In the formal returns (_cartae_) made to the exchequer in 1166 by the tenants-in-chief (_barones_) of England, of which the official transcripts are preserved in the _Liber Niger_ and the _Liber Rubeus_, we have our earliest glimpse of the organization of that purely feudal host among whom our lands had been parcelled out to be held, as I shall show, by military service. We have, therefore, in them our best starting-point for an inquiry into the origin and growth of military tenure in England. It may be well perhaps, at the very outset, to contrast these _cartae_ of 1166 with those of the Domesday Inquest eighty years before.[40] For the essentially feudal character of the former is at once, by the comparison, thrown into relief. The original returns of the Domesday Inquest were made Hundred by Hundred; those of 1166 were made fief by fief. The former were made by the jurors of the Hundred-court; the latter by the lord of the fief. Thus, while the one took for its unit the oldest and most familiar of native organizations, the other, ignoring not only the Hundred, but even the shire itself, took for its unit the alien organization of the fief.[41] The one inquest strictly continued, the other wholly repudiated, the Anglo-Saxon system. It is consequently worse than lost labour to examine these two inquests, based as they are on opposite systems, and giving us as they do a cross-division as if they were but successive editions of the national register or rate-book. The first point to be considered is this: What was the information which the tenants-in-chief were called upon to supply in these returns? It was _not_, as Dr Stubbs and others have supposed, the amount of 'service' due from each fief to the crown.[42] The information asked for was _the number of 'milites' actually enfeoffed_ by each 'baron' and his predecessors in title, with the number of 'servitia' due from each such 'miles' to the 'baron'. In this distinction, missed by Dr Stubbs, we find the key to the problem. The crown, we shall see, must previously have known the total amount of 'service' due from each fief; but what it did not know, and what it wished to know, was the number of knights' fees which, up to 1166, had been created on each fief. Although there is great diversity in the form of return adopted--a diversity which imparts to the _cartae_ a pleasant flavour of character--it may fairly be assumed that, as in similar cases, they were called for throughout the realm by one uniform writ. If we may deduce the purport of that writ from the collation of those returns which refer to it most explicitly, we must infer that the information asked for was to be given under four heads: (1) How many knights had been enfeoffed before the death of Henry I? (2) How many have been enfeoffed since?