HEAVY gases can be used to increase simulation Reynolds number. For this reason, and to test transonic modeling, interferograms were taken of 10% nonlifting biconvex airfoils in gases with specific heat ratios 7 of 1.101.67. The airfoils were mounted in Ludwieg/shock tubedriven solid-wall test sections and run at chord Reynolds number of O.SOxlO 6 to 5.5xl0 6 with nominal Mach numbers M of 0.70 to 0.80. No 7 effect was detected at subcritical M, and the data were in agreement with inviscid theory. A 13% systematic decrease in peak negative pressure coefficient and forward shock movement were observed as 7 increased in supercritical flow, contrary to the predictions of inviscid theory. A model incorporating a viscous displacement correction explained these results. Contents New simulation problems can arise from changing test gases.l For the present study, real gas effects are calculated to be unimportant compared with a change in the specific heat ratio 7. This parameter appears in the similarity rules obtained from the small-disturbance form of the inviscid equations of motion, but it persists to first order only in transonic flow. Recent experiments have shown that some correlation may be achieved at supercritical M through use of transonic similarity.2 However, the boundary-layer development depends on local properties, which change with 7. Thus the displacement thickness, transition point, separation point, and shock/boundary-layer interaction may also exhibit 7 dependence. This investigation primarily used a Ludwieg-tube-d riven staged nozzle with a 30x40 cm cross-section solid-wall test section.3 Biconvex airfoils 10% thick, with chords c of 9.8 and 7.3 cm, were pin mounted in windows centered 35 cm upstream of the choke. The airfoils had cross-sectional area blockages of 2.5 and 1.8%, tunnel half height H to c ratios of 2.0 and 2.7, and aspect ratios of 3 and 4, respectively. They were tested in unheated A, N 2, or CO 2 (7 = 1.67, 1.40, 1.31). In order to avoid condensation, a shock-tube-dri ven 7.5 x 10 cm solid-wall test section was used for the Freon - 12 and SF6 7=1.10 experiments. For these a 10% biconvex airfoil with c = 2.5 cm was pin mounted in windows centered 30 cm upstream of the choke, giving H/c and aspect ratio identical to