More plausibility lay in the etymology of geréfa adopted by Spelman; this rested upon the assumption that geréfa was equivalent to gereáfa, and that it was derived from reáfan, to plunder; this view was strengthened by the circumstance of the word being frequently translated by _exactor_, the levying of fines and the like being a characteristic part of a reeve’s duties. But this view is unquestionably erroneous: in the first place geréfa could not have been universally substituted for the more accurate ggereáfa, which last word never occurs, any more than on the other hand does réfan for reáfan. Secondly, an Anglosaxon geréfa, if for gereáfa, would necessarily imply a High-dutch garaupjo, a word which we not only do not find, but which bears no sort of resemblance to krávo and grávo which we do find[361]. Lambarde’s derivation of geréfa from gereccan, _regere_, may be consigned to the same storehouse of blunders as Lipsius’s _graf_ from γράφειν. Again, as words compounded with _ge-_ and ending in _-a_, often denote a person who participates with others in something expressed by the root, geréfa has been explained to be one who shares in the roof, _i. e._ the kings roof: and this has been supported by the fact that graf is equivalent to _comes_, and that at an early period the comites are found occupying the places of geréfan. But a fatal objection to this etymon lies in the omission of the _h_ from geréfa, which would not have been the case had hróf really been the root. Grimm says, “I will venture another supposition. In old High-dutch rávo meant _tignum_, _tectum_ (Old Norse rǽfr, _tectum_), perhaps also _domus_, _aula_; garávjo, girávjo, girávo, would thus mean _comes_, _socius_, like gistallo, and gisaljo, gisello (Gram. ii. 736)[362].” There is however a serious objection to this hypothesis: were it admitted, the Anglosaxon word must have been gerǽfa, not geréfa for geróefa, that is, the vowel in the root must have been a long ǽ, not a long é, springing out of and representing a long ó. I am naturally very diffident of my own opinion in a case of so much obscurity, and where many profound thinkers have failed of success; still it seems to me that geréfa may possibly be referable to the word róf, _clamor_, róf, _celeber_, _famosus_, and a verb rófan or réfan, _to call aloud_: if this be so, the name would denote _bannitor_, the summoning or proclaiming officer, him by whose summons or proclamation the court and the levy of the freemen were called together; and this suggestion answers more nearly than any other to the nature of the original office: in this sense too, a reeve’s district is called his mánung, _bannum_[363]. In this comprehensive generality lay the possibility of so many different degrees of authority being designated by one term; so that in the revolutions of society we have seen the German markgraf and burggraf assuming the rank of sovereign princes, while the English