216 L. H. N. Cooper 



circles, so that the law of attenuation would need to be worked out not from plane 

 but from spherical trigonometry. However, in the Eastern North Atlantic the source 

 of internal waves is postulated not as a point but as ridges or lines more than 600 miles 

 long. Consequently the energy in waves from these lengthy sources arriving at the 

 European continental slope may be considerable. 



Let us now consider what may happen when such a wave system meets a continental 

 slope. If it approaches head on at a smooth gentle slope, it is likely to run straight 

 up it and to be reflected back. A standing wave might well develop off slope, but the 

 conditions are unlikely to lead to a lot of mixing. From the usual snap oceanographical 

 observations such a standing wave might be interpreted as evidence for a strong along- 

 slope current. If the wave system approaches a smooth slope at a wide angle, the 

 waves may swirl along it, and somewhat more mixing would result. 



The really interesting case is when an internal wave system approaches a highly 

 dissected continental slope or borderland at a glancing angle. The slope abreast of 

 the English Channel is likely to provide an excellent site for this process, while the 

 Southern Californian continental borderland may provide an even better one. Then, 

 as I see it, there should be a very great deal of sloshing about indeed, much vertical 

 mixing and a tendency for homogenization of all properties as far down as the internal 

 waves occur. This means that nutrients will be brought up to such depths that the 

 classical methods of wind-driven upwelling can bite into the enriched water and bring 

 it right up to the surface. If, and when, cold Arctic winters produce large boluses and 

 consequent internal waves of large amplitude, then upper-water enrichment with 

 nutrients over continental slopes would be favoured. 



Furthermore, we may be concerned not only with the conventional nutrients, but 

 with associated organic growth accessory factors which we only dimly understand. 

 Much dead plankton, faeces and organic detritus sink in the sea and decompose. 

 This should happen to the greatest extent in the oxygen poor layer. Oxygen defect 

 (saturation value less content of oxygen observed) provides a very rough but ready 

 measure of the extent of decomposition products. During periods of minimum internal 

 wave activity the maximum oxygen defect may well be more than in periods of great 

 activity. But paradoxically, the depth range of the oxygen poor layer should be less. 

 The inflexion should be more pointed. This is because the homogenizing action of 

 internal waves beating against a dissected slope should spread the oxygen defect both 

 above and below the point of inflexion. The upward spread would bring not only the 

 oxygen defect but also the accompanying organic substances nearer the surface. 



In this group of interlocking hypotheses, some may be rejected without gravely 

 imperilling the rest. The concept of homogenization is not one of these for, if it 

 falls, the others become useless. To illustrate the argument, let us assume that some 

 property such as phosphate content is linearly proportional to depth, and is subject 

 to a uniform process of homogenization from surface to an inert bottom. In mid 

 water, exchanges of phosphate will occur, but the later state will remain analytically 

 indistinguishable from the earlier. If a layer was labelled with radiophosphorus, 

 homogenization would spread this up and down so that the result might be seen as a 

 form of eddy diffusion. At the surface, exchange with the air being impossible, a 

 homogeneous layer enriched from below would begin to build up. As the process 

 proceeds, both depth and phosphate content of the homogeneous layer would increase. 

 Similarly at the bottom, homogenization will increase the thickness of a uniform 



