JARS v54n3 - In Search of Native Azaleas

In Search of Native Azaleas
Henry T. Skinner's Historic Trip, 1951
Henry T. Skinner
Reprinted with permission from the Morris Arboretum Bulletin, April 1955, Vol. 6, No. 2.

The first part of Skinner's account appeared in the Spring 2000 issue of the Journal American Rhododendron Society.

New York and Westward
A check of the flowering guide at this stage indicated the need for northern samples of the early species and for more westerly collections of both these and R. calendulaceum . Accordingly on May 26th a route was chosen via the Pocono Mountain area of northern Pennsylvania to the Finger Lake region of central New York for Pinxterbloom and Rose Shell Azaleas - or for what passes as these two species after their too rapid or too sociable post - glacial trek to the Carolina Hills of Ithaca. They were here in abundant bloom, in excellent color and in oft-proved hardiness but both species are a little more like one another than are R. roseum and nudiflorum of Virginia  - a fact which has worried both botanist and azalea growers of the north on more than one occasion. Good collections of New York State Pinkshell were also made in the entertaining company of Dr. C. G. (Rhododendron) Bowers in the hills above Binghamton.

Travelling southwest into Pennsylvania the same azaleas were now getting past bloom except for some very showy specimens of R. nudiflorum in the high and late-season plateau area of Somerset County, Pennsylvania. On further into West Virginia R. roseum and nudiflorum were in flower at higher elevations and in the region south of Morgantown and Elkins R. calendulaceum was abundant in the early, large flowered form in shades of yellow to deep orange. Recognizing a possibly parallel situation to that in certain of the native blueberries Dr. W. H. Camp had earlier suggested that orange R. calendulaceum in this evidently tetraploid early flowering phase may quite logically represent a species originally derived from early hybridization between diploid red and yellow progenitors. Towards furnishing proof for this hypothesis an evident need was the field discovery of such suitable red and yellow diploid azaleas, if they should still be in existence. The Oconee Azalea is a red with suitable characters except perhaps, its time of flowering, and at this stage of collecting it was hoped that a small-flowered, fairly late, clear yellow azalea might perchance be found on some secluded slope of these westerly hills of the Virginias, Kentucky or in the Ozarks. Anticipating our story we can say that a late flowering diploid yellow was never found, and probably never existed but the diligent though fruitless search for it covered many square miles of territory and was always interesting. A number of unusual yellows actually turned up in West Virginia, in Kentucky, Georgia and Tennessee but always they were solitary plants and usually the product of hybridization between a late phase of R. calendulaceum and either R. arborescens or viscosum .

Hybrids
One of the very interesting hybrid swarms was found on June 14 on the eastern slope of Spruce Knob Mountain in West Virginia. The plants were scattered through an abandoned pasture in a region where R. calendulaceum , nudiflorum and roseum all grew and bloomed together. The progeny of these triple matings were bizarre in the extreme - short and tall bushes bearing large or small flowers in every color from coral pink through salmons to rich lavender, pale yellow or pure white. The last was large flowered and otherwise identical with the Flame Azalea. Such happenings, exciting to the horticulturist, could obviously be most confusing when unexpectedly encountered in an herbarium where such specimens customarily lack any reference to flower color or to peculiarities of their occurrence. White-flowered hybrid progeny seem relatively frequent when parental R. nudiflorum is involved.

Returning to Virginia, final collections were made from the highest elevations of the Pinnacle Mountain transect before again striking southwest for later investigation of the more southern species. A later flowering and somewhat redder phase of Flame Azalea was found in partially open bud on White Top Mountain and High Nob in southwest Virginia, at a time when the last blossoms of the normal large and orange-flowered R. calendulaceum were scattered on the lower slopes. This same joint occurrence was likewise met on June 6th on Big Black Mountain in Kentucky, only here at the higher elevation of over 4,100 feet. The later "Camp Red" phase of the summit would obviously not be at its best for another two weeks or more.

In Quest of R. cumberlandense
Planning a return to the interesting azaleas of Black Mountain, we headed northwest for a general Kentucky reconnaissance in an eleven county circular swing to Yahoo Ridge, type locality for R. cumberlandense at the Kentucky end of the Cumberland Plateau.

