lHay-J une, 1963] WHITE-THACHEARY ELEMENTS OF FERNS. I. 447 ---, H. Q. SrEvt;NSON, AND A. E. Knun. H)58. products in Maryland tobacco. Arch. Biochem. Limits and consequences of multiple ullopolvploidy Biophys. ,43: 269-285. in Nicotiana. The Nucleus 1: 205-222. TUHNER, B. L., AND R. ALSTON. 19;'i\J. Segregation and SWAIN, M. L., A. EISNlm, C. F. WOODWAIW, AND B. A. recombination of chemical constituents in a hybrid BRICI'. 1049. Ultraviolet absorption spectra of swarm of Baptisia laeoicaulis X B. viridis and their nicotine, nornicotine and some of their derivatives. taxonomic implications. Arner. Jour. Bot. 4G:678-686. Jour. Amer. Chem. Soc. 71: 1341-1345. WADA, E., T. KISAKI, AND M. I HIDA. 1959. The to­ '1'80, T. C. 1962. Some novel concepts on the bio­ bacco alkaloids in the root and sap of some Nicotiana synthesis and biogenesis of tobacco alkaloids. Bot. plants. Arch. Bioehem. Biophys, 80: 258-267. Bull. Acad. Sinica 3: 61-71. WILLITS, C. 0., M. L. SWAIN, J. A. CONNEI,LY, AND ---, AND R. N. J~;FFREY. 1953. Paper chroma­ B. A. BRICE. 1950. Spectrophotometric deter­ tography of alkaloids and their transformation mination of nicotine. Anal. Chern. 22: 430-433. TRACHEARY ELEMENTS OF THE FERNS. 1. FACTORS WHICH INFLUENCE TRACHEID LENGTH; CORRELATION OF LENGTH WITH EVOLUTIONARY DIVERGENCE I RICHARD A. WHI'fE2 Department of Botany, University of Michigan, Ann Arbor, Michigan ABSTRACT WHITE, It. A. (U. Michigan, Ann Arbor.) Tmcheary elements of the ferns. 1. Factors which influence tracheid length; correlation of length with evolutionary divergence. Amer. Jour. Bot. 50(5): 447-455. 1963.-'1'0 determine the variability of tracheid lengths in ferns, several factors were surveyed which might modify them. On the basis of average lengths based on from 100 to 300 measurements, the following factors were found to affect fern tracheid length significantly: location in the plant, stage of maturation of the plant, length of the internode, ploidy, and habitat. These factors were considered when a statistical analysis was made of the relationship between tracheid length and taxonomic divergence among the species of ferns. A Pearson's correlation coefficient was computed, based on average tracheid lengths from 221 species of ferns, and a di­ vergence index based on 8 characters accepted as indicators of specialization or primitiveness in the ferns. A statistically significant correlation (at the 5% level) between length and divergence was found for root tracheids, but not for rhizome or petiolar traoheids. A better understanding of fern evolution is necessary before a more precise picture of length correlation and divergence in the ferns can be obtained. ll)EAS of the evolutionary relationships of ferns These later workers have incorporated numerous are presently in a state of controversy and flux, morphological characters into their systems in an and there are many differing views. The early attempt to determine more accurately the rela­ classifications of Diels (1899), Bower (192:3, 1926, tionships in this group. The fact that there are 1928) and Christensen (1938) have been modified such differing views, however, indicates a need and revised by the more recent work of Ching for additional characters of comparative value (1940), Copeland (1947), and Holttum (1949). which can be used in correlation with those al­ ready known. 1 Received for publication August 15, 1962. Problems of evolutionary relationships in the Part of a dissertation presented to the Horace H. Rack­ angiosperms have been approached frequently ham School of Graduate Studies, University of Michigan, through the use of comparative anatomical data in partial fulfillment of the requirements for the degree of Doctor of Philosophy. The advice and assistance of Dr. (Bailey, 1949). Studies of the xylem have been one Paul S. Dwyer, Mathematics Dept. and Statistical Labora­ of the chief sources of information regarding trends tory of the University of Michigan, with the statistical of diversification within the higher groups. The portions of this study, and of Drs. Charles B. Beck and works of Bailey and Tupper (1918), Frost (1930a, Warren H. Wagner, Jr., throughout the course of this work is gratefully acknowledged. Opportunities to collect b), and Cheadle (1943) are major attempts to de­ research materials were made possible by an H. H. Rack­ termine morphological trends in xylem elements ham School Summer Grant, The H. H. Bartlett Fund which can be applied to the broader problems of and N.S.F. Grant G-I0846. evolution in those groups studied. Their conclu­ 2 The author is presently (1962-63) a National Science Foundation Postdoctoral Fellow in Botany at the Uni­ sions have had a far-reaching effect on our under­ versity of Manchester, England. standing of the systematics of the flowering plants. 