas, the cross. Yellow
X White, Z. albo-maculata was made. All the progeny bore pale cream
flowers, 4 out of 17 being nearly white. The leaf shape of these hybrids
varied at random between the limits of the parent types.

Granite Series of Chrysanthemums

FoUowing the collection of the wild Chrysanthemum sibiricum in the Pouk
Han mountains of Korea by E. M. Meader in November, 1947, seeds were
mailed to A. F. Yeager in Durham, New Hampshire. There the new plants
proved fully winter hardy. Thus it became possible for crosses to be made
with garden varieties. Male-sterile plants of C. sibiricum were caged with
Nashua, Sunapee, Welcome, and several other varieties in the greenhouse.
House flies which were released within the cage accomplished easily what
would otherwise have been difficult cross pollinations. Reciprocal crosses and
backcrosses were also made by a similar procedure. Thus it was possible
with but little effort to grow several generations involving hundreds of
seedlings.

At no time was winter protection given to the field-grown seedlings and
those propagated for testing had to withstand at least three successive
winters in the open field before being considered for final selection. Those
having blossoms readily injured by the first faU frosts were discarded as

One of the Granite series chrvsantheniunis.

well as hundreds of plants, many with beautiful flowers, which did not prove
fully winter hardy.

After this rigid selection for hardiness the following ten clones having
attractive blossoms were named and introduced by the New Hampshire
Agricultural Experiment Station in 1955 as the Granite Series. The first
seven are long stemmed varieties suitable for cut flowers while the last
three have low-growing plants.

Granite Glow — Single, yellow flowers having two rows of petals markedly
resistant to fall frosts.

Granite Buttercup — Burnt gold pompom flowers.

Granite Harvest — Light orange, semidouble flowers with yellow center.

Granite Ranger — Semi-double, brick-red flowers.

Granite Promise — A single flower with several rows of brick-red petals.

Granite Enchantment — Intermediate pompom flowers, orange in color.

Granite Gem — Strong-growing, cushion type plant, with double lemon-
yellow flowers.

Granite Bronze — Double flowers, resistant to frosts.

Granite Pageant — Very floriferous with rusty, red, single flowers with
long narrow petals.

In general the Granite Series of varieties at Durham, N. H. have had
good foliage even though neither fungicides nor insecticides have ever been
used in their cultivation.

Two additional selections meeting the requirements for winter hardiness
of the Granite Series were named March 22, 1957. One (NH #LR) having
brick-red color with a 2''^-inch fully double flower has been called Granite
Pleasure. The other, an early-blooming double white (NH 4^W-9) has been
named Granite Truth. It is not fully frost hardy in the fall but starts flower-
ing August 15 and gives at least one month of garden beautification before
heavy frosts. Plants of the Granite Series are available from nurserymen.

Colchicine Treated Coleus

To teach a class of students how to use colchicine in plant breeding, a com-
mon ornamental coleus plant, Coleus blumei Benth., grown in the green-
house, was treated with this material in 1953. The original plant had leaves
with a green border next to a red area and with maroon color at the leaf
center. The apical growing points were removed from two plants 8 inches
tall and then they were immersed in one half of one percent colchicine
solution for about 20 hours. The idea was to see what the effect of doubling
the chromosomes might have on the color of the leaves, if any. From these
two plants, 42 lateral shoots were removed and rooted. Among the 42
rooted cuttings there were 7 color variations, all distinctly different from
the original plant and from each other.

These seven selected plants were grown for a year and increased by di-
vision to 20 plants of each clone. During this time, they retained the differ-
ences originally noted. Chromosome counts were made from root tips from
plants of each clone by a graduate student, Dr. D. L. Craig. It was found
that the original plant had 43 chromosomes. Of the 7 variants, 2 had 48
chromosomes, 3 had 72 chromosomes, one had 96 chromosomes and an-
other approximately 96. Thus what resulted from the colchicine-treated
plants was two new color patterns with no increase in chromosome number,
three new ones which had half the chromosomes doubled producing trip-
loids, and two which had all chromosomes doubled and were tetraploids.
Those which had the doubled number
of chromosomes were weak plants
with narrow irregular leaves and
rather poor root systems. The ones
with 72 chromosomes were much
stronger even than the original plants.
Of the ones which had the same
chromosome number as the original,
one was weak and died.

