14 hours should not be too difficult to find. For these conditions the signifi- 

 cant wave height would be nearly double and the E value would be nearly 

 four times those observed for the 18=7 knot wind according to the results 

 of Neumann, 



If such distorting effects as differential shrinkage and column noise could 

 be eliminated by proper choice of film and prelinainary studies of their causes 

 it would then be possible to allow a four-fold increase in the white noise vari- 

 ance without seriously affecting the results. This would permit the planes to 

 fly higherj thus covering a larger area in one stereopair and providing better 

 resolution and more degrees of freedom. An appendix written by Simeon 

 Braunstein, the Research Division photographer at New York University, 

 follows these recommendations and conclusions. In it is given a discussion 

 of the stability of different types of film bases and of film processing methods 

 which should eliminate the effects of differential shririkage. 



In such an experiment more careful attention should be paid to the wind 

 field, and winds should be recorded at least every hour for as long a time as 

 possible prior to the observations, A decrease of wind speed in the wind 

 field should be avoided. The winds, if possible, should be measured at 

 several heights. 



Moreover, now that a fairly good method of analysis for the results 

 has been developed, it should be possible to program additional operations 

 on the spectrum to carry out in just a few minutes all the computations 

 made in Part 11. 



255 