Within a few miles the first little red and red-orange flowered azalea plants were found on a ridge of Pine Mountain in Letcher County, an azalea which in "best" forms makes a low, twiggy bush, often quite stoloniferous, with glossy green leaves often glaucous beneath and which may or more probably may not be quite the same as the late azalea of Black Mountain. At least on Pine Mountain this is undoubtedly R. cumberlandense of E. L. Braun's description and its smaller, thin-tubed flowers are immediately suggestive of a diploid if the earlier, coarsely large-flowered Flame Azalea is truly tetraploid - a point to be tested by later cytological examination of living plants collected for this purpose.

Leaving Pine Mountain there was an interval of several miles in which only normal R. calendulaceum , past bloom, was seen, but again at higher elevation in Owsley County beautiful little geranium-red azaleas were in shining bloom on a rocky cliff face; they remained with us in Clay County, in Laurel County and in fact seemed quite common throughout these wooded hills of southeast Kentucky, all the way to Yahoo Ridge where the type locality for R. cumberlandense was revisited with the aid of detailed directions kindly furnished by Dr. Braun. Unfortunately the station where Braun had collected some time after logging operations in 1935 was now so rapidly reforesting that the shade was becoming heavy and the azaleas poor - a rotation which was frequently observed on this journey. Again and again the most striking displays of azaleas were in open woodland which had obviously been logged, cleared or burnt a few years previously. Presumably it is the scattered parent plants which burst into bloom with the sudden sunlight, set abundant seed and populate the forest floor before young trees again almost shade them out. In the long view one gains the impression of ephemeral, constantly shifting populations, except perhaps in the case of conservative R. prunifolium of West Georgia or R. speciosum of the Savannah River. By the average plant age the latter species seem to have occupied the same territory for many years. They reproduce sparingly.
As noted in this region perhaps the finest single Kentucky collecting point for the Cumberland Azalea was on a fire tower hill in west central Knox County. The road up this hill was one of those eroded rock and mud affairs which may have been passable to a jeep in good weather but which caused the Chevrolet to rest quietly near the main highway during an attack on foot. The hill was covered with open deciduous forest and towards the top, flowing over the ridges and down the sides of steep gullies was a multicolored riot of azaleas. It must have been a fairly old growth for while some of the flat-topped bushes were only waist-high others were well above eye level, indicating that fair height is attained by this species, at least in partial shade. Under these conditions, and compared with normal Flame Azalea, the flowers seemed especially thin-tubed and delicate and with a color luminosity, in the filtered sunlight, which the other wholly lacks. The shades of color were infinitely and widely variable from pale straw yellow through yellow-orange to red, and from salmon through pink to translucent cerise as lively as shot silk. Such diversity was often later found, although a constant leaning toward orange-red and red suggests that the latter may possibly have been the original color of this azalea.

Having confirmed the Kentucky occurrence of this distinct phase of former R. calendulaceum the next obvious task was to determine with some accuracy the limits of its distribution. So far it had been confined to the northern heights of the dissected west escarpment of the Cumberland Plateau so that a logical course was to follow this westerly escarpment southward - and the decision proved a wise one. Leaving Kentucky on the 9th of June, this same little azalea was followed in comparative abundance into the upland woods of Scott County, Tennessee, into Fentress County, Overton County, Van Buren County, Sequatchie County, both west of the Sequatchie Valley and to the east on Signal Mountain. From here we were headed for Georgia - and the azalea was there too on Fort Mountain in Murray County. Continuing at about 3,000 feet elevation (in contrast to early R. calendulaceum of the lower mountain slopes) it was found towards the summit of Mt. Oglethorpe in Pickens County, and on Branch Mountain in Dawson County. It was on this mountain that a spot of brilliant red, like a scarlet tail-light, shone from the top of a cliff bordering the new highway 136. This little beacon was too fascinating to pass up, even though the only approach lay by way of a long flanking climb. But the reward was a tiny, twiggy, rock-clinging azalea plant 6 inches high, a foot across, gray leaved and covered like a pin cushion with its little red bells - as extreme a form of this R. cumberlandense as one could hope to find and a gem for the garden if its habitat is not unduly altered by cultivation. Travelling northeast into adjacent Lumpkin County the azalea stays with us near Woody Gap. In Union County it is especially abundant on the mountain slopes above Lake Winnfield Scott and not far away, just east of Wolfpen Gap in Vogel State Park, it covers a hillside in a billowy patchwork of clear yellow, orange, orange-red, cerise and all shades of salmony pink to apricot - both colors and plants so reminiscent of those of our fire tower hill in Kentucky that even before making a detailed check of less obvious characters one could scarcely doubt that this was the same Kentucky azalea. But was it? This particular spot happened to have been sought out by design for it is the type locality of R. Bakeri described by Lemmon and McKay in 1937, four years before R. cumberlandense was named by Braun from Yahoo Ridge. Since both descriptions fit these plants with reasonable accuracy it would seem that this gay little bush of the Cumberland Plateau must soon shed its dual personality to be recognized by the single, prior, though less happily descriptive name of R. Bakeri . Here in Georgia its color may tend slightly more toward the yellow and yellow-orange and its flowers may be slightly larger than when it was seen in Kentucky but such differences would seem to be of very minor consequence.

