C 1998 The Japan Mendel Society Cytologia 63: 329-339, 1998

A Chromosome Phylogeny of the by Using CMA-DAPI Fluorescent Banding

Yoshikazu Hoshi and Katsuhiko Kondo1

Laboratory of Chromosome and Gene Stock, Faculty of Science, Hiroshima University, Higashi-Hiroshima City 739-8526, Japan

Accepted May 21, 1998

Summary vesiculosa L., Dionaea muscipula Ellis, 20 species of L. and Dros- ophyllum lusitanicum Link. were investigated for a metaphase chromosome analysis by using the se- quentially fluorescent CMA and DAPI staining methods. and the Drosera species showed numerous small-sized chromosomes and some middle-sized chromosomes with no primary constriction. Dionaea muscipula had the middle-sized chromosomes with the localized cen- tromeres and CMA-negative and DAPI-positive pericentric bands. lusitanicum dis- played quite large-sized chromosomes with well-defined localized-centromeres. Computer-aided measurements of metaphase chromosomes stained with CMA and DAPI showed that Aldrovanda vesiculosa and the Drosera species displayed the common features such as localization of CMA-pos- itive and DAPI-negative satellites and non-staining region. Key words CMA, DAPI, Diffused centromere, Aldrovanda vesiculosa, Dionaea muscipula, Drosera, Drosophyllum lusitanicum, Droseraceae, Localized centromere.

The Droseraceae consists of four genera; the monotypic Aldrovanda L., Dionaea Ellis and Drosophyllum Link., and the polytypic Drosera L. with approximately 90 species (Diels 1906, Marchant and George 1982). The Droseraceae is a monophyletic family with remarkable differ- ences of chromosomes in size (Rothfels and Heimburger 1968, Stebbins 1971) and in localized and diffused centromeres (Behre 1929, Rothfels and Heimburger 1968, Kondo et al. 1976, Kondo and Segawa 1988, Sheikh and Kondo 1995, Sheikh et al. 1995, Hoshi and Kondo 1998). Drosera has a diffused centromeric chromosome which is disadvantage to designate conventional karyotype (Kondo et al. 1976). Some species of Drosera have a difficulty of chromosome measurements due to small-sized chromosome with obscure euchromatic ends (Rothfels and Heimburger 1968). Extensive cytological studies have been done in Aldrovanda vesiculosa (Kondo 1969, Kress 1970), Dionaea muscipula (Behre 1929, Sato 1947, Kondo 1970), various species of Drosera (e.g. Shimamura 1941, Wood 1955, Kondo 1966, 1969, 1970, 1973, 1976, 1984, Kondo et al. 1976, Kondo and Olivier 1979, Ficini et al. 1980, Beuzenberg and Hair 1983, Kondo and Segawa 1988, Hoshi et al. 1994, Kondo et al. 1994, Sheikh and Kondo 1995, Sheikh et al. 1995, Hoshi and Kondo 1998), and Drosophyllum lusitanicum (Behre 1929, Rothfels and Heimburger 1968). Recently some Drosera studies have demonstrated that fluorochromes such as chromomycin A3 (CMA) and 4',6-diamidino-2-phenylindole (DAPI) were useful dyes to obtain well-stained chromosomes with clear ends and characteristic bands (Sheikh and Kondo 1995, Hoshi and Kondo 1998). Furthermore, computer-aided metaphase measurements offered objective and accurate chro- mosomal analysis (Hoshi and Kondo 1998). In this study, CMA and DAPI fluorescent staining karyotypes in Aldrovanda vesiculosa, Dion- aea muscipula, 20 species of Drosera, and Drosophyllum lusitanicum are analyzed by using a com-

1 Corresponding author: E-mail: [email protected] Contribution no. 58 of Laboratory of Plant Chro- mosome and Gene Stock. 330 Yoshikazu Hoshi and Katsuhiko Kondo Cytologia 63 puter to justify phylogenetic relationships and to speculate speciation and evolution in chromo- somes of the Droseraceae.

