Treatment of erect stems of Prosopis with near phytotoxic levels of 2,4-D or 2,4,5-T causes the formation of an unusual wood with narrow, thick-walled vessels and axial parenchyma in which cell wall thickening is inhibited. Although reduced in diameter, the vessels formed during 2,4-D and 2,4,5-T treatment are so numerous that there is no significant difference between phenoxyacetic acid and control seedling groups with regard to total area of xylem occupied by vessels. The preferential maturation of xylem vessels over parenchyma and the transformation of fusiform initials into septate parenchyma strands in phenoxyacetic acidtreated Prosopis resemble the structural changes reported to occur after girdling in the cambial tissue of other arborescent angiosperms. Bending experiments indicate that tension-wood fibers of Prosopis differentiate in response to an auxin deficiency. However, xylogenesis in erect stems treated with TIBA is affected such that a significantly higher proportion of the cambial cell population becomes axial xylem parenchyma. MANY WELL-DOCUMENTED attempts have been made to account for the killing actions of 2,4dichlorophenoxyacetic acid and the other phenoxyacetic acids since the advent of their utilization. Early explanations cited the disruption of the phloem or the depletion of stored carbohydrates (Tukey, Hamner, and Imhofe, 1946; Loustalot and Muzik, 1953). However, as further explanations were sought, it became apparent that many factors earlier attributed to causing death of the plants represented only part of the total action of the phenoxyacetic acid herbicides. Recent investigations have shown that many aspects of plant development, including cell expansion and proliferation, are affected by herbicides (Kiermayer, 1964). Cardenas et al. (1968) report that 2,4-D causes an abnormal increase in axis growth at the expense of other organs, especially leaves and young roots. Picloram has been shown to alter sink activities such that there is preferential movement of assimilates to the more mature parts of the plant axis (Leonard, Glenn, and Bayer, 1972) at the expense of young leaves (Sharma and Vanden Born, 1971). In seedlings of Pinus resinosa treated with picloram or 2,4-D both root growth and needle extension are inhibited, but tissue proliferation in the stem, hypocotyl and upper root is promoted (Wu et al., 1971). Abnormal axis growth following 2,4-D treatment is associated with high levels of aberrant 1 Received for publication 17 November 1972. We thank Elsie D. Morey for valuable technical assistance and Dr. William R. Atchley for help with the statistical analysis. This work was supported by the Noxious Brush and Weed Control Program and the Institute for University Research of Texas Tech University. nucleic acid and protein synthesis (Cardenas et al., 1968), and, as might be expected, a wide variety of anatomical modifications accompany these physiological changes. In general, the least differentiated tissues of the axis are most subject to proliferative activity following herbicide treatment. Proliferation of cambial, phloem, cortical and other tissues may result in tumor formation (Murray and Whiting, 1947) in which the cells may follow a number of developmental pathways, including differentiation as tracheary elements or organization into root primordia. In some tissues 2,4-D may affect cell differentiation by inhibiting cell wall synthesis (Bourke, Butts, and Fang, 1962). The study reported here on Prosopis, though limited to effects on wood formation, attempts to further explain the killing action of the phenoxyacetic acid herbicides and, more importantly, to expand the work in this area on woody plants. In addition it was realized early in this study that normal wood formation in Prosopis was in itself of considerable interest. For example, xylem parenchyma may comprise up to 75 % of the axial system of cells, and the fiber component, even in erect stems, is largely of the tension-wood type. The experimental studies on Prosopis reported herein reveal information on the mechanisms controlling cell differentiation of those xylem elements not easily accessible to study in other systems. MATERIALS AND METHODS-2,3,5-triiodobenzoic acid (TIBA) and 2,4-dichlorophenoxyacetic acid (2,4-D) in anhydrous lanolin (2.0 %, w/w) were applied in a ring around the midpoint of the first or third internode of erect fourto six-week