A remark on the compositeness criteria in the lee model