Materials and methods Plant materials Aldrovanda vesiculosa, Dionaea muscipula, 20 species of Drosera and Drosophyllum lusitan- icum used are listed in Table 1. of the Drosera species were purchased from Western Aus- tralia, Australia and sowed and germinated both in vitro and in vivo in the Laboratory of Plant Chromosome and Gene Stock, Faculty of Science, Hiroshima University. For in vitro culture of the Drosera species seeds were surface-sterilized and sown on sterilized 1/2 strength of Murashige and Skoog basal medium (MS; Murashige and Skoog 1962) supplemented with 0.2% gelrite without

Table 1. Source of the materials of Aldrovanda vesiculosa, Dionaea muscipula, 20 species of Drosera and Drosophyllum lusitanicum investigated 1998 Chromosome Phylogeny of the Droseraceae 331 any growth substance. Taxonomic treatment of Drosera followed basically Diels (1906).

Slide preparation Growing root tips in vitro were harvested and pretreated with 0.002 M hydroxyquinoline for 2 hr at 18C before fixation with 45% acetic acid for 5 min and then, hydrolysis with a mixture of 1N hydrochloric acid and 45% acetic acid (2 : 1) at 60C for 7 sec. Their root meristems were cut and squashed in 45% acetic acid. The preparations were air-dried for 48 hr at room temperature after removal of coverslips with dry ice.

Fluorescent staining The method of sequentially fluorescent staining with distamycin A and chromomycin A3 (DMA-CMA staining) followed Schweizer (1981) with slight modification. The air-dried slides after the coverslip was removed were used for sequential DMA-CMA staining: They were preincu- bated in Mcllvine's buffer (pH 7.0) for 30 min, treated with 0.1 mg/ml DMA (Sigma) in Mcllvine's buffer for 10 min in a humid chamber and rinsed in Mcllvine's buffer supplemented with 2.5 mg/ml MgSO4 for 10 min. After rinsed, they were stained with 0.1 mg/ml CMA (Sigma) in Mcllvine's buffer supplemented with Mg504 for 10 min. Then, they were mounted with glycerol and stored at 4C for 12 hr to prevent fading. These chromosomes stained with DMA-CMA were irradiated with BV (blue violet) filter cassette and fluoresced yellow. The method of sequentially fluorescent staining with actinomycin D and DAPI (AMD-DAPI staining) followed Schweizer (1976) with slight modification. After DMA-CMA staining, the slides were used for sequential AMD-DAPI staining: The slides were destained in 45% acetic acid for 30 min, rinsed in distilled water for 5 min. They were dipped in Mcllvine's buffer for 30 min, treated in 0.25 mg/ml AMD (Sigma) in Mcllvine's buffer for 15 min in a humid chamber and rinsed in Mcllvine's buffer for 10 min. After rinsed, they were mounted in glycerol and observed immediate- ly. These chromosomes stained with AMD-DAPI were irradiated with UV (ultra violet) filter cas- sette and fluoresced blue. Photographs were taken with the Professional T-MAX film (ISO 400) in a Nikon fluorescent microscope.

Chromosome measurements Chromosome classification followed Hoshi and Kondo (1998). Additionally, super large-sized chromosomes of Drosophyllum lusitanicum were designated karyotype formulae. The karyotype formulae were based on measurements of 30 metaphase cells. Chromosome sizes were defined as: Super large (LL) >15.00 um2 in area, large (L) 6.00-14.99 um2, middle (M) 1.50-5.99 pm2, and small (5) <1.49 um2. Degree of karyotype symmetry at mitotic metaphase was estimated according to Stebbins' classification (1971). Ratio of the largest chromosome to the smallest chromosome was calculated by average area of the largest chromosome/average area of the smallest chromosome. In- terchromosomal asymmetry index proposed by Romero Zarco (1986) was calculated by standard deviation of average area of chromosome complement/average area of chromosome complement. This index provides a way of estimating variation of chromosome area.

Results and discussion Figs. 1-3 illustrate the results of DMA-CMA and AMD-DAPI staining in the chromosomes of Aldrovanda vesiculosa, Dionaea muscipula, 20 species of Drosera, and Drosophyllum lusitanicum of the Droseraceae. Table 2 and Fig. 4 show the results of mitotic chromosome numbers, total chro- mosome area and standard deviation, individual chromosome area from the largest to the smallest chromosome, average area of chromosome complement+ standard deviation of each species and 332 Yoshikazu Hoshi and Katsuhiko Kondo Cytologia 63 1998 Chromosome Phylogeny of the Droseraceae 333 334 Yoshikazu Hoshi and Katsuhiko Kondo Cytologia 63

