Fructans are polysaccharides that are made up mainly of fructose. They are non-digestible carbohydrates and act as prebiotics to selectively promote the growth of colonic bifidobacteria, thereby improving human gut health. Fructans are present in the grain of wheat (Triticum aestivum L.), a staple food crop. Until now, there has been no research on genetic improvement of the concentration of fructans in wheat grain, partly because it has been difficult to accurately measure. One aim of this research project was to develop a simple and effective method to measure the fructan concentration in wheat grain. This was achieved by modifying a method that involves extraction of fructans from wheat grain followed by enzymatic hydrolysis to break down fructans into monosaccharides and quantification by anion-exchange liquid chromatography coupled with pulsed amperometric detection. The modified procedure is reliable and allows the handling of large numbers of flour samples at a relatively low cost, and can therefore be useful for assessing large numbers of wheat breeding lines. Using this method, grain samples taken from a diverse set of 117 wheat cultivars and breeding lines, including parents of mapping populations, were analysed for grain fructan concentration. There was significant genotypic variation among these materials, with grain fructan concentration ranging from 0.3 to 2.3% of grain dry weight. There was no evidence of strong genotype-byenvironment interaction; the fructan concentrations of the same genotypes were positively correlated over different environments in Australia. Genetic mapping was carried out to detect and map loci affecting grain fructan concentration in wheat using a doubled haploid population derived from a cross between Berkut (high fructan) and Krichauff (low fructan). Grain samples were obtained from two field sites in South Australia and one in Kazakhstan. Fructan concentration varied widely within the population (0.6-2.6% of grain dry weight), with heritability estimated as h² = 0.71. A linkage map of 528 molecular markers covering 21 wheat chromosomes was used for locating quantitative trait loci (QTL). Genetic mapping identified two major QTLs on chromosomes 6D and 7A, with the (high fructan concentration) alleles contributed from Berkut, contributing to a 30-40% increase in wheat grain fructan compared to the Krichauff alleles. Effects of these chromosome regions were validated in additional environments and in another mapping population, Sokoll/Krichauff, with the favourable alleles contributed from Sokoll. The major QTL on chromosome 7A was in the same region with a reported fructosyltransferase orthologue (AB029888), while the major QTL on chromosome 6D seemed to be co-located with a reported gene encoding for a fructan-degrading enzyme 1-exohydrolase (1-FEHw2). It is concluded that grain fructan concentration of wheat can be improved by breeding and that molecular markers could be used to select effectively for favourable alleles in two regions of the wheat genome. / Thesis (Ph.D.) - University of Adelaide, School of Agriculture, Food and Wine, 2009

Fructans are polysaccharides that are made up mainly of fructose. They are non-digestible carbohydrates and act as prebiotics to selectively promote the growth of colonic bifidobacteria, thereby improving human gut health. Fructans are present in the grain of wheat (Triticum aestivum L.), a staple food crop. Until now, there has been no research on genetic improvement of the concentration of fructans in wheat grain, partly because it has been difficult to accurately measure. One aim of this research project was to develop a simple and effective method to measure the fructan concentration in wheat grain. This was achieved by modifying a method that involves extraction of fructans from wheat grain followed by enzymatic hydrolysis to break down fructans into monosaccharides and quantification by anion-exchange liquid chromatography coupled with pulsed amperometric detection. The modified procedure is reliable and allows the handling of large numbers of flour samples at a relatively low cost, and can therefore be useful for assessing large numbers of wheat breeding lines. Using this method, grain samples taken from a diverse set of 117 wheat cultivars and breeding lines, including parents of mapping populations, were analysed for grain fructan concentration. There was significant genotypic variation among these materials, with grain fructan concentration ranging from 0.3 to 2.3% of grain dry weight. There was no evidence of strong genotype-byenvironment interaction; the fructan concentrations of the same genotypes were positively correlated over different environments in Australia. Genetic mapping was carried out to detect and map loci affecting grain fructan concentration in wheat using a doubled haploid population derived from a cross between Berkut (high fructan) and Krichauff (low fructan). Grain samples were obtained from two field sites in South Australia and one in Kazakhstan. Fructan concentration varied widely within the population (0.6-2.6% of grain dry weight), with heritability estimated as h² = 0.71. A linkage map of 528 molecular markers covering 21 wheat chromosomes was used for locating quantitative trait loci (QTL). Genetic mapping identified two major QTLs on chromosomes 6D and 7A, with the (high fructan concentration) alleles contributed from Berkut, contributing to a 30-40% increase in wheat grain fructan compared to the Krichauff alleles. Effects of these chromosome regions were validated in additional environments and in another mapping population, Sokoll/Krichauff, with the favourable alleles contributed from Sokoll. The major QTL on chromosome 7A was in the same region with a reported fructosyltransferase orthologue (AB029888), while the major QTL on chromosome 6D seemed to be co-located with a reported gene encoding for a fructan-degrading enzyme 1-exohydrolase (1-FEHw2). It is concluded that grain fructan concentration of wheat can be improved by breeding and that molecular markers could be used to select effectively for favourable alleles in two regions of the wheat genome. / Thesis (Ph.D.) - University of Adelaide, School of Agriculture, Food and Wine, 2009

Bibliography: leaves 234-256. / ix, 257 leaves, [12] leaves of plates : ill. (chiefly col.) ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Genetic studies of tolerance of durum wheat (Triticum turgidum L. var durum) to high concentrations of boron (B) were undertaken to identify genetic variation in response to B, the mode of gene action, number of genes and chromosomal locations of genes controlling tolerance. Results demonstrated that tolerance to B is under simple genetic control as observed in bread wheat. High levels of tolerance can be transferred into sensitive commercial varieties via backcrossing and selection can be performed during seedling growth at early generations. / Thesis (Ph.D.)--University of Adelaide, Dept. of Plant Science, 1996

Errata slip inserted. / Bibliogaphy: leaves [93-109]. / x, 92, [17] leaves, [20] leaves of plates : ill. ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Identifes and maps markers associated with the quality traits of grain protein content, milling yield and flour colour, and with genes resistant to stem- and leaf-rust diseases in wheat. / Thesis (Ph.D.)--University of Adelaide, Dept. of Plant Science, Waite Agricultural Research Institute, 1998

This study comprised three major parts: a comparative physiological study of drought responses under controlled conditions; a genetic study to construct the skeleton map of a doubled haploid (DH) population; and a quantitative trait loci (QTL) analysis to identify QTLs associated with drought tolerance traits in the field. In the first part (Chapter 3), three cultivars of wheat (Triticum aestivum L.) adapted to South Australian conditions were tested for drought tolerance under cyclic drought in growth rooms and glasshouse. Extensive physiological traits, including stomatal conductance, chlorophyll content and fluorescence, ABA content, water status traits (e.g. osmotic adjustment, RWC and leaf water potential), water soluble carbohydrates (WSC) and carbon isotope discrimination (Δ¹ ³C) were measured during experiments. Through these experiments, the drought responses of the three cultivars were physiologically dissected and the likely processes contributing most to drought tolerance were identified. In the South Australian wheatbelt, cyclic drought is a frequent event, represented by intermittent periods of rainfall which can occur around anthesis and post-anthesis in wheat. Three South Australian bread wheat cultivars, Excalibur, Kukri and RAC875, were evaluated in two growth room experiments under cyclic water-limiting conditions. In the first experiment, where plants were subjected to severe water stress, RAC875 and Excalibur (drought tolerant) showed significantly (P < 0.05) higher grain yield under cyclic water availability compared to Kukri (drought susceptible), producing 44% and 18% more grain yield compared to Kukri, respectively. In the second growth room experiment, where plants were subjected to a milder drought stress, the differences between cultivars were less pronounced, with only RAC875 showing significantly higher grain yield under the cyclic water treatment. Grain number per spike and the percentage of aborted tillers were the major yield components that affected yield under cyclic water stress. Excalibur and RAC875 adopted different morpho-physiological traits and mechanisms to reduce water stress. Excalibur was most responsive to cyclic water availability and showed the highest level of osmotic adjustment (OA), highest stomatal conductance, lowest ABA content and most rapid recovery from stress under cyclic water stress. RAC875 