These results have indicated a be-
havior of plants following colchicine
treatment different from what is or-
dinarily expected. They tend to con-
firm the findings of the South Dakota
Experiment Station with Sorghum
where somatic segregation was found
to have resulted from colchicine treat-
ment. Seemingly colchicine offers
some possibilities in breeding other
than the doubling of chromosomes

which some some have assumed to Colchicum in a border planting. The
be its direct effect. bulbs of this plant furnish colchicine.

Dwarf Evergreens

An effort is being made to find dwarf evergreens suitable for use in northern
New England. Present day architectural styles demand such material and
the list of such shrubs hardy enough for our climate is very limited. A
search is being made among our native evergreens for dwarf and unusual
forms that may have ornamental and other horticultural qualities.

Many of the dwarf evergreens now in use originated as witches brooms.
Witches brooms are unusual or abnormal forms of growth usually caused
by fungus or other infections. These growths may occur on almost any
woody plant but are most striking and often very beautiful on coniferous
evergreens. Cuttings taken from such growths usually maintain the original
characteristics such as dwarf stature or dense foliage or whatever other
qualities may have made the witches broom notable.

A little searching of the roadsides and pastures of New Hampshire and
Maine have given us a number of these witches brooms from several species
which we are now testing for ease of propagation and general horticultural
properties.

The following have been located and are being tried.

Wilche-i Broom in a larch tree. This is typical of many such growllis

found on conifers.

8

Juniperous horizontalis Moench. (This is the Seaside Juniper ■ — not to
he confused with the Common Pasture Juniper of New Hampshire). Many
witches brooms on this species were found in eastern Maine of which nine
were selected for trial. As found these growths ranged from one to three
feet in height and were of varied growth patterns from narrow columnar to
broadly pyramidal. All were of dense growth with the foliage of the juvenile
needle-like type rather than the mature scale-like form. Propagation of these
forms has been successful and limited numbers of them are in the field
to observe growth and general behavior as independent plants.

Larix laricina Mill. Numerous witches brooms were also found on this
species in eastern Maine and although it is deciduous it is a beautiful
conifer. Several of these witches brooms are of a very dense desirable growth
up to 7 or 8 feet in diameter. Three selections have been tried for propa-
gation in the cutting bench and as yet none have been successful. Other
rooting teachniques or even grafting may be necessary.

Pinus resinosa Ait. — Red or Norway Pine. An unusually fine witches
broom was found on this species which measured approximately 12 ft.
wide by 5 or 6 ft. high. From hundreds of cuttings none rooted following
a variety of treatments. Pines are possibly the most difficult of all ever-
greens to root from cuttings and other methods may have to be used.

Pinus strobus L, — White Pine. Five witches brooms have been found
on this species three of which have been collected and propagation trials
made. One was unsuccessful and two others are still being tried. Again
other techniques may be necessary.

Abies balsamea Mill. — Common Balsam Fir. Witches brooms are very
common on this species but almost always have some serious defect for
ornamental purposes. However of the few attractive ones found two have
rooted successfully and are being grown in the field for observation. A very
recent acquisition shows the most desirable shape and foliage characters
of any we have seen. Dwarf forms of this evergreen should be among the
best for severe climates.

Picea glauca (Moench) Voss. P. rubens Sarg. and P. mariana (Mill.)
B.S.P. Spruces; White, Red and Black. Witches brooms on these species
also usually have serious ornamental defects possibly due to the fact that
the organism causing the abnormal growth is often the parasitic dwarf
mistletoe which is itself undesirable.

Partial success has been had with one specimen from White Spruce and
two from Black Spruce.

In addition to dwarf evergreens originating as witches brooms there are
also other possibilities, the genetic dwarfs and those which might be called
physiological dwarfs. The genetic dwarfs are found as naturally occurring
seedlings either in nurseries or in the wild.

