The persistent assertion that the _Cartae Baronum_ were connected with, and preliminary to, the _auxilium ad filiam maritandam_ of 1168 is undoubtedly to be traced to Swereford's _ipse dixit_ to that effect. He distinctly asserts that the aid was fixed (_assisum_) in the thirteenth year (1167), that the returns (_cartae_) were made in the same year (1167), and that the aid was paid and accounted for in the fourteenth year (1168).[117] Modern research, however, has shown that the returns were made quite early in 1166, while the youthful Matilda, we know, was not married till October 1168. This throws an instructive light on Swereford's _modus operandi_. Finding from the rolls that the payments made in 1168 were based on the returns in the _cartae_, and not being acquainted with the date of the latter, he jumped to the conclusion that they must have been made in 1167, it being his (quite unsupported) thesis that all levies were fixed in the year preceding that in which they were accounted for on the rolls. Proceeding further, we find him explaining (p. 9) that he omits the aid of 1165, 'quoniam probata summa auxilii propter hoc non probatur numerus militum'. And yet this aid, the last to be taken before the returns of 1166, is of special value and importance for the very purpose he speaks of. It is, indeed, an essential element in the evidence on which I build; and this compels me to discuss the point in some detail. Those who contributed towards this aid either (1) gave arbitrary sums for the payment of _servientes_--whose number was almost invariably some multiple of five--or (2) paid a marc on every fee of their _servitium debitum_. We are only here concerned with those who adopted the latter course. Now let us take the case of those who adopted this alternative in the counties of Notts and Derby, and compare their payments with their _servitium debitum_ as known to us from other sources. PAYMENTS (1165) SERVICE (1166) _marcae_ knights Hubert fitz Ralf 30 30 Ralf Halselin 25 25 Robert de 'Calz' 15 15 Roger de Burun 10 10 In this case there is no doubt as to the _servitium debitum_, for it is ascertained from the _cartae_ themselves. Having then proved, by this test, the exact correspondence of the payments, I turn to the case of Devonshire. PAYMENTS (1165) SERVICE (1166) _marcae_ knights[118]