was more ‘conservative’ in its responses, with moderate OA, high leaf waxiness, high chlorophyll content and slower recovery from stress. Within this germplasm, the capacity for osmotic adjustment was the main physiological attribute associated with tolerance under cyclic water stress, which enabled plants to recover from water deficit. In the second part (Chapter 4), the genetic linkage map of a DH population including 368 lines, which was developed from a cross between ‘RAC875’ and ‘Kukri’, was constructed. The genetic linkage map consisted of about 500 molecular markers including ~300 DArT (Diversity array technology) and ~200 SSR (Microsattelite markers). In the third part (Chapter 5), Quantitative Trait Loci (QTLs) linked to plant phenology and production traits under irrigated and drought stress conditions were mapped by means of a DH population. To phenotype the population, 368 DH lines were cultivated in two replicates in five environments (three sites across South Australian wheatbelt in collaboration with Australian Grain Technology (AGT) in 2006, and two trials in Mexico in collaboration with CYMMIT, 2007). Data of grain yield, yield components, maturity related traits and some morpho-physiological traits such as leaf chlorophyll content, leaf waxiness, plant height, peduncle length, flag leaf and spike length were measured. Raw data were then analysed for spatial variation for each single trial using the REML procedure in GenStat (version 6). The DH lines showed significant variation for plant phenology, grain yield and yield components under irrigated and drought stress conditions. QTL analyses were performed using QTLCartographer and QTLNetwork for each trait in each site. Two major QTL for maturity traits were identified on chromosomes 2BS and 2DS corresponding to Ppd-B1 and Ppd-D1, respectively. A region was identified on chromosome 7A that harbored major QTL for grain yield, number of grains per square meter, number of grain per spike and spike fertility under drought stress. For yield data in the irrigated trial, two major QTL were identified on chromosome 3B which were not detected in drought stress environments. By using different datasets in the QTL analysis (splitting the population into two subpopulation based on heading time and also adjusting the phenotypic data for heading time to eliminate heading time effect), a QTL for grain yield was consistently detected on chromosome 7A in drought-affected environments. The coincidence of a drought response index QTL on this chromosome indicated that it might be a QTL for yield response under drought. This study demonstrated that the region on the long arm of chromosome 7A identified for grain yield and yield components is a drought response QTL which is closely linked to, but separate from, a heading time QTL. This QTL cluster on chromosome 7A could be used as a good target for positional cloning and gene isolation. However further work would be required to confirm and validate the identified QTLs in this preliminary QTL analysis. / http://proxy.library.adelaide.edu.au/login?url= http://library.adelaide.edu.au/cgi-bin/Pwebrecon.cgi?BBID=1340056 / Thesis (Ph.D.) -- University of Adelaide, School of Agriculture, Food and Wine, 2008

This study comprised three major parts: a comparative physiological study of drought responses under controlled conditions; a genetic study to construct the skeleton map of a doubled haploid (DH) population; and a quantitative trait loci (QTL) analysis to identify QTLs associated with drought tolerance traits in the field. In the first part (Chapter 3), three cultivars of wheat (Triticum aestivum L.) adapted to South Australian conditions were tested for drought tolerance under cyclic drought in growth rooms and glasshouse. Extensive physiological traits, including stomatal conductance, chlorophyll content and fluorescence, ABA content, water status traits (e.g. osmotic adjustment, RWC and leaf water potential), water soluble carbohydrates (WSC) and carbon isotope discrimination (Δ¹ ³C) were measured during experiments. Through these experiments, the drought responses of the three cultivars were physiologically dissected and the likely processes contributing most to drought tolerance were identified. In the South Australian wheatbelt, cyclic drought is a frequent event, represented by intermittent periods of rainfall which can occur around anthesis and post-anthesis in wheat. Three South Australian bread wheat cultivars, Excalibur, Kukri and RAC875, were evaluated in two growth room experiments under cyclic water-limiting conditions. In the first experiment, where plants were subjected to severe water stress, RAC875 and Excalibur (drought tolerant) showed significantly (P < 0.05) higher grain yield under cyclic water availability compared to Kukri (drought susceptible), producing 44% and 18% more grain yield compared to Kukri, respectively. In the second