Non-ohmic comes to mind, but that gives an impression that the conductor never satisfies Ohm's law. Is there a better term which describes a conductor which might satisfy Ohm's law at certain voltages and temperatures, but not at others.
I've heard such things as thermistors called "non-linear resistors":
The idea is that a linear resistor has a linear V-I curve, but a non-linear resistor does not. In the case of a thermistor, it's due to heating, but in the case of, say, a varistor, it could be due to voltage:
Since an "ohmic" conductor is defined as one with constant resistivity, and since all purely elemental conductors exhibit a non-zero temperature coefficient of resistance, all elemental conductors' resistivities change when the current through them changes their temperature.
It follows then, that - since resistivity of elemental conductors isn't invariant and that for a conductor to be ohmic, by definition, its resistivity must be constant - there are no elemental conductors which are ohmic.
A better term which describes a conductor which might satisfy Ohm's law over certain temperature ranges, but not over others, might be: "partially ohmic" or, in some cases, "superconducting."
There are, however, some alloys which come close to being ohmic over wide temperature/current spans with Manganin, I believe, leading the pack.
I think there is no specyfic term that describes materials or circuits which "satisfy Ohm's law at certain voltages and temperatures, but not at others".
"ohmic" and "non-ohmic" are terms describing voltage-current relation.
An ohmic contact is a non-rectifying junction: an electrical junction between two conductors that has a linear current–voltage (I-V) curve as with Ohm's law. Low resistance ohmic contacts are used to allow charge to flow easily in both directions between the two conductors, without blocking due to rectification or excess power dissipation due to voltage thresholds.
By contrast, a junction or contact that does not demonstrate a linear I-V curve is called non-ohmic. Non-ohmic contacts come in a number of forms (p–n junction, Schottky barrier, rectifying heterojunction, breakdown junction, etc.).
I think when you describe material or circuit - you should keep in mind circumstances/conditions and purpose of your description. All depends on context. Are you considering DC or AC? Do you want your description to describe steady state only or should include transient response?
For example - lightbulb in some circumnstances can be described as "ohmic". If current is very low and filament temperature changes can be negligible. However - lightbulb that works in normal (for bulb) conditions is "non-ohmic", because filament resistance changes a lot.
So, if you want to describe some material clearly - you may have to use more words, like "thermistor behaves like ohmic for constant temperatures".
Sorry about my weird examples (nobody considers lightbulb behavior at constant filament temperature or thermistor at constant temperature). I had no better ideas.
First and formost, to avoid confusion, it would not be called a 'conductor'. The term conductor is generally interpreted as a 'device' that either a (for the design) negligable resistance, or at least a small, predictable and ohmic resistance. So, if the range in which it is (low) ohmic fals within the specification of the design, it will just be a conductor. As such, in any other circumstance, such a device would basically be called something else... Like a thermistor, semiconductor, etc. There isn't a general term for it, apart form 'not a conductor' or 'a device not following ohm's law'.
I think, the best method for desribing devices with voltage-current relations that do not follow Ohm`s law is to use the mathematical description (formula) for this voltage-to-current relationship. Well-known example: pn diode with an exponential expression.
However, it is to be noted that, of course, Ohm`s law can and must be applied (in a specialized form): (1) Static resistance for quiescent values: R,st=V/I and (2) Dynamic/differential resistance for small-signal values: r,dyn=v/i