This thesis examines the situation of Maori women in Otago District between 1874 and 1936. Their situation is indeed unique in that there was a higher level of intermarriage during this period than in the North Island. The focus is on population decline and its recovery, health, alienation of the land, poverty, education and what effect all these had on the women. It has been extremely difficult to expand further because of limited sources of information, mainly the Appendices to the Journals of the House of Representatives and pieces of information scattered in official reports. Oral history interviews have been combined with appropriate photographs to illustrate the issues raised. The conclusion rested on the fact that a lot of these women were little different from North Island Maori women at this period, despite the high level of intermarriage. This is because they lost most of their land and became economically marginalised like the North Island Maoris.

Alluvial gold placers in Otago and northern Southland occur at several stratigraphic horizons within the Late Cretaceous - Recent sedimentary sequence. The gold is derived ultimately from primary sources in the Otago Schist. Poor correlation between the distribution of placers and that of the known primary deposits reflects repeated recycling of gold in the present drainage network and two precursor networks, each with substantially different architecture. The previous drainage networks were inundated and buried during marine or lacustrine transgression. There has been local addition of first cycle gold and immature detritus during each recycling phase. Most of the placer deposits are fluvial in origin but colluvial placers occur locally along the margins of several Central Otago ranges. Aeolian placers and marine placers are rare. Most of the gold in placers north of the Caples/Torlesse Terrane boundary is Au-Ag alloy, except in vicinity of the Hyde-Macraes Shear Zone, where α-Au-Ag-Hg alloy is also present. Conversely, α-Au-Ag-Hg alloy dominates in placers south of the Caples/Torlesse Terrane boundary, except those whose headwaters lie, or lay, in the Torlesse Terrane. These systems that cross the terrane boundary contain Au-Ag and Au-Ag-Hg alloys in the reach downstream of the boundary, and placers with their source in Aspiring Lithologic Association also contain both alloys. Textural and compositional maturity of the placer host generally increase with decreasing age, reach maxima at the Waipounamu Erosion Surface, and generally decrease in tandem with age above this surface. Exceptions occur in northern Southland, where quartz pebble conglomerate placers are actively forming. Fluvial quartz pebble conglomerate placers have not formed in a single sedimentary cycle. Instead, they have formed from precursor sediment where a high water table drives alteration of the labile component, and when uplift and erosion rates, topography and stream gradients are all sufficiently low to drive sedimentary recycling without significant input of low-grade basement detritus. The maturity of pre-marine examples (Taratu, Papakaio and Hogburn Formations) was enhanced by wave-reworking before final inundation during marine transgression. Colluvial placers in alluvial fans at the margins of Quaternary schist antiforms are repeatedly recycled into younger fan sediments during range growth. These ranges grow in width, as well as length and height, at the expense of the intervening basins, which become progressively narrower. The colluvial placers are ultimately reworked into a fluvial placer in an axial river between two ranges, which concentrates all of the colluvial gold into an incised channel once the widening ranges meet. Aeolian placers have formed from fluvial precursors in the semi-arid parts of the rain shadow east of the Southern Alps, particularly on the lower slope of ranges exposed to westerly winds. Silcrete and less common greywacke ventifacts are commonly associated with these placers. Progressive changes in gold particle shape by flattening during transport in fluvial systems has been the most important process in the concentration of gold in placers. Flattening changes the hydrodynamic behaviour of gold particles by increasing their surface area to volume ratio, thereby making them easier to entrain and enabling transport to lower energy parts of the fluvial system. Gold particle flatness determines whether transport or concentration occurs and there is a predictable relationship between particle flatness and transport distance. This relationship explains the typical occurrence of placers immediately downstream of terminal moraines, the confluence with steeper tributary streams and the mouth of incised gorges. In each case, gold with sub-critical flatness is deposited from a higher energy system or reach into a lower energy system or reach, and must be flattened to a critical state before further transport can occur. Chemical mobility of gold in groundwater occurs during uplift, commonly in association with sedimentary recycling. Secondary gold overgrowths are common in some placers and stitch or overgrow transport-induced features such as folds and abrasion marks. Chemical mobilisation and re-precipitation of gold is of minor importance, however, and results in volumetrically insignificant amounts of secondary gold. Increases in gold grain size upward through the section on both the local and the regional scale does not result from gold 'growth', but instead from preservation of progressively more proximal reaches of the host placers with decreasing age.