growth room experiment, where plants were subjected to a milder drought stress, the differences between cultivars were less pronounced, with only RAC875 showing significantly higher grain yield under the cyclic water treatment. Grain number per spike and the percentage of aborted tillers were the major yield components that affected yield under cyclic water stress. Excalibur and RAC875 adopted different morpho-physiological traits and mechanisms to reduce water stress. Excalibur was most responsive to cyclic water availability and showed the highest level of osmotic adjustment (OA), highest stomatal conductance, lowest ABA content and most rapid recovery from stress under cyclic water stress. RAC875 was more ‘conservative’ in its responses, with moderate OA, high leaf waxiness, high chlorophyll content and slower recovery from stress. Within this germplasm, the capacity for osmotic adjustment was the main physiological attribute associated with tolerance under cyclic water stress, which enabled plants to recover from water deficit. In the second part (Chapter 4), the genetic linkage map of a DH population including 368 lines, which was developed from a cross between ‘RAC875’ and ‘Kukri’, was constructed. The genetic linkage map consisted of about 500 molecular markers including ~300 DArT (Diversity array technology) and ~200 SSR (Microsattelite markers). In the third part (Chapter 5), Quantitative Trait Loci (QTLs) linked to plant phenology and production traits under irrigated and drought stress conditions were mapped by means of a DH population. To phenotype the population, 368 DH lines were cultivated in two replicates in five environments (three sites across South Australian wheatbelt in collaboration with Australian Grain Technology (AGT) in 2006, and two trials in Mexico in collaboration with CYMMIT, 2007). Data of grain yield, yield components, maturity related traits and some morpho-physiological traits such as leaf chlorophyll content, leaf waxiness, plant height, peduncle length, flag leaf and spike length were measured. Raw data were then analysed for spatial variation for each single trial using the REML procedure in GenStat (version 6). The DH lines showed significant variation for plant phenology, grain yield and yield components under irrigated and drought stress conditions. QTL analyses were performed using QTLCartographer and QTLNetwork for each trait in each site. Two major QTL for maturity traits were identified on chromosomes 2BS and 2DS corresponding to Ppd-B1 and Ppd-D1, respectively. A region was identified on chromosome 7A that harbored major QTL for grain yield, number of grains per square meter, number of grain per spike and spike fertility under drought stress. For yield data in the irrigated trial, two major QTL were identified on chromosome 3B which were not detected in drought stress environments. By using different datasets in the QTL analysis (splitting the population into two subpopulation based on heading time and also adjusting the phenotypic data for heading time to eliminate heading time effect), a QTL for grain yield was consistently detected on chromosome 7A in drought-affected environments. The coincidence of a drought response index QTL on this chromosome indicated that it might be a QTL for yield response under drought. This study demonstrated that the region on the long arm of chromosome 7A identified for grain yield and yield components is a drought response QTL which is closely linked to, but separate from, a heading time QTL. This QTL cluster on chromosome 7A could be used as a good target for positional cloning and gene isolation. However further work would be required to confirm and validate the identified QTLs in this preliminary QTL analysis. / http://proxy.library.adelaide.edu.au/login?url= http://library.adelaide.edu.au/cgi-bin/Pwebrecon.cgi?BBID=1340056 / Thesis (Ph.D.) -- University of Adelaide, School of Agriculture, Food and Wine, 2008

Bibliography: leaves 203-212. / xiii, 212 leaves : ill. ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / This study investigated the genetic diversity for tolerance of durum wheat (Triticum turgidum L. var durum) to micronutrient deficient soils with an emphasis on manganese. 69 genotypes were studied under field conditions at Marion Bay (Lower Eyre Peninsula) and Coonalpyn. Durum genotypes, notably Stojocri, were identified as having higher tolerance than commerical durum varieties. / Thesis (Ph.D.)--University of Adelaide, Dept. of Plant Science, 1999

