(c) Various simple theories of baroclinic instability (e.g. Phillips, 
1951; Robinson, 1963), when applied to the laminar thermocline 
theories based on a laminor interior regime (following Sverdrup), 
indicate that the interior solution as given by the thermocline theories 
of Stommel, Robinson, and Welander is dynamically unstable. 
The immense complication connected with the theoretical problem of 
computing the fully developed geostrophic turbulent processes in the 
oceanic thermocline and the very incomplete observational descrip- 
tion of such processes preclude further development of the unstable 
thermocline theory at present. When more observational guidelines 
are available, it seems probable that the theory can proceed, numeri- 
cally if necessary. Of course, it is not at all clear whether the im- 
portant property-transferring eddies owe their existence to instabili- 
ties of the thermocline or coastal boundary currents or to irregularities 
of bottom topography or applied wind stress. 
(d) Early theories of oceanic circulation (Rossby, Hidaka, Stock- 
mann) placed much emphasis on the hypothetical existence of large, 
lateral eddy transports. Sverdrup banished them from open oceans, 
and Munk found that their influence might be limited to the western 
sides of oceans and computed fields of transport in the oceans which 
bear considerable resemblance qualitatively to observed geographic 
mean distribution of ocean currents. The magnitude and role of eddy 
processes envisaged in the Munk theory is purely hypothetical. 
One possible array is that 30 buoys be placed within a 4° square, 
centered at 32° N., 53° W. (an abyssal plain area). On each buoy 
there will be velocity and temperature measuring units at depths of 
25, 50, 100, 200, 500, 1,000, 1,500, 2,000, 3,000, and 4,000 meters. Read- 
ings are taken at each point every 20 minutes for a year so that a total 
of 8X 10° velocity vectors and 8 X10 ® temperatures will be measured. 
Much of this information will be needed simply to filter out short- 
period components such as tides and their harmonics and short-period 
internal gravity wave phenomena. 
4. OCEANWIDE NET AS AN EXAMPLE; QUARTER-OCEAN NET 
Professor John Isaacs has proposed experiments with a much larger 
network of buoys—perhaps 80—covering a significant portion of the 
Pacific in order to monitor long-period changes in circulation. This 
type of network might be compatible with meteorological moored 
buoys of the World Weather Watch and might eventually be merged 
with that service. 
5. HOW TO GET STARTED 
Each of the three examples described above is a complicated ex- 
pensive operation. 
116 
