Fig. 1. The two subspecies of G. paraguayense growing side by side in the Huntington Botanical Gardens.

Photo by Daryl Koutnik.




Graptopetalum paraguayense (N.E.Br.) Walth. first appeared about 1904 as volunteer seedlings in the glasshouse of Frank Weinberg in New York. According to Weinberg, the seeds came in with cacti exported from Mexico by J. A. McDowell. First known under the unpublished names of Echeveria weinbergii and E. arizonica, it was not validly named until 1914, when N. E. Brown, believing it to be from Paraguay, named it Cotyledon paraguayensis. It has also been placed in Sedum and in its own genus, Byrnesia, and by now has a synonymy of ten binomials. No native locality was known until 1979, when Alfred Lau collected a variant on Cerro Bernal, Tamaulipas, Mexico. His collection, described here as subsp. bernalense Kimn. & Moran, has smaller yellowish green leaves, in contrast to the larger greyish violet leaves of sub­sp. paragyayense.


Ideally, a new plant should be discovered by a botanist who quickly publishes the pertinent facts and gives it a name it can go by forever, but in botany, as elsewhere, all is seldom ideal. The botanist may study the rest of the family for many years before he gets around to naming the plant, and meanwhile the grower has to be content to label it, say, MJX 763945-P. (Has your MJX 763495-P flowered this year?) Sometimes a plant appears in cultivation with no record of its source and goes unnoticed by botanists, and so nameless, for years. Or it may be given several names, which cause confusion until someone finally studies the group and decides what name is proper. Or three people may study it and decide on three different names. With more knowledge of relationships, or perhaps with less, someone may move the plant to a different genus, which may or may not improve the classification but, alas, will cause another name change. The grower grumbles and wonders why the bloody botanists don't make up their bloody minds; while the offending botanist, if he thinks about it at all, excuses the name change by saying that, ah well, the next generation will learn the new name and never even know about the old one. By that time, of course, there may be an even newer name or two. As L. H. Bailey used to say, there is no finality in taxonomy.

Surely a classic case of botanical confusion is the history of Graptopetalum paraguayense, the so-called ghost plant. For 80 years it has been spreading through the gardens of the world, winning friends and admirers but meanwhile acquiring a dozen names under five genera. Its source was for many years a complete mystery and even now can only be called an incomplete mystery. It has been given a source some 4000 miles from what must be its true home, and it is doomed to bear a name that glaringly perpetuates this error. The best pub­lished clues to its checkered history are in a nearly forgotten article by Ellen Rooksby (1941b) in Desert Plant Life. We now give you, so far as we can trace it, the intricate story of Graptopetalum paraguayense.(*1)

At the turn of the century, Frank Weinberg, from Breslau, Germany, was a notable grower of cacti and other succulents at Woodside, Long Island, New York (Rooksby, 1941a). His impressive 76 page illustrated catalog of 1906 suggests that he may scarcely have been exaggerating when he called his the "most complete commercial collec­tion of its kind in the world". (In any case, the phrase "of its kind" gave him a way out.)

The plants of G. paraguayense now grown so widely and for so many years all trace back to Frank Weinberg at Woodside; but details are murky. Some 30 years afterwards, Weinberg wrote that the plant first appeared as volunteer seed­lings on his greenhouse bench with some cacti received in 1904 from J. A. McDowell, a cactus exporter in Mexico City. In a letter to Mrs. Rooksby (1941b), he told how "in 1905 Dr. Rose saw those plants at my greenhouse and took half a dozen along with him to Washington... and on a later visit told me that the flowers on the plant were somewhat different [from those] of Echeveria and allied species. Later on he wrote me... that he... named it Echeveria Weinbergii. This original let­ter of Dr. Rose I have turned over to Mr. William Hertrich of the Huntington Botanical Gardens for his files and future references."

Fig. 1. The earliest known photo of G. paraguayense subsp. paraguayense, probably taken Dec. 1909 in Washington, D.C.

