century did any great divergence of speech exist among those who naturally used French in this Island. The erroneous conception can, however, be accounted for to some extent, especially if we bear in mind that continental Norman was not a homogeneous dialect, and that in phonetic development it lay across the border line of Western and North-Eastern French. Even to-day French chat is pronounced ka (for older kat) in the greater part of Normandy; whereas Fr. chant is pronounced kã (older kant) only along the eastern fringe. On the other hand for Fr. chasse, the pronunciations kash (older kache) and shas (older chace) are very evenly distributed. In the same way Fr. cerise is pronounced shriz or sheriz (older cherise) in the greater part of the province, but for Fr. cercle the pronunciation shekl or sherkl (older chercle) is confined to a few isolated areas.[1] From these examples we can infer that, as a rule, the pronunciation more widely spread in Normandy alone survived in England; or if the rarer pronunciation was also retained, it was restricted to a special meaning, e.g. Engl. 'cant' by the side of 'chant'. It is interesting to note that the form chace (chase) alone occurs in literary texts and became fashionable, while cache continued to find favour among the menial classes; hence the double pronunciation and double meaning in Modern English.
Although the spelling of scribes is a poor guide, there is reason to suppose that the pronunciation soon became more uniform in England than it ever was in Normandy. The phenomenon is not without parallel in European history. At home the Romans spoke a variety of dialects—the divergence of Italian dialects is notorious and in many cases goes back to great antiquity—but they imposed on their vast dominions a language which was practically uniform. Recent history furnishes us with an example not less
