The lodes are parallel and sub-parallel over a width of 150m. West Olympus has a strike length of 350m and extends to 225mRL with parallel and sub-parallel lodes over a width of 130m. A nominal 0.3g/t cut-off and minimum 2m downhole width was used to produce sectional interpretations in Surpac.

In the open pit, the closer spaced grade control data was used; away from the pit, data was extrapolated half the drillhole spacing up or down dip. The interpreted surface geology has numerous NW striking late faults; the interpreted sections were projected to these faults and terminated. A total of 9 separate wireframes were interpreted.

Leapfrog software was used to create two nested grade-based shells, at nominal cut-offs of 0.3/t (Low Grade) and 0.5g/t (high grade). Drillhole intersections >0.3g/t were extracted from the database and used to define the mineralisation in the drillholes. To control the shapes of the shells, the centreline from the previous manual wireframe interpretation was digitised into a curved surface, with additional points to honour the intersections from recent Kalamazoo drilling.

This curved surface was used as an anisotropy to allow the program to model around the structural flexure. For Zeus, the mineralisation trends were digitised from the previous manual interpretations and formed into a single dipping surface. Leapfrog was used to create nested grade shells at nominal 0.5g/t (low grade) and 0.8g/t (high grade) cut-offs.

Reverse circulation drilling to industry standards was used to obtain samples at 1m intervals with a rig-mounted static cone splitter, from which 3kg was pulverised to produce a 30g charge for fire assay. The splitter apparatus was cleaned regularly with compressed air via the sample hose between 1m samples and by washing with water at the end of each hole as a minimum. Any 4m composite samples of approximately 3kg were collected with a sampling tube from the 1m bagged RC drill cuttings.

Wet, damp, or dry sample condition was recorded for each metre. Diamond core drilling to industry standards was used to obtain diamond core from which a half core sample between 0.5m and 1.2m length was pulverised to produce a 30g charge for fire assay. Diamond core was logged and sample intervals selected based on the presence and character of mineralisation with minimum and maximum interval lengths of 0.5m and 1.2m respectively.

The core sample interval was marked with a cut line by the logging geologist to define an approximate even distribution of mineralisation on each side. The core was then cut to the line with a standard core cutter and half-core sampled. Both reverse circulation and diamond core samples are sorted at ALS Laboratory in Perth and weights recorded in LIMS. Following drying at 105°C to constant mass, all samples below approximately 3kg are totally pulverised in LM5's to nominally 85% passing a 75µm screen.

The few samples that are above 3kg are riffle split to <3kg prior to pulverisation. The same or similar sample preparation is stated in previous Resource Estimates or otherwise assumed for older pre-Kalamazoo samples. Kalamazoo field QC procedures involve the use of high, medium and low- grade gold certified reference standards inserted at a ratio of 1:20 and crushed feldspar blanks at 1:25 for standard sampling (1m for reverse circulation or 0.5m - 1.2m for diamond core).

For 1m resampling of composited intervals Kalamazoo use high, medium and low- grade gold certified reference standards inserted at a ratio of 1:20 and crushed feldspar blanks at 1:25 A Surpac block model was created to cover the volume of the Mt Olympus and West Olympus deposits, sub-blocked to honour the volume of the wireframes. Au grades were estimated into blocks inside the domain wireframes, constructed during the geological interpretation, using ordinary kriging where kriging parameters were optimised using the Kriging Neighbourhood Analysis option in Supervisor software. 1m downhole composites of Au and S were extracted for each interpreted domain.

Some domains had large populations of composites; others were much smaller. After an inspection of the means of each of the smaller domains it was decided to group the Mt Olympus domains into Group 1, Group 2 and Group 3; and to group all of the West Olympus domains into a single group. Each group of Au assays (except for Domain 22) required top- cutting to reduce the excessive variability.

Top cuts were chosen from inspection of log-probability and mean and variance plots; the top cuts selected gave the best reduction in variability, as measured by the co-efficient of variation, whilst not reducing the mean by more than 5%, except in the case of West Olympus where a single extreme value required severe cutting. Parent cell sizes used were 10 x 20 x 5m with sub-cell block sizes chosen at 2.5 x 2.5 x 2.5m. For the material outside the domain wireframes, an indicator approach was chosen, using a 0.5g/t cut-off.

1m composites outside the domains were set to 1 if their grade was > 0.5g/t, and the indicator value kriged to estimate the proportion in the block as a value between 0 and 1. For reporting purposes, the proportion was converted into a block ore tonnage. The ore tonnage in the block was assigned a grade of 2.2 (Mt Olympus) or 1.6 (West Olympus), these being the mean grade of the composites > 0.5g/t for the two areas. A Surpac block model was created to cover the volume of the Peake deposit sub-blocked to honour the volume of the wireframes.

Gold grades were estimated into blocks inside the domains using ordinary kriging and kriging parameters were optimised using the Kriging Neighbourhood Analysis option in Supervisor software. Variograms were modelled and in general the experimental variograms are poorly- structured and required a normal scores transformation for modelling. Block sizes chosen were 10 x 20 x 5m for the parent blocks and sub-cell blocks at 1.25 x 2.5 x 0.625m.

Due to the small numbers of composites, all domains in Peake were combined into a single estimation domain. A top cut was applied to reduce the variability. It is worth noting that the denser drilling inside the pit returns a higher mean grade than the wider spaced drilling below and along strike of the pit.

Waugh and Zeus Separate Surpac block models were created to cover the volume of the Waugh and Zeus deposits, each sub-blocked to honour the volume of the wireframes. The 1m downhole composites were extracted from the resource dataset and selected by the nested Leapfrog shells. For both, top cuts were chosen from inspection of log-probability and mean and variance plots; the top cuts selected gave the best reduction in variability (as measured by the CV) whilst not reducing mean by more than 5%.