°266 BUREAU OF AMERICAN ETHNOLOGY [Bull, 177 
near the end of the 'Taruma Phase, sites that seriate early in the 
sequence can be eliminated from consideration. This rules out E-25. 
K-20 produced too few sherds to be congruent with its identification as 
a village site. Since the archeological data agree better with the 
map than with the field notes, it must be assumed that the former is the 
more accurate of the two sources. 
If we may conclude that the Taruma villages seen by Farabee are 
located below Manakakashin and above the mouth of the Kassikaityu, 
the choice then lies between Sites H-16, E-17, and E-19. Since the 
two villages are described as “a short day’s journey” apart, E-19 is 
the most logical candidate for Farabee’s “Tohi,” except that there are 
rocks in the river at this point and Farabee specifies “no rocks.” Once 
again, however, the field notes contradict the published map, which 
shows rocks in the vicinity of the village, and again the archeology 
agrees better with the map. The description of E-19 equates in gen- 
eral with Farabee’s description of “Tohi.” He notes that it is on 
“high ground”; E-19 is on a bank 18 to 20 meters above the low water 
level. He estimated the clearing as “5 or more acres”; we estimated 
it as 500 by 200 meters, all of which was probably not cleared at the 
same time. He describes the houses as “near the river side of the 
clearing”; we found the sherd refuse to be toward the front of the 
area of secondary growth. The major point of disagreement is Fara- 
bee’s statement that “the clearing is 200 yards from the river,” whereas 
we judged it to extend to within 10 meters of the edge of the bank. 
This difference can be reconciled in one of two ways: (1) the clearing 
was extended forward after Farabee’s visit, a conclusion quite within 
the realm of possibility; or, (2) Farabee’s occasional looseness in 
estimating distance (cf. pp. 267-268) makes some of his figures 
unreliable. 
The relatively large size of the habitation area at E-19 is in keep- 
ing with Farabee’s (MS., Notebook A) description of the village in 
1914: 
Here there are two houses—two shelters for casava work ete. and a visitors 
house. The two houses are each about 85 ft. diam. and circular. One has side 
wall 4% ft. high—the other comes down to ground like great straw stack. The 
roofs are all made of braided thorny palm leaves. Houses are about 20 ft. high 
with a long central pole extending from above the roof. 
The visitors house is 12 ft. in diameter & same height—6 side posts & one 
central post. The roof comes within 5 ft. of ground & no walls builtin... 
Only a very small bare ground about the houses, about double the area of 
houses. 
If we may tentatively identify Site E-19 with Farabee’s “Tohi,” 
then we are left with two possibilities for the second place, referred 
