wave. In the hollow the fluctuations are of much smaller intensity and occur on the 

 high velocity side of the wave. The frequency of these fluctuations is about six times 

 that of the primary wave which had a frequency of 145 cycles per second. Frequencies 

 this high would indicate that these high intensity fluctuations are probably not a con- 

 sequence of non-linear effects but are more in the nature of a secondary instability. 

 However, more study is called for in this connection. Preliminary measurements have 

 indicated that these high frequency fluctuations at the peak and in the hollow occur at 

 the same time, indicating that the smaller intensity fluctuation may be just an excitation 

 caused by the very intense pattern at the peak, and that the initial breakdown of 

 laminar flow has a very short extent in the z-direction which might be described loosely 

 as point-like. These localized regions are spaced along the wave front giving the 

 impression of streets along which a succession of bursts were occuring. 



In Figure 2 there is shown the breakdown of laminar flow from the beginning 

 of transition to where it becomes fully turbulent. It was too inconvenient to move the 

 hot wire through the transition region; instead the oscillograph traces shown in the 

 figure were obtained by moving the transition region over the hot-wire by continuously 

 increasing the ribbon amplitude. The hot-wire was at a position corresponding to a 

 peak in the intensity distribution shown in Figure 1. At first we see the rather pure 

 wave somewhat distorted with a flat-top appearance on the high velocity side, then 

 the appearance of high-frequency fluctuations which consume more and more of the 

 primary wave, and then the onset of fully turbulent flow. This process, i.e. from the 

 beginning of transition to where we have fully developed turbulent flow, takes place 

 within a distance equivalent to 50 boundary layer thicknesses. 



An attempt was made using the china clay technique to determine whether the 

 3-dimensionality associated with the intensity of the wave in the region where the wave 

 was still quite pure could possibly be due to Gortler-type vortices. The drying pattern 

 on the china-clay surface which was initially sprayed with oil of wintergreen is shown 

 in Figure 3. The dark line at the left of the figure is the vibrating ribbon. The limita- 

 tions of the method prevented any definite conclusions as to the existence of Gortler- 

 type vortices in the region before transition. As far as could be determined the street- 



FORGING AMPLITUDE INCREASING 



TRANSSTSON FROM FORCED OSCILLATIONS 

 Figure 2. 



368 



