A problem posed in an influential textbook is analysed in more detail than is given there. The textbook answer to the problem has been represented as counterintuitive, as is probably the case; however, it is here shown that it depends critically on an assumption. If this is relaxed, then a wide variety of possible answers is available. All the same, if the special assumption made is replaced by one of a number of more realistic ones, then the order of magnitude of the approximation given by the author of the textbook is preserved.