To determine the greatest possible error for the measurement 0.991 g, we should consider the significant figures. The measurement has three significant figures, which implies that the smallest unit of measurement is in the hundredths place.
When rounding, the greatest possible error is typically expressed as half the value of the smallest increment of the measurement. Since the last significant digit in this case is in the thousandths place, the smallest increment would be 0.001 g.
Therefore, half of 0.001 g is 0.0005 g.
Thus, the greatest possible error for the measurement of 0.991 g is 0.0005 g.