The only way I could see that happening would be if the points and condenser were not grounded. There should be zero volts at the negative side of the coil with the points closed, and battery voltage with them open. What you are decribing would be consistant with a twelve volt system with a ballast resistor, except that the opening and closing of the points is backwards and the readings taken from the + side of the coil rather than the -(distributor side).