Introduction
For several years I have been intrigued by the appearance and the behaviour, from late December to late February, of the flowers of my Graptopetalum paraguayense (Ghost Plant). Not only does the season of its bloom seem odd, as does the fact that it blooms in spite of my cold, dry, seasonal treatment, but the manner and sequence of opening of each bud seems special, somehow. The apparent preference for blooming in the cooler seasons I have heard explained as a characteristic of some of the Crassulaceae. For example; in our Dalhousie University greenhouse just now (February '02), with the temperature ranging between 8^oC and 18^oC, our larger Crassula argentea  plants are loaded with buds. Echeveria gibbiflora too is (reluctantly) pushing out its large inflorescernce and Pachyphytum bracteosum is doing the same. The photos of Ghost Plants in this article are dated March and January. 

Some years ago I was able to pause long enough to watch the process more closely, and I was rewarded with one of Nature's delightful small dramas. Gordon Rowley's splendid "Illustrated Encyclopedia of Succulents" (New York, 1978) and Claude Chidamian's excellent "Book of Cacti and Other Succulents" (New York 1958) were very helpful in making the whole exercise educational as well as pleasant. Several corrections, clarifications and additions in this 2002 revision are due to a very profitable checking of my text throughout with Jules Janick's "Horticultural Science", Fourth Edition, (New York, 1986). What follows is my interpretation, as an amateur, of what I saw, and I offer all this in the hope that those who detect my errors will point them out for our general benefit.

Description

Monocasial Cyme Inset is after G. Rowley (1)

The inflorescence of G. paraguayense first becomes visible in the axil (upper angle between leaf & stem) of about the tenth or eleventh rosette leaf below the youngest as a pointed, oval bud surrounded by four successive, small, narrow, succulent bracts (modified axis leaves). The bracts are later seen to be arranged spirally around the peduncle (stalk or axis) at intervals of about 120 degrees. As the peduncle elongates between bracts, and the bud pushes out further on its own pedicel (flower stalk), a new branch is seen to form in the axil of the topmost of the four bracts which (bract) then persists for a time while the other three gradually shrivel and eventually drop off. Three more bracts are produced on the peduncle and the axil of each produces a new branch. The same sequence is repeated behind each new bud as it appears, "...the oldest flower terminates the axis and the new flowers arise from branches beneath it." (Rowley, p.32). Such an inflorescence is called a "cyme" and, in this case since the bracts are arranged spirally, no two of them are at the same stage of development on any one branch, only one axil on each branch becomes active at one time, and branching is therefore "one-sided". Such an inflorescence is called a "monochasial cyme."

As the bud swells, the sepals' rate of growth gradually slows down, evidently stopping entirely when about 6mm long, about half the length of the bud. At this stage, the petals begin to separate, and the maroon colour of the anthers begins to show through. Soon an anther emerges between each pair of petals and, within a few hours, the maroon colour begins to turn darker, the anther begins to shrink, splits along two sides and exposes the off-white pollen. This takes about 20 minutes, and all five anthers "ripen" like this within a few hours. Meanwhile, one can see another five anthers inside the bud.

Depending upon the heat, light and moisture available to the plant, the flower gradually opens over a period of eight hours or so. By the time it has opened about half way, one can count some of the major parts. Although not all the details listed here can be seen without taking the flower apart, the parts are, counting from the centre: 

• The gynoecium, consisting of five, rust-coloured, "free" carpels in the first of five whorls, each carpel carrying, at its tip, a stigma (pl.= stigmata). The fused bases of the carpels form the ovary. 
• Five very small, golden, stubby, leaf-like nectaries, one at the base of each of the 5 Carpels, forming the second whorl; 
• Ten stamens, Five of which arise from the pedicel (flower base and stem) and are fused for 3mm to the inner surface of the seams between pairs of petals; and five others, each arising out of the centre of the base of each petal, and fused with it for 3mm; in the third whorl; 
• Five petals (collectively the corolla ="small crown") in the fourth whorl; and 
• Five sepals (collectively the calyx = Gk."husk") in the fifth.

