In their recent paper, Singh & Negi, (This journal, 57, 741-744) contend that if thd slope of the empirical linear relation between heat flow and heat production is interpreted as the decay-length of an exponential depth-distribution of sources, a discrepancy rises, whereas if it is interpreted as the depth of a step distribution, it does not. I should like to point out that their discrepancy follows from their arbitrary assumption of one of a range of physical possibilities unconstrained by the observations; with an equally valid alternate assumption (Lachenbruch 1970) the discrepancy disappears. In any case such discrepancies are probably minor compared to physical difficulties that arise from the step model, and to uncertainties introduced by other assumptions in any simple model.-Author