In Rose's greenhouse notebook at the U.S. National Herbarium, we find that in October 1907 Rose logged in five cultivated plants from Weinberg. Number 07.575 was entered as "Echeveria", with "Weinbergii" later written after it in another hand. Still later, "expatriata Rose n. sp." was written above; then this and "Echeveria" were crossed out, and "Byrnesia" was substituted, leaving the final name as Byrnesia weinbergii. There is noth­ing about Weinberg's source for the plant. Of the other four plants received from Weinberg at the same time, two were entered as "Echeveria" and one as "Cereus donkelari", none with locality or other information. The fifth plant was listed as "Opuntia elata" and noted as "Native of 'Paraguay'", with "Paraguay" in quotation marks as if to sug­gest doubt. That is all. Nothing here suggests that the plants were collected or imported together. In fact, Cereus (Selenicereus) donkelarii is from Yucatan, and no echeverias are known from there or from Paraguay. These five cultivated plants, then, just happened to come from Weinberg to Rose at the same time.

So here the name of Paraguay turns up inno­cently on the same page of a notebook with the name of the ghost plant. Nothing ominous here, you might think. But wait! Watch how from this juxtaposition the plant comes to be permanently tagged as Paraguayan.

In the manuscript collection of the Huntington Library is a series of 40 letters from Rose to Weinberg. These letters shed some further light (though leaving us in the dark, and very curious, about what Weinberg wrote back to Rose): "About October, 1907, you sent me some Echeverias, one of which I numbered 07.575. This has just flowered, but it is entirely unlike anything I have seen here­tofore. I wonder if you can give me any information as to the origin of this plant. I enclose a leaf which may help you to identify it" (Jan. 8, 1908). "I wish to say that I would hesitate very much to describe your Echeveria without knowing more of its his­tory. I know very little about the south African Cotyledons. Your specimen may be one of those African plants and yet it looks to me more like an Echeveria. I will take another look at the plant but I do not believe we had better publish it as new" (Feb. 11, 1908). "In 1907 you sent me a curious Echeveria which I was not able to identify at that time. This plant is now coming into flower and I write to inquire if you have ever obtained a satis­factory name for it. Will you tell me from where your material came and if you know anything of its origin" (Dec. 15, 1909). "Thank you for your note upon the Echeveria. I will give this careful con­sideration and decide what is best to do about it" (Jan. 3, 1910). "No, I have never been able to iden­tify that curious Echeveria of yours" (July 27, 1911). "Please call that peculiar Echeveria of yours E. Weinbergii, although I may be obliged to make some other disposal of it in time" (Sept. 20, 1911).

At the bottom of the last letter Weinberg wrote in pencil: "5 years for a name!" But the name Rose gave him in 1911, Echeveria weinbergii, was never officially and botanically published: it was a ghost name for the ghost plant.

Weinberg's memories disagree slightly with what Rose wrote in his letters. Weinberg said he gave the plants to Rose at Woodside in 1905, but Rose wrote that Weinberg sent them in 1907 — as also his notebook shows. It is possible, of course, that Rose got the plant from Weinberg twice, but if so, wouldn't he have remembered the first time and commented in his letter of 1908? Or did Weinberg perhaps, after so many years, mis-remember the details? (Or does it matter at all now?) Apparently Weinberg did confuse some related details when he said (Rooksby, 1941b) that Lenophyllum pusillum and L. weinbergii "originated as seedlings at the same time with Echeveria weinbergii in my greenhouses at Woodside, Long Island". Rose (Britton and Rose, 1905: 28) told of receiving from Weinberg the type plants of L. pusillum, which came up with seedlings of Echeveria agavoides. But Britton and Rose (1904:162) gave the origin of L. weinbergii as "cracks of rocks on high mountains in the northeastern part of Mexico, State of Coahuila, from the collections of Mr. McDowell in the City of Mexico, transmitted to the New York Botanical Garden, December, 1903, by Mr. Frank Weinberg".

Fig. 2. Graptopetalum paraguayense subsp. bernalense on Cerro Bernal, with a Zapotec boy, Miguel Barron.

Figs. 3 and 4. Subsp. bernalense with tillandsias, Ficus and Sedum palmeri. Photos by A. Lau.

Whatever the details, it does seem entirely credible that the graptopetalum first appeared as volunteers in Weinberg's greenhouse. If the seeds did indeed come in with the cacti from McDowell, that doesn't help us; for we have no way of know­ing what the cacti were or where they originally came from. A minor mystery is how McDowell could have gotten the plant from the wilds to Weinberg without apparently noticing it himself. Perhaps he mainly noticed cacti: yet he did send the less attractive Lenophyllum weinbergii to the United States. But never mind—we didn't promise the full story.

