NEW WORLD PALEO-INDIAN—ROBERTS 421 
the overlying silts and the thickness of their stalagmitic cover. The 
high-water level in Lake Confins, from which the silt-bearing flood- 
waters are believed to have come, at present is 65 feet below the bottom 
of the cave. Throughout historic times there has not been sufficient 
water to produce overflowing into the cave, and consensus is that only 
near the end of the Pleistocene or corresponding pluvial period was 
there adequate precipitation to cause such phenomena as the deposition 
of the silts and the formation of the stalagmite. At present knowledge 
of that period is too incomplete to provide a basis for chronological 
estimates on the age when the pluvial came to an end in the area. Some 
of the animals represented in the faunal assemblage seem to have sur- 
vived into relatively recent times and consequently are not much help 
as acriterion. The Confins man was unquestionably “early” insofar 
as his relationship with subsequent inhabitants of the region is con- 
cerned, but just how ancient he actually was is a problem still waiting 
to be solved. 
The Sambaquis, shell heaps scattered along large sections of the coast 
of Brazil and on the banks of certain rivers in the interior, should be 
_ mentioned, although their value as evidence for an early occupation is 
debatable. There is considerable variation in the size of these forma- 
tions. Some of them measure 300 to 350 feet in length and 50 to 80 
feet in depth. In most cases, however, they are not as extensive and not 
more than 10 to 20 feet deep. Many are located some distance from the 
present shore line. This indicates that they were formed when the 
sea was at a higher level, and for that reason they are regarded as 
being of considerable age. They have been referred to as evidence for 
a fairly early and dense population in the area because many people 
have believed that all were of human origin despite suggestions from 
time to time to the effect that most of the large ones probably were the 
product of natural forces. Recent investigations tend to substantiate 
the latter point of view (Serrano, 1938). None of the shells in the 
larger heaps have been opened, and such traces of human activity as 
have been discovered apparently were intrusions and not contem- 
poraneous material. Various small heaps and occasional portions of 
some of the medium-sized ones, however, consist of kitchen-midden 
deposits or refuse and definitely are man-made. The small heaps lo- 
cated farthest from the sea in the southern provinces, especially in the 
Sao Paulo district, yield the oldest-appearing artifacts. A simple cul- 
ture in strong contrast with that of historic times is represented. Stone 
and bone implements comprise most of the complex. The stone tools 
are as a rule roughly chipped, but they are polished along the edges. 
Potsherds are sometimes found on the surface or in the upper part of 
the top layer of such mounds and the evidence suggests a late pre- 
ceramic group just beginning to acquire pottery. This, in combination 