Confusion in Macon County
Still anxious to find where else this little late red azalea might be, another visit was next paid to the Nantahala region of North Carolina, a few miles across the Georgia border. The first plants were found in a deep valley on the approach to Wayah Bald from the west. The plants were 2 feet high in stoloniferous patches deep red in color and just coming into bloom at this higher elevation. But they were not alone. On all sides were bushes in a bewildering array of colors, of heights to 15 feet or more and of flower sizes to 6 centimeters across the "wing" petals. Either R. Bakeri had gone crazy or it had met up with something else. The latter seems probably the better guess, for not far away were a few late-flowering individuals of normal, early Flame Azalea. The sampling and collecting of this amazing population consumed a full half day during which time the characteristics of these intermediates became reasonably familiar. Finally heading to Nantahala Lake and Wayah Bald, imagine our astonishment at discovering that the fast opening azalea display around the lake and well up the slopes of the mountain was composed not of R. Bakeri or "nomal" calendulaceum but entirely of recurring batches of these vari-colored intermediates which eventually settled down to something resembling a reasonably uniform "type" of their own.* Other collections were made on later visits to this region, and many more in principally orange and orange-red colors were subsequently found at higher elevations (above 3,000 feet) north through the mountains and right back again to southwest Virginia and Kentucky. A seeming third phase of the R. calendulaceum complex presents a puzzling pattern which will need much further study for elucidation of its true nature and origin; but the fact of its existence begins to shed light on the confusing flowering-time behavior of R. calendulaceum from different collection sources.

And Into Alabama
Having followed the Cumberland Azalea to Georgia and North Carolina there remained the possibility that it might also occur in Alabama - since the Cumberland Plateau enters into the northeastern part of this state. On the 18th of June the Chevrolet was consequently headed towards Jackson County, Alabama. Along the way some excellent Sweet Azalea, R. arborescens , was found in full flower, white with pale yellow blotch, growing with Catawba Rhododendron in a moist valley near Cloudland in DeKalb County and again not far away, while after crossing the Tennessee River and making a sharp climb of the steep ascent of the plateau north from Scottsboro in Jackson County abundant Cumberland Azalea was still in flower in open forest near Kyles on Crow Mountain. It was certainly in northern Alabama and on reflecting the matter in camp that night there came a wild thought of Alabama's highest point, isolated Mt. Cheaha, a hundred miles south in Talladega County. Pulling a long shot, we packed lunch next morning, took to the mountain road which became poorer and very dusty through the forest climb up Cheaha, and by noon were enjoying this lunch seated amid Cumberland Azalea right on the summit of the mountain! They were a little past bloom but there was ample color to aid recognition of this gratifying find at a lone point so far from Kentucky and Yahoo Ridge. But this was no large batch of azaleas; beneath the wind bent oaks were perhaps a few hundred plants in this colony which must have been isolated for a very long time. It was hoped that they might show something more of original characters or flower color but upon superficial examination they were similar indeed to the little azalea we had followed so far. It is of passing interest that Rehder did not record the existence of a calendulaceum-like azalea in Alabama and that the reference in the Eighth Edition of Gray's Manual should properly refer to the Cumberland Azalea rather than to R. calendulaceum proper.