This study comprised three major parts: a comparative physiological study of drought responses under controlled conditions; a genetic study to construct the skeleton map of a doubled haploid (DH) population; and a quantitative trait loci (QTL) analysis to identify QTLs associated with drought tolerance traits in the field. In the first part (Chapter 3), three cultivars of wheat (Triticum aestivum L.) adapted to South Australian conditions were tested for drought tolerance under cyclic drought in growth rooms and glasshouse. Extensive physiological traits, including stomatal conductance, chlorophyll content and fluorescence, ABA content, water status traits (e.g. osmotic adjustment, RWC and leaf water potential), water soluble carbohydrates (WSC) and carbon isotope discrimination (Δ¹ ³C) were measured during experiments. Through these experiments, the drought responses of the three cultivars were physiologically dissected and the likely processes contributing most to drought tolerance were identified. In the South Australian wheatbelt, cyclic drought is a frequent event, represented by intermittent periods of rainfall which can occur around anthesis and post-anthesis in wheat. Three South Australian bread wheat cultivars, Excalibur, Kukri and RAC875, were evaluated in two growth room experiments under cyclic water-limiting conditions. In the first experiment, where plants were subjected to severe water stress, RAC875 and Excalibur (drought tolerant) showed significantly (P < 0.05) higher grain yield under cyclic water availability compared to Kukri (drought susceptible), producing 44% and 18% more grain yield compared to Kukri, respectively. In the second growth room experiment, where plants were subjected to a milder drought stress, the differences between cultivars were less pronounced, with only RAC875 showing significantly higher grain yield under the cyclic water treatment. Grain number per spike and the percentage of aborted tillers were the major yield components that affected yield under cyclic water stress. Excalibur and RAC875 adopted different morpho-physiological traits and mechanisms to reduce water stress. Excalibur was most responsive to cyclic water availability and showed the highest level of osmotic adjustment (OA), highest stomatal conductance, lowest ABA content and most rapid recovery from stress under cyclic water stress. RAC875 was more ‘conservative’ in its responses, with moderate OA, high leaf waxiness, high chlorophyll content and slower recovery from stress. Within this germplasm, the capacity for osmotic adjustment was the main physiological attribute associated with tolerance under cyclic water stress, which enabled plants to recover from water deficit. In the second part (Chapter 4), the genetic linkage map of a DH population including 368 lines, which was developed from a cross between ‘RAC875’ and ‘Kukri’, was constructed. The genetic linkage map consisted of about 500 molecular markers including ~300 DArT (Diversity array technology) and ~200 SSR (Microsattelite markers). In the third part (Chapter 5), Quantitative Trait Loci (QTLs) linked to plant phenology and production traits under irrigated and drought stress conditions were mapped by means of a DH population. To phenotype the population, 368 DH lines were cultivated in two replicates in five environments (three sites across South Australian wheatbelt in collaboration with Australian Grain Technology (AGT) in 2006, and two trials in Mexico in collaboration with CYMMIT, 2007). Data of grain yield, yield components, maturity related traits and some morpho-physiological traits such as leaf chlorophyll content, leaf waxiness, plant height, peduncle length, flag leaf and spike length were measured. Raw data were then analysed for spatial variation for each single trial using the REML procedure in GenStat (version 6). The DH lines showed significant variation for plant phenology, grain yield and yield components under irrigated and drought stress conditions. QTL analyses were performed using QTLCartographer and QTLNetwork for each trait in each site. Two major QTL for maturity traits were identified on chromosomes 2BS and 2DS corresponding to Ppd-B1 and Ppd-D1, respectively. A region was identified on chromosome 7A that harbored major QTL for grain yield, number of grains per square meter, number of grain per spike and spike fertility under drought stress. For yield data in the irrigated trial, two major QTL were identified on chromosome 3B which were not detected in drought stress environments. By using different datasets in the QTL analysis (splitting the population into two subpopulation based on heading time and also adjusting the phenotypic data for heading time to eliminate heading time effect), a QTL for grain yield was consistently detected on chromosome 7A in drought-affected environments. The coincidence of a drought response index QTL on this chromosome indicated that it might be a QTL for yield response under drought. This study demonstrated that the region on the long arm of chromosome 7A identified for grain yield and yield components is a drought response QTL which is closely linked to, but separate from, a heading time QTL. This QTL cluster on chromosome 7A could be used as a good target for positional cloning and gene isolation. However further work would be required to confirm and validate the identified QTLs in this preliminary QTL analysis. / http://proxy.library.adelaide.edu.au/login?url= http://library.adelaide.edu.au/cgi-bin/Pwebrecon.cgi?BBID=1340056 / Thesis (Ph.D.) -- University of Adelaide, School of Agriculture, Food and Wine, 2008

