The use of neon as a liquid refrigerant for high-temperature superconducting devices is briefly discussed, together with the problems posed for the liquefaction process by the small temperature difference between neon's normal liquefaction and melting points. The paper describes experiments which indicate how a Gifford - McMahon refrigerator can be made self-regulating at a temperature above the triple point of neon, so that it can be employed to condense neon used as a secondary refrigerant, without the danger of solidification.