Rose also said in those letters that his plants had flowered in Washington about December 1907 and again the next winter; yet he later wrote (1922) that the plant was an extremely shy bloomer, flowering only once in Washington (December 1909) and once in the New York Botanical Garden (February 1921) during the fifteen years he had it under observation." Fig. 1 is a never-published photograph from a collection of Rose's photos and manuscripts in the Huntington Library. On the back is written "07.575. Echeveria, cultivated, Fr. Weinberg". Though we don't know whether this photo was taken in 1907 or 1909, it is the earliest we have seen of this species.

Both Weinberg and Rose did their part in spread­ing the yet nameless waif around the world; and as it went, it gathered names. Weinberg wrote (Rooksby, 1941b): "In 1908 I shipped one hundred of these plants to a nursery in Erfurt, Germany, who cata­logued it as Echeveria Weinbergii... The plant grew and multiplied... rapidly and I presume, not having a ready sale, the name was changed to Echeveria arizonica. When I visited Kew Gardens in London in 1910, there were... a few plants in pots under that name from the Erfurt nursery." It may well be that this nursery (probably that of F. A. Haage, Jr.) was the first to publish the name Echeveria weinbergii unofficially, but we have not seen their catalogs from that time. The earliest German catalog in which we have found it listed is that of Haage and Schmidt (also of Erfurt) for 1915.

Rose sent the plant, or photos of it, to many per­sons in America and Europe, seeking information (Rose, 1922). About 1909 he sent it, apparently as E. weinbergii, to the Theodosia B. Shepherd nur­sery in Ventura, California. In answer to a later inquiry, Mrs. Shepherd wrote him July 6, 1919: "We got this from you about ten years ago... labeled [Echeveria weinbergii], and as soon as we could propagate it we began listing it in our cata­logue". In her catalog for 1912 under Echeveria we find: "Weinbergii (New). Very handsome of bluish gray color. 25c and 50c." The word "new", also used after some very old names, clearly meant newly offered in the trade and did not mean that Mrs. Shepherd was proposing to name a new species. The name is thus "mentioned incidentally" in the sense of the ICBN (Stafleu, 1983), and this men­tion is not a valid publication of E. weinbergii.

According to Sir David Prain, then Director of the Royal Botanic Garden at Kew (letter to Rose, September 29,1919), Rose sent the plant to Kew in 1912 labeled "Paraguay, F. Weinberg, n. 575". Flowering at Kew, it came under the careful scrutiny of N. E. Brown, a botanist noted mainly for his work on African plants. Brown (1914) described it under its first legitimate botanical name, Cotyledon paraguayensis, using Cotyledon in the broad sense of Bentham and Hooker (1865) and Schönland (1890) — a concept we now consider artificial. Brown's statement of the origin was: "Paraguay. Without precise locality, F. Weinberg. Described from a living plant which flowered in Kew in April 1914." Brown seemed to imply here that Weinberg had actually collected the plant in Paraguay.

Perhaps it was not until 1919, with the letter from Prain, that Rose learned of N. E. Brown's 1914 publication. Rose had just sent the plant to Kew for the second time, though now labeled "Byrnesia sp., no. 07.575". Prain told him this plant was "singularly like, if not identical with" the one he had very kindly sent in 1912 and was there­fore the plant described by N. E. Brown as Cotyledon paraguayensis. Prain might have noted — but may have hesitated to mention—that it even bore the same number.

Rose (1922) published a handsome color plate of the plant together with its second validly pub­lished botanical name, Byrnesia weinbergii. In so doing, he treated the name Echeveria weinbergii Hort. as validly published in the Shepherd catalog of 1912 and thus antedating Cotyledon paraguayensis N. E. Brown of 1914. And he then founded for it the genus Byrnesia, named for Edward M. Byrnes, long-time Superintendent of Gardens and Grounds for the U. S. Department of Agriculture in Washington. An expert grower of Crassulaceae (Britton and Rose, 1905: 20), Byrnes probably supervised the care of Dr. Rose's greenhouse plants. Rose briefly compared Byrnesia with his closely related genus Graptopetalum. He then knew that genus by only two species, both differing from G. paraguayense in their short stems, small dense rosettes and conspicuously spotted petals. As with the Cactaceae, so also with the Crassulaceae, Rose took a rather narrow view of genera.