Fig. I. Mitotic-metaphase chromosomes in somatic cells of Drosera, subgen. Rorella, sect. Rossolis stained with DMA-CMA (a, c, e, g, i, k, m, o, q) and AMD-DAPI (b, d, f, h, j, 1, n, p, r). a, b) D. aliciae. c, d) D. capensis. e, f) D. collinsiae. g, h) D. dielsiana. i, j) D. hilaris. k, 1) D. madagascariensis. m, n) D. montana. o, p) D. trinervia. Two middle-sized chromosomes were observed (arrowheads). q, r) D. vil- losa. CMA-positive and DAPI-negative sites were observed in D. aliciae, D. capensis, D. dielsiana, D. hilaris, D. madagascariensis, D. montana, D. trinervia and D. villosa (arrows). Bar= 5 pm. 1998 Chromosome Phylogeny of the Droseraceae 335

Fig. 2. Mitotic-metaphase chromosomes in somatic cells of Drosera stained with DMA-CMA (a, c, e, g, i, k, m, o, q, s, u) and AMD-DAPI (b, d, f, h, j, 1, n, p, r, t, v). a, b) D. burmanni, sect. Thelocalyx. c, d) D. sessilifolia, sect. Thelocalyx. e, f) D. adelae, sect. Arachnopus. g, h) D. indica, sect. Arachnopus. i,j) D. prolifera, sect. Arachnopus. k, l) D. hamiltonii, sect. Stelogyne. m, n) D. binata, sect. Phycopsis. o, p) D. cist(ora, sect. Ptycnostigma, q, r) D. pauciflora, sect. Ptycnostigma. Two middle- and six small-sized chromosomes (arrowheads) were more than 1.00 pm2 area, while 32 small-size chromosomes less than 0.99 um2 area. s, t) D. auriculata, sect. Polypeltes. u, v) D. peltata, sect. Polypeltes. CMA-positive and DAPI-negative sites were observed in D. burmanni, D. sessilifolia, D. adelae, D. indica, D. prolifera, D. hamiltanii, D. binata, D. cistiflora, D. pauciflora and D. peltata (arrows). Bar=5 pm. Cytologia 63 336 Yoshikazu Hoshi and Katsuhiko Kondo

Fig. 3. Mitotic-metaphase chromosomes in somatic cells of three monotypic genera of the Droser- aceae stained with DMA-CMA (a, c, e) and AMD-DAPI (b, d, f). a, b) Aldrovanda vesiculosa. The chro- mosomes did not obviously visualize any localized-centromere and showed CMA-positive and DAPI- negative sites (arrows). c, d) Dionaea muscipula. The chromosomes showed CMA-negative and DAPI- positive bands at their localized-centromeres. e, f) Drosophyllum lusitanicum. The chromosomes exhibit- ed well-differented, localized centromeres. Weakly CMA-positive and DAPI-negative sites were observed (arrows). Bar=5 um