Doctor of Philosophy / The responses of 92 barley genotypes to selected P. hordei pathotypes was assessed in greenhouse tests at seedling growth stages and in the field at adult plant growth stages to determine known or unknown resistances. On the basis of multipathotype tests, 35 genotypes were postulated to carry Rph2, Rph4, Rph5, Rph12, RphCantala alone or combinations of Rph2 + Rph4 and Rph1 + Rph2, whereas 52 genotypes lacked detectable seedling resistance to P. hordei. Five genotypes carried seedling resistance that was effective to all pathotypes tested, of which four were believed to carry uncharacterised resistance based on pedigree information. Field tests at adult plant growth stages indicated that while 28 genotypes were susceptible, 57 carried uncharacterised APR to P. hordei. Pedigree analysis indicated that APR in the test genotypes could have been derived from three different sources. The resistant responses of seven cultivars at adult plant growth stages were believed to be due to the presence of seedling resistance effective against the field pathotypes. Genetic studies conducted on 10 barley genotypes suggested that ‘Vada’, ‘Nagrad’, ‘Gilbert’, ‘Ulandra (NT)’ and ‘WI3407’ each carry one gene providing adult plant resistance to P. hordei. Genotypes ‘Patty’, ‘Pompadour’ ‘Athos’, ‘Dash’ and ‘RAH1995’ showed digenic inheritance of APR at one field site and monogenic inheritance at a second. One of the genes identified in each of these cultivars provided high levels of APR and was effective at both field sites. The second APR gene was effective only at one field site, and it conferred low levels of APR. Tests of allelism between resistant genotypes confirmed a common APR gene in all genotypes with the exception of ‘WI3407’, which based on pedigree information was genetically distinct from the gene common in ‘Vada’, ‘Nagrad’, ‘Patty’, ‘RAH1995’ and ‘Pompadour’. An incompletely dominant gene, Rph14, identified previously in an accession of Hordeum vulgare confers resistance to all known pathotypes of P. hordei in Australia. The inheritance of Rph14 was confirmed using 146 and 106 F3 lines derived from the crosses ‘Baudin’/ ‘PI 584760’ (Rph14) and ‘Ricardo’/‘PI 584760’ (Rph14), respectively. Bulk segregant analysis on DNA from the parental genotypes and resistant and susceptible DNA bulks from F3 lines using diversity array technology (DArT) markers located Rph14 to the short arm of chromosome 2H. Polymerase chain reaction (PCR) based marker analysis identified a single simple sequence repeat (SSR) marker, Bmag692, linked closely to Rph14 at a map distance of 2.1 and 3.8 cM in the populations ‘Baudin’/ ‘PI 584760’and ‘Ricardo’/‘PI 584760’, respectively. Seedlings of 62 Australian and two exotic barley cultivars were assessed for resistance to a variant of Puccinia striiformis, referred to as BGYR, which causes stripe rust on several wild Hordeum species and some genotypes of cultivated barley. With the exception of six Australian barley cultivars and an exotic cultivar, all displayed resistance to the pathogen. Genetic analyses of six Australian barley cultivars and the Algerian barley ‘Sahara 3771’, suggested that they carried either one or two major seedling resistance genes to the pathogen. A single recessive seedling resistance gene, Bgyr1, identified in ‘Sahara 3771’ was located on the long arm of chromosome 7H and flanked by restriction fragment length polymorphism (RFLP) markers wg420 and cdo347 at genetic distances of 12.8 and 21.9 cM, respectively. Mapping resistance to BGYR at adult plant growth stages using a doubled haploid population derived from the cross ‘Clipper’/‘Sahara 3771’ identified two major QTLs on the long arms of chromosomes 3H and 7H that explained 26 and 18% of total phenotypic variation, respectively. The QTL located on chromosome 7HL corresponded to the seedling resistance gene Bgyr1. The second QTL was concluded to correspond to a single adult plant resistance gene designated Bgyr2, originating from cultivar ‘Clipper’.

Title from handwritten title supplied by author, on folder cover sheet. Includes bibliographical references A classification of South Australian wheat varieties. Reprinted from: Bulletin / Dept. of Agriculture of South Australia. no. 266 (1932), pp. 1-19 -- Cytology of wheat x rye hybrids of the 5th and 6th generation. Reprinted from: The Australian journal of experimental biology and medical science, vol. xi (1933), pp. [123]-137 -- Cytology of rye (work carried out under the Ernest Ayers Research Scholarship in Botany at the University of Adelaide, 1931). Original typescript. 15 leaves + 4 leaves of plates