While in the organising of administration the national character of the Celts was respected so far as was at all compatible with the unity of the empire, this was not the case as regards language. Even if it had been practicable to allow the communities to conduct their administration in a language, of which the controlling imperial officials could only in exceptional cases be masters, it undoubtedly was not the design of the Roman government to erect this barrier between the rulers and the ruled. Accordingly, among the coins struck in Gaul under Roman rule, and monuments erected on behalf of any community, there has been found no demonstrably Celtic inscription. The use of the language of the country otherwise was not hindered; we find as well in the southern province as in the northern monuments with Celtic inscription, written in the former case always with the Greek,[52] in the latter always with the Latin[53] alphabet; and probably at least several of the former, certainly all of the latter, belong to the epoch of Roman rule. The fact that in Gaul, outside of the towns having Italian rights and the Roman camps, inscribed monuments occur at all in but small number, is in all probability to be accounted for mainly by supposing that the language of the country, treated as dialect, appeared just as unsuited for such employment as the unfamiliar imperial language, and hence the erection of memorial-stones did not become generally adopted here as in the Latinised regions; the Latin probably may at that time in the greater part of Gaul have had nearly the same position, as it had subsequently in the earlier Middle ages overagainst the popular language of the time. The vigorous survival of the national language is most distinctly shown by the reproduction of the Gallic proper names in Latin, not seldom with the retention of non-Latin forms of sound. The facts that spellings like _Lousonna_ and _Boudicca_ with the non-Latin diphthong _ou_ found their way even into Latin literature, that for the aspirated dental, the English _th_, there was even employed in Roman writing a special sign (Ð), that Epadatextorigus is written alongside of Epasnactus, and Ðirona alongside of Sirona--make it almost a certainty that the Celtic language, whether in the Roman territory or beyond it, had in or before this epoch undergone a certain regulation in the matter of writing, and could already at that time be written as it is written in the present day. [Sidenote: Evidences of continued use of Celtic.]