ARCHEOLOGICAL INVESTIGATIONS IN MISSOURI 29 



None of these species exists in the Missouri River today, owing to 

 the heavy burden of silt, which renders the stream wholly unfit for 

 shellfish. If a like condition obtained in aboriginal times, as is in 

 all likelihood true, mussels must have been taken out of the small 

 tributary streams, such as Line Creek, which are, or formerly were, 

 generally characterized by clear water and sandy or gravelly beds. 

 In some of these creeks, shellfish still occur, though in greatly re- 

 duced numbers. 



All the species named above were taken presumably for food pur- 

 poses, but under favorable conditions they are also pearl-bearing. 

 Because the suitable watercourses are nearly all quite small, and since 

 at best only a relatively small fraction of mussels grow pearls, it may 

 be questioned whether sliellfish were ever numerous enough in the 

 Line Creek district to have returned a worthwhile yield to possible 

 aboriginal pearl-gatherers. The evidence also indicates that the use 

 of shells for artifacts and ornaments here was almost negligible. 



Several species of small land snails were also found as shells at 

 the Renner site. Since such a situation would meet their normal 

 habitat requirements, it is improbable that they were present because 

 of any possible utility as food. The shells, thin and fragile, were use- 

 less for the making of artifacts, nor would they offer any induce- 

 ment to pearl-hunters. In short, they are probably no index to the 

 food or other cultural habits of the natives. The list includes the 

 following species: Anguis'pira alternata (Say), Htiicodiscus paral- 

 lehts (Say), Polygyra clansa (Say), Polygyra mAdtUineata (Say), 

 Polygyra profunda (Say), and Zonitoides arhoreus (Say). 



POTTERY 



Broken pottery, though not very abundant on the surface of the 

 site, was found underground in practically all parts tested in our 

 excavations. Greatest concentration was noted in and about the 

 garbage-filled pits, with lesser quantities scattered to depths of nearly 

 30 inches throughout the refuse mantle of the old living surface. 

 No complete vessels were recovered, other than two miniature jars, 

 and restorable specimens likewise numbered only two. The descrip- 

 tion that follows is based on a series of 2,827 body fragments and 

 621 rimsherds, together with two restored and four partly restored 

 vessels. It is impossible to state with certainty the number of pots 

 represented originally by the sherd series. From the wide variation 

 seen in the rim fragments in respect to shape, size, decoration, and 

 other particulars, I should consider 400 different vessels a conservative 

 estimate. 



Petrographic analysis has not been attempted, and the present dis- 

 cussion is based chiefly on observations made with a hand lens. With 



