Observations based on a study of more than 1350 species distributed among 165 families of angiosperms are presented. The tracheary elements which mature later than the helical ones in the protoxylem‐metaxylem transition are described in terms of a 2‐phase wall deposition process. These elements have a helical framework (first‐order secondary wall) between the gyres of which is deposited additional secondary wall material in the form of sheets or strands or both (second‐order framework). This is indicated both by the sequence of mature elements throughout the primary xylem and by the ontogeny of later maturing elements. Elements in which the second‐order system of the secondary wall is deposited more or less synchronously and in which the sheets or strands are restricted to cell edges, i.e., lines of intersection of adjacent cell faces, are interpreted as primitive. Elements in which the second‐order sheets and strands appear nonsynchronously and with less regard for cell edges are interpreted as advanced. Alternate pitting results from the appearance of oblique second‐order strands with subsequent wall deposition maintaining strand orientation such that pit axes are tilted. In certain elements second‐order strands are deposited before, and wall deposition continues after, the cessation of cell elongation. This results in an alternate pit pattern and may explain certain irregular patterns. Branching of the first‐order helix seems to be relatively insignificant in the development of more elaborate wall patterns. It is more significant in perforation plate elaboration. “Open pits” occur in a number of dicotyledons. These are pit‐like openings which are extended laterally as the thin areas between the gyres of a helix or comparable openings in a reticulum. They constitute a conspicuous feature of the entire protoxylem‐metaxylem transition in certain species. The simple perforation plate of only certain angiosperms seems to be the result of bar breakdown in a multipored plate. Reduction in pore number is also the result of fusions in the first‐order framework lateral to a multipored plate. In dicots this trend rarely culminates in a simple perforation plate, but it frequently does so in monocots. This type of pore number reduction and enlargement frequently accompanies bar breakdown in dicots and certain monocots. The perforation plate is often simple as the result of a terminalization on the cell, in which case the pore does not intersect the first‐order framework. This type of perforation plate occurs in species with and without more obliquely oriented simple perforation plates subject to a breakdown interpretation. Complex multiperforate plates are interpreted as falling into 3 categories : Plates in which a reticulum has resulted from introduction of second‐order secondary wall strands at various orientations and with variable amounts of distortion following deposition; plates in which a reticulum has resulted from ramification and fusion of the strands of the first‐order framework ; and plates which are multiperforate as the result of the presence of a number of separate loci of breakdown within a single pore membrane. Possible ontogenetic complexities in the development of perforation plates subject to breakdown interpretation are discussed. Protoxylem‐metaxylem transitions are described in terms of the sequence of types of perforation plates. Most sequences with various types of plates support the concept of progressively earlier ontogenetic expression of specialized features with progressive elimination of primitive ones. The concept is contradicted by those species in which nearly all perforation plates are simple. Non‐simple plates in these species are found from early protoxylem through mid‐metaxylem but not in the earliest protoxylem. If non‐simple plates are uncommon in a species, they may have a different ontogenetic history in terms of procambial divisions and apical cell growth. In the monocots with a variety of perforation plate types, the probability that a given element will be imperforate or perforate in one way or another will depend on its diameter, not on its position within the sequence. The occurrence of vessels of very limited extent is discussed. It has been calculated that vessels in the stem of Scleria average from 2 to nearly 50 cells in length depending on their diameter.

No abstract

No abstract