New proofs of some results of Nielsen