IN an article entitled 'The Morality of Killing'1 T. Goodrich seeks to show 'that common sense morality is less clear about killing than at first appears'. Conflicts are cited between, for example, the duty not to kill and the duty to prevent the spread of barbarism (the citizen's dilemma faced with Nazi aggression), the duty not to kill versus the duty to prevent prolonged, incurable suffering (a doctor considering a case for euthanasia), the duty not to kill a foetus versus the duty to save the health and sanity of a mother (a possible case for abortion). Duties to animals (to reduce their suffering, certainly, and not to kill-well, possibly) add further complications. Further, for good measure, Mr. Goodrich raises questions about duties to possible future generations of men and animals, but finds no answers to any of the questions raised. Ethical theories are brought in (two, to be precise, viz. Hare's universalizability thesis and hedonistic utilitarianism) in an attempt to save us from complete scepticism; but their application to these issues, according to Mr. Goodrich, only 'reveals the inadequacies of many philosophical theories about morals.'2 His conclusion is 'All this is rather unsettling. I would like now to give all the correct answers. Unfortunately I do not know them. However it is something to have shown that arguments usually taken as valid are in fact fallacious.'3 I wish to argue that the situation is not quite as desperate as Mr. Goodrich paints it. Ethical theories are available which might have helped him more than those he chooses to discuss. Nor does he play fair with those he does call to his aid.3 Puzzles, however, still remain and it is because of their importance that Mr. Goodrich's paper has some interest. I should like to try to isolate and illuminate what seem to me to be these crucial issues. Mr. Goodrich might have posed his problems in the context of existing ethical theory in several ways. He is not the first to discuss conflicts of duty. Sir W. D. Ross's distinction between 'prima facie duties' and 'actual duties'5 could at least have provided the terminology for talking about the conflicts faced by a citizen in wartime, a doctor considering euthanasia or abortion, etc. Ross does not tell us in detail, of course, how to discover an 'actual duty' between conflicting 'prima facie duties'; but I doubt whether anyone who has appreciated Ross has ever subsequently been tempted to think that any of the common rules such as 'do not kill' or 'eliminate suffering' might be absolute rules, or that even unsophisticated common sense ever seriously thought they were. Furthermore, numerous writers in ethics have made the distinction between subsidiary moral rules and ultimate moral principles, or perhaps the ultimate moral principle.6 The search for the ultimate principle of morality has been one of the primary motives among moral philosophers. One thinks of Mill's 'greatest happiness principle' and Kant's 'categorical imperative', especially its second formulation. Clearly the injunctions not to kill and to eliminate or reduce suffering can qualify only as 'subsidiary moral rules' and not as 'ultimate moral principles'. Had Goodrich seen that they do not even compete for this latter classification and had devoted his enquiry instead to explicating the ultimate principle(s) from which these subsidiary rules derive his results might have been more fruitful. It is not too much to suggest that the principle(s) for which he