What may be confusing is that the myelin is wrapped around the membrane of the axon.
The easiest way to see this is in cross-section:
The axon is indicated by #1 in the diagram and the myelin sheath is #4
The intracellular space (as represented by the horizontal resistors in the case of your model of the axon) is the fluid, replete with ions, that is contained within the membrane of the axons.
At the nodes of Ranvier, there is a break in the myelin sheath, exposing the underlying membrane of the axon. In membrane in those gaps of the sheath, there are sodium channels (voltage gated) and potassium channels that facilitate the influx and efflux of ions that "refresh" the action potential as it travels down the axon. The extracellular space is simply the milieu of fluid in the space between the axon and any adjoining cells. For the purposes of analyzing the currents of the neuron, the sodium and potassium ions (and the conductances of the channels) are all that you really need to be concerned with.
What is confusing about your diagram is that it represents the conductance -- the inverse of the resistance of the channels -- of the channels (inwards and outwards) as a single vertical resistor in between the horizontal resistors of the intracellular membrane. This oversimplification is handy when doing the cable theory-type analysis, but it betrays the actual physiology of the neuron.
From a model standpoint, I'm breaking all the rules here, but you can think of things like this:
where one "rung" of your resistor ladder has $r_m$ broken out into the 2 components and drawn differently.