SummaryInteraction of the stress field of a favourably oriented edge dislocation with the magnetostriction in a ferromagnetic domain wall in magnetite causes the dislocation to act as a potential well for the domain wall. The coercivity of 20 pm. grains, in which the domain structure is considered to be particularly simple, can be explained on this basis if the dislocations are arranged so that the effects of several of them are additive. The required density of edge dislocations is 109 cm-2, which is entirely reasonable; screw dislocations are not effective in magnetite. To explain the variation of coercivity with grain size it appears necessary to assume that the arrangement of dislocations is neither regular nor random but is partially ordered.

The growth in the usage of CT scanners in Australia between 1994 and 2002 is reviewed. The parallel growth in the per caput dose from CT procedures is derived by first estimating the mean dose for broad categories of Medicare Benefits Schedule procedures and then applying these estimates to counts of the number of procedures in each category. It is estimated that the annual dose per capita from CT procedures has increased from about 0.37 mSv in 1994 to about 0.9 mSv in 2002. From 1994–2002 the growth in CT usage has been mainly amongst adults over 35 years of age while usage amongst infants in the 0–4 year age group has remained within 10% of that in 1994.

A nationwide survey of Australian homes was conducted to determine the average annual dose equivalents to the Australian population from exposure to Rn and gamma radiation. The exposure to Rn was measured using solid-state track detectors (SSTD), while the gamma radiation dose was concurrently determined using thermoluminescent dosimetry. Dosimeters were placed in approximately 3,400 randomly distributed homes (representing about 1 in 1,400 occupied dwellings) for 12 mo. The average Rn concentration in Australian homes measured over a year is 11 Bq m-3. Using appropriate conversion factors, the annual average effective dose equivalents to the Australian population were determined to be 0.5 mSv and 0.9 mSv for Rn and gamma radiation exposure, respectively.

Estimates are made of the risks to the Australian population as a function of age and gender for mortality or morbidity for all solid cancers after exposure to radiation. Excess relative risk (ERR) and excess absolute risk (EAR) models are used. The model coefficients are re-evaluated for radiation doses expressed as effective dose using data from the Japanese Life Span Study. Life-table methods are used throughout and the risk measures studied are: the risk of exposure related death, RERD and the risk of exposure related cancer, RERC. Australian life-table data and the agespecific cancer incidence and mortality rates of Australian males and females are taken from recent published tables. No dose and dose-rate effectiveness factor is applied. Sources of uncertainty used to calculate the confidence regions for the estimated risks include the statistical uncertainties of the model parameters and of the extrapolation of the risks beyond the period supported by the epidemiological data. Summary values of the risks are reported as averages of those calculated from the ERR and the EAR models. For males, the mortality risks per sievert range from 14% for 0-9 year age group, 7% at 30-39 years and 4% at 50-59 years. Corresponding values for females are 20%, 10% and 6%. Incidence risks are higher: for males the estimates are 32% for the 0-9 year group, 12% at 30-39 and 5% at 50-59. Corresponding values for females are 56%, 20% and 8%. The 90% confidence regions are about r 50% of these values. Estimates are given for the risks from CT whole-body scanning or virtual colonoscopy which could be used for cancer screening. If used at 3 year intervals and the effective dose per procedure is 10 mSv, then the RERD for males beginning screening at 40, 50 and 60 years is 0.4%, 0.3% and 0.1%, respectively and for females, 0.6%, 0.4% and 0.2%, respectively. RERD estimates for a 5 year interval between screens are about one-third smaller.