CYTOLOGICAL AND GROWTH CORRELATIONS IN THE FLOWER BUD AND ANTHER OF LILIUM LONGIFLORUM

Abstract

CYTOLOGISTS WHO have worked with flowering plants are familiar with the fact that in some species anthers containing microsporocytes in meiosis or microspores in mitosis can be found with some degree of certainty by selecting buds of particular sizes. This suggests, at the least, that there is some consistent correlation between the general growth of the flower bud and the nuclear processes preceding, and involved in, the formation of pollen. A careful study of the relationship of morphological and physiological changes which occur in development of the anther, to the cytological processes of microsporogenesis and pollen formation, may contribute as much to our understanding of cell division, as may the continuation of the traditional intensive studies of the dividing cells. It was with this in mind that the present attempt to sketch in the broad outlines of the growth of the flower bud and anther in its relationship to the cytology of the anther was undertaken. The first part of this paper will deal with growth curves of the bud and anther, and their allometric relationships, the second with the mitotic and meiotic divisions which take place within the anther locules. Some attention will also be given to the growth of the pistil, megasporogenesis and embryo-sac formation. The correlations which are demonstrated in this paper have considerable interest in themselves. They will also have practical value as the basis for future cytological studies from a developmental and physiological point of view. Such a study has already been published (Erickson, 1947), in which it was shown that meiosis of the microsporocytes, and mitosis of the microspores, each coincides with a marked temporary drop in the rate of oxygen uptake of the anther as a whole. This observation was partially anticipated by Guilcher (1937), who presented respiratory data for Lilium croceum anthers and pistils, but no cytological information. It has recently been confirmed and extended for Trillium anthers during the microspore stage by Stern and Kirk (1948). Stern's (1946) demonstration of an increase in permeability of Trillium microsporocytes and microspores during meiosis and mitosis might also be cited. Physiological studies of the anther and its contents, such as these, hold great promise of throwing light on fundamental cytological problems. 1 Received for publication June 28, 1948. The greater part of this work was done in the Department of Botany of the University of Rochester. A small portion of the work, and completion of the manuscript was done in the Department of Botany of the University of Pennsylvania, with the financial support of the SloanKettering Institute for Cancer Research, and the Cancer Commission of the University of Pennsylvania. The author recalls with pleasure the stimulating discussions with Professor David R. Goddard which led to this study. BUD GROWTH.-Lilium longiflorum var. "Floridii" has been used for this study. Like other varieties of the Easter lily, it is commerciallv propagated by vegetative means, and appears to have the genetic uniformity of a single clone. Bulbs were obtained from the American Bulb Company, New York, N. Y., and kept in cold storage. A few were withdrawn each week and planted in mixed loam and peat moss in 61/2-in. pots. The pots were kept in the greenhouse. In this way a continuous supply of flower buds in various stages of development can be maintained over an extended period. During the summer of 1945, the spring of 1946, and the spring and summer of 1947, the length of each flower bud on two-five plants has been measured each day, with occasional omissions, at 10:00 A.M., EST, from the time at which the bud could first be measured conveniently until the flower had opened and withered. On either side of the midrib of each petal (inner perianth segment) is a prominent groove into which the margin of a sepal (outer perianth segment) fits. The measurements of bud length were made in millimeters and tenths with a steel vernier calipers, by placing one jaw of the calipers in contact with the tip of the bud, or, after anthesis, with the straightened tip of one of the petals, and adjusting the other jaw to the basal end of one of the above-mentioned grooves. When bud lengths are plotted on a linear axis against time in days, curves are obtained which appear to be exponential, from the time of the first measurements, until a day or so before anthesis. This supposition is supported by plotting the logarithms of bud lengths against time, when approximately straight lines are obtained. Some illustrative data are plotted in this way in fig. 1. The small arrow beside each curve indicates the day on which the bud had started to open. The exponential equation