That's a common fallacy in many text books and taught in schools.
Many of the most primitive flowering plant that are not monocotyledons have
floral parts in whorls of three.
Also, the number of locules (carpels) in the fruit of many genera of dicots
do not correspond to the number of floral parts (or lobes in sympetalous
genera) in a whorl.
"HesperAloe" <hesperaloe at aol.com> wrote in message
news:20010215001708.20895.00001241 at ng-fn1.aol.com...
> I just bought a loofah sponge (Cucurbitaceae family) and noticed it had 3
> compartments (carpels). I thought dicots always had floral parts in
> of 4's and 5's...Does this also apply to the carpel visible when the fruit
> sliced transversely?