- 4 - 



average speed of these two which is precisely a mean solar day 

 per 360°. Thus, the second diurnal drum was run at one cycle 

 per minute and the other two at 30.5 seconds and 64.5 seconds 

 per cycle respectively. This combination produced a succession 

 of great diurnal tides yielding to lesser semi-diurnal tides at 

 intervals of about two "weeks," a progression characteristic of 

 this region of the Gulf of Mexico (fig. 12). 



The winds were generated by tank- type vacuum cleaner blowers 

 mounted around the model at the cardinal and intercardinal points. 

 Each was wired to a switch at a central control point and its 

 rate of pumping governed by a Variac, Thus, it was possible to 

 vary the wind speed and direction to reproduce either the detailed 

 or average wind as it occurred during the periods of interest to 



the sponsors and in accordance with the records of the U. S. 



2/ 

 Weather Bureau reports from Grand Isle, Louisiana. Velocity cali- 

 bration of the wind machines was accomplished by observing the 

 rate at v*iich the surface layer of water in the model was pro- 

 pelled under wind stress at the steady state. It was assumed on 

 the basis of field experience and values derived from empirical 

 equations for wind stress that in more than 10 miles of fetch the 

 stress on the surface is such as to produce an equilibrium sur- 

 face current speed between 2 to 3% of the wind speed. By thxs 



2/ See Appendix V attached to Copy No. 1. 



^ Sverdrup, Johnson & Fleming, The Oceans . Prentice Hall, New 

 York, 1942, pp. 489-503. 