The only recorded genetic dwarfs of the red pine were found in Wolfe-
boro, New Hampshire, 50 years ago. These are now beautiful plants 15 feet
across and nearly as tall. As yet they have not been successfully propagated.

Tsuga canadensis L. (Carr.) — Hemlock. Gentic dwarfs and variants
of this species appear with some frequency but few of them have been se-
lected and propagated. We have one from Franklin, New Hampshire which
has rooted successfully. We have located a number of others and are
searching for more. This species widely regarded as the finest of the coni-
fers holds great promise of dwarf horticultural forms for northern gardens.

Picea species — Spruce. One genetic dwarf natire spruce has been found
with superior ornamental qualities that shows some promise of propa-
gation. We would like to locate more specimens.

Juniperus communis L. — Common Pasture Juniper. Although regarded
by farmers as a pasture weed this species has unusual ornamental possi-
bilities because of its variation in stature from prostrate to columnar. A
very fine group of the columnar shape has been located in New Hamp-
shire and is being given field trial. Propagation is quick and easy. As a
hedge plant it should be superior.

Juniper hybrids. A large group of naturally occurring hybrids between
J. vinriniana crebra Fern. & Grisc. (Red Cedar) and /. horizontalis (Sea-
side Juniper) has been located on the Maine coast. This is a most remark-
able group of plants that vary in height from a few inches to 5 or 6 feet
and in width from 3 to 15 feet or more. Foliage color varies from dark
green to blue gray. Many selections are now in the propagating benches
to test for rooting response.

A genetic variant of the White Pine (Pimis strohus) of a handsome
columnar shape and fine foliage has been located in Bedford. New Hamp-
shire and cuttings from it are now in the propagation bed.

Of many forms that may be found whether witches brooms or genetic
variants only a limited number respond to present propagation techniques
and of these only a fraction will have enough other horticultural qualities
to make them valuable.

However continued search and trial can eventually give us a list of dwarf
and other horticultural forms of native evergreens that will be suited to our
climate and needs, and should be second to none in beauty.

Geraniums

A seedling geranium of unknown parentage, grown by P. J. Durkee at
the New Hampshire Agricultural Experiment Station greenhouses, was
named White Mountain, and distributed to New Hampshire florists in 1955.
It has double white flowers on a clean, bushy plant that has good growth
habits under glass.

Another seedling having purple red flowers has been found valuable for
landscape purposes. It has been named Purple Heart. Compared with the
British Horticultural Color Chart, the petals shade from Solferino purple,
4^26 — 26/1, in the center to Rose Bengal #25 at the margins while the
inner halves of the two upper petals are Vermilion #18 veined with Rose
Madder #23.

Gladiolus

During the period 1947 to 1955, about twenty-three thousand hybrid gladi-
olus seedlings were grown and evaluated in an attempt to develop a multi-
ple stalked, small flowered, disease resistant variety for home garden use
in New Hampshire. Gladious improvement requires the growing of large
populations. From these seedlings, a few selections having one or more of
the desired qualities were made. Several of these selections have been sent
to other gladious breeders in the New England area for their use in con-
tinued breeding.

10

Gloxinia

In 1953 Mrs. J. H. Sanborn, Seed Analyst, New Hampshire Agricultural
Experiment Station, pollinated the flowers of Gloxinia, Sinninga speciosa
Benth. and Hook, with pollen from African Violet, Saintpaulia ionantha
Wendl. Repeated attempts to get seed by self-pollinating this particular
gloxinia plant had failed. When the gloxinia flowers were pollinated with
pollen of the African violet, they set pods and produced seed. This seed
when planted produced thirty plants. They gave no evidence of being hy-
brids yet were not as vigorous as normal gloxinia plants. When grown to
maturity, they produced typical gloxinia flowers of several colors. An ex-
amination of the anthers showed that no good pollen was produced by any
of the plants. None of them produced any seeds.

Root tip chromosome counts by R. B. Pike showed them to be haploids.
Attempts were made to double the chromosome number by soaking the leaves
in colchicine solution without success.