The Texas Azalea
Two further geographic possibilities remained for this fascinating plant, the southward extension of the Cumberland Plateau south of the Tennessee River in North Central Alabama and - a very long shot - the Ozark Mountains of Arkansas and eastern Oklahoma which has azaleas and certain interesting representatives of other eastern plants. The next day, June 20th, was spent in the R. alabamense hills of Cullman and Winston Counties, but only late forms of the latter species were found, no Cumberland Azalea since the hills are perhaps too low and, strangely, not even any R. arborescens , which had been expected.

No collections were made on the long drive through Hot Springs, Arkansas, to Mt. Ida for the night but luck was better during the next two days coverage of the principal mountain peaks of the Ouachita and Boston Mountains, the length of Rich Mountain and adjacent LeFlore County, Oklahoma, impressive Magazine Mountain, Flat Top Mountain and northerly White Rock Mountain in Franklin County, Arkansas. There was no Cumberland Azalea as had been vaguely hoped but local R. oblongifolium , the Texas Azalea, was found in several places with sufficient plants still in flower for at least representative collections. This generally white and rather small flowered species is confusing in that it so frequently grows side by side with a pubescent-leaved pink azalea akin to R. roseum , and evidently breeds with it; the white form may be more adapted to moist valley sites and the other to drier hillside slopes but the line of preference is not strong. There is needed an earlier season and more careful study of these Ozark plants than was possible in this too rapid survey.

On the third morning in Arkansas, on June 24th, the car was again headed back towards the now-passing eastern azaleas. It chanced to be a Sunday, with Sunday drivers in slow lines on the highways but nevertheless nearly 600 miles were covered before nightfall in eastern Kentucky. No azaleas were collected; none was seen and they were doubtless sparse to nonexistent over most of the rich agricultural land traversed.

Back to the Alleghenies
The next week was spent in a run north through the mountains of Virginia and West Virginia and back south into Tennessee and North Carolina in quest of late forms of R. calendulaceum and of northerly R. arborescens and viscosum , wherever it might occur. A last visit was paid to the late red azaleas of Kentucky's Big Black Mountain and more of the Cumberland Azalea was found in Wise County, Virginia, but father east at elevations of 3,000 feet and above it gave way to the late phase of R. calendulaceum , mentioned earlier, which was also of plentiful occurrence on the high points of the Alleghenies from Grandfather Mountain and Mt. Pisgah, west to the Tennessee border. Throughout this tour the Sweet Azalea was fairly plentiful along stream sides of the upland valleys and in some places, as at Mountain Lake, Virginia, and on Great Pisgah in North Carolina, it was hybridizing freely with R. viscosum to produce variable and often pink-flowered hybrids quite similar to entire populations seen in northern Pennsylvania a month later.

The Sweet Azalea tends to be quite variable in certain characteristics and throughout its range from New England to Georgia and Alabama. It may be variable in habit from low, wide spreading and bushy in open places to tall and leggy in denser woods; its foliage may be glaucous beneath or entirely green; its corolla may be pure white or carry yellow blotches of varying intensity and in flower size it may be a plant of mediocre attraction to one of quite outstanding quality. A clone with especially large and showy flowers was found on the east fork of the Pigeon River in Pisgah National Forest but others almost equally good were seen at intervals. Such individuals from the horticultural stand point were quite superior to over-extolled var. Richardsonii of Wayah Bald whose flowers are medium in size and whose dwarfness seems a product of windswept exposure which is not expressed in the forest shelter at a few feet lower elevation.

