History of classification

The botanical term "angiosperm", from Greek words angeíon (ἀγγεῖον 'bottle, vessel') and spérma (σπέρμα 'seed'), was coined in the form "Angiospermae" by Paul Hermann in 1690 but he used this term to refer to a group of plants which form only a subset of what today are known as angiosperms. Hermann's Angiospermae including only flowering plants possessing seeds enclosed in capsules, distinguished from his Gymnospermae, which were flowering plants with achenial or schizo-carpic fruits, the whole fruit or each of its pieces being here regarded as a seed and naked. The terms Angiospermae and Gymnospermae were used by Carl Linnaeus with the same sense, but with restricted application, in the names of the orders of his class Didynamia.^[2]

The terms angiosperms and gymnosperm fundamentally changed in meaning in 1827 when Robert Brown established the existence of truly naked ovules in the Cycadeae and Coniferae.^[3] The term gymnosperm was from then on applied to seed plants with naked ovules, and the term angiosperm to seed plants with enclosed ovules. However, for many years after Brown's discovery, the primary division of the seed plants was seen as between monocots and dicots, with gymnosperms as a small subset of the dicots.^[4]

In 1851, Hofmeister discovered the changes occurring in the embryo-sac of flowering plants, and determined the correct relationships of these to the Cryptogamia. This fixed the position of Gymnosperms as a class distinct from Dicotyledons, and the term Angiosperm then gradually came to be accepted as the suitable designation for the whole of the flowering plants other than Gymnosperms, including the classes of Dicotyledons and Monocotyledons.^[4] This is the sense in which the term is used today.^[5]

In most taxonomies, the flowering plants are treated as a coherent group. The most popular descriptive name has been Angiospermae, with Anthophyta (lit. 'flower-plants') a second choice (both unranked). The Wettstein system and Engler system treated them as a subdivision (Angiospermae). The Reveal system also treated them as a subdivision (Magnoliophytina),^[6] but later split it to Magnoliopsida, Liliopsida, and Rosopsida. The Takhtajan system and Cronquist system treat them as a division (Magnoliophyta).^{[citation needed]} The Dahlgren system and Thorne system (1992) treat them as a class (Magnoliopsida). The APG system of 1998, and the later 2003^[7] and 2009^[8] revisions, treat the flowering plants as an unranked clade without a formal Latin name (angiosperms). A formal classification was published alongside the 2009 revision in which the flowering plants rank as a subclass (Magnoliidae).^[9]

The internal classification of this group has undergone considerable revision. The Cronquist system, proposed by Arthur Cronquist in 1968 and published in its full form in 1981, is still widely used but is no longer believed to accurately reflect phylogeny. A consensus about how the flowering plants should be arranged has recently begun to emerge through the work of the Angiosperm Phylogeny Group (APG), which published an influential reclassification of the angiosperms in 1998. Updates incorporating more recent research were published as the APG II system in 2003,^[7] the APG III system in 2009,^[8][10] and the APG IV system in 2016.

Traditionally, the flowering plants are divided into two groups,

• Dicotyledoneae or Magnoliopsida
• Monocotyledoneae or Liliopsida

to which the Cronquist system ascribes the classes Magnoliopsida (from "Magnoliaceae") and Liliopsida (from "Liliaceae"). Other descriptive names allowed by Article 16 of the ICBN include Dicotyledones or Dicotyledoneae, and Monocotyledones or Monocotyledoneae, which have a long history of use. In plain English, their members may be called "dicotyledons" ("dicots") and "monocotyledons" ("monocots"). The Latin behind these names refers the observation that the dicots most often have two cotyledons, or embryonic leaves, within each seed. The monocots usually have only one, but the rule is not absolute either way. From a broad diagnostic point of view, the number of cotyledons is neither a particularly handy, nor a reliable character.^{[citation needed]}

Recent studies, as by the APG, show that the monocots form a monophyletic group (a clade) but that the dicots are paraphyletic. Nevertheless, the majority of dicot species fall into a clade, the eudicots or tricolpates, and most of the remaining fall into another major clade, the magnoliids, containing about 9,000 species. The rest include a paraphyletic grouping of early branching taxa known collectively as the basal angiosperms, plus the families Ceratophyllaceae and Chloranthaceae.^{[citation needed]}