INSCRIPTIONS ci and the new fact which is supposed to be proved is set roHing, and draws after it other inferences still more uncertain. A possible deduction from the inscriptions, such as the doubling of the Athenian tribute in the year 425, or the transfer of the common treasury from Delos to Athens in 454 (resting only on the circumstance that in this year the quota lists begin), is repeated at second or third hand as a great historical discovery. In the absence of contemporary, we are satisfied with later, evidence ; and the older history of Athens is interpreted by inscrip- tions of the second or third century, and inscriptions of the second or third century are explained by the older history of Athens. Where singular forms of grammar occur only once or twice, e.g. o-ww for o-oWw', or the omission of the article, we are not quite certain how much is to be attributed to the carelessness of the engraver. On the other hand, from the frequent repetition of it, there can be no doubt that the form of the third person plural imperative, -oaOwv for -iaOm', is a real variety of inflexion. The uncertainty in the use of several letters, even in the same inscription, or the inconsistency of the writing and the subject (C. LA. 8, 93, 283), suggests doubts as to the limits within which this undoubtedly valid argument of date may be employed. The considerable differences which occur in the interpretation and reading of the text, often incomplete, as given by various critics, are another element of uncertaint}'. All these are reasons for hesitation. They show that we must not indulge in sanguine or exaggerated language, but must confine ourselves to general results. And general results, when they relate to the history of the past, are by no means to be despised. Though we cannot rewrite the history of Greece out of her stones, is it a small thing to know that inscriptions of the fifth century before Christ confirm and illustrate the great literary works of the same
- Cp. Meisterhans, § 143, i.
