earthquake protection; the idea of an equal risk being borne by the population was accepted on humanitarian grounds.It was later recognized that the risk was not uniform and that zoning was justified for economic reasons and for property insurance, especially for more frequent shocks of lower magnitude.Zoning did not take into account readily differences in the frequencies of incidents. For short return periods, of the order of 1&50 years and therefore spanning the life of a structure, frequencies should not affect zoning. Nor should long return periods do so. It was in the intermediate range, for return periods of 50-250 years, that engineers justifiably could take the time factor into account in design; moreover they would expect to receive numerical evidence to support their assumptions, irrespective of the reliability of the evidence.Tectonically, New Zealand's structure was complex. There was a major fault system across both islands. However, unlike the situation in California, the epicentres were not tied closely to the faults; over much of its length in the South Island the Alpine Fault was remarkably free of earthquakes.Moreover, the pattern of seismicity could vary with time.Good instrumental records had been obtained since 1940, reasonably good records from about 1900 and rather inadequate records, based on felt information, since 1830. For still earlier events, it was necessary to fall back on history, legend and geological evidence. The principal value of geological evidence, however, was to gauge areas of equal risk, since movement on faults could take place by creep and not all faulting was evident on the surface. Apparently stable gaps on a fault system between groups of known earthquakes should be regarded as the positions of likely earthquakes in the future.Recently Dr Warwick Smith, working at the New Zealand Seismological Observatory, had made a thorough survey of instrumental records and had correlated these events with observed intensities of motion. He found that the relationships between intensity, magnitude and epicentral distance varied markedly with the tectonic structure. Near the subduction zone on the east coast of the North Island the intensity decreased linearly with the logarithm of distance, but the slope of the correlation varied with the magnitude. With shallower shocks over much of the country, intensity decreased more slowly, leading to widespread felt effects. However, in the south-west of the South 8t 9Downloaded by [ UNIVERSITY OF EXETER] on [11/09/16].