Fig. 4. Histograms of total chromosomes measured in Aldrovanda vesiculosa, Dionaea muscipula, 20 species of Drosera and Drosophyllum lusitanicum, D.= abbreviation of Drosera. Total chromosome areas of D. aliciae and D. sessilifolia both with 2n=80 were approximately twice as large as those of the species of Drosera with 2n=40. Among the total chromosome areas of the 23 species in four genera in the Droseraceae studied, the total chromosome area of D. burmanni with 2n=20 was the smallest. 1998 Chromosome Phylogeny of the Droseraceae 337 karyotype formulae. The chromosome numbers of Drosera dielsiana Exell ex Laudon, D. hilaris Cham. et Schlechtd., D. montana St. Hil., D. pauciflora Banks ex DC., D. trinervia Spreng. and D. villosa St. Hil. were reported here for the first time, and those of the other species in the Droseraceae studied here verified the previous counts. The largest chromosome of 40.02 ,um2was observed in Drosophyllum lusitanicum (2n=12) (Table 2, Fig. 3e, f), while the smallest chromosome of 0.21 um2 was observed in Drosera pauciflo- ra (2n=40) (Table 2, Fig. 2q, r). In other words, the largest chromosome in the Droseraceae studied here was approximately 190-fold as large as the smallest chromosome. In contrast, the largest total chromosome area of 337.28 tm2 in Drosophyllum lusitanicum was about 25-fold as large as the smallest total chromosome area of 15.36 um2 in Drosera burmanni Vahl. (2n=20) (Table 2, Figs. 2a, b, 3e, 0. The differences of total chromosome area might be reflected on wide range of DNA vaues in genome size. However, Rothfels and Heimburger (1968) estimated that the Droseraceae could have 60 to 80-fold range of DNA values in genome size. Then, the disagreement of estima- tions between Rothfels and Heimburger's DNA values (1968) and the present chromosome size speculated that a degree of chromosomal condensation in Drosera might be different from that in Drosophyllum lusitanicum. Drosera species may have more decondensed chromosomes than Dros- ophyllum lusitanicum because genome size estimation using DNA amount showed greater range than that of chromosome size. Among the chromosome numbers of the Drosera species studied here the aliquot numbers by ten were found in the members of sections Arachnopus Planch., Rossolis Diels, Ptycnostigma Diels and Thelocalyx Planch. (Table 2, Figs. 1, 2a-d) . All species studied in sections Rossolis except for D. trinervia Spreng. and Thelocalyx displayed commonly symmetrical karyotypes with lower aver- age interchromosomal asymmetry index less than 0.28. In contrast, all species studied in sections Arachnopus and Ptycnostigma displayed asymmetrical karyotypes with higher ratio of the largest chromosome to the smallest chromosome more than 4.7 (Table 2, Fig. 2e-j, o-r). The basic chro- mosome number of x=10 has been suited for the species of section Rossolis, series Eurossolis Diels (Kondo 1971, Kondo and Segawa 1988). The present study strongly suggested that the basic chromosome number of x=10 fitted to the species of section Thelocalyx as well as section Rosso- lis, but did not to the species of sections Arachnopus and Ptycnostigma. The fluorochromes of CMA and DAPI were used to detect base-specific DNA segment on chromosomes in the Droseraceae. CMA binds specifically to guanine regions in helical DNA (Ward et al. 1965) or heterochromatin (Deumling 1981, Deumling and Greilhuber 1982), while DAPI binds specifically to AT base pairs in the minor groove of DNA (Portugal and Waring 1988). There- fore, a CMA-positive or DAPI-negative (CMA+DAPI-) segment is high GC-content. CMA+DAPI- sites were observed in chromosome complements of Aldrovanda vesiculosa and most of the Drosera species studied (Table 2, Figs. 1-3a, b). All CMA+DAPI- sites in Drosera species were speculated to hold the nucleolar organizing region (NOR) due to the following reasons: (1) CMA+DAPI- sites were observed in or near-by nucleoli in the interphase nuclei (Figure not shown); (2) CMA+DAPI- sites were elongated in certain chromosomes during prophase to early- metaphase (Figure not shown); (3) Two CMA+DAPI- sites were observed as the satellites at an end of certain metaphase chromosomes in many Drosera species (Figs. 1, 2); (4) Generally, NORs in- clude 45S rDNA sequence having high GC contents, which could be detected by sequentially stain- ing with CMA and DAPI (e.g. Hizume et al. 1988, 1989, 1991); and (5) NORs localized at an end of two chromosomes were demonstrated in D. petiolaris R. Br. ex DC. by using the silver staining method (Kondo and Lavarack 1984). In contrast, Dionaea muscipula and Drosophyllum lusitanicum showed different CMA+DAPI- banding patterns from Aldrovanda vesiculosa and the species of Drosera described above. Dionaea muscipula did not show any CMA+DAPI- site (Fig. 3c, d), and Drosophyllurn lusitanicum had ex- 338 Yoshikazu Hoshi and Katsuhiko Kondo Cytologia 63 tremely weak CMA+DAPI- sites on the secondary constriction located at the interstitial region near the centromere (Fig. 3e, f, arrows). Thus, the CMA and DAPI staining pattern suggested that Dros- ophyllum lusitanicum had two NORs at the secondary constrictions of two chromosomes. The diploid (2n=20), tetraploid (2n=40) and octaploid (2n=80) species of Drosera, which has basic chromosome number of x=10, always showed two CMA+DAPI- satellites (Table 2), ex- cepting D. collinsiae showed no CMA±DAPI- satellite (Fig. 1e, f) and D. dielsiana showed four CMA+DAPI- (Fig. 1g, h). These results led us a speculation that a half reduction of satellites in the tetraploid and octaploid might occur after a polyploidization of the diploid species, and might be due to avoid an abnormal meiotic division with trivalent and tetravalent chromosomes. Two notable middle-sized chromosomes in D. trinervia (Fig. 1p, arrowheads) supported this speculation. The South African and South American members of section Rossolis studied here were the tetraploid (2n=40) and octaploid (2n= 80) species with small-sized chromosomes (Table 2), while the northern hemisphere's members of section Rossolis studied previously were the diploid (2n=20) and tetraploid (2n=40) species with middle-sized chromosomes, excepting D. spathulata Labill. was an intraspecific polyploid species (Kondo 1971, Hoshi and Kondo 1998). An origin of the polyploidization of Drosera with the basic chromosome number of x=10 has been discussed by Rothfels and Heimburger (1968) and Kondo (1971). Rothfels and Heimburger (1968) inferred that polyploidization of the might be originated from the diploid (2n=20) with all small-sized chromosomes. However, Kondo (1971) hypothesized that the tetraploid D. spathulata which had 40 small-sized chromosomes might be arisen from the diploid D. spathulata which had 20 small-sized chromosomes. These investigations suggested that the diploid species which carried 20 middle- sized chromosomes and the tetraploid species which carried 40 small-sized chromosomes might have different genome sizes but be occurred from an common ancestral species which carried 20 small-sized chromosomes by increase of chromosome size or number, respectively. No chromosome with the primary constriction was observed in all Drosera species (Figs. 1, 2). The diffused-centromeric chromosome has been proposed in Drosera (Kondo et al. 1976, Kondo and Lavarack 1984, Sheikh and Kondo 1995, Sheikh et al. 1995, Hoshi et al. 1998). The chromo- some nature of Aldrovanda vesiculosa was quite similar to that of the Drosera species. In contrast, the localized-centromeric chromosomes with CMA-DAPI+ pericentromeric heterochromatin were shown in Dionaea muscipula (Fig. 3c, d) and the well-differentiated, localized-centromeric chromo- somes were observed in Drosophyllum lusitanicum (Fig. 3e, f).

