out wherever possible. In some cases, internal data processing 

 may be required such that the final record is of the parameters 

 dBsired rather than those which are actually sensed by the instru- 

 ment. There should be no loss of useful information in this pro- 

 cess, however. 



In the second place it is my feeling that the final material 

 records of all instruments should be "modular" in nature as far 

 as is possible. I realize that I am using a meaningful word in a 

 rather unusual sense but I am unable at the moment to find another 

 one that expresses the rather elusive thought. What I am trying 

 to say is that the economy of storage, transcribing, and processing 

 cannot stand too great a variety of final record. We camot afford 

 a special curve tracer or transcriber for each instrument nor can 

 we afford to employ an excessive variety of storage and handling 

 equipment. Moreover, the collation of records demands that the 

 utmost in uniformity of record be achieved. 



So much for broad objectives. Now perhaps a word is in 

 order as to how the NODC fits into this picture. 



Many technical and commercial fields have provided adequate- 

 ly for a centralized data processing facility and a d ecentralized 

 operational and research activity. They most commonly fail, 

 however, to provide the function that serves to knit the two together 

 into a unified operating system. This function is obviously 

 COMMUNICATION. NODC intends to do its part to provide this 

 function. Communication is obviously a two-way proposition. In 

 the first instance it means that adequate provision must be made to 

 disseminate to the research or operational individual the materials 

 he requires in a rapid and economical fashion and in the form most 

 suitable for his use. In the second instance it means that adequate 

 provision must be made for the communication oi requirements 

 between the research and operational interests and the data center. 

 This must be accomplished in both the short and long term senses. 



1/ (Continuation from page 61) 



first transcribed the word from her notes it came out reluctant . 

 I don't think she realizes how appropriate the mistake actually was. 

 In fact, the term was so apropos of the discussion I was almost 

 tempted to add a section on "reluctant data. " Certainly we have 

 numerous data in our files that are aptly described as "reluctant" 

 and I have no doubt but that some of you could have offered sug- 

 gestions as to how we might cope with some of this "reluctance. " 



62 



