The nuclear wave functions, obtained from a shell model with variable strength of spin-orbit coupling, are used to compute M1 and E2 transition widths. Comparison with experiment is made for Be8, B10, B11, and C12. The agreement is not nearly so good as was that obtained for energy level schemes. The pure M1 transitions are in good agreement with experiment. The values computed for E2 transition strengths are found to be generally low, though about the right order of magnitude. This suggests the need for adding some collective behavior to the model.