References Behre,K. 1929.Plfsiologische and zytologische Untersuchungen uberDrosera. Planta (Berl.) 7: 208-306. Beuzenberg,E. J. and Hair,J. P. 1983.Contributions to a chromosomeatlas of the NewZealand flora 25 miscellaneous species.New Zealand J. Bot.21: 13-20. Deumling,B. 1981.Sequence arrangement of a highlymethylated satellite DNA of a plant,Scilla: a tandemlyrepeated in- vertedrepeat. Proc. Natl. Acad. Sci. USA 78: 338-442. -and Greilhuber, J. 1982. Characterization of heterochromatin in different species of the Scilla siberica group (Liliaceae) by in situ hybridization of the satellite DNAs and fluorochrome banding. Chromosoma (Berl.) 84: 535-555. Diels, L. 1906. Droseraceae. In: Engler, A. (ed.): Das Pflanzenreich IV, Leipzig. Verlag von Wihelm Engelmann, pp. 136. Ficini, G., Garbari, F., Giordani, A. and Tomei, P. E. 1980. Numeri cromosomici per la Flora Italiana. Inf. Bot. Ital. 12: 113-116. Hizume, M., Abe, K. and Tanaka, A. 1988. Fluorescent ,chromosome bandings in the Taxodiaceae. La Kromosomo 11-50: 1609-1619. - , Kitazawa, N., Gu, Z. and Kondo, K. 1991. Variation of fluorescent chromosome band in Picea brachytyla var. com- planata collected in Yunnan, China. La Kromosomo 11-62-64: 2149-2158. - , Ohgiku, A. and Tanaka, A. 1989. Chromosome banding in the genus Pinus II. Interspecific variation in fluorescent banding pattern in P. densiflora and P. thunbergii. Bot. Mag. Tokyo 102: 25-36. Hoshi, Y., Hizume, M. and Kondo, K. 1994. Genomic in situ hybridization to improve a hypothesis on natural-hybrid Drosera spathulata 'Kansai type'. La Kromosomo 11-75-76: 2619-2623. 1998 Chromosome Phylogeny of the Droseraceae 339

