For those of you with a little background in computer science or mathematics or programming, raceme and cyme can be defined recursively as follows (the rest of you don't worry if you didn't know what that meant):
[R] F | | | | | | [F]----|----[F] [C]----|----[C] | | | | | | | Raceme Cyme Here F = flower, R = raceme, C = cyme. If the letter is in brackets, it means 0 or 1 of that unit.
In English, here is what those diagrams mean:
A raceme is a stem from which one or more pedicels develop at a node, and which continues to another node, from which one or more pedicels develop, and which continues to another node, from which one or more pedicels develop, and which continues...
On a raceme the lowest pedicels and their associated flowers develop first, followed by the next lowest, and so on. A raceme tends to peter out in one or more buds which don't fully develop.
A cyme is a stem terminating in a flowerbud, but which might have, below the flowerbud, a node from which one or two cymes develop. This definition sounds circular (and that is what is meant by the word "recursively" above), but really isn't. After some number of branchings, the stem terminates in a single flowerbud, and there the circularity ("recursion") stops.
On a cyme the terminal flowerbud develops first, followed by the terminal flowerbuds of the cyme's node's cymes, followed by the terminal flowerbuds of the cyme's node's cyme's node, ad not quite infinitum. Trust me, look at the branching of a multiflowered stalk on an African violet or streptocarpus, and all will become clear.
The basic gesneriad inflorescence is a
cyme, but not just an
Rather, it is a
+-----F | | +-----[PF] F | | | | | | [PFC]----|----[PFC] [C]----|----[C] | | | | | | | Pair-flowered Cyme Normal Cyme Again, F = flower, C = cyme, and something in brackets means 0 or 1 of the unit. PF = pair flower. PFC = pair-flowered cyme. In the pair-flowered cyme, the top flower and the pair flower usually are in a plane perpendicular to that of the secondary cymes -- that is, they would be coming toward you out of the screen. A pair-flowered cyme can consist of ** a single flower (both [PFC] and [PF] absent), ** two flowers ([PF] present but both [PFC] absent), ** more than two flowers ([PF] and one or more [PFC] present). What usually does not happen is for one or more [PFC] to be present without the pair-flower [PF] being present.
The best way to get the idea of a cyme (and its pair-flowered variant) is to closely examine a saintpaulia flowerstalk or a multi-flowered streptocarpus flowerstalk. Because the flowerstalks are larger than in other gesneriads, it is easier to make out the structure.
It isn't as obvious, but the pair-flowered cyme is the basic inflorescence in Sinningia too.
Somewhat ironically, one of the best pictures I have of a pair-flowered sinningia cyme is on a hybrid. This picture of Sinningia reitzii x cardinalis shows the end flower opening first, then the pair flower directly beneath it, and lastly, one on each side, the two still-developing one-flowered "cymes".
Sinningia eumorpha and Sinningia sp. "Black Hill" normally have one flower per axil. How can one even speak of an inflorescence in that case, with just one flower?
As shown on the Black Hill page, a plant of that species sometimes has more than one flower per axil, and when it does have a second flower, it is in the pair-flower position.
There is an interesting article by Anton Weber in the one and only issue of Gesneriana on the development of pair-flowered cymes. It has scanning-electron microscopy pictures to show the sequence in which the various flowers and bracts are initiated and their position on the inflorescence bud. Two of the species used in his investigation are sinningias: S. bulbosa and S. lineata.
The pair-flowered cyme is the base inflorescence type in the Gesneriaceae. It is found throughout every branch of the family.
However, it is not restricted to the Gesneriaceae. This picture shows an inflorescence on Penstemon tenuis, which is native to the south central United States. The central flower has opened first, and its pair flower is just about to open. There are two secondary cymes, each with one terminal flowerbud plus a pair flowerbud. Penstemon used to be classified in the Scrophulariaceae, but has been moved to the related family Plantaginaceae, still not far from the Gesneriaceae. Other genera in the Plantaginaceae (such as Tetranema) also have pair-flowered cymes. [References: Anton Weber's treatment of the family Gesneriaceae in K. Kubitzki and J. Kadereit (eds.), The families and genera of vascular plants, vol. 7, 63–158, Springer-Verlag, Berlin, Germany, 2004 (which I have not seen), and personal communications, A. Weber and A. Chautems.]
This is a drastic oversimplification, but there are five general ways in which the flowers can be organized in a sinningia. The first three types below characterize indeterminate growth (no predetermined number of nodes per stem), while the last two characterize determinate growth (predetermined number of nodes per stem).
Please note that these are descriptive categories, not developmental ones. I'm not a professional botanist and have no assurance that all sinningias that form (for instance) "platform" inflorescences do so in the same way. Furthermore, inflorescence type does not correspond well to the taxonomy of the genus Sinningia (that is, species with the same inflorescence type are not grouped together in the family tree), so that inflorescence type seems quite mutable during the process of evolution.
Strictly speaking, all sinningias have axillary cymes. That is, the flowers are borne on inflorescences which are cymes and originate in the axils of nodes. There are no racemes in Sinningia and no inflorescences which do not originate in axils. In the present context, however, we use the phrase axillary cyme to refer to inflorescences originating from the axils of fully-developed leaves on stems with indeterminate growth.