448 AMERICAN JOURNAL OF BOTANY [Vol. 50 TABLE 1. Comparison of average lengths' of tracheids from long and short internodes. Both examples indicate a relationship between tracheid length and internode length N ephrolepis spp, M atteucia struthiopteris (Davalliaceae) (Aspidiaceae) N. cordifolia N. hirsutula S.D. S.D. S.D. Rhizome erect 1.21 0.17 2.30 0.22 1.63 0.4!l procumbent 5.46 1. 78 6.42 1.84 3.22 1.06 Root 3.52 1.00 1.79 0.23 Petiole 2.86 0.74 5.64 1.59 9.54 2.Hl a Traeheid length in mm. In contrast to this extensive work on xylary tis­ made of the variation patterns in the length of sue in the angiosperms, nothing of a similar scope fern tracheids. has heretofore been done for the ferns. It would be MATERIALS AND METHODs-The materials used valuable to know whether the evolutionary pat­ in this study of ca. 210 species in approximately terns described for tracheids and vessel members 91 genera and 18 families were made available to of the higher groups are also present in the ferns. me by Mr. A. Murray Evans, Professor Warren Whether or not such trends are found, a broader H. Wagner, Jr., Professor Ralph H. Wetmore, knowledge of the variation of the morphology of Mr. Wing Yew Kwan, Miss Anne Johnson, the the tracheary elements of the ferns might provide University of Michigan Herbarium, and the Uni­ additional information of systematic and evolu­ versity of Michigan Botanical Gardens. Personal tionary significance. collections were made in the Hawaiian Islands in A primary aim of this research, therefore, was to the summer of 1961. Vouchers have been main­ compare the variation in morphology of tracheary tained for most of this work. cells over a wide sample of fern species. All of the Both dried and fixed specimens were used. relevant characters regarded as significant by Where sections were made, the killing and fixing Bailey and Tupper (1918), Frost (1930a, b), and were followed by dehydration in a tertiary butyl Tippo (1946) were examined. A primary require­ alcohol series and embedding in paraffin, as de­ ment, according to Frost (1930a), for a compara­ scribed by Johansen (1940). Serial longitudinal tive study from which broad evolutionary trends and transverse sections were cut (10 J.L) and stained are to be established is the presence of sufficient with a combination of safranin and fast green. diversity in the character to be used as the basis For maceration, the technique of Jeffrey (1917) for conclusions. The discovery of vessels in the was found to be the most useful. Better results roots of Marsilea (White, 1961) and presence of were obtained with dried material if it was pre­ tracheids which are morphologically intermediate treated in an aqueous solution of Tergitol prior to between vessel members and tracheids in widely maceration. A 1% solution of Bismark Brown in separated genera of ferns, to be described sub­ 70% alcohol was used as the stain. sequently, seem to satisfy this requirement. Where possible, 100 tracheids each were meas­ In order to ascribe comparative value to any ured from the rhizome, root, and petiole of each trends which may be found in tracheid length species. In evaluating the results of these obser­ and morphology in the ferns, one must consider vations, the mean and standard deviation were factors which affect these features. Those con­ calculated. An ocular micrometer was used to ditions that are induced by external influences are measure tracheids chosen randomly from each of less value than those which arc genetic in nature. slide. The maceration slide preparation technique Previous studies of gymnosperms and angio­ was developed to the point where there was little sperms have revealed the presence of several ge­ breakage of tracheids during the process; hence a netic, environmental, and positional factors which normal representation of both larger and shorter modify tracheid length. tracheids was assured. The frequency distribution A single pattern of length variation which is of tracheids in the samples (i.e., slides) approxi­ uniformly applicable to plants in general has not mated a normal curve. yet been established. A Pearson's correlation coefficient was com­ Most workers studying the normal variation of puted (Dixon and Massey, 1957) to measure the cells and tissues have utilized the secondary xylem extent of linear relationship between the degree of of woody plants in which cambial initials play an "relative divergence" of the family and the aver­ important role in the establishment of tracheid age traeheid length. The family index number was length. The only recent comparative studies of applied to the genera and species which make up primary xylem have been made by Cheadle (1943) each family, in order to compute the correlation and Biorhorst (1960), and no studies have been coefficient at these levels. Specific collection data, May-June, 1963] WHI'l'E-THACHEARY ELEMENTS OF FERNS. 1. 449 TABLg 2.