Lilacs

James Macfarlane Lilac

The common lilac, Syringa vulgaris L., is well adapted to New Hamp-
shire. In some years, though the plants are fully winter hardy, they either
fail to bloom or else produce ragged clusters of flowers because of late
spring frosts which injure the blossom buds before they open. Some other
species of lilacs bloom rather late, hence escape such damage to the flowers.
In our lilac trials, one of the best late-blooming varieties is Royalty which
is a variety of the hybrid species Syringa prestonae McKelvev.* Open-
pollinated seeds from Royalty have given many desirable plants. One of the
best has been named James Macfarlane in memory of the former well-
known florist at the University of New Hampshire.

Lilac 'James Macfarlane' has beautiful bright pink flowers. The florets
are single, the clusters large and branching. The variety blooms at an early
age, often on one-year-old rooted cuttings. It is particularly valuable as a
large shrub in landscape plantings or as a specimen shrub. Some years James
Macfarlene has produced a second flush of blossoms in July about a month
after the normal season for flowering. Propagation of this variety from two-
leaf, softwood cuttings taken when the plants are in full bloom is fully sat-
isfactory.

Nellie Bean Lilac

Nellie Bean was a sister seedling to James Macfarlane and resembles
the latter variety in many respects. Nellie Bean has purple flowers instead
of pink. This variety was named in honor of a former leader in the State
Federation of Garden Clubs which through its Lilac Foundation Fund sup-
ports the lilac research work. Nellie Bean can be propagated readily from
softwood cuttings.

* Royalty is commonly listed as a cultivar in the group 5. josiflexa P. but records
of the Dominion Experimental Farm, Morden, Manitoba, Canada, indicate Royalty to
be a hybrid of S. villosa x 5. reflexa selected at that station from seeds produced in 1922.

11

The James Macfarlane lilac. It is a late pink.

Anna Anihoff Lilac

This variety, named for a woman who for many years was a leader of
the Portsmouth Garden Club, is the third in the series of new late-blooming
varieties descended from Royalty. The flowers of Anna Amhoff when fully
open are white. Before the flower buds open, they are light pink. This
variety is more difficult to propagate from softwood cuttings than its sister
varieties.

Miss Kim Lilac

So far as we know Miss Kim lilac is the first named variety of the species
Syringa velutina Komarov ( Syn. 5. palihiniana Nakai). One of seven seed-
lings raised from seeds collected by E. M. Meader in the Pouk Han Moun-
tains of Korea in 1947, Miss Kim was named in 1954. It is a late lilac,
hence escapes frost damage when common lilacs are injured. Its fragrant
blossoms are lavender in color. One outstanding characteristic of Miss Kim
lilac is its small stature which makes it more desirable than other lilacs

12

for general landscape work. Morover, the leaves turn a burgundy-red color
in fall. This variety can be propagated readily from softwood cuttings.

Maybelle Farnum Lilac

This variety was named June 18, 1958. It is a seedling of James Mac-
farlane pollinated by a very dense cluster selection of our own breeding
descended from Royaltv. It blooms about two weeks later than common
lilac. The dense shouldered clusters are about nine inches in length. The
florets are single and of a lilac purple color. Propagation from softwood
cuttings is possible but it is not as easily rooted as James Macfarlane.

Other Late Lilacs

From crosses between James Macfarlane and Anna Amhoff, several
desirable plants have been chosen for further testing.

Since Anna Amhoff has been found difficult to propagate from softwood
cuttings, a white-flowered selection which can be propagated readily in
this manner is the most likely to be named.

Mary Blanchard Lilac

This variety belongs to Syringa vulgaris, the common lilac. It is an open
pollinated seedling of Congo and was distributed for trial as N. H. 4^70.

The flower clusters average eight inches in length, the florets are one inch
in diameter, single and the buds are purple when first open but become
violet in full bloom. It was named June 3, 1958. It may be propagated by
softwood cuttings.

