When was this continuity disturbed? From the earliest times the sea-coast of Germany seems to have been Frisian, and from the earliest times the tribes of the interior seem to have moved from the inland country towards the sea. Their faces were turned towards Britain; or, if not towards Britain, towards France, or the Baltic. I believe, then, that as early as 100 B.C. the displacement of some of the occupants of the Frisian area had begun; this being an inference from the statement of Cæsar, that the Batavians of Holland were, in his own time, considered to have been an immigrant population. From these Batavians have come the present Dutch, and as the present Dutch differ from the Frisians of A.D. 1851, so did their respective great ancestors in B.C. 100--there, or thereabouts. But the encroachment of the Dutch upon the Frisian was but slow. The map tells us this. Just as in some parts of Great Britain we have _Shiptons_ and _Charltons_, whereas in others the form is _Skipton_ and _Carlton_; just as in Scotland they talk of the _kirk_, and in England of the _church_;[2] and just as such differences are explained by the difference of dialect on the part of the original occupants, so do we see in Holland that certain places have the names in a Dutch, and others in a Frisian form. The Dutch compounds of _man_ are like the English, and end in -_n_. The Frisians never end so. They drop the consonant, and end in -_a_; as _Hettema_, _Halberts-ma_, &c. Again--all three languages--English, Dutch, and Frisian--have numerous compounds of the word _hám_=_home_, as _Threekingham_, _Eastham_, _Petersham_, &c. In English the form is what we have just seen. In Holland the termination is -_hem_, as in _Arn-hem_, _Berg-hem_. In Frisian the vowel is _u_, and the _h_ is omitted altogether, _e.g._, _Dokk-um_, _Borst-um_, &c. Bearing this in mind, we may take up a map of the Netherlands. Nine places out of ten in Friesland end in -_um_, and none in -_hem_. In Groningen the proportion is less; and in Guelderland and Overijssel, it is less still. Nevertheless, as far south as the Maas, and in parts of the true Dutch Netherlands, where no approach to the Frisian language can now be discovered, a certain per-centage of Frisian forms for geographical localities occurs.[3]