Abstract 150 years ago, the carefully-planned Presbyterian settlement of Dunedin was torn apart by the discovery that nearly every stream in Otago was laden with gold. The population exploded, adding the accents of Greece, Tipperary, Victoria, California, Guangdong and the King Country to the Scots burr which had been predominant. Almost immediately a myth of identity emerged, typified by goldfields balladeer Charles Thatcher’s ‘Old Identity and New Iniquity’ and boosted by the histrionics of a press enamoured of the romanticised machinations of the Otago goldfields ‘digger’. This popular mythology conflates the imagery of California, Victoria and early Gabriel’s Gully to perpetuate stories of desperate, gold-mad miners swarming across the province fighting, drinking and whoring away sparse winnings in a vast and lawless land, where bodies float down the Clutha, diggers battle corrupt police and vast fortunes are won and lost. This thesis seeks to construct a de-mythologised account of the rush for Central Otago gold, examining the engineering processes, social dynamics and communal relationships implicit in the development of claims, the construction of goldfields structures, the growth of towns and the emergence of financial networks. This explains and reveals the social, technological and economic developments of the gold rush that wrought a profound change on the Otago landscape and to New Zealand’s history. Focussing on the New Zealand Department of Conservation’s historic reserve at Bendigo as an exemplary site, this thesis focuses on the people of the goldfields who left traces of themselves in archives, letters, newspapers, court records and in the heritage landscape to explain their mining, commercial and family lives, and concludes by exploring the remnants of their existence in the relic-strewn ghost-town. By elucidating the depth and breadth of relationships, processes and lives of the residents, miners and merchants, I refute the pervasive myth of innocent simplicity around the era to replace it with a surprisingly complex reality. This complexity is revealed in the new conclusions I draw around the myriad processes behind identity formation, rush events, water race construction, quartz mine development and labour relations, merchant finances and heritage remnants.

This thesis is an attempt to assess the degree of differentiation between two regions, Otago and Southland, to be found in the styles of four types of artefact; Bird-spear points, One-piece fish-hooks, Composite hook points, Adzes. In order to assess the significance of these differences the comparison has been made not only between the two regions mentioned above but also with a set of samples from the northern North Island used as a bench-mark. The data has been collected in the form of non-metrical (presence/absence) and metrical (continuous or ratio-type) variables specific to each artefact type. The method of analysis of the data is concerned with the study of the relative frequencies of these ranges of variables. This is supported by Chi� and Student�s T tests. As well as seeking to establish the degree of differentiation between the material cultures of the regions the interpretation also seeks to distinguish between causal factors for these differences (for example, variations in functional requirements, differing or limited access to material types, etc.).

Summary: Waikouaiti is a small village and seaside resort thirty-two miles north of Dunedin. One mile from the township is the beautiful beach stretching for three miles from the headland of Matanaka in the north to the Merton River at the southern extremity. No one who walks the beach at the present time would realise that this was the scene of a seige during the Maori Wars in the South Island, or that the beach and the banks of the Merton River was the site of a whaling station from 1835 to 1845. What is now called Waikouaiti is not the Waikouaiti of early days, which was the name given to the whaling station at the mouth of the Waikouaiti River. The name of the river has also been changed to the Merton River. The word Waikouaiti has been spelt in various ways. Some maintain it should be Wai-kau-iti (water become little) or Wai-ko-wai-iti (water, the end of the little stream). Tamati Parata, an old chief at Puketeraki, near Waikouaiti considers that Waikouaiti (properly Waikowai-iti) means the end (Ko) of the water or stream (wai iti running into water or sea (wai). The name Merton is now given to this locality at the mouth of the little stream, now the Merton River. The present spelling Waikouaiti was fixed in 1843, but before that date it was spelt Whikowhiti, Whykowite, Waikoaite, Whykowat, Waikooti, Waikowaiti, and Whycauity.

