In Carl's experiment, the variable that should be kept constant is:
B. the amount of water that goes into each pot
This ensures that any differences in the amount of water that flows out can be attributed to the permeability of the different soil types, rather than variations in the initial amount of water used. Other variables, such as the type of soil, will naturally differ as that is the focus of the experiment.