Of these names he has made two lists, the one derived from the names found in ancient charters, and so perfectly trustworthy, the other inferred from existing names of places which appear to be in the same form. The latter list is of course subject to considerable correction and deduction, inasmuch as it depends entirely upon the ancient forms in which these names would appear whether they would come under this category or not. Thus, if a name were anciently Billing_a_ham, it would be "the home of the Billings," while if it were Billingham, it would simply be the home of an individual man called Billing. And in looking through this list, a few names will be found, which a comparison with his own index of place-names shows to be incorrectly assigned. Thus he infers Impingas from Impington in Cambridgeshire, and Tidmingas from Tidmington in Worcester, whereas it appears from his index that the ancient name of the one was Impintun, and of the other Tidelminctun, both being thus from the name of an individual and not of a tribe or family. Sempringham again in Lincolnshire, whence he derives Sempringas, I find to have been Sempingaham, and so used already for Sempingas. I also feel very great doubt about names taken from places ending in _by_, _thorp_, and _toft_, in Lincolnshire and the ancient Denelaga, as being Scandinavian, and given at a distinctly later period. Indeed I have a certain amount of distrust of all names taken from the North of England, in the absence, as far as I know, of any distinct proof in any one case. Northumberland would perhaps be the county to which, as containing the greatest number of such forms, any such doubt would the least strongly apply. Moreover, I do not feel at all sure that _ing_ is not in some cases simply a form of the possessive, and that Dunningland, for instance, is not simply Dunn's land. This doubt is considerably strengthened when the name is that of a woman, as in Cyneburginctun (now Kemerton in Glouc). Cyneburg is certainly a woman's name, and as such could not, I should suppose--though the question is one for more experienced Anglo-Saxon scholars--form a patronymic, in which case Cyneburginctun can only be "Cyneburg's tun." And if it be so in one case, it may of course be so in others. Mr. Kemble's second list, then, requires to be used with a certain amount of caution, though in the main his deductions may be taken as trustworthy.