I noticed that in example demonstrating the calculation of the marginal likelihood on page 42, after taking out the common factors, the exponents of the last two theta values (i.e., the number of successes) change from 7 to 8 (see equation 1.38 below)