We should remember that the atmosphere and the ocean represent 

 the two largest heat reserviors with which we are associated. In 

 reality, they are vast heat engines. The surface of the ocean to 

 which we have referred both critically and humorously is the region 

 through which energy exchanges occur between these two dynamic 

 systems. We are just beginning to get a glimmering of basic under- 

 standing, and we need a great deal more solid instrumentation. 

 New theories, too, are needed. We need to do a great deal more 

 work in this interfacial region because it covers such an impor- 

 tant proportion of the earth's surface, where energy may be most 

 easily exchanged. 



We should take very much to heart the warnings and comments 

 made yesterday with respect to the problems of the National Ocean- 

 ographic Data Center. Dr. Jacobs pointed out, it is not difficult 

 to overload almost any data processing system if the data is not 

 in the right form or selected with the proper discretion. Close 

 cooperation between the instrument designer, the oceanographer , 

 and the data custodian is extremely important. I was surprised 

 to hear that the oceanographer of the future might not have any 

 firsthand contact with his medium but would come to depend upon 

 data from data centers. This I doubt. 



We must examine our position on expendable instruments 

 carefully. I would like to emphasize that if we are going to deve- 

 lop a successful, expendable BT, for example, we must look at 

 both the design problem and the packaging problem very critically. 

 Do not assume that our present techniques or past practices are 

 adequate. Let us look hard at possible things which we can use, 

 things we can eliminate, and get down to what is really necessary. 

 Remember that huge volumes of these items will be required. Do 

 not design it for only small production quantities. The design 

 should permit large -volume production. This is the only way that 

 we can get unit costs down. We heard some rather high figures 

 on unit costs for observations given this morning. These are not 

 excessively high. These are the facts. The cost of the little B'J' 

 slide record to the Navy, not counting labor, is somewhere between 

 $10 and $25 an observation delivered on deck. However, we should 

 use this figure with caution. 



I do not believe any reference was made to the fact that re- 

 peated surveys are going to be in order. We are not going to 

 be able to go once over the ground lightly or even thoroughly and 



232 