Little is known of the biology of the parasite Dendrogaster (Crustacea: Ascothoracida), an obligate endoparasite of sea stars. There are some 25 species of Dendrogaster reported worldwide. The highly modified females live within the coelomic cavity of the host, and heterochronous males are parasitic within the mantle cavity of the female. Two sea star species from the Otago Coast were found to be infected with two species of Dendrogaster. These parasite populations were studied from July 1993 to December 1994 to investigate specific aspects of their biology and the interactions between these parasites and their hosts. The parasite found in Allostichaster insignis (Asteridae) was identified as Dendrogaster argentinensis Grygier and Salvat 1984, and Asterodon miliaris (Odontasteridae) was found to contain a previously unknown species, herein described and named Dendrogaster otagoensis n. sp. The structure and function of the mantle of female Dendrogaster sp were determined using light, scanning and transmission electron microscopy. The ultrastructure of the mantle is modified, sharing characteristics found in Rhizocephala (Crustacea: Cirripedia), copepods, and Ulophysema oeresundense (Ascothoracida: Dendrogasteridae), which enable these crustaceans to absorb nutrients. Such modifications include a very thin, crenulated epicuticle, a soft, unscleratinised procuticle, and infoldings of the apical plasma membrane. In Dendrogaster the gut is likely to retain the capacity to absorb nutrients. In general, aspects of infestation of Dendrogaster argentinensis and D. otagoensis are the same. Prevalence of Dendrogaster sp. in Otago sea stars is high and parasitism occurs throughout the year. Seventy-seven point five percent of Asterodon miliaris are infected with D. otagoensis and 97.4% of Allostichaster insignis are infected with D. argentinensis. Up to 144 females are found in a single host, and as many as 34 males in a single female. Mean number of parasites per host changes little over time, but began to decrease in D. argentinensis towards the end of the sample period. The female population is dominated by small individuals, with maximum size likely to depend on space available within the host. Males achieve a large size quickly, but their growth is also restricted by the size of the female they are within. The size and number of females per host are not related to size of the host. Females acquire new males and males grow larger with an increase in female size. Males occur in females of all reproductive stages, but occur with the greatest frequency in brooding females. Reproduction is spread evenly through females in less heavily infected hosts, and there is a reduction in mean size of females as the number of females per host increases, indicative of the crowding effect. Three indices of aggregation show that both sexes of Dendrogaster are highly aggregated in their hosts, with few hosts harbouring most of the parasites. Dendrogaster is a parasitic castrator. Infected hosts have parasite loads ranging from 0.1% - 112% total sea star wet weight. Infected sea stars have smaller pyloric caecae and gonads, however, no gonads were present in uninfected Allostichaster insignis. Effect on the gonads and pyloric caecae increases as parasite load increases. Rather than directly affecting the germinal tissue of the host, Dendrogaster castrates its host indirectly through crowding and/or competitive castration. This castration is variable among hosts in Asterodon miliaris, but is important in Allostichaster insignis. The presence of D. argentinensis in A. insignis is associated with an increase in the incidence of asexual reproduction among larger (R>25mm) individuals. Specific growth rates of intact and regenerating arms in fed and starved A. insignis are lower in infected individuals. Larvae are brooded in the mantle cavity of the female until reaching the infective, swimming ascothoracid II stage. The larval development of Dendrogaster argentinensis and D. otagoensis is abbreviated. Larvae moult directly from the egg into the metanauplius. Metanauplii moult into ascothoracid I stage, which moult into the ascothoracid II stage immediately upon release from the female. Ultrastructure confirms the putative sensory function of the lattice organs. Other larval cuticular structures are also sensory. Histological examination revealed that Dendrogaster has unique system of ovulation among Crustacea. The ovarian membrane contains primary oocytes, early vitellogenic and vitellogenic oocytes, but not maturing oocytes, which are found outside the ovarian membrane. The ovarian membrane forms follicles from which vitellogenic oocytes burst into the surrounding mesodermal tissue. These oocytes mature within the mantle wall tissue until entering the mantle cavity, where they are fertilised. Ovaries are absent in most brooding females. Spermatogenesis takes place within the males posterior protrusions and is continuous. Dendrogaster are likely to be labile hermaphrodites, as sperm are found in 20% of ovaries. Female Dendrogaster display asynchronous reproduction, with females of all sizes at different stages throughout the year. Females within one host are also at different stages. Spermatogenesis is present in males of all sizes. The presence of spent males in some females is indicative of a females outliving the males.

This research has examined the historical vegetation of two Eastern Otago sites below the regional treeline, with the aim of addressing questions about the distribution and spread of native tussock grasslands prior to human arrival in New Zealand c. 800 yr BP. Pollen and phytolith (plant opaline silicate) proxies have been extracted from peat cores at Swampy Summit and Clarks Junction to provide a record of vegetation spanning the Holocene. Using multiple proxies and two sample sites has allowed for comparisons of the record of vegetation from within sites and between sites. A record of the modern pollen rain was also gathered from localised moss polsters at Swampy Summit in order to reconcile modern pollen assemblages and transport patterns with historical findings. It became clear from the research that the record of vegetation inferred from phytoliths was not analogous to the pollen-based records, which supported the hypothesis that vegetation reconstructions based solely on pollen may be unreliable. Good pollen preservation in the sediments allowed for the identification of over 50 taxa, although only Chionochloid forms were identifiable to a family level in the phytolith records. Poaceae pollen was abundant throughout the Clarks Junction record, suggesting grassland had persisted at this site during the Holocene, while Poaceae pollen at Swampy Summit was minimal and sporadic. Phytoliths at Swampy Summit show grasses have persisted at the site thoughout the Holocene, at times in much greater proportions than the pollen record would suggest, while Chionochloid phytoliths only become common near the top of the record, possibly reflecting increasing dominance of this taxa after human disturbance. In contrast to Swampy Summit, the Clarks Junction phytolith record reflects a more stable presence of grasses throughout the Holocene, with Chionochloid forms present throughout. Phytoliths appear to be a more reliable proxy for local vegetation, with both sites indicating a Holocene presence of grasses below the regional treeline prior to human arrival in New Zealand. In comparison, the pollen record appears to indicate a more regional pattern of vegetation, with the grassland pollen record complicated by pollen dispersal and deposition factors.