A herbarium sheet at US bears a rosette and inflorescence of Rose 07.57 5 from the flowering of December 1909, together with the photo shown here as fig. 1. This specimen is labeled as a new species of Byrnesia, with the epithet "expatriata" and with Rose as author. (Since the combination is unpublished, we prefer not to use it here or in our synonymy.) The only other specimen of the species at US, labeled Byrnesia weinbergii, is also from Rose 07.575 but consists only of four leaves; it is undated but has the Addisonia reference. Apparently Rose at first planned to name the plant as a new species; and had he done so, presumably he would have named US 3049576 as the type. When, in­stead, he transferred the catalog name of Echeveria weinbergii to Byrnesia (Rose, 1922), he did not name a type for a plant that supposedly was already named. But since the name of E. weinbergii was illegitimate, Rose was really proposing a new name if, according to his own synonymy, he was thereby creating a new synonym for Cotyledon paraguayense N.E. Brown. Since Brown's name and Rose's name are based on plants of the same collection (Rose 07.575) (another plant of which is beauti­fully illustrated in Addisonia plate 243), there is no ambiguity. At the risk of being unnecessarily tidy, however, we designate US 3049576 as lec-totype for Bymesia weinbergii Rose.

We can only wonder how, or whether, Weinberg ever answered Rose's repeated questions about the origin of the plant. If Weinberg could remem­ber 30 years later that it volunteered with some Mexican cacti, why wouldn't he have said so to Rose? Yet after 15 years and 40 letters to Weinberg, Rose (1922) still could say only that it had come from Weinberg "with some Paraguay plants" and that Weinberg had said he knew nothing of its origin. Rose went on to quote a 1920 letter from Alwin Berger, who said he didn't believe it came from Paraguay but thought it looked like a Mexican plant. But meanwhile, Rose had sent it to Kew labeled "Paraguay" and so was indirectly re­sponsible for N. E. Brown's misleading name of Cotyledon paraguayensis.

In his generic revision of the Crassulaceae, Berger (1930: 446) made Graptopetalum a section of Sedum, with Byrnesia a synonym, and so treated our plant as Sedum weinbergii. Walther (1930) placed it in Graptopetalum, as G. weinbergii. He commented that the origin of the plant was un­certain but, according to Weinberg, almost certainly not Paraguay; he must have gotten this opinion directly from Weinberg. Von Poellnitz (1936: 263) listed Echeveria paraguayensis Hort, but as a synonym of Sedum weinbergii. Considering the 1912 catalog publication of Echeveria weinbergii invalid, Bullock (1937) took up N. E. Brown's 1914 name instead and made the combination Sedum paraguayense. And finally (we can only hope), Walther (1938) put the correct epithet with what we would call the right generic name to make Graptopetalum paraguayense, the name we use today.

Rooksby (1933) wrote that Britton and Rose gave the type locality of Echeveria weinbergii as Coahuila, Mexico — though in fact these authors never wrote jointly about this plant The explanation seems to be that in the synonymy of E. weinbergii she included Lenophyllum weinbergii Britton, a different plant, whose type locality they did in­deed give as Coahuila.

Although G. paraguayense seemed to be Mexican, it was 75 years before anyone again found any­thing like it in Mexico. In 1979 that tireless collector Alfred B. Lau found similar plants near Gonzalez, east of Ciudad Mante, Tamaulipas, Mexico. They grew at 700 - 800 meters near the top of Cerro Bernal, a solitary peak looming above a largely forested but partly cultivated plain. The peak is shown by Glass and Foster (1981: 79) as the type locality of their new Mammillaria anniana, found by Lau on the same trip. The graptopetalum grew in vertical rock fissures (see figs. 3 & 4) and on more or less level ground, with Sedum palmeri and Echeveria walpoleana; other associates included three different tillandsias, Dolichothele sphaerica, a ficus and two orchids (Schomburgkia and Oncidium).

