PROBLEMS IN HISTORY OF GREEK SCULPTURE 613 



guish master from master form of tear-duct, of ear-lobule, or 

 whatever it be may be due to the copyists and therefore valueless 

 for the purpose desired. Indeed, the subjection of these inconspicu- 

 ous details to the law of habit, which makes them useful as identi- 

 fying marks, renders it unlikely that they would be reproduced 

 save in copies of superlative accuracy; and copies of superlative 

 accuracy are unfortunately very rare. Hence that method of con- 

 noisseurship which examines, as one means toward recognizing the 

 individual master, the treatment of inconspicuous details must be 

 regarded as largely inapplicable in dealing with Roman copies, or at 

 least as of dubious probative force. 



Again, the problem of recognizing, whether in originals or copies, 

 the works of a single master is not merely the problem of recognizing 

 decisive similarities. An artist's productions may vary greatly in 

 different periods of his career, or even in one and the same period. 

 If we are trying with our bits of evidence to make out the achieve- 

 ments and so the personal style of a great Greek sculptor, we need 

 a theory as to the limits of the variation which we may in reason 

 attribute to him. How are we to form such a theory? Judgments 

 on this point commonly have an air of a priori dogmatism. Some 

 one proposes to attribute two works to the same artist. The objector 

 says, "No. The differences between the two are too great." No 

 proof is offered, but such a verdict, in spite of its air of intuitive 

 certainty, is doubtless derived more or less consciously from one's 

 knowledge of art and artists generally in the past and in the present. 

 No\v I think that what is needed is a more thorough-going stud}' 

 directed to this very point. The work of artists of modern times 

 lends itself to the purpose. Only when we have satisfied ourselves 

 as to the widest limits of variation shown by any one of them are 

 we in a position to form so much as a legitimate guess as to whether 

 two Greek works are too unlike to have been conceived by a single 

 brain and executed by a single hand. 



Let me illustrate. There exist in Dresden two closely similar 

 Athena figures, one headless, the other with head partially pre- 

 served. By combining, on the strength of convincing proof, a head 

 in Bologna with the headless Dresden figure, and by supplying what 

 else is missing in one from the other, two complete and substantially 

 identical statues have been won. 1 It is argued that in these we possess 

 copies of the Athena Lemnia of Phidias. Certainly the original 

 must have been a work of extraordinary merit and one of the Phidiac 

 age and school. There is some literary evidence, based chiefly upon 

 the absence of a helmet from the head, for believing it to lie by 

 Phidias himself. While this external evidence is far from satisfactory, 

 it appears to me to establish a considerable probability that the 

 1 Furtwangler, Masterpieces of Greek Sculpture, p 4 ff. 