A Mountain-top Marvel
A fine Fourth of July found the Chevrolet headed towards headquarters of the Great Smoky Mountains National Park in Gatlinburg, Tennessee, but wisdom prevailed in time for the effective substitution of a collecting detour over little-travelled Max Patch Mountain. Holiday makers had thinned somewhat by the next day, permitting a visit to Park Headquarters for the advice of Arthur Stupka, Park Naturalist, concerning the azaleas of the park and especially of now famous Gregory Bald on the park's southern rim. Representatives of the Gregory Bald population, as collected earlier by W. H. Camp, had been inconclusively studied previously and there was a real need to discover what this puzzling situation might actually be. Mr. Stupka was helpful indeed in providing access to the Park Herbarium and in giving suggestions on approaches to Gregory. A day's food supplies were laid in, the distance was covered to Cades Cove, a pack was made up with temporary presses, photographic equipment and a blanket and the stiff 4½ mile climb to the mountain top was started in rather late afternoon. A cabin at about 4,900 feet elevation was reached at dusk, leaving just time to complete the distance to the summit for a preview of the azalea display before cooking supper and turning in for a night's rest. Unfortunately the "night's rest" was enjoyed in the damp cold against which one blanket afforded little protection. It was also amid rather noisy wildlife from inquisitively reconnoitering mice to larger creatures, perhaps bears, which by morning had spirited away my lone loaf of bread. But there was enough food for a good breakfast and with warming sunlight and the azaleas, discomforts were soon forgotten.
The azaleas of Gregory Bald are at first glance bewildering and almost unbelievable. The mountain is a true "bald" having a broad, grassy summit fringed by scrub trees leading quickly into vigorous deciduous forest. The origin of the bald is unknown but it is probably man-made, resultant from earlier Indian grazing. It is the marginal region between trees and the grass sod which supports a peripheral band of a bizarre collection of azaleas - thousands of plants in every imaginable hue from pure white to pale yellow, salmon yellow, clear pink and orange-red to red. Many of the flowers are yellow blotched, many of the bushes are stoloniferous and foliage varies from normal to deep glossy green, often glaucous beneath. Obviously it is a complicated hybrid swarm dating, in the older plants, to perhaps thirty years ago when some happening such as a brush or forest fire may have been responsible for the start of this strange and fascinating collection. Assuming that these were in fact hybrids the evident procedure was to search for the species which might have been involved in the hybridization process. The red and red-orange colors were an obvious lead and bushes of a Bakeri -like late flowering phase of R. calendulaceum were soon recognized, particularly on the west and southern sides of the bald. Such azaleas had been observed lower on the trail on the approach to the summit. From previous experience the clear pinks suggested hybridization between a red and white and white clones gave surer evidence of a parent of this color. Glossy leaves and glabrous shoots suggested that one such white might well be R. arborescens . A quick search for this species was unsuccessful but it too had been seen on the approach to Gregory Bald. That it actually grew in the vicinity of the Bald has since been confirmed by F. C. Galle of the University of Tennessee who has made a special study of this population. In the search for R. arborescens a visit was paid to nearby Parson Bald a mile to the south and on this mountain was found a splendid growth of a second white azalea, the dwarf and stoloniferous form of the Swamp Azalea, R. viscosum var. montanum whose characteristic small, sticky flowers and suckering root system had also been recognized as being carried by many of the Gregory hybrids. R. viscosum may have been growing on Gregory itself or its pollen could easily have been carried this short distance by flying insects. Certainly it, with the other two species mentioned, was involved as an original parent of these plants. These three species are the only ones likely to be met at this elevation and in this particular region. Full collections were made of the Gregory population for later detailed study. While hybrid swarms involving as many species are not rare among eastern azaleas, no other yet seen has equaled this one in impressive size and effect. While many of the plants are beautiful from the horticultural standpoint it is fortunate that they are protected by National Park authorities for all to enjoy. Each color could be simply reproduced by cross-pollinating the same species under artificial conditions.

The study and photographing of this collection was still not complete by dusk, necessitating a second night on the mountain which was rendered slightly more comfortable by a harvest of fern fronds for softness and a little warmth. Dry cereal provided a slim breakfast next morning but work was completed in time for a rapid descent starting by noon.

Being reasonably close to Knoxville, Tennessee, a visit was paid the University to discuss azalea problems with members of the Department of Botany and to review briefly a fine set of herbarium material well worthy of later study. This was July 9th and since more late material was still needed from the mountain areas a route was chosen via Wauchecha Bald in Graham County, North Carolina, Robbinsville, a return to Wayah Bald via the Winding Steps road, a northern swing over Cowee Bald, memorable for dew-laden red azaleas and a brilliant sunrise over its cloud filled valleys, and thence to Highlands, North Carolina. The principal collections of this tour consisted of R. arborescens , late specimens of the red-orange Flame Azalea including an especially fine stoloniferous clone in full bloom near Nantahala Lake on July 10th, and R. viscosum var. montanum , also in excellent bloom, pure white, in low thicket growth in open woods near Highlands. Occasional hybrids of this plant with the Sweet Azalea can be striking with their large pink flowers, as are similar hybrids with viscosum itself at lower elevations.