Lau's collection differs from the familiar long-cultivated ghost plant in its smaller size and dif­ferent coloration. We place it in the same species but name it below as a new subspecies. The source of the original subspecies remains a mystery—but now perhaps a lesser mystery. We suppose it must grow on some other peak in the same general part of Mexico—certainly not in Paraguay. Only further exploring can show where it does grow and whether yet other forms grow on yet other peaks.

It would be hard to prove that a plant did not come from Paraguay or couldn't have. Certainly, no one has combed Paraguay for this one, nor would not finding it prove it was not there. Con­sidering how casually the name of Paraguay became linked with the ghost plant, however, it would seem a remarkable coincidence if the plant really had come from Paraguay. The subject appears quite dead; but let us count the nails in the coffin. First, no Crassulaceae are known from Paraguay. Second, the genus Graptopetalum is otherwise known only from Arizona and Mexico. Third, Weinberg eventually stated that the plant volun­teered with some Mexican cacti. And finally, as we have seen, a close relative has now turned up in Mexico.

So Graptopetalum paraguayense really comes from Mexico? Then how absurd to saddle it with such an unfitting name! From the viewpoint of the plant—if it may be allowed a viewpoint— paraguayense may be no more absurd than weinbergii or any other name, for what plant feels the need to be classified and named? But if the absurdity of paraguayense does not offend the plant, we can't help feeling that it is somewhat demeaning to the Science of Botany. Calling a Mexican plant paraguayense is Science? But then, if we think paraguayense absurd, why not change to a more appropriate name, like weinbergii? Well, under our system of rules, the International Code of Botanical Nomenclature, we can't do that. A main object of the Code is stability of names—and God knows we have little enough of that at best. The rules are to prevent just such arbitrary changes: for if we don't like paraguayense (and we don't), someone else may not like weinbergii and might want to change to some other name. And if every­one were free to choose his own name for a plant, the system would fail and communication about plants would break down. So we follow the rules and keep the offensive name of paraguayense in what seems a good cause, trying to stabilize names.

Fig. 6. Graptopetalum paraguayense subsp. paraguayense, HBG 494.

1. Rosette, 0.5.

2. Leaf, with cross-sections, X 0.5.

3. Apical portion of stem with inflorescence, X 0.25.

4. Flower, apical and lateral views, X 1.5.

5. Petals and stamens, X 3.

6. Gynoecium, X 3.

7. Nectaries, greatly enlarged.

We can try to be philosophical about it by saying that a name is just a handle and that any nonsense word would do. In fact, all plant names are non­sense words to many who use them—though they are more interesting and easier to remember when we know their meanings. So hooray! We have struck a blow for stability: this plant can remain Grap­topetalum paraguayense for ever and ever, and no one will ever have to learn a new name for it— unless, of course, some botanist decides to put it in another genus, or combine it with an earlier named species, or change its rank. Stability is wonderful!

If anyone has any questions, please hold up your hand. But we didn't promise the whole story, remember?


Graptopetalum paraguayense (N.E. Brown) Walther.

Caudex decumbent to pendent, branching only near base, to 3 dm or more long, 5-12 mm thick, at first glaucous-white, becoming yellowish green and later tan, with prominent leaf scars 4-10 mm apart. Rosettes lax, (3-)4 -12(-16) cm wide, of 15 -25(-30) spreading leaves over 4-8 cm of stem, only the youngest crowded. Leaves cuneate-obovate, short acuminate, 2.5 - 6(-8) cm long, 1.5 - 4 cm wide and 7-10 mm thick above, 5-10 mm wide and 5-7 mm thick at base, the basal attach­ment elliptic, 1 - 2.5 mm wide, with a single bundle scar, the upper surface flattish, often slightly channelled in basal third, the lower surface strongly convex, eccentrically low-keeled, the margins obtuse or in lower third acute, the surface smooth, at first rather dull with a slight waxy bloom, grayish violet to grayish white or yellowish.

