NEPHRITE CELT 
201 
in with the monoxide bases; which is in accord with Penfield’s 
interpretation of the role of hydroxyl and fluorine in the amphi- 
boles. In all four the amount of CaO is slightly above the ratio 
CaO : MgO =1 ' 3 , The close correspondence in refractive 
indices is noted above. The densities of the three crystallized 
tremolites are somewhat higher than is that of the nephrite, which 
may well be accounted for by the texture of this last. 
The nephrite pebble from Amargoza shows the same general 
chemical characters as does our celt, although the analysis is far 
from good. In the Amargoza pebble the silica is lower and the 
lime is higher; the much higher alumina is probably to be con¬ 
nected with the correspondingly lower magnesia, the magnesia 
not having been completely separated from the alumina by repre¬ 
cipitation— a very common analytical error. The high PgOg 
in the Amargoza nephrite is in accord with Hussak’s observation 
that apatite is an almost unfailing accessory mineral in the speci¬ 
mens from this locality examined by him. None of this mineral 
could be detected in my thin sections, and not a trace of phos- 
phomolybdate was precipitated in my test. This would seem 
to be the essential chemical point of difference between the two 
nephrites. 
Amargoza is mentioned by both Hussak and Da Silva as the 
only known locality in Bahia where nephrite occurs in place, 
as boulders and pebbles, and where large numbers of nephrite 
celts and other artifacts have been found. Its former name was 
« 
Baetinga, used by Hussak, which means “a white object” in the 
native Tupi-Guarani language, referring to the white crust pro¬ 
duced by weathering of the stones. Amargoza lies within the 
great belt of gneiss which traverses Bahia ^ from north to south. 
It lies about 100 miles west of the city of Bahia, and about 30 
miles south of Tapera, on the Paraguassu River. 
The Amargoza nephrite, according to Da Silva and Hussak, 
shows the same green color, dense texture, and toughness as 
does our celt. The chief differences are that the Amargoza rock 
weathers to white and, according to Hussak, contains some apatite 
as a common accessory mineral, about one per cent in that analysed 
by him. In spite of these differences, however, it may be held 
as probable that the Paraguassu celt came from the Amargoza 
7 Cf. J. C. Branner, Bull. Geol. Soc. America, Vol. XXX, p. 234, 1919. 
