Few studies of the abundance of arsenic in seawater have attempted to describe its distributiΒ°n between the arsenite and arsenate species' SΒ°rne recent investigatΒ°rs I 4 report measurable quantities of arsenite in seawater, although thermodynamically arsenate is by far the more stable form 5. Thus, there is some discrepancy between the theoretical calculations of the expected arsenate/arsenite ratio and the measurements which have been made. There have been no reported investigations concerning the mode or rate of arsenic oxidation and/or reduction in seawater aside from the observations of Armstrong and Harvey 6. They state that when sodium arsenite is added to seawater it is oxidized to arsenate, "... rapidly in some waters, slowly in others". The lack of information on this subject may in part be due to the lack of appropriate methodology. The available procedures for examining arsenate/arsenite ratios, at such low concentrations as exist in seawater, are both complex and timeconsuming.
Basically there have been two methods applied in the investigations of arsenic speciation. The modified Gutzeit procedure 7 is reported 8 to suffer from difficulties in the quantitative evolution of small quantities of arsine gas. Furthermore, any reduction of arsenate during gas generation will bias the results towards high arsenite values. The results obtained by Gorgy e t al. 1 with this method appear to have been greatly in error compared with the data of more recent investigators 2 -4.8-13
In the method of Sugawara and Kanamori 14, the molybdenum complexes of phosphorus and arsenic are separated by solvent extraction. The method has the advantage of determining the arsenate and arsenite fractions independently but the experimental procedure is so complex as to render it inefficient for processing a large number of samples. Several attempts to follow this procedure have so far been unsuccessful, for reasons that are not completely understood.
We have developed a relatively simple method, based on the work of Johnson 13, which allows for routine'analysis of arsenite, arsenate, and phosphate by spectrophotometric measurement of the arsenic and phosphate molybdenum blue complexes. As in the Johnson procedure, a reducing reagent is used to lower the oxidation state of any arsenic present to + 3, which eliminates any absorbance caused by molybdoarsenate, since arsenite will not form the molybdenum complex. This results in an absorbance value for phosphate only. The mixed (color) reagent added to an "untreated" sample produces a color from both molybdophosphate and molybdoarsenate formed from any arsenate present. The difference between the two absorbance readings, when corrected for suitable blanks, is proportional to the arsenate concentra-