Figure 1: Quantized levels of charge
Figure 1 shows as a function of 1/r'. The plot clearly indicates the quantization of charge, because only certain 'levels' in the values of are seen. The problem for the further analysis is assigning a specific number of elementary charges to each of the lines (determining n in equation 8), so that can be determined from . Assigning the quantum numbers is the essence of he whole experiment, because that directly yields the quantization of charge.
A naive approach to this step would be to assign n=1 to the lowest level. However, the magnitude of the elementary charge may very well be so small that the electric field of the capacitor plates would be too weak to overcome the gravitational pull on droplets of the size involved in this experiment. A better approach is to estimate an average of the lowest line, and also to look at the average spacing between the proceeding lines. Since is approximately twice as big as the spacing between the consecutive lines, it is fair to assume that q' on the lowest level corresponds to 2e'. Thus the quantum number 3,4..... can then be assigned to the consecutive lines. By dividing through with the determined quantum numbers, the graph in figure 2 is obtained. In the graph, lines with have been omitted, because uncertainties don't make the lines as distinct as they ideally should be. Thus, even though the lines are not clearly separated, they would, if they were included, contribute to determining the elementary charge without showing any indication of such an elementary charge themselves.
Figure 2: Linear fit yielding the elementary charge
By fitting a line to the points in figure 2 and extrapolating it to , the value is obtained. The main sources of error associated with this number are estimated to be: 2% uncertainty in the voltage across the capacitor plates, 1% uncertainty in the viscosity due to variations in the temperature, 2% uncertainty in the timing of the motion and 2% uncertainty due to the fit itself. Since all these parameters enter the formula for (equation 7) in the first order, the total uncertainty is found by taking:
To get the value for e the value obtained by the fit must be raised to the 3/2 power. This also multiplies the error with a factor 1.5, giving a total error of The following value for the elementary charge is thus obtained:
The experimental value is in agreement with the tabulated value, .