analysis had been an order of magnitude too small. The workers at NYU 

 worked out a metliod for leveling the data that would take about one and one- 

 half hours of UNIVAC time. We hoped to get this done on the DTMB UNIVAC, 

 Later the job was done commercially at New Yorko While working out the 

 details of the programiining of the data Mehr upped the estimate of total 

 UNIVAC time to over 20 hours.''' This made things look a little dark for us 

 as far as getting the job done at DTMB was concerned. 



On 26th August our letter to DTMB asking them for UNIVAC time was 

 answered. The Model Basin wanted to program the analysis Jxistead of having 

 it done at NYU and wanted to examine tlie data to see whether or not it 

 would be practical to do the work on their UNIVAC. Their desire to do 

 their own programming was uaderstandable since an improperly jirogram- 

 med operation could run up the total time used by the UNIVAC considerably ,^ 

 However, Mehr had already done most of the programming at NYU. An- 

 other conference appeared, to be in orders so on 19 September Dr. Polachek 

 and Mr. Shapiro and Mr. St, Denis of DTMB and Mr. Mehr of NYU and my- 

 self met a.t DTMB to discuss the problem.. Mr. Mehr had finished program- 

 ming the a.nalysis. All that rem.3,ined to be done was to "de-bug" his pro- 

 gramming setup and then run tb.e analysis. He estimated that about 20 good 

 UNIVAC hours would be required for each of the three sets of data. Con- 

 sidering their other high priority commitments for the UNIVAC this was 

 way out of line with what we hoped to get. I agreed to try to find funds to 

 de-bug and run the first set of data commercially, =• the plan being then to 

 *For each set of data. 15 