The Plumleaf Azalea
It being now the 12th of July a reference to the collecting map indicated that the late red azalea of Georgia, R. prunifolium , should be in flower. Its type locality is near Cuthbert in Randolph County, southwest Georgia, and in this direction the Chevrolet was turned from the hillsides, the rhododendron forests and the delightful climate of Highlands, North Carolina. The only detour made in crossing Georgia was in search of the Sweet Azalea at the southernmost part of its range in Upson County of central Georgia. The search consumed nearly a full day but azalea was at last found in splendid quantity and, strangely enough, in full bloom in spite of this low elevation so far south. In pure white flower it followed the banks of a small Moccasin-infested tributary of the Flint River, the same azalea by all outward characteristics as its counterpart of 800 miles away in West Virginia.

Fort Gaines, Georgia, is a sleepy little town on the banks of the Chattahoochee, the river separating Georgia and Alabama which, 75 miles farther south, joins the Flint River (from Upson County) to become the Apalachicola of northern Florida. Both Fort Gaines and Cuthbert, 20 miles northeast, are situated in a region where the clays of the rising Coastal Plain have been cut into deep gullies by small meandering streams. The sides are often so steep that the only access is by wading the stream, and one is almost forced to do this (in spite of the Water Moccasins) by the dense cat-briar tangles of the wooded surroundings.

It is in these gullies of a few Georgia and Alabama counties, generally centering on Fort Gaines, that the Georgia late red azalea, R. prunifolium , is at home. Here, on steep slopes, wherever enough light has penetrated to permit flowering, it is found in round-topped bushes up to 12 feet high in reds, red-oranges, apricots and orange-yellows. The color range is not far different from that of the Cumberland Azalea and after seeing the latter for so long one is impressed by the similarity between the two. They both have those characteristic ridged flower tubes in the bud stage; they are both late, both red, and in more detailed morphology have little to show reason why they could not be quite logically and quite possibly regarded as high and low elevation derivatives from a common ancestor. By its lateness of bloom and geographic isolation R. prunifolium has not had the opportunity for recent gene exchange with other species. Thus it lacks the aggressive adaptability of its mountain counterpart, so that now, even in its chosen locale, young seedlings are seen so infrequently that one wonders how much longer it may persist without more effective protection than it now receives.

The type locality for this species, 2¼ mile northeast of Cuthbert, is now a golf course with no azaleas evident but a good collection was made in a small ravine 1¼ miles distant. The visit to Fort Gaines also provided an opportunity to discuss mutual interests with that authority on southern azaleas, Mr. S. D. Coleman, who not only showed me his own unusual plants, so well tended and arranged, but who also provided a valuable lead to a curious little May-flowering white azalea of Central Georgia and Alabama, hitherto overlooked. The next day or so was spent in following this low growing plant, now past bloom, as far as Mississippi. On a basis of characteristics which lie somewhere between R. viscosum , serrulatum and oblongifolium , it has not yet been taxonomically assigned as a previously described entity and should certainly be credited to Mr. Coleman if a new designation is warranted.

Hammocksweet
Returning to Mississippi on a stifling 18th of July with the thermometer hovering around 104°F, the first plants of true R. serrulatum , the Hammocksweet Azalea, were found in flower the day following on the edge of a wooded swamp in Jones County. Through the following week and a half, it was chased in equally good flower into southeastern Louisiana, east around the Gulf Coast to within a few miles of Lake Okeechobee in South Central Florida, back to its type collecting locality in Lake County, Florida, north again to the edges of the Okefenokee Swamp and again east to Folkston, Georgia, and the type locality of Rehder's R. serrulatum var. georgianum . Throughout this thousand miles and more the Hammocksweet Azalea showed no excessive variation. At times it is true that its leaves or dormant buds became more silky pubescent, its flower pedicels varied from pale green to deep red in color and its flowering season was obviously prolonged in lower Florida where single individuals may bloom from July to October or later, but essentially it remained the same sticky-tubed and rather inconspicuous little white azalea of the bog tussocks and the cypress island of the southern waterways. At times it formed rounded bushes 10 feet tall, but it was often low or producing but a few rangy stems seeking light through a dense cover of vine-covered holly or palmetto. The very late flowers of individual specimens could well be a characteristic worthy of exploitation in some future race of garden hybrids.

