A partial answer to a conjecture of B. Simon concerning unique continuation