cleaned vials and jars with Teflon-lined lids. In some cases the liver 
was dissected and stored frozen in solvent-cleaned aluminum foil. 
Cooked samples for chemical analysis . In order to obtain a limited 
evaluation of the effect of cooking on contaminant dose, selected samples 
were subjected to a "standardized" cooking procedure. Based on interview 
results, pan-frying was the most common method of fish preparation. In 
order to minimize variations in cooking, a Teflon-coated electric fryer 
(wok) was used. This device had thermostatic control and a curved bottom 
that allowed minimal volumes of oil to be used. Blank analyses were 
conducted on the cooking oil (Wesson brand), the electric fryer, and the 
utensils used during cooking (Teflon-coated forceps). Preweighed fish 
samples (15-24 g) were placed in 50 ml of oil preheated to 200 +/- 10 C. 
The cooking proceeded for 3-5 minutes, and was halted when the appearance 
of the fish sample indicated complete cooking. The cooked fish was allowed 
to drain, and a cooked weight was then determined. Cooked samples were 
divided between organics analysis and metals assay. Procedures for analysis 
were identical to those used for raw fish (see succeeding sections). 
Trace metals analysis . For trace metal measurement, 0.3-0.5 g dried 
sample was accurately weighed and transferred into a 50 ml Teflon beaker. 
The sample was initially digested on a hotplate after adding 10 ml Ultrex 
HNO^ and covering the beaker with a Teflon cover. This treatment was 
enough to decompose most of the organic matter. To assure complete digestion, 
2 ml Ultrex HNO^ and 1 ml HCIO^ were added to the sample and the digestion 
continued to near dryness. If violent reactions were observed, the sample 
was cooled, an additional portion of HNO^ was added, and then the digestion 
was continued carefully. 
The final sample was diluted with 0.5 M HNO^ for instrumental analysis. 
The lipid content of the sample was not ashed completely and showed as a 
drop of oil on the surface of the diluted sample. The drop was removed to 
avoid interference. The instrumental analysis was carried out by flameless 
atomic absorption spectrophotometry (AA) for Ag, Cd, Cu, and Pb. 
For Hg measurement, 1-2 g wet tissue was accurately weighed and placed in 
glass bottles equipped with glass stoppers. After the bottles were chilled 
in ice water, 2 ml concentrated H 2 S 0 ^ and 2 ml 6 % KMNO^ solution were added 
sequentially to the samples under continuous stirring (Toffaletti and Savory, 
1975). The bottles were then capped and allowed to stand overnight to 
complete the digestion. Mercury was reduced with NaBH. and measured as 
cold vapor. ^ 
Neutron activation analysis (NAA) was conducted by standard comparison, in 
which samples of both known and unknown composition are irradiated together, 
and the elemental concentrations in the unknowns determined by comparison 
with the knowns. National Bureau of Standards reference materials were 
used as standards. 
114 
