set up to take evidence on the matter (Shakespear, Walker, and Rowan 1887). The descriptions recorded of sludge impacts could be linked to either a property or a general locality and thus could be mapped. Where it was possible to get estimates of depths of sedimentation or erosion from the evidence these were collated and mapped. It was also possible to infer rates of change from these data, which were also mapped. Interpretation After the dispersed disturbance from prospectors searching for gold, there was the sudden focus of attention to specific sites where gold had been discovered. Shallow shafts were excavated to the contact with palaeo riverbeds. Coarse sediment was left close to the shafts, and fine sediment, which was more likely to contain placer gold, was removed for processing. Riverbeds and banks were also dug up. Rivers were employed as an industrial processing stage to winnow away the fines and to leave the gold. Different techniques were used including cradling, puddling, sluice boxes and ground sluicing. These techniques needed water to a greater or lesser extent. Ground and box sluicing predominated in the northeast of Victoria in areas such as the Ovens River. Puddling was more common in the drier more seasonal rivers in the west, (Figure 6). The different techniques influenced both the volume of sediment being delivered into the river and its size distribution. Puddlers supplied a much finer range of sediments, whilst sluicing tended to have slightly coarser sediment delivered to the stream. Our calculations suggest that 58% of the sludge came from ground and box sluicing (Davies et al. 2018a). The cumulative effect of large numbers of small-scale mining operations massively changed both the area of mining and the rivers downstream. The volume of sediment from mining operations would have, on its own, completely changed the sediment budget of the river systems. An added factor was the change in the size of sediment being delivered during low and medium flows. Sands, silts and clays predominated, rather than the usual mix of fine to coarse sediment. The result of this constant delivery of sediment was that channel beds aggraded. Waterholes that were metres in depth filled up first, and then the channels as a whole started to fill. This loss of channel capacity, sometimes combined with an upstream diversion of water and the loss of vegetation, meant that channels spilled onto their floodplain more frequently. Floodplains aggraded 1 m a-1 on average by 1886, and up to a maximum of 2.4 m (Figure 8). Extensive areas of floodplain were covered, up to 4–5 miles (6–8 km) laterally on the Ovens River. This inundation homogenised the floodplain surface and, in some places, the sludge was enough to kill the surface vegetation. This could be because the rate of deposition was too great, or the sludge baked hard after it was deposited creating a crust that inhibited plant growth. The combination of infilling and reduced vegetation roughness would have resulted Figure 7.  VICProd and VICMine data of primary and placer gold mines for the period 1864–1960. Feature Mining to mud 48 PREVIEW JUNE 2019