A series of experiments was performed with a steady-state electric analog simulating a cylindrical segment of the aquifer underlying the plains of the Punjab region of West Pakistan. In most of the experiments recharge was assumed to be from the surface, within a specified radius of influence, and distributed uniformly over the area within this radius. Experiments were made with different anisotropies (ratios of lateral to vertical resistance) so that various possible combinations of aquifer thickness and effective radius or radius of influence and combinations .of lateral and vertical permeability could be included in the models. Flow nets were constructed to show distribution of potential in the vertical section and intersections of stream surfaces with the vertical plane.
The series of experiments in which the screened interval is in the upper part of the aquifer shows that flow decreases and stream tubes shift progressively toward the upper part of the aquifer as anisotropy increases.
Another series illustrates that total yield increases and yield per foot of screen decreases as screen length increases.
The experiments indicate that, under conditions prevalent in the Punjab, the Distance-drawdown method for determining permeability gives results with an error of 10 percent or less provided that at least one piezometer or observation well is within a few feet of the pumped well and that no observation well or piezometer used is more than 100 feet from the pumped well.
Relative traveltime for each of 10 stream tubes is given for three models. Relative traveltimes for one-fourth and one-half the effective radius are given for selected stream tubes. By substituting values for the aquifer parameters, actual traveltimes are computed from the relative-traveltime data.