Interspecific Lilac Hybrids

What was thought to be a hybrid between the common lilac. Syringa
vulgaris L. and 5. amurerjsis japonica (Maxim) Fr. and Sav., mentioned
in New Hampshire Agricultural Experiment Station Bulletin 383, has proved
to be only a seedling of the late-flowering species. Other attempts to make
the cross since have given some plants but probably none are hybrids.
Also attempts have been made to cross Miss Kim, and S. amurensis japonica.
If crosses between the late-blooming species can be made, some most inter-
esting seedlings should result.

Moiiarda

In 1955, Monarda 'Granite Pink' was released by the New Hampshire Agri-
cultural Experiment Station to nurserymen and home gardeners. This mem-
ber of the mint family, growing three feet tall and as wide across, and
covered with soft pink flowers from July to September, is an open-pollinated
seedling of a taller variety, 'Croftway Pink'.

New Hampshire florists have found 'Granite Pink' useful for cut flowers
when field grown. In the greenhouse it will produce a spring cut flower crop
from cuttings taken in January from dormant stock plants forced in a 50°F.
greenhouse. The color is a true pink which does not clash with the colors

13

A sample of Monarda in the breeding
block.

of other flowers in mixed bouquets.
Plants of 'Granite Pink' were sold in
most all states of the United States in
1958.

'Granite Purple' was introduced in
the summer of 1958. The flower color
is an official "Cyclamen Purple 30"
by the British Horticulture Color
Chart and the calyxes are green. The
leaves are pubescent and taper
straight-sided from a base half as
broad as the total leaf length. The
leaf margins have large single teeth
from the base to the tip and the
leaves are outstanding in their dis-
ease resistance and long life on the

plant each season. The husky, square-
angled stems grow about four feet
tall and bear up to six large flowers
per stem. Good plants, two years old from cutting or divisions will be
four feet in diameter and in full bloom during the first two weeks in August.

Modern hybrid Monardas (Bee Balms) contain genes from two native
North American wild flowers; the bright red Oswego Tea, M. didyma L., of
moist, shady woodlands and Wild Bergamot, M. fistulosa L., the lavender-
flowered species of dry, open fields. Until recently all garden hybrids have
been produced in England where seeds of our native species were sent as
early as 1744. 'Croftway Pink' was introduced from England during World
War II. 'Granite Pink' was produced at Dover. New Hampshire in 1953,
and 'Granite Purple' was raised in the same location a year later.

Monardas with flowers of other colors are now being developed with
emphasis on plants having attractive, disease free leaves and good com-
pact growth habit. This hardy perennial is a minimum-maintainence plant
adaptable to most soil types. Wherever they are grown, the Bee Balms at-
tract bees, humming birds and the interesting humming-bird sphinx moths
in large numbers.

Rambler Roses

The canes of most varieties of rambler roses now available to home gardeners
will freeze and die back when the minimum temperature falls below -15°F
in the winter. It is, therefore, an exceptional winter in nearly all parts of
New Hampshire when unprotected rambler rose canes do not freeze and
die. Loss of flowers results the following summer. These canes may be pro-
tected by burying then under a foot of garden soil, however this is not
always possible and, at best, is not convenient. The objective of the breed-
ing program on rambler roses has been the development of canes which
would withstand temperatures as ?ow as -20° or -25°F. without injurv.
Among the seedlings raised, selection has been made also for disease re-
sistant leaves, vigorous growth, ease of propagation, and a tendency to
repeat blooming all summer long.

14

To accomplish this purpose, a collection of species and hybrid roses from
various parts of the world was made at the Horticulture Farm of the N. H.
Agr. Exp. Station starting in 1950. It was known that ramblers or long,
flexible caned roses differ genetically from bush or short, stiff caned roses
by only one, or very few genes. It was also known that garden and wild
roses are complex hybrids, often isolated from breeding with one another
by various types of incompatibility. Thus the number and viability of seeds
to be expected from the cross breeding of any two roses and the probable
nature of their offspring are unpredictable.

The following account describes three named cultivars resulting from
this project and gives an idea of the general nature and breeding behavior
recorded for the many rose species and varieties involved.

