150 BULLETIN OF THE ESSEX INSTITUTE. 



nothing to do with a bow so far as drawing the arrow was 

 concerned. It was important, however, to settle detinitely 

 this question. Derived from early classic times it seemed 

 reasonable to believe that, if it were associated with archery 

 in any way, a representation of it would certainly be found 

 on figures of soldiers or hunters in antique bronzes and 

 marbles. A categorical statement of the objects repre- 

 sented in the hands of these ancient figures would show 

 among other implements, Utensils, weapons, etc., such as 

 the cestus, discus, strigil, shield, spear, sword, cymbal, 

 pipes and even the bow ; and such Ornaments as arm-brace- 

 lets, clasps, etc., all details of the sau dal, and manner of 

 fastening ; and yet an examination of hundreds of these 

 figures fails to show any object remotely resembling the 

 bow-puller. An extended examination of the decoratiou on 

 ancient vases did not reveal any object of this nature. The 

 figures are depicted as holding in their hands various weap- 

 ons, flowers, tablets, branchof a tree,flask, staff,club, jump- 

 ing weight, double flute, oil-jug, rillet, helmet and an infinite 

 variety of other objects, yet no evidence of this implement 

 is found either in the hand or upon the person. Wall paint- 

 ings in Etruscan tombs while showing a variety of weapons 

 do not depict the bow-puller. Surely if this object was 

 associated with man as an implement or was utilized in 

 any way by a soldier, a hunter, or an archer, we ought in 

 some single case to find a trace of it. What more natural 

 than to show the insignia of an archer on the hand, or se- 

 cured to his person ? Yet figures of archers, and fragments 

 of hands in the attitude of drawing the bow have been re- 

 peatedly found and no such applianceas the bow-puller is 

 depicted. Its entire absence in these ancient representa- 

 tions is certainly overwhelming proof, if no other evidence 

 were needed, to show that this object has been wrongly 

 named. 



