726 
Telopea Vol. 6(4): 1996 
and the lack of a subtending bract beneath the first flower to open reveals its terminal 
position. Since the subtending bracts of the remaining flowers are displaced, the 
situation becomes complicated. Nevertheless, such inflorescences must not be called 
racemes, which would be misleading, when comparisons with other taxa are attempted. 
In the inflorescences of Lasiopetaluiii (Classen 1988) and Keraudrenia, both monochasial 
and dichasial ramifications exist, whereas in Thamasia and other Lasiopetaleae, 
monochasia (cincinni) are found. The basic, fully ramified, dichasial form is found in 
Keraudrenia, which often is described as a terminal cyme (see, e.g., Hutchinson 1967, 
Jessop 1986). If structures such as those exemplified by K. hermanniifolia (Fig. 1) are 
meant, this is not correct, because in true cymes alt ramifications arise from prophyllar 
axils. Since the terminal flower of a terminal inflorescence never has prophylls, the 
first order lateral flowers cannot arise from prophyllar axils. In this case, in which 
the partial inflorescences are cymose, the whole structure can be termed a cymoid 
(Troll 1964, Briggs & Johnson 1979). Yet it differs from a 'normal' cymoid in the 
constant occurrence of an additional sterile bract which is incepted below the 
branching off of the lateral dichasia but ontogenetically is displaced distally beyond 
them. Sterile bracts in inflorescences are usually called 'Zwischenbliitter' (Schumann 
1890, Nordhagen 1937) or 'metaxyphylls' (Briggs & Johnson 1979); they are known 
to occur in various determinate inflorescences (see Troll 1964) and are defined as 
phyllomes 'situated between the ultimate pherophyll(s) (or the prophylls) and the 
flower' (Briggs & Johnson 1979; 244). Flowever, the sterile bract in the inflorescence 
of Keraudrenia is different because it is situated between fertile phyllomes. Although 
it is a constant structure in the inflorescences of very many representatives of the 
Malvales (Bayer 1994), in this context and until we know more about its origin we 
refrain from creating a new term for it. 
If we compare the cymoid of Keraudrenia with the monochasial inflorescence widespread 
in the Lasiopetaleae (exemplified in Fig. 3 B), again recaulescent shifts of the bracts are 
noticeable. It would, however, be grossly misleading to equate the fertile bracts on the 
axes bearing the main flowers of the inflorescences in Fig.s 1 B and 3 B. Instead, the 
single 'supernumerary' sterile bract (a in Fig. 1) provides an appropriate fixed point 
for comparison. This allows recognition that the triad of bracts that form the epicalyx 
below flowers 1^ in Fig. 3 B correspond to the bracts a, b, c in Fig. 1. Thus, the 
epicalices of the flowers within monochasial inflorescences of Lasiopetaleae is 
hypothesized as having originated through the reduction of distal ramification 
(Fig. 4). The three appendages forming the epicalyx are individual, sterile bracts. 
Indeed, distal ramifications including a supernumerary, sterile bract are typical of 
the inflorescences of the Sterculiaceae and Tiliaceae (Bayer 1994). In these families 
an increasing reduction of these ramifications is observed, which leads to flowers 
surrounded by a sterile epicalyx (Fig. 4 C). In Sterculiaceae (e.g. Dombeyeae, 
Fremontodendreae) and Tiliaceae, the occurrence of an epicalyx is considered as an 
advanced condition, while in the Malvaceae and Bombacaceae it appears to represent 
the basic character state (Bayer 1994). 
Our interpretation of the epicalyx contrasts with the view of Classen (1988), according 
to which the epicalyx consists of a single leaf organ with its stipules. She preferred the 
latter interpretation in spite of having noted that the foliage leaves of Lasiopetaluiii are 
devoid of stipules and that the lateral appendages do not show the precocious 
development typical of stipules. If occasionally only a single bract is found beneath the 
flower (in some species of Lasiopetahwi, Classen 1988), to our mind this corresponds to 
one of the three bracts of the Keraudrenia inflorescence. Therefore Classen's view (1988) 
that in Lasiopetalum the trimerous epicalyx is derived from a simple bract is rejected. 