White Mountains

Distributed for testing as N. H. #5217, this vigorous rambler resulted
from the self-pollination of Skinners Rambler in 1952. Its small white
flowers are borne in large clusters along the previous years' canes in late
June. The individual flowers have 45 petals and resemble the "sweetheart"
type rose in the bud. These flowers have remained on the plant in good
condition for as long as 37 days, turning greenish white with age. The canes
are slender, thorny and vigorous but not rampant. Additional plants are
easily propagated from Summer cuttings and all plants have been completely
winter hardy at Durham, N. H. The leaves are of the R. multiflora type,
bright green and disease free for the most part. "Wbite Mountains" has
crossed readily with other roses. This rambler produces large numbers of
small fruits, green in color, that mature viable seeds early in the fall.

White Mountains hardy climbing rose.
15

Skinners Rambler, the parent variety, was generously donated for this
work by Dr. Frank L. Skinner, plant breeder and nurseryman, of Dropmore,
Manitoba, Canada. It is a vigorous seedling of Rosa maximoivizciana Regel
having five pink petals, flowers in clusters in June on the preceeding years'
canes, and sets viable seed with a number of unrelated pollen parents. It
has 14 chromosomes and is therefore a diploid rose.

Grafton Pillar

Bright red, double flowers in large clusters produced continuously from
June until severe frost characterize this rose. Distributed for testing as N. H.
4^551, it is a second generation seedling of Skinners Rambler x Gruss an
Aachen, (see chart). The flowers have 30 petals and are in terminal clusters
on canes two to eight feet long which places it in the classification of a
pillar rose needing the support of a fence or post but not growing as long
canes as a rambler rose does. The canes are hardy at -20° F but the
leaves have a mild susceptability to both mildew and blackspot diseases.
Spraying with a good fungicide is recommended as a regular part of the
care of this variety. Leafy stem cuttings taken in July root and grow very
easily.

Gruss an Aachen, the first floribunda rose, is a dwarf, everblooming rose
whose canes are not hardy. It is a triploid rose having 21 chromosomes as
verified in this project, yet it produced the viable pollen used on the 14
chromosome. Skinners Rambler to produce a 14 chromosome seedling desig-
nated as SR3GAA. From this, 78 plants were grown in the F2 generation.
One of these had double flowers (30 petals I and two were everblooming.

Grafton Pillar rose is double flowered, everblooming;. and red.

16

Durham Pillar

Distributed for testing as N. H.
4^536, this is an open pollinated seed-
ling of Chevy Chase rambler. The
unique feature of Durham Pillar is
its apparent freedom from both black-
spot and mildew diseases. The leaves
are a glossy, olive green and are
thick and relatively rigid. In the
greenhouse the plant is an evergreen
and continues growth all winter at
70°F but will not flower under glass.
In the garden it blooms continuouslv
from June to frost in large clusters
terminal to the pillar-length canes.
The flowers have five petals and are
red. The canes winter-kill to the soil
line at temperatures below 17°F but
growth and flowering are satisfactory
from the short basal canes as from
a fforibunda rose when it has been
protected by mounding with soil.
This variety seems to have some
drought resistance and seems to grow
more vigorously in warm seasons.

Durham Pillar has 14 somatic chromosomes. It is a second generation
seedling of Rosa souUeana Crepin by a dwarf red China Rose hybrid which
had the appearance of a Polyantha variety.

It is suggested that Durham Pillar be used in a sunny exposure trained
on a low fence, against a stone wall or on a trellis or post where a total
plant height of four to eight feet is desirable. No fungicide is needed and
it is hoped that rose breeders will make use of this cultivar to produce a
race of disease free garden roses.

Durham Pillar rose, a hardy, ever-
blooniing, disease-resistant variety.

Other Seedlings

The following numbered seedlings are included in this record of rambler
rose breeding because they have been distributed for testing or have been
referred to in one or more publications in various ways. It is possible that
some of these may be useful for further breeding work.