―and Kondo, K. 1998. Chromosome differentiation in Drosera, subgenus Rorella, section Rossolis. Cytologia 63: 199-211. Kondo, K. 1966. Meiosis in PMC of three species of Drosera. Chrom. Inf. Service 7: 23-24. ―1969 . Chromosome numbers of carnivorous . Bull. Torrey Bot. Club 96: 322-328. ―1970 . Chromosome numbers in Drosera and Dionaea in North Carolina. J. Jpn. Bot. 45: 11-16. ―1971 A review of the Drosera spathulata complex. J. Jpn. Bot. 46: 321-326. ―1973 . Chromosome numbers of some Drosera taxa. J. Jpn. Bot. 48: 193-198. ―1976 . A cytotaxanomic study in some species of Drosera. Rhodora 78: 532-541. ―1984 . Three new species of Drosera L. from Australia. Boletim Da Sociedade Broteriana 57: 51-60. ―and Lavarack, P. S. 1984. A cytotaxonomic study of some Australian species of Drosera L. (Droseraceae). Bot. J. Lin- nean Soc. 88: 317-333. ―and Olivier, M. C. 1979. Chromosome numbers of four species of Drosera (Droseraceae). Ann. Missouri Bot. Gard. 66: 584-587. ―and Segawa, M. 1988. A cytotaxonomic study in artificial hybrids between Drosera anglica Huds. and its certain closely related species in series Drosera, section Drosera, subgenus Drosera, Drosera. La Kromosomo 11-51-52: 1697-1709. ― ,―and Nehira, K. 1976. A cytotaxonomic study in four species of Drosera. Mem. Fac. Integrated Arts and Sci. Hiroshi- ma Univ. Ser. 4 2: 27-36. ― ,Sheikh, S. A. and Hoshi, Y. 1994. New finding of another 2n=6 species in the angiosperm, Drosera roseana Marchant. Chrom. Inf. Service 57: 3-4. Kress, A. 1970. Zytotaxonomische Untersuchungen an einigen Insektenfängern (Droseraceae, Byblidaceae, Cephalotaceae, Roridulaceae, Sarraceniaceae). Ber. Dtsch. Bot. Ges. 83: 55-62. Marchant, N. G. and George, A. S. 1982. Drosera. pp. 9-66. In: Briggs, B. G., Barlow, B. A., Eichler, H., Pedley, L., Ross, J. H., Symon, D. E, and Wilson, P. G. (eds.). , Vol. 8. Australian Government Publ. Serv., Canberra, pp. 420. Murashige, T. and Skoog, F. 1962. A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol. Plant 15: 473-497. Portugal, J. and Waring, M. J. 1988. Assignment of DNA binding site for 4',6-diamidine-2-phenylindole and bisbenzimide (Hoechst 33258). A comparative study. Biochem. Biophys. Acta 949: 158-168. Romero Zarco, C. 1986. A new method for estimating karyotype asymmetry. Taxon 35: 526-530. Rothfels, K. and Heimburger, M. 1968. Chromosome size and DNA values in sundews (Droseraceae). Chromosoma (Berl.) 25: 96-103. Sato, D. 1948. The karyotype of the insectivorous plants. Oguma Comm. Vol. Cytol. Gen. pp. 25-28. Schweizer, D. 1976. Reverse fluorescent chromosome banding with chromomycin and DAPI. Chromosoma (Berl.) 58: 307-324.

―1981 . Counterstain-enhanced chromosome banding. Hum. Genet.57:1-14. Sheikh, S. A. and Kondo, K. 1995. Differential staining with orcein, Giemsa, CMA and DAN for comparative study of 12 species of Australian Drosera (Droseraceae). Am. J. Bot. 82: 1278-1286. ― ,―and Hoshi, Y. 1995. Study on diffused centromeric nature on Drosera chromosomes. Cytologia 60: 43-47. Shimamura, T. 1941. Cytological study of Drosera obovata Mert. et Koch with special reference to its hybridity. Bot. Mag. Tokyo 55: 553-558. Stebbins, G. L. 1971. Chromosomal evolution in higher plants. Addison-Wesley Publ. Co., Reading, Mass. pp. 216. Wood Jr., C. E. 1955. Evidence for the hybrid origin of Drosera anglica. Rhodora 57: 105-130.