A similar formula in terms of the pitot tube reading shows that the overshoots indicated in Fig. 4 of reference 2 could be explained by a 1-5 per cent rise in static pressure pu in front of the pitot mouth. The preceding conjecture as well as the general up stream effect can be checked by taking simultaneous readings of pitot pressure at each y position and of static pressure from wall orifices under and slightly upstream of the pitot tip. If there is a demonstrable effect of upstream influence when a relatively large probe is outside of the layer, it would follow that similar or even larger effects could occur throughout the boundary layer and that some effect should be present for smaller probes. The presence of this effect and that due to the leading edge makes the near agreement with theory (which takes account of neither) in references 1, 2, and 3 the more remarkable. The question remains as to the non-dimensional parameters properly characterizing the above interference, with or without separation. Clearly, the probe height h has to be non-dimension alized with respect to some characteristic layer thickness in order to account for the observed effect of relative size. Just how this should be accomplished is not clear since it involves ques tions of pressure disturbance strength as function of y and the upstream viscous influence. The width of the probe must be significant since its increase means closer approach to the ob served two-dimensional critical pressure disturbance for sepa ration, P for instance, the upstream distance, s, through which a given (linearized) pressure disturbance at wall diminishes by a factor e1, measured with boundary layer thick ness as a unit, depends on the parameters approximately as fol lows :