P van Rijckevorsel writes
A wheel has come off here?
Aren't you the one who is supposed to answer such questions, rather than
pose them?

Stewart Robert Hinsley schreef:
If there's a question on phylogeny I can refer to Judd et al, or MOBOT,
or grab sequences off EMBL and analyse them myself, but when it comes to
more traditional bits of botany such as floral anatomy I haven't yet got
myself properly trained.

+ + +
You might take it to taxacom? Don't know if this will help, but would allow
Peter Stevens to pitch in if he feels like it.
+ + +

Not sure this question has a real answer: the "Basal tricolpates" are
something of a mess in this respect with Menispermaceae having a 3-merous
perianth, as well as Berberidaceae, etc.

That's why I was asking. I didn't find it obvious whether a 4- or
5-merous, biseriate, differentiated, perianth was basal to the tricolpates,
with derived flower types within many basal taxa, or whether it is a
homoplasy shared by several clades therein. Someone might have done a study
on this, perhaps supported by developmental evidence.

+ + +
The APG system is very grand about making sweeping statements, but a lot of
detail still needs to be filled in. Someone might indeed have done a study
.... or not.
+ + +


I will play it safe and quote from Judd & al (Plant Systematics, a
phylogenetic approach, 2nd ed, 2002):

I tried the 1st edn for an answer. Large chunks of Ranunculales have a
5-merous, biseriate, differentiated perianth, but the rest of
Ranunculales seems to be 3-merous. Platanaceae are 3- to 7-merous, and
Proteaceae are uniseriate.

Moving into the core eudicots, Vitales does has a 4- or 5-merous,
biseriate, differentiated perianth, but Caryphyllales are uniseriate,
except for Cactaceae, which is multiseriate. Within Polygonales
Droseraceae is 5-merous, but Polygonaceae is 6-merous. Saxifragales,
Rosids and Asterids are mostly 4- or 5-merous, but Santales don't
regularly fit this pattern.

+ + +
Isn't Santalales (see also http://www.science.siu.edu/parasitic-plants/) one
of those orders that is notorious for not fitting? With up to 6 cotyledons,
IIRC?
+ + +

The Angiosperm Phylogeny Site at MOBOT has Saxifragales+Vitales+Rosids as
a weakly supported clade. Which sort of suggests it's a homoplasy of
Ranuculaceae (in part), Droseraceae, Saxifragales+Vitales+Rosids, and
Asterids.


"Core Eudicots:
[...] These plants have usually 4-5 merous flowers with the perianth
differentiated into a calyx and a corolla; placentation is ususally axile"
p307-308

The book's got bigger since the 1st edn. Core eudicots start at page
238, and the only synapomorphies mentioned are cpDNA sequences.
--
Stewart Robert Hinsley

+ + +
Yes, I have yet to track exactly what the extra hundred pages are used for,
but it is distinctly bigger. Times change ...
PvR