Floral stems February to May, arising ca. 1 - 2 cm below stem apex, usually 1 or 2 and if 2 usually on opposite sides of stem, the peduncle 5 -10 cm long, 2-5 mm thick, pinkish, constricted at base and readily detaching, the leaves 1-5 (-12), ascending, the lower like rosette leaves but smaller, 1-2 cm long, 3-9 mm wide, 3-4 mm thick, the upper ones and the bracts relatively narrower, branches 2 - 4 (-6), often constricted at base, bifurcate or mostly simple, not circulate, 3 - 12cm long, 5 - 15-flowered; pedicels 5-12 mm apart, ascending, 4-16 mm long, 1-2 mm thick, often with 1 or 2 bracteoles 1-3 mm long. Calyx 3-5 mm long, 4 - 8 mm wide, the segments erect against corolla, nearly equal, triangular-ovate, acute, 1.5 -3 mm wide. Corolla white or with few red dots, yellowish or greenish on keels outside, 11-19 mm wide, 10-13 mm long closed, the tube 3-5 mm long, 5-7 mm wide above, the segments spreading, triangular-ovate, acute, channelled, 6-8 mm long, 2.5 - 4.5 mm wide. Stamens 10, at first erect, later recurving against or between petals, the fila­ments white, 7 - 9 mm long from corolla base, ca. 1 mm wide at base, adnate ca. 3 - 4 mm to corolla, the antesepalous thicker radially, adnate ca. 1 mm to intercarpellary tissue; anthers light pink to red, ca. 1-1.2 mm long and 0.5 - 0.9 mm wide. Nectar glands yellow, erect, quadrate, 0.6 - 1.0 mm high and wide, 0.3 - 0.5 mm thick. Gynoecium yellowish green or somewhat reddish, 6 - 9 mm high, 3-5 mm thick above, 1.5 - 2.5 mm thick at base, the pistils connate ca. 1 mm, erect, at first appressed, later separating above middle, somewhat keeled, the styles 0.5-1 mm long; ovules ca. 85, clavate or cylindric, 0.6 - 0.7 mm long, 0.2 mm thick Follicles ascending with tips ca. 6 mm apart, light tan, acutely keeled, 5-nerved. Seeds reddish brown, cylindric, ca. 0.7 mm long and 0.2 mm thick, with ca. 5 irregular longitudinal ridges. Chromosomes: n= 68.

The nearest ally of G. paraguayense appears to be G. amethystinum Rose, of the Sierra Madre Occidental, in western Mexico. That is a plant of similar size and habit, growing pendent from cliffs, and it also has turgid lavender leaves. A sugges­tion of relationship comes from the chromosome number (Uhl, 1970): these two are the only members of the genus with numbers based on 34; all others have 30 - 33 or multiples. However, G. amethystinum is amply distinct, the leaves fewer and more turgid, with broadly rounded margins and with the apex obtuse to rounded, the inflorescence more com­pact, with generally more and shorter branches, the corolla pale yellow, closely crossbanded with dark red. These species belong to Graptopetalum

sect. Byrnesia (Rose) Moran, in which the caudex is elongate and the leaf bases are generally of the Echeveria type, with a single vein and a reduced attachment

Key to subspecies

Leaves 4 - 7(-8) cm long and 2 - 3(-4) cm wide above, grayish violet..........................ssp. paraguayense

Leaves 2.5 -3.5 cm long, ca. 1.5 cm wide above, yellowish or whitish green...............ssp. bernalense

Fig. 7. Graptopetalum paraguayense subsp. bernalense, HBG 42730.

1. Rosette, X 0.5.

2. Leaf, with cross-sections, X 0.5.

3. Apical portion of stem with inflorescence, X 0.25.

4. Flower, apical view, X 1.5.

5. Petals and stamens, X 3.

6. Gynoecium, X 3.5.

7. Nectaries, greatly enlarged.

Drawings by Diana Jacobs, 1983.

G. paraguayense subsp. bernalense Kimn. et Moran, subsp. nov.

Quam subsp. typica omnibus partibus minor, rosulis 4 -6 cm latis, foliis 2.5-3.5 cm longis, 10-16 mm latis, griseo-glaucis, corolla 11-15 mm lata, segmentis 2.5-3 mm latis, perpaucimaculatis immaculatisve. Typus: A. Lau 089.

Caudex 5 - 6 mm thick Rosettes 2.5 - 6 cm wide, the leaves 2.5 - 3.5 cm long, 10 -16 mm wide above, yellowish or greenish cream. Corolla 11-15 mm wide, the tube 3 mm long, the segments 2.5-3 mm wide, white or with 1 or 2 tiny dots near margin.

