and functional form he chooses. But, then the comments do not go beyond noting that the coefficients are significant and the overall R2 is high; or, at best, that the coefficients are of the right signs. The values of the coefficients are neglected because they are not meaningful. For instance, in the present case, the author observes: "The picture is the same when we consider the cow-bullock ratio instead of the cow density. Other things remaining the same, this ratio tends to be high in areas with high bullock density and to be lower in regions where the calving rate and/or the survival rate of young milch [presumably should be male] cattle is high. ... Interestingly, there is no indication of regions with high bullock density economising on the number of cows they need to maintain by achieving higher calving rate and/or lower death rate of young male cattle" (pp 73-74). The basis of the last observation is not clear. The author does not seem to see that it is disconcerting to his main thesis. For him, it is only interesting. If we tell him that his data do not support his hypothesis, probably he will reply: "Maybe. But isn't it interesting?" Evidently, his is, what is called, a robust hypothesis. A more meaningful analysis of the data should be possible. For instance, the arithmetical relation between the ratio of adult male stock (AMC) and the adult female stock (AFC) and other three factors is as follows:
