It was anticipated that the former effect would be small, since in the primary experiments the total-pressure range was from 25 psia to 100 psia. The Reynolds Number of the smallest probe at the lowest pressure was 435, based on free-stream conditions and the perfect-gas results of reference 3. At this Reynolds Number, the probe correction was indicated to be less than onehalf per cent,4 which is inside the experimental error. Since there were small differences in probe geometry between these probes and those discussed in reference 4, a supplementary set of experiments was conducted at a total temperature of 600°F., over a total-pressure range of 5 psia to 100 psia. These data were assumed to be independent of vibrational lag and spanned the Reynolds Number range of the primary tests. The results indicated that the effects of low Reynolds Number were inside the experimental error for these tests, as expected. The following parameter was used to calculate the departure from equilibrium at each total-pressure level. (iYAPoi) measured (P02/P01) frozen a = (P02/P01) equilibrium ~ (^02/^01) frozen The number-seven probe was assumed to give the frozen values of P02/P01 while the number-one and -two probes were assumed to give equilibrium values. Note that a varies from zero to unity. The former indicates frozen flow, and the latter corresponds to equilibrium flow. The factor a was related to the probe geometry by a dimensionless parameter, K. This parameter is made up as follows: pj\ T is proportional to the collision frequency; exp \/T1^ is proportional to the number of collisions required to reach equilib­ rium; and d/U is a measure of the time that is available for a particle traversing the distance between the bow shock and the stagnation point. This factor can be generalized to include the gas properties, and, since this is an impact-point experiment, it can be written in terms of equilibrium impact properties. In engineering units, K = (36.5 po Wair/?W eXp I — 30(w/Wair) X [O/1013)/2']2}l/3