Type collection: Common on north-facing porphyritic rocks of Cerro Bernal, 700 - 800 m, 39 km E of Ciudad Mante, Tamaulipas, Mexico (near 22° 46' N, 98° 35' W), Alfred B. Lau 089, Huntington Botanical Gardens 42730; flowering at HBG and in San Diego (HNT, holotype; BH, MEXU, SD, US, isotypes).

Distribution: Known only from the type collection.

From a plant of the type collection, Charles H. Uhl of Cornell University made a chromosome count of n = 68, as in the typical subspecies.

G. paraguayense (N.E. Brown) Walther subsp. paraguayense

Echeveria weinbergii Hort., catalog of T. B. Shepherd, p. 37. 1912. Nom. nud.

Cotyledon paraguayensis N. E. Br., Kew Bull. 1914: 208. 1914.

Byrnesia weinbergii Rose, Addisonia 7: 37, pl. 243.1922.

Echeveria arizonica Hort. ex Rose, I.c.. Pro syn.

Sedum weinbergii Berger in Engler, Nat. Pflanzenf., ed. 2, 18a: 446. 1930.

Graptopetalum weinbergii Walther, Journ. Cact. Succ. Soc. Amer. 1: 184. 1930.

Graptopetalon byrnesia Walther, sphalm. Journ. Cact. Succ. Soc. Amer. 2: 414. 1931.

Echeveria byrnesia Hort. ex Rooksby, Des. PI. Life 5: 5. 1933. Pro syn.

Echeveria paraguayensis Hort, ex von Poelln., Repert. Sp. Nov. 39: 263. 1936. Pro syn.

Sedum paraguayense Bullock, Kew Bull. 1937: 276. 1937.

Graptopetalum paraguayense Walther, Cact Succ. Journ. (USA) 9: 108. 1938.

Caudex 8-12 mm thick. Rosettes 7 - 12 (-16) cm wide, the leaves 4 - 7 (-8) cm long, 1.5 - 3 (-4) cm wide above, usually lavender-tinted. Corolla 13 -19 mm wide, the tube 4 - 5 mm long, the segments 3 - 4.5 mm wide, white with scattered red dots.

Type collection: Cultivated plant received from Frank Weinberg, J. N. Rose 07.575 (K).

Distribution: Known only in cultivation, but doubtless from northeastern Mexico.

Uhl (1970) reported chromosome counts of n = 68 for this subspecies, based on four cultivated collections presumably all tracing back to the type collection. Getting only six seedlings from three attempted self-pollinations, he wrote that self-fertility seemed low, and he thought it likely that most or all plants then grown were of one original clone. (It seems possible, however, that Weinberg grew more than one seedling or that his plant later produced more seed and hence that there is more than one clone existing today). However, Uhl got more than 850 seedlings from 27 crosses with 23 other species. He found meiosis normal and pollen 97% normal in ssp. paraguayense and fertility good in some crosses.

Because of its large, attractively colored leaves and easy cultivation, subsp. paraguayense has been widely used in hybridization. It is a parent of numerous bigeneric hybrids, among them X Graptoveria (with Echeveria), X Grapsonella (with Thompsonella), X Graptosedum (with Sedum), X Graptophytum (with Pachyphytum), X Lengraptophyllum (with Lenophyllum) and X Cremnopetalum(*2)(with Cremnophila).

Although Rooksby did not refer to them in her article of 1933, two popular names, "Ghost Plant" and "Ghost Flower", were widespread by the time she discussed the plant in 1941. The earliest printing we have seen of "Ghost Plant" is in the 1936 catalog of the Johnson Cactus Gardens, where the leaves are described as white (green­house cultivation often diminishes the lavender tint). Even though the color of ghosts seems to be a matter of conjecture, Rooksby remarked: "The pale amethyst color gave good reason for this com­mon name". As the petals are also white and nearly unmarked, they could have been the inspiration for the name "Ghost Flower", which later may have been corrupted to "Ghost Plant"—or per­haps it happened the other way around. No one seems to know the origin of either name. Harry Johnson, who coined many popular names for suc­culents, may well have thought up "Ghost Plant", though he has recently told us he doesn't remember.

Many have noted how easily the ghost plant grows. It survives under varied conditions, lux­uriating with ample watering, marking time when dry, soon recovering from frosts. Britton (1931) said it was the best of the Crassulaceae for West Indian gardens, where it thrived on St. Thomas, Cuba and Puerto Rico, with 50 inches of rain or more a year. He believed this to be evidence that it was native to "a hot and wet region".

