Note that the left-hand side of(1.5) is independent of the choice of Haar measure dg. The general weight formula (that is, without the restrictions (1.4)) is similar to (1.5) except that an extra averaging over classes in the genus of A is necessary [$2]. If A is indefinite in 3 or more variables and is of square-free determinant, then by a well-known result of Meyer [Me] there is only one class in the genus of A, explaining the form of (1.5) in this case. We give a new proof of Siegel’s mass formula for indefinite quadrics (n 1) in 4 or more variables and then observe that this, together with Dirichlet’s class number formula, is sufficient for proving that the Tamagawa number of any special or- thogonal group is 2. The general case of Siegel’s formula then follows from a formal computation ofadelic volumes with respect to the Tamagawa measure, for example as in Weil’s paper [W1]. Our proof is by comparing two methods for counting the asymptotic number of integer points on the intersection of the quadric with a ball in Rm.