It is the purpose of this paper to produce some evidence on the accuracy to be expected from normal assay procedure using the usual first class equipment, and to examine the effect that many assaying inaccuracies have on reported metallurgical recoveries.Generally speaking, the evaluation of the gold content of a mineral ore sample by the fire assay method is a relatively simple process.A weighed quantity of finely ground ore is fused with suitable fiuxes including litharge and a reducing agent to give a button of lead and a quantity of slag. The shower of lead produced by the reduction of the added litharge collects the gold and silver present in the ore, concentrating the values into the lead button.The button of lead, after separation from the slag, is, on heating in an absorbent cupel in a current of air, reoxidised to litharge and absorbed into the cupel leaving a gold and silver bead in the cupel. This bead' is heated in a solution of nitric acid to dissolve the silver, leaving the gold undissolved to be dried and weighed. Knowing the original weight of the ore and the weight of the gold prill the value of the ore is easily calculated.Summarising1. Fusion Process.-Ore plus fiuxes give a slag plus lead button.2. Cupellation Process.-Lead button is re-oxidised to litharge leaving a gold and silver bead.