GRAGG. — TUE GREEK EPIGRAM BEFORE 300 B. C. 45 



The above table covers the occurrence of these terminations in all 

 epigrams of the early period, whether inscribed or not. 



It is evident from this table that the favorite positions for such ter- 

 minations are j ust before the feminine caesura of the third foot and in 

 the fifth and sixth feet of the hexameter. The number of cases where 

 they stand in other positions is so small that we can hardly say that 

 one position is preferred over another. 



The chief result of the preceding investigation has been to show 

 how little we can really state positively from the material at our com- 

 mand. We know enough about Attic epigrams to be able to say with 

 comparative certainty when we may expect to find Ionic or epic forms, 

 but in all other cases our knowledge is so limited that we can draw no 

 conclusions which would enable us to make corrections or even to jus- 

 tify MS. tradition in those epigrams for which we have not the testi- 

 mony of the stones. Reitzenstein's general statement i®^ ig the best 

 we can do till we have more material to examine. 



Appendix 



The following epigrams are grouped by centuries : in each century 

 the sepulchral epigrams stand first, followed by the dedicatory and 

 epideictic. Within each group the inscriptions precede the epigrams 

 which are preserved in MSS. only. The inscriptions are arranged accord- 

 ing to their provenience to facilitate comparison of dialect. It is quite 

 possible that some epigrams written at the beginning of a century have 

 been wrongly grouped among those belonging to the end of the preced- 

 ing century and vice versa, ^^^ but I trust that no such mistake has been 

 made in any case where it would aft'ect the historical conclusions which 

 have been drawn. I have made no attempt to give all the sources for 

 an epigram nor to indicate all the collections where it may be found, 

 but when an epigram is included in the collections of Kaibel, Preger, 

 Hoffmann, or Bergk, and when it is cited by an author earlier than 300 

 B. c, I have so indicated. In the case of inscriptions the first reference 

 is to a facsimile and the restoration adopted is by the editor of the fac- 

 simile unless it is otherwise stated. Where the reputed author of an 

 epigram is mentioned it is merely for purposes of citation and is not 

 meant as an assertion of the correctness of the attribution. 



"3 See p. 40. 



^^* This is especially likely in the case of Simonides. Doubtless some of the epi- 

 grams attributed to him belong to the sixth century, but there is no way of distin- 

 guishing them from those of the fifth century, and it has seemed simpler to group 

 them all together except such a^ are obviously later than the fifth century. 



