616 CYCLES OF ORGANIC AND INORGANIC SUBSTANCES 



mixing proceeds they will develop what can only be called a honey- 

 comb structure of pockets of rather varying composition. As time 

 passes and the journey proceeds, the earlier components will tend 

 to homogenize but fresh complexity will become built into the 

 bolus as it descends into deeper and deeper water. Ultimately the 

 bolus will arrive at a level of potential density in the deep ocean 

 equal to its own. It will descend no further. 



Some slender evidence that in the deep ocean the resident water 

 consists of a series of plates of different waters each in neutral 

 adiabatic equilibrium at closely defined potential densities has 

 been given. If a bolus has a potential density precisely that of one 

 of the resident water masses, it should quickly amalgamate with 

 this and be lost to sight. The greater probability is that the bolus 

 will have a density somewhere in one of the gaps between the densi- 

 ties of the several resident water masses (Fig. 8) ; that is, the density 

 discontinuities form a series of platforms at sigma-theta 27.831, 

 27.862, 27.892, and 27.903. Consequently, a descending bolus will 

 tend to come to rest on one of these platforms where its thickness 

 — as a static system — makes it unstable. It will tend to spread 

 widely in the density gap as a thin plate of water of thickness 

 measured in meters or even centimeters. 



No two boluses are likely to have quite the same history. They 

 are not likely all to have the same size when they are born. The 

 mixing history of each is likely to be different. Consequently, not 

 only are they likely to come to rest on different discontinuity 

 platforms but each is likely to ha\e rather different temperature, 

 salinity, and oxygen contents. They will all flatten out to form thin 

 plates of water between the main plates of the resident water 

 masses and show a considerable range in composition (Fig. 10). 



These thin plates are likely to differ in density ever so slightly 

 from each other and so would build up into a complex laminated 

 structure occupying the density gap between the resident water 

 masses. 



This then is a speculative explanation of the failure to ration- 

 alize the data around, potential temperatures of 2.5°, 3.1°, and 

 3.8°. I suggest they represent a situation so complex as to defy the 

 effort we have so far brought to bear on it. 



If the thesis is accepted, some riders follow. 



