The Aircraft Reactor Experiment was designed for operation at temperatures in the region of 1500ºF (1100 K) at a power of 1-3 MWt with a fluoride-salt fuel circulating in a heterogeneous core. The moderator was hot-pressed BeO blocks cooled by circulating sodium. The heat produced was dissipated in water through hot liquid-to-helium-to-water heat exchange systems. All sodium and fuel circuit components were made of Inconel fabricated by inert-gas (Heliarc) welding. The system was heated to design temperature by means of electrical heating units applied over all parts of the system. Instrumentation and control of the experiment were fairly conventional. For the most part, standard instruments were modified slightly for the hightemperature application. The reactor system was constructed and operated in a building specifically provided for the purpose.The Aircraft Reactor Experiment (ARE) to be described in this paper was originally conceived in a very different form from that in which it was ultimately constructed. In fact, the original high-temperature reactor design did not employ a fluoride-salt fuel, and, in order to understand some of the features incorporated in the ARE, a certain amount of historical background is necessary.By 1950, at various places in the country, work had progressed on the handling of hightemperature sodium metal to the point that it was being seriously considered as a coolant for nuclear reactors. Accordingly, a group of engineers and physicists at ORNL started design work on a solid-fuel-pin sodium-cooled reactor, with the fuel consisting of 235 U (as UO 2 ) canned in stainless steel. It was decided to make this a thermal reactor and to use BeO blocks as the moderator. The circulating sodium was to extract heat from the fuel pins and at the same time to remove heat from the moderator blocks. The design of this solid-fuel-pin, BeO-moderated, sodium-cooled reactor proceeded to the point of purchase of the BeO moderator blocks. These blocks, as fabricated for the original reactor design, are shown in Figure 1.The solid-fuel-pin thermal reactor design was found to possess a serious difficulty when the design concept was projected to cover a relatively high-power reactor. The problem was the positive temperature coefficient of reactivity associated with the cross section of xenon at elevated temperatures. This xenon instability was considered to be serious enough to warrant abandoning the solid-fuel design concept, because of the exacting requirement placed on any automatic control system by this instability.An obvious way to avoid the control problem would be to incorporate a liquid fuel that would have a large density change for a given change in temperature. If, upon heating and expanding, a portion of the fuel could, in effect, be made to leave the critical lattice, a selfstabilizing reactor would result.