Vegetative reproduction by stem or leaf is extra­ordinarily easy. Every year or two, when the stems become too leggy, the terminal rosettes can be broken off and laid on the ground, where they soon take root. Fallen or detached leaves do likewise and furnish a most striking example of propa­gation of succulents by leaf-cuttings, as was charm­ingly described by Mrs. Rooksby (1933); "From the garden pathway one day I picked up a number of leaves and put them in my smock pocket think­ing to plant them a little later. But the matter escaped my mind and I hung up the smock in the studio on the lower floor. Three months afterward I noticed the smock and preparatory to sending it to the laundry emptied the pockets and found among labels, nails, seed-packages, matchboxes, handkerchiefs, a dime and several pennies, my little leaves with plantlets an inch or more in length, attached! Surely anything that will work for you cheerfully under such inhospitable conditions is deserving of this belated tribute."


Bentham, G., and J. D. Hooker, 1865. Genera plantarum 1, part 2. London.

Berger, A. 1930. In Engler, Nat. Pflanzenf., ed. 2, 18a: 352-483.

Britton, N. L. 1931. Dr. Britton's Notes on Byrnesia weinbergii. Journ. Cact. Succ. Soc. Amer. 2: 414.

Britton, N. L. and J. N. Rose. 1904. Lenophyllum, a New Genus of Crassulaceae. Smithsonian Misc. Colls. 47: 159-162, figs. 18, 19, pl.xx.

_______________. 1905. Crassulaceae. In North American Flora 22: I: 7-74.

Brown, N. E. 1914. Decades Kewenses. Kew Bulletin 1914: 208-209.

Bullock, A. A. 1937. Section paraguayense. Kew Bulletin 1937: 276.

Glass, C., and R. Foster. 1981. Mammillaria anniana, a New Species from Tamaulipas, Mexico. Cact. Succ. Journ. (USA) 53: 79-80.

Poellnitz, K. von. 1936. Zur Kenntnis der Gattung Echeveria DC. Fedde, Repert. sp. nov. 39: 193-270.

Rooksby, E. 1933. Common Succulents. "Echeveria Weinbergii". Desert 5:5. [Article signed "R", the magazine edited by E. Rooksby.]

_______________. 1941a. Bits from a Collector's Life [about Frank Weinberg). Desert Plant Life 13: 173-175. [Article unsigned but just before Rooksby 1941b in a magazine edited by E. Rooksby.]

_______________. 1941b. Graptopetalum paraguayense, "The Ghost Flower" from Mexico. Desert Plant Life 13: 175-176.

Rose, J. N. 40 typed letters to Frank Weinberg, 1903-1912. Located at: Manuscript Department, Hun-tington Library, San Marino, California.

_______________. 1922. Byrnesia Weinbergii. Addisonia 7: 37-38, pl. 243.

Schönland, S. 1890. Crassulaceae. In A.Engler and K. Prantl, Nat. Pflanzenfam., ed. 1. III, 2a: 23-28.

Shepherd, T. 1912. Descriptive catalogue, 1912: Rare plants, choice flower seeds, cacti and succulents. Theodosia B. Shepherd Co., Ventura, Calif.

Stafleu, F. A., et. al. 1983. International Code of Botani­cal Nomenclature. Utrecht.

Uhl, C. H. 1970. Chromosomes of Graptopetalum and Thompsonella (Crassulaceae). Amer. Journ. Bot 57(9): 1115-1121.

Walther, E. 1930. A New Species of Graptopetalum. Journ. Cact. Succ. Soc. Amer. 1: 183-186.

_______________. 1938. Notes on Crassulaceae. Cact. Succ. Journ. (USA) 9: 108.


*1) We are grateful to Deborah A Bell at the Smithsonian Institution for finding Rose's specimens and the crucial page in his notebook, to William Dress for checking old nursery catalogs in the Bailey Hortorium, and to Hal Johnson for tracking down early appearances of the name "Ghost Plant".

*2) This name is herewith established for hybrids be­tween Cremnophila and Graptopetalum.

© Cactus & Succulent Journal of America, 1986