Bose-Einstein condensation of atoms in a trap

We point out that the local density approximation (LDA) of Oliva is an adaptation of the Thomas-Fermi method, and is a good approximation when \ensuremath{\epsilon}=\ensuremath{\Elzxh}\ensuremath{\omega}/kT\ensuremath{\ll}1. For the case of scattering length a\ensuremath{\gtrsim}0, the LDA leads to a quantitative result, Eq. (14\ensuremath{'}), easily checked by experiments. Critical remarks are made about the physics of the many-body problem in terms of the scattering length a. \textcopyright{} 1996 The American Physical Society.