HOLMES] GROUPS OF POTTERY 21 



the art in past o-enerations, althouoli meagei' enough, are not appealed 

 to in vain, as will be amply shown in subsequent sections of this paper. 



The iirst requisite on taking up the study of a field so extensive and 

 varied is a means of classifying the phenomena. We soon observe 

 that the pottery of one section differs from that of another in material, 

 form, color, and decoration, and that groups may be defined each 

 probably representing a limited group of peoples, l)ut more conven- 

 ientlj' treated as the product of a more or less well-marked specializa- 

 tion area. By the aid of this grouping it is easy to proceed with 

 the examination of the ware, and a reasonably clear idea of the art of 

 the regions and of the whole field may readily be gained. 



First in importance among the groups of ware is that called in 

 former papers the Middle Mississippi Valley group, (xeographically 

 this group presents some interesting featurt>s, which will l)e considered 

 in detail later. The margins of the area it occupies are not well defined, 

 and occasionally pieces of the ware are found far outside its ordinary 

 habitat and associated with strangers. This area has a central posi- 

 tion in the Mississippi valle}^ and other varieties of pottery lie to 

 the north, east, and south, with overlapping and often indefinite out- 

 lines. On the north is the area characterized l)y wari^ to which I 

 have for convenience given the name Upper Mississippi oi' North- 

 western group. In the Ohio valley we have varieties of ware to which 

 local names may be attached. The New York or Iroc[uoian pottery 

 occupies the .states of New York and Pennsylvania, extending in places 

 into other states and into Canada. We have Atlantic Algon((uian 

 ware, South Appalachian ware, and several groujjs of Gulf (.'oast 

 ware. Many of these groups are so clearly dift'erentiated as to make 

 their separate study easy. Within the limits, however, of their areas 

 are numerous subgroups which do not jtossess such strong individu- 

 .ality and such clear geographic definition as the larger ones, but which 

 may well be studied separately and may in time be found to have an 

 ethnic importance quite equal to that of the ])etter-defined groups of 

 ware. Although they are confined to such definite geographic ureas 

 we are not at all sure, as has been pointed out, that these groups of 

 wai-e will be found to have any intelligible correspondence with the 

 .stocks of people that have at one time or another occupied the 

 region, for varieties of art phenomena are often regional rather 

 than ethnic. Besides, many important groups of people have not left 

 great accumulations of art products, and great groups of pi-oducts 

 maj' have been left by comparatively insignificant conmmnities. Sep- 

 arate groiqjs of people may have practiced nearly identical arts, and por- 

 tions of a single peojjle may ha\e practiced very different arts. In 

 view of these and other uncertainties hampering the correlation of 

 archcologic data with peoples, we can not do better than at first 