N. H. 4^5211 (Skinners Rambler self-pollinated). A sister seedling of
White Mountains, this one has great clusters of double pinkish magenta
flowers, and makes an extremely vigorous and rampant growth. The stout,
thorny and bristly canes bear Rosa niultiflora type leaves not fully as shiny
as those of 'White Mountains'. Propagation of N. H. #5211 from cuttings
is easy. It is also a good seed parent.

N. H. #541 (an open-pollinated seedling of N. H. #.5211). Having
double white flowers like White Mountains, this one has dark, shiny green
Rosa multiflora type leaves that are free of blackspot and mildew diseases.

17

Thus the leaves remain on the plant long after frost. So far N. H. :^541
has not produced viable seeds. It was selected primarily for its excellent
foliage. The flowers are of poor quality.

N. H. :^5425 (Skinners Rambler x Yellow Pinocchio). Handsome glossy
foliage, rampant canes, and 25-petaled flowers of apricot color and hybrid
tea shape characterize this rose. The canes are not hardy at Durham, N. H.
without winter protection. The flowers are sterile, a condition typical of
offspring from crosses involving the diploid Skinners Rambler and tetra-
ploid hybrid tea varieties. Prepared slides of N. H. ^^5425 root-tips when
examined under the microscope revealed more or less equal numbers of
cells having 16, 17, 18, 19 and 21 chromosomes. The anticipated number was
21. Since it blooms but once in June, introduction as a named variety is
doubtful.

N. H. :^5414 (open-pollinated seedling of White Mountains). A botanical
curiositv, this thick caned rose bears a few twice-compound leaves. The
canes are covered with bristles but lack all thorns. It has 14 chromosomes
in the root tips.

N. H. #524 (Skinners Rambler x Betty Bland). Rosa blanda Ait. is a
thornless, extremely hardy species native to this area. Its genes were intro-
duced into a cross with Skinners Rambler through the use of the variety,
Betty Bland. Seedling N. H. :^524 is virtually thornless and has red canes
that are completely winter hardy. They are, however, subject to attack by
cane girdlers. The rather unattractive flowers have 29, small, magenta and
white-striped, irregular petals. The few seeds which have been produced
have failed to germinate. The leaves are dull, pale green and drop from
the canes early in the autumn.

N. H. :^5314 (Skinners Rambler x Rosa virp;iniana) . A single seedling
resulted from this cross made in 1953. This seedling has all the taxonomic
features of the pollen parent, Rosa virginiana Mill, but has produced no
seeds. A single seedless fruit produced in 1957 was like those of Skinners
Ramb'er, the seed parent. Apparently no other visible characteristics were
inherited from the maternal parent and this is the only seedling of Skinners
Rambler that has inherited bush form in the Fj. Like both its parents,
N. H. 4^5314 has 14 chromosomes. Taxonomically, the pollen parent is typi-
cally 7?. virginiana but is a diploid. Most R. virgin'ana are tetraploid.

Other Crosses Involving Skinners Rambler

Skinners Rambler has repeatedly been pollinated by the Apple Rose, Rosa
pomifera Hermm. Only two seedlings have resulted. One of them, N. H.
#5246 has proliferated carpels within the double pink flowers. The other
seedling resembled a wild R. muh'flora Thunb. rose and from it an F2
population of 76 plants has been grown. All were similar to their parent
and showed no evidence of hybridi;^y involving the Apple Rose.

The Apple Rose is one of the Caninae group which produces pollen bear-
ing seven chromosomes and egg cells having the seven mates plus 14 extra
chromosomes, thus giving the plants a somatic tetraploid number of 28.
Very different offspring would thus be expected from a cross in which the
Apple Rose was the seed parent than if it were the pollen parent.

