This essay is itself a tentative opinion; invited as a comment and written in consequent haste, it cannot pretend to the grounding in research that underlie the essays Professors Anthony and McGarity have written. It does not attempt to say how to tell an interpretive rule from a legislative rule-rather, just to indicate that the stakes seem both higher than and different from those that the cases and Professor Anthony have generally discussed, and that a "binding in practical effect" approach will sweep with undesirable breadth. It will have succeeded if it promotes thought about the benefits of receiving advice, and the costs of inhibiting that by proceduralisms-thought that at the moment seems to me missing from the analysis. The courts seem to be seeing particular "trees," individuals who have complaints about the impact on them of advice they do not like. They seem not to be seeing the general advantages of receiving advice, the benefits of bureaucratic regularity, and the ways in which their approach may undercut both. Undoubtedly, it is desirable for agencies to engage in consultation as they develop important interpretations-advice repeatedly given, for example,, by the Administrative Conference of the United States-and indications are that this is often done. Consultation can be helpful to an agency that may wish its advice to be well informed, respected, and understood. But perhaps no Federal Register notice appears, or in some other ways a shortcut is taken past the usual regimes of section 553 or the tertium quid. Imposing the procedural requirement in the latter terms will have consequences beyond promoting good practice. Undoubtedly too, cases will remain in which courts will conclude that legislative rulemaking is required for work that an agency is trying to accomplish by publication rules. The Supreme Court's enigmatic decision in Morton v. Ruiz may be an example of a case in which the agency had not sufficiently employed legislative rulemaking to permit the use of publication rules; or it may reflect, more simply, a failure of adequate publication of what might have been a proper publication rule (a manual provision embodying the challenged interpretation). In my judgment, however, its most important thrust lies in its insistence that ad hoc decisionmaking by low-level bureaucrats must be avoided. The regrettable and perverse impact of strongly discouraging publication rulemaking would be sharply to diminish the effective resources available to control the exercise of low-level discretion.