TESTS AT CORAL BAY


at Magens Bay in that they contain more sand, are reddish or
brownish (not black or gray) in color, do not have irregular sur-
faces, and none have the bent inward rims of convex or concave
bowls (Fig. 3). Previously the Hull Series have represented about
10 percent but at Coral Bay they account for 25 percent of the
sherds. On the other hand, sherds of the Magens and Bay Series
are extremely rare.
 Rim points and painted plates are present but rather rare while
Botany or Bay Horned, Lugged, Side Appliqued, Incised or Punc-
tated Lip, Incised Casuela, and Adorned are entirely absent. An
important feature is the presence in rather large quantities of Bot-
any Wide Handled sherds. Rare in our collection, probably be-
cause of surface erosion, but undoubtedly a prominent feature of
the complex are various white-painted Coral Series sherds.
 Three unique sherds should be mentioned. One is the unique
zoned red sherd (P1. XIV, d) described in the typological section.
It came from our surface collection. The other two are thick (2
cm.), coarse, sand-tempered, griddle fragments which exhibit in-
cision on their upper surfaces. Rouse (1952a: 343) lists this as a
trait of the Ostiones period of Puerto Rico. These sherds and the
one Botany Narrow Handled, one Botany Painted Plain, and the
two Hull rim points may indicate brief reoccupation during the
M.agenI, II period.
 The ceramic complex at Coral Bay-vessels of the Coral Series
plus utilitarian Botany Wide Handled, undecorated Hull con-
tainers, and Botany and Hull Griddles-is the one briefly mentioned
earlier under the Magens I period. Due to stratigraphic suggestions
from Maigcn( Bay, particularly those recorded by Hatt (1924. 32-33),
from Cinnamon Bay, and from Coral Bay and the lack at Coral
Bay of pottery types deemed to be relatively late, the culture repre-
sented by the midden at Coral Bay is seen to be the earliest known
ceramic culture of the islands.
 This ceramic complex is almost identical to that called Cuevas
in Puerto Rico (Rouse 1952a: 336-40; P1. 2) and similar to that
called Cedros (Rouse 1947) in Trinidad. The major difference,
apparently, is the lack at Coral Bay of fine cross-hatched incising.
Such sherds may later be found in Hatt's collections at the Danish
National Museum but they are not illustrated or mentioned by