The Ballet
The most curious aspect - the one which strikes me as most fascinating, at any rate, is the concerted behaviour of the two quintets of stamens as the flower opens - a "Ballet of the Stamens"so to speak. Just as the bud begins to open, the first five to appear curve gracefully out between the petals. Meanwhile, the other five are pressing their anthers (pollen glands atop the filaments) outward, against the inner faces of the petals. As the petals continue to open out, the recurved stamens slowly straighten out, eventually to rotate each of the now whitish pollen-bearing anthers to within about 1mm of, and between, each pair of stigmata at the tips of the carpels. There, some pollen is released. Although evidently no stigma has "ripened" yet one sees, eventually as some pollen drops away, that the anthers have shrunk and turned black, and the filaments (stamen stalks) are still a glistening white.

When the petals have opened out fully, forming a short tube about 5mm long and 6mm in diameter, and the flared, 5-pointed corolla (little crown of petals) about 20cm in diameter, the five anthers nestled against the petals each rise a little off the petals, split and expose their pollen. Then their filaments gradually straighten up and swing these anthers inward to join the ring of stigmata at the tips of the carpels.

Then begins the Ballet Finale. Within an hour or so the first five stamens swing back to curve gracefully out again between and beyond the petals, and the second set follows suit, a few hours later, to rest again against the faces of the petals. The tips of the carpels also swing out a bit to form an inner ring. Later, one notices that the tips of the carpels stretch out a very little bit, and a white, sticky blob forms at the tip of each. These are the ripening stigmata, and the flower becomes ready for pollination. Note that all of the flowers' own anthers are held as far away from the stigmata as possible. This improves the chances that preferably windblown or insect-borne pollen from a flower of another clone, descendent of a different seed, will reach the stigmata first. Failing that, however, since Graptopetalum is a self-pollinating genus, the pollen of its own flower, or that of another plant (ramet) of the same clone will do just as well and can produce seed.

The whole dance takes from 24 to 30 hours, from the opening of the bud, and the flower retains this disposition until it dries, without much change in appearance, although the stamens move about a little. It strikes me that a 'movies' enthusiast would find all this an intriguing subject for time-lapse photography.

Flowers of most other genera play out their splendid little performances more privately or more slowly, or unnoticed under the dazzling spectacle of their corollas. Ghost Plant literally broadcasts its performance by moving accented parts noticeably against the bright but unobtrusive backdrop of its corolla. It comes under one's notice subtly - as much by graceful, subtle change as by the beauty of its form and disposition. Each flower lasts a long time and the whole inflorescence repeats this excellent performance time and time again for weeks. It is a joy to behold indeed!

Other Observations
Other interesting things were observed during this refreshing little exercise. For example; G. paraguayense is a "winter grower" and it responds very quickly to changes in growing conditions. While kept cold (4^oC to 10^oC), dry, in good light in the greenhouse, it grows very slowly, very compactly, with its leaves very closely imbricated (overlapped) to conserve moisture, beautifully suffused a delicate pink, glaucous (grayish, powdery blue) and very shiny. Within 36 hours of bringing it into the heat, higher moisture and relative dark of the house, one can notice a certain 'loosening' of the rosette, and soon the central, new rosette leaves push out. Within another day or so, one notices a definite internodal growth, the rosette becomes pyramidal and somewhat etiolated (pale, feeble). This can be counteracted by replacing the plants gradually into the strongest sunlight available but, in a severe winter with little sunshine for long periods, even a southern exposure was not quite adequate and I had to return the plants to the stronger light and the cold of the greenhouse. This keeps them compact but still growing and blooming - more slowly - and it becomes more difficult to water them correctly, so I tend to water them much less. Consequently the older leaves shrivel, drop a little earlier and the rosettes are a little smaller. Nevertheless after having closely observed them, for curiosity's sake, I find them even more delightful, now, probably because I know them that much better. At any rate, I find them always very beautiful! 

(An earlier version of this article was published in 1980 in the Amateurs' Digest (http://www.theamateursdigest.com) . 

Paul J. Brunelle Feb 23, 2002