200 ISLAND CULTURE AREA OF AMERICA [etii. anx. 34 



the end of the arm. A cross section of an elbow stone near the boss 

 following the elbow band is, as a rule, about the same as that near 

 the pointed pole of an ovate slender collar. The surface of an elbow 

 stone, esjjecially the boss, is generally I'ough, but several examples 

 have the remaining parts finely polished. 



" So close are the general likenesses between the boss and the arms 

 or .shoulders of collars and elbow stones that an identification of the 

 latter with broken collars is most natural. In order to explain minor 

 differences in the two types, it is held by some of those who entertain 

 this opinion that a broken collar has been subsequently fashioned 

 into an elbow stone and its surface redecorated to fit it for secondary 

 use. So radically diiferent, however, are the carvings and symbols 

 on the surfaces of these two types of objects that this conclusion 

 seems unreasonable. 



" Other archeologists believe that the elbow stone is a fragment 

 of a type of collar differently ornamented from any that have been 

 found entire. The resemblances are believed by them to be close 

 enough to indicate identity and the differences are looked upon as 

 special rather than as general characters. 



" The belief that the elbow stone belongs to a distinct type is far 

 from the thought that there is any utilitarian difference between 

 the two classes of objects. All indications tend to show a like use, 

 and that if we could satisfactorily explain the meaning of one type 

 we should be in a fair way to interpret the other. It is, in fact, pri- 

 marily to shed some light on the significance of the stone collar that 

 the author presents the following results of his comparative studies: 



" Elbow stones, like stone collars, may be divided into right- 

 handed and left-handed, or right-armed and left-armed, according 

 to the position of the decorated arm. When an elbow stone is placed 

 so that the panel will show, this feature will be seen either on the 

 right or the left hand, thus determining the designations ' right- 

 handed' and 'left-handed' elbow stones. The significance of the 

 difference in this feature is not known ; it may mean nothing, but it 

 would ajjpear that its very occurrence in both collars and elbow 

 stones has some important bearing on the function of the objects. 



" The style of ornamentation furnishes data for a classification 

 of elbow stones on other grounds. Two distinct varieties of these 

 objects can be i-eadily recognized accordingly as a head, face, or 

 body is sculptured on the outer surface of one of the arms. This 

 sculpture, when it appears, is generally in low relief, and always 

 represents human features, never those of an animal. In elbow 

 stones on which such a sculptured figure does not appear there is 

 always a panel with a shallow, oval, concave pit hollowed in the 

 middle, in which is sometimes a secondary depression, as shown in 

 figure 36 [see pi. 99. B], The arm bearing this panel with its pit cor- 



