684 ' ON THE CHARACTER AND INFLUENCE 



them in being accompanied by their wives. Whether this was the 

 case elsewhere in England, I do not know. I am inclined to think 

 that savagery was no great recommendation, nor heathendom 

 either, to a Christianised female population in those days ; and that 

 the reluctance which would on these grounds interpose itself to 

 prevent intermarriage between Romano-Britons and Saxons, sets 

 up as great an a 'priori improbability against the theory which 

 assumes that such intermarriages did take place, as the difficulty 

 of bringing wives over in the ships of those days sets up in its 

 favour. 



Indeed, on the hypothesis of much intermarriage, the actuality 

 of our Anglo-Saxon language is a very great difficulty. We do 

 speak a language which, though containing much Celtic and a good 

 deal of Norman-French, is nevertheless ' English.' Now we know, 

 from finding cremation urns of the Anglo-Saxon type all over 

 England nearly, that the whole of the country was overrun by a 

 heathen population ; to thus overrun it, this population must have 

 been (relatively at least) numerous : add to the two conditions of 

 heathendom and multitude, which may be considered as proved, the 

 third condition of isolation, which may be considered as matter 

 for dispute, and then the fourth, of this heathendom and isolation 

 lasting from the time of Hengist to that of Augustine, — and the 

 present fact of our language being what it is is explained. 



For proving anything as to the period of which I have been 

 speaking, a period which is rendered prehistoric not so much by 

 conditions of time as by conditions of space — the absence of contem- 

 porary historians having been entailed by geographical and political 

 isolation — arguments of two kinds, literary arguments and natural 

 history arguments, must be employed. Neither the one kind nor 

 the other is sufficient by itself. The empires of the natural sciences 

 and of literature touch at many isolated points, and here and there 

 they lie alongside of each other along lengthy boundary lines. But 

 empires need not be hostile though they be conterminous ; and 

 that the empires of which we have just spoken may be united 

 happily and in a most efficient alliance from wants in common, may 

 be seen from the title-page of that most excellent German periodical, 

 the ' Archiv fiir Anthropologic,' where we have the name of the 

 physiologist Ecker coupled in editorship with that of the antiqua- 

 rian Lindenschmit. The necessity for a combination of the two 