This thesis is an attempt to assess the degree of differentiation between two regions, Otago and Southland, to be found in the styles of four types of artefact; Bird-spear points, One-piece fish-hooks, Composite hook points, Adzes. In order to assess the significance of these differences the comparison has been made not only between the two regions mentioned above but also with a set of samples from the northern North Island used as a bench-mark. The data has been collected in the form of non-metrical (presence/absence) and metrical (continuous or ratio-type) variables specific to each artefact type. The method of analysis of the data is concerned with the study of the relative frequencies of these ranges of variables. This is supported by Chi� and Student�s T tests. As well as seeking to establish the degree of differentiation between the material cultures of the regions the interpretation also seeks to distinguish between causal factors for these differences (for example, variations in functional requirements, differing or limited access to material types, etc.).

Mineralised shear zones in Otago are often truncated by regional low-angle faults, which juxtapose schist of different metamorphic grade. The Footwall Fault and the Hyde-Macraes Shear Zone are one example for this kind of tectonic setting, and are the subject to this study. Although, the mechanisms for the development of the mineralised thrust-origin shear zones are well studied, the relationship to the truncating faults is still poorly understood. Currently, the truncating low-angle faults are assumed to be related to crustal extension, starting in the early Cretaceous after the schist passed the ductile-brittle transition. This study presents new kinematic data for the Footwall Fault, suggesting development of normal sense movement under ductile conditions due to an abundant shear band cleavage in the footwall, which dynamically recrystallises quartz grains. However, brittle high-angle normal faults truncating these shear bands indicate either reactivation of normal sense movement after passing the ductile-brittle transition or continuous normal sense movement during the transition. Furthermore, this study presents a model, which suggests a regional scale rolling hinge development, consisting of an array of individual low-angle normal faults along the boundary of the textural zone change from TZ IV to TZIII, and strike-parallel high-angle faults at the NE margin of the Otago schist.

The deformed Permian Dun Mountain Ophiolite Belt (DMOB) forms the basement of the Dun Mountain-Maitai terrane and is traceable through the entire length of New Zealand. The DMOB contains a variably serpentinised mantle portion and a crustal portion containing gabbros, dolerites, cross cutting dikes and extrusives, together they are similar to oceanic crust. The initial crustal portion, however, is atypical when compared to other ophiolites, being thin and lacking a sheeted dike complex, but has well spaced inclined intrusive sheets and sills. At least four post-Permian deformation periods affect the DMOB; collision and rotation during emplacement of the DMOB on the Gondwana margin, compression during Mesozoic orogenies, extensional deformation during the Gondwana break-up and transpressive deformation related to the modern plate boundary through New Zealand. Structural work in the Northern Bryneira Range focused on well preserved outcrops to investigate crustal growth and contemporaneous deformation during the Permian. Structural evidence of Permian deformation was determined by examination of pseudostratigraphy, structures constrainable to the Permian, and the geometric relationships with the overlying Maitai sedimentary sequence. Crosscutting by intrusive phases was used to determine a chronological order of crustal growth and deformation episodes. It was concluded that all deformation was extensional and that two major phases of magmatism were separated by a period of deformation and were followed by ongoing syn-sedimentary deformation during the deposition of the Maitai Group. After removal of Mesozoic rotation, the resulting orientations of paleo-horizontal markers and diverse orientations of intrusive sheets were analysed. Two hypothesises were tested to assess the origin of inclined intrusive sheets: a) that the diverse orientations were the result of tectonic rotation coeval with the intrusion of dikes. b) that primary orientations of the sheets had been diverse. Results show that the sheets were intruded with diverse orientations, probably related to variation in the principle horizontal stress over time. Further rotation of the assemblage of sheets occurred during the last stages of magmatism and during the subsequent period of sedimentation. The last stage probably relates to large scale normal faulting during the development of the sedimentary basin. iii