While I thank Paul Kavanagh for his response (T.N.M., May 14-20) to my contribution to Odds ‘n’ Sods (T.N.M., April 22-28), I am driven to question his recollection about the accuracy with which drill holes predicted gold content in the gravels of the Klondike area.
In my article, I referred to a graph (see image below) that I prepared to show the variation between estimates and recovery. Each of the 60 or so points on the accompanying graph represents the experience from a dredge during one season. The horizontal axis represents pre-mining estimates of gold content in terms of cents per cubic yard, while the vertical axis represents post-mining experience in terms of the ratio of “recovery over estimate.”
Were the estimates accurate, then, all the spots would lie on a horizontal line running through the figure on the vertical scale, which is clearly not the case. I believe the production estimates referred to by Paul Kavanagh were those that were reached after application of the so-called “management factor.”
The phenomenon shown on the graph does not reflect a fault in the drilling procedure but is an example of the “regression effect” that procedures of geostatistics attempt to correct. Whether or not the application of geostatistics would work in the case of the dimensions of the variations experienced here in placer ground is questionable, at least in my view.
Similar examples of this phenomenon were reported between 1914 and 1920 in the long-defunct Scientific and Mining Review from several operations in the U.S. These examples reflect similar experience when plotted.
In a contribution to the International Tin Symposium in 1974, British engineer D.J. Batzer reported monthly estimates and recoveries from a tin dredging operation in Malaysia. These figures plot to show roughly the same effect.
Brian Hester
Vineland, Ont.
Be the first to comment on "Letter to the editor: Hester vs. Kavanagh, round II"