As an example suppose that we seek a 90% confidence interval estimate of the mean number of cracks per 1,000 linear feet in a pipeline. We observe that in 6,000 feet there were a total of 30 cracks so that n = 6 and t_0 = 30. Solving (3) using the spreadsheet software (procedure described below) we found an exact 90% confidence interval with θ_L= 3.496 and θ_H = 6.655, which gave p_1= .063 and p_2 = .037 and satisfied the conditions in (8). Figure 5 shows the geometric representation for the Poisson problem.