Another Caninae rose, Rosa eglanteria L., the Sweetbriar, has been used
in rose breeding to introduce genes for fragrant foliage. It is a pentaploid
species having a total somatic chromosome number of 35 but the pollen

18

carries only seven of these chromosomes. In this project nine seedlings were
grown from the cross R. ei^lanteria x N. H. i^5211 rambler and one from
Skinners Rambler x Lady Penzance (7?. es;lanteria x R. joetida hicolor
(Jacq.) Willm.). These have not been fully evaluated but do exhibit the
usual first generation characteristics of Sweetbriar hybrids. Seedlings in
which the maternal parent was Sweetbriar strongly resemble that species
with such minor variations as stem and flower color, extra petals, etc., and
the fruity fragrance of the foliage has been retained. Seedlings in which
Sweetbriar was the pollen parent exhibit little if any noticable resemblance
to that species. Those few second generation seedlings of Sweetbriar that
anyone has been able to grow since the first hybrids were produced around
1895 are said to resemble that species in all respects.

Attempts to cross both the Damask Roses and French Roses, hybrids of
Rosa damascena Miller and Rosa gallica L. respectively, with Skinners
Rambler have resulted in very few seeds and no worthy seedlings. Oratam
is a hybrid Damask rose having large flowers and a strong perfume. Nine-
teen seedlings from the cross Skinners Rambler x Oratam inherited the fra-
grance, copper flower color, leaf type and armature of Oratam, but also
extreme susceptability to blackspot. All of them are sterile. Eleven other
seedlings having no Oratam features resembled vigorous, wild species, R.
multiflora.

Skinners Rambler was crossed with Bonfire, a double flowered, red,
multiflora-type rambler and an F^ population of 63 seedlings was grown.
All grew vigorously until attacked by mildew and all were discarded.

Skinners Rambler x Rosa ruhrifolia Vill. yielded only seven seedlings
following repeated pollinations and none of these had any resemblance of
the Redleaf Rose. R. ruhrifolia is another of the Caninae roses which is
tetraploid but has only 7 chromosomes in the pollen.

Perhaps the most luxurious, wrinkled, dark green foliage of any rose
is that of Max Graf, a hybrid of Rosa riigosa Thunb. and Rosa wichuraiana
Crep. It has long flexible canes and large, single, pink flowers. Max Graf
has 14 somatic chromosomes and is nearly sterile. Skinners Rambler x
Max Graf yielded an F^ population of 24 plants, all thin caned, small
leaved ramblers that are not particularly resistant to mildew disease.

N. H. #500 {Rosa wichuraiana x Betty Bland). This is believed to be
the first successful cross involving the genes of Rosa wichuraiana and R.
blanda. A thin, red caned rambler with single pink flowers that rarely pro-
duce seeds, N. H. #500 is extremely winter hardy. This hardiness comes
from the pollen parent, Betty Bland, which N. H. #500 resembles in many
of its characteristics. It has 14 somatic chromosomes.

19

GRAFTON PILLAR ROSE

N. H. Agricultural Experiment Station, 1958)
red, pillar, 2n

r

SR3GAA (seedling)

(N, H. Agr. Exp. Sta., 1953)

pink, rambler, 2n

SR3GAA (seedling)

(N. H. Agr. Exp. Sta., 1953)

pink, rambler, 2n

Skinners Rambler
(Skinner, 1955)
pink, rambler, 2n

unknown double flowered
garden rose

Gruss an Aachen
(Geduldig, 1908)
white, dwarf floribunda, 3n

Rosa maximowicziana Regel

(N. E. Asia, cult. 1880)

white, sprawling, tall bush, 2n

Frou Karl Druschki
(P. Lambert, 1901)
white, hybrid perpetual, 4n

Franz Deegan
(W. Miner, 1901)
yellow, hybrid tea

1

Merveille de Lyon

(Fernet pere, 1882)

white, hybrid perpetual

Mme. Caroline Testout
(Pernet-Ducher, 1890)
rose, hybrid tea

Kaiserin Auguste Viktoria
(P. Lambert, 1891)
white, hybrid tea

t

Baroness Rotschild sport
(Pernet pere, 1868)
rose, hybrid perpetual

Mme. de Tartas

I
Lady Mary Fitzwilliar

(Bennett, 1882)

pink, hybrid tea

Coquette de Lyon

t

Souv. de la Reine
de Angleterre sport

Devoniensis
(Foster, 1841)
white, tea

Victor Verdier
(Lacharme, 1852)
rose, hybrid perpetual

E. B. Risley, 1958