glitter pattern on the sea surface is taken, it is possible to deduce the angular distribu- 

 tion of the long wave components from diffration analyses of two small areas of the 

 photograph. (In practice, two independent photographs of the required areas would 

 be taken.) This technique has some theoretical and practical limitations, perhaps the 

 most important being that it is only applicable to the long-wave part of the spectrum. 

 It has not so far been applied to a practical example. 



It is possible that Barber's simple diffration analysis could be applied to an 

 artificial "photograph" of the sea surface prepared from stereoscopic photographs in 

 such a way that the optical density is proportional to the corresponding surface eleva- 

 tion. 



Pierson (brief remark in his 1955 reference) is analysing a number of stereo- 

 scopic photographs of the sea surface. Perhaps he will tell us later how he is analysing 

 the measurements and something about the results. The first step is to convert the 

 photograph into "maps" of the elevation of the sea surface. Marks (1954) has described 

 one way in which these maps may be analysed by averaging surface elevations along 

 a series of parallel lines to produce a one-dimensional record which contains contri- 

 butions from components traveling in a known narrow range of directions. By altering 

 the direction of the lines of averaging, a 2-dimensional analysis may be made. Methods 

 using stereoscopic photographs are unlikely to become routine, since the labor of analysis 

 is great. 



A related method is based on recording surface profiles along different courses 

 using an airborne radio altimeter. Perhaps Dr. Longuet-Higgins can inform us about 

 the present status of this technique? 



Cox and Munk (1954) have shown that it is possible to obtain some directional 

 information from a simple photometric analysis of an out-of-focus photograph of the 

 sun's glitter pattern (see Fig. 1). Their method gives only the second harmonic of the 

 spectrum of the surface slopes integrated over all frequencies, and as such is of limited 

 utility. (Their principal aim was to determine the statistical distribution of slopes.) 

 They found a beam width of ±45°. 



Another class of methods depends on the use of directional primary measure- 

 ment. Eckart (1953) and Crombie (1955) have shown that if a beam of radio waves 

 is directed at the sea surface at almost glancing incidence, the back-scattering is due 

 to the component of the sea waves with half of wave-length of the incident radiation 

 and which is traveling in the direction of the receiver. Since the sea-waves are moving, 

 the back-scattered radiation has a Doppler shift in frequency which enables it to be 

 separated from direct coupling between the transmitter and receiver and from back- 

 scattering due to static objects. The range of directions included in the back-scattering 

 depends on the beam-width of the transmitting aerial, and this can only be made small 

 if the aerial is several wavelengths long. The method is therefore probably practical 

 only for short waves. 



Groves (private communication) is examining the possibility of using towed 

 electrodes as a directional wave measuring device. The use of towed electrodes for 

 measuring ocean currents is now a routine matter (Longuet-Higgins, Stern and Stommel, 

 1954). The average value of the recorded e.m.f. is due to the average motion of the 

 cable in the Earth's magnetic field. Wave frequency oscillations, now smoothed, might 

 give useful information. Groves finds that with normal electrode separations the results 

 are rather difficult to interpret, particularly when the device is near the magnetic 

 equator. If a cable with a length of several wavelengths is used the results become 

 easier to interpret, but the device may then be impracticable. The cable has to be kept 

 near the water surface and sufficiently slack to enable it to move with the water par- 

 ticles, but at the same time it must be kept in a straight line. 



To sum up, through several ways of obtaining information about the direction of 

 wave travel has been suggested, the more practicable methods give very limited infor- 

 mation, and no satisfactory way of obtaining a detailed 2-dimensional spectrum is yet 

 available. 



54 



