Wall Thickening in Sieve Elements.

Because sieve elements are generally regarded as the specialized conducting cells in the phloem tissue, their structure in relation to function has been the subject of many studies and discussions. ' Both the protoplast and the walls of sieve elements have distinctive characteristics, some of which are difficult to relate to the rapid flow of materials-at the rate of up to 100 cm. per hour-recorded in the phloem tissue. We have recently stressed one such feature2 not previously observed: in many dicotyledons having relatively long cambial initials, various anticlinal divisions of the cambial derivatives result in sieve elements shorter than such initials. Thus reduction in the potential length of the sieve elements introduces additional end walls, and the moving solution must pass through a greater number of sieve plates. The present study is concerned with another feature that would seem to interfere with a free flow of materials through the sieve elements-the remarkably thick walls that occur in sieve elements of many species of dicotyledons. The material included 142 species in 121 genera of 58 families of dicotyledons. In 67 per cent of the species, each was represented by more than one collection, usually two, one taken during dormancy, the other during the early or late part of the growth season. Part of the collections was made in California, part in Rhode Island. Only secondary phloem was examined, mostly from branches over ten years old. The methods of preparation of material for microscopic study were the same as described by Cheadle and Esau.I To obtain some notion of the relation of wall thickness to cell size, the sieve elements and their lumina were measured in transections along their tangential and radial diameters, 10 measurements of each for each plant. The cell diameters were measured from middle lamella to middle lamella, the lumina from wall to wall. The difference between the diameter of the cell and that of the lumen, divided by 2, gave the thickness of the wall. The product of the tangential and radial diameters was used to estimate the transectional area of the sieve element and of its lumen. It should be emphasized that these measurements and calculations have only a rough comparative value. The material used was in killed condition, and in such material the walls are somewhat thinner than in fresh sections. Further, the transections of sieve elements and their lumina are not rectangular (see figures), and thus the products of tangential and radial diameters give only approximate size values of these areas. The walls of sieve elements of different species show a wide range of variations in thickness. Some species have walls approximately 1 1A thick (Figs. 1, 2, 5); in others the wall almost occludes the lumen of the cell (Figs. 3, 11). Many intermediate thicknesses occur in various other species (e.g., Figs. 7, 8). There is also a variation in the appearance of the walls. In some species the walls seem to be rather homogeneous (Figs. 1, 2, 5); even the middle lamella may not be distinguish-