The Northward Return
With good collections of R. serrulatum one could feel with fair satisfaction that the gamut of southern azaleas had been about run, until such time as return visits to puzzle areas might be called for in another year. A northward return was thus in order, so planned as to catch any further outliers of the R. serrulatum complex together with a fairly detailed survey of its northern counterpart, R. viscosum , which should now be in scattered bloom well into New England.

Leaving Folkston on the 28th of July, our route headed towards Savannah and the Georgia side of the Savannah River where late azaleas had been observed during the R. speciosum season. The only collections this day were of fine specimens from a northerly distribution of Befaria racemosa , the curious ericaceous Tar Flower which, with its spikes of pink blossoms, is suggestive of a primitive azalea form. Here in coastal Georgia it grows on dry soils of the pine-palmetto forest. Farther south the scattered clumps of this single North American representative of a Central and South American genus is a frequent sight along the Florida roadsides.

Occasional Hammocksweet Azaleas were seen on the way to Savannah while, bypassing this city, the first low white azaleas resembling R. viscosum rather than R. serrulatum were found at a woodland edge in Effingham County. Farther along, in Screven County, there was found a swamp near Oliver where the swamp tussocks were covered with quite normal R. serrulatum , the swamp margins with a very variable dwarf and stoloniferous azalea, sometimes highly pubescent in its buds and leaves which was clearly much more akin to R. viscosum than the other species. On drier land an outer circle of R. canescens , past bloom, completed the azalea picture. This was the last collection of R. serrulatum , which does not seem to spread north of the Savannah River. It is clearly a region where the two late white azaleas meet and as such it is likely that gene exchange with resultant variability could be expected here in East Central Georgia.

Crossing the Savannah River on Route U.S. 301, this road was followed north to Baltimore, as it parallels the coast some 100 miles inland. Throughout the distance of the Carolinas, Virginia and Maryland, R. viscosum was mass-collected, usually in good bloom, at intervals of approximately 60 miles. From Baltimore it was followed past Philadelphia into New Jersey, across New Jersey to Connecticut, and across Connecticut and Massachusetts to Cape Cod and even to the island of Martha's Vineyard where it was flowering on August 8th. This was another thousand mile run in which the variation of one species could be observed, step by step, until it became a fascination that terminated only as the last plants were collected. From the dwarf, twiggy and semi-evergreen bushes of the marshes of South Carolina to the tall, gray leaved and large flowered shrubs on the pond margins of Cape Cod, the Swamp Azalea is much more changeable than its sister of the Gulf Coast. Rehder has divided it into eight varieties and forms. One could make these many more, or less, depending upon the viewpoint of the observer. It seems certain that not a little of the trouble is due to R. viscosum and arborescens having met on occasion in the northern states, as was strongly suggested by the last New York State and Pennsylvania collections on the return to Philadelphia. In some of these northern swamps genes have been so freely exchanged between these two species that nomenclatural assignment of present populations becomes virtually impossible. The situation is similar to that previously noted with regard to R. roseum and nudiflorum . But in spite of these local happenings, R. viscosum can still be regarded as "good" a species, though variable, as R. roseum , nudiflorum or serrulatum .

This last run from central Pennsylvania to Philadelphia was on Sunday, August the 12th, and thus ended, after twenty-one weeks and 25,000 miles of almost continuous collecting, our quest for native azaleas. A few additional collections have since been made, as doubtless there will be others in the future. From this major field survey were secured 8,000 herbarium specimens and 500 living plants whose study should throw much new light upon the nature and the behavior of these plants. The herbarium specimens, now mounted and catalogued, are deposited in the herbarium of the Morris Arboretum of the University of Pennsylvania, at which institution the collection of living plants is also maintained for future observation and for their use in current cytological studies.**

* Though not realized at the time, distinctive qualities of the Flame Azalea of the Nantahala region have previously been pointed out by Braun in The Red Azalea of the Cumberlands , Rhodora, 43:33, 1941.

** Dr. Skinner also prepared, as a matter of record, a listing of herbarium and living collections.

JARS Editor's Note: Several azaleas in Dr. Skinnner's report have since been reclassified. Consult the RHS The Genus Rhododendron, Its classification & synonomy , 1996, published by the Royal Botanic Garden Edinburgh.