This paper addresses the hypothesis that land plants have a biphyletic
origin as the product of an endocellular mutualism between a green al
ga and a tip-growing, fungus-like organism, culminating in the acquisi
tion of part of the latter's genome by the host alga (Atsatt, P.R., 19
88, Are vascular plants 'inside-out' lichens? Ecology 69, 17-23). Acco
rding to this hypothesis, the tip-growing symbiont's capacity for inva
sive growth was exploited during the further evolution of the holobion
t for the development of various specialized plant cell types, but esp
ecially those displaying tip growth. Here, noting the recent discovery
of the dependence of pollen tube tip growth on flavonoids, this hypot
hesis is refined and extended by suggesting that a symbiotic relations
hip was advanced by the evolution of UV-protective flavonoids in the a
lga, followed by the evolution of a growth response by the tip-growing
symbiont to the presence of those flavonoids, allowing the symbiont t
o continue to live with the alga in its new, high-light habitat. This
growth response then evolved into a dependence on flavonoids in the co
ntext of an obligate, mutualistic relationship progressing toward endo
symbiosis and incorporation of the endocytobiont's genetic capacity fo
r cell polarization, tip growth and their control into the host alga's
genome. Land plants and advanced charophycean algae (which are the cl
osest green-algal relatives of land plants) are likely products of thi
s process, while a primitive charophycean alga (lacking both tip growt
h and cell polarization) is proposed to have been the likely host for
the endocytobiont. A series of tests of this hypothesis, based mainly
on the identification and molecular phylogenetic analysis of appropria
te genes, are proposed. Whether the endocytobiont could have been a re
lative of the earliest endomycorrhizal fungi is assessed.