curve for linear calibration. To operate the gage after the gage 

 has been made linear, the only adjustment necessary is that with the 

 trimmer in the tank circuit of the fixed R. F. oscillator. 



Figure 5 shows the error that can be expected due to wave shape. 

 Since this gage was thought to be more accurate in the measurement of 

 small waves than any other means available, sheet metal shaped in the 

 form of a train of water waves was used to find the probable error 

 due to wave shape. The shaped sheet metal was put on a carriage 

 which ran on a track under the gage. The gage was made adjustable 

 as to its height above the shaped sheet metal. The formed sheet 

 metal waves ranged from 1/4 to 1 inch in height and 1 to 5 feet in 

 length. At one end of the carriage there was a flat plate over which 

 the gage was calibrated; it was then assumed that the calibration 

 would be the same over still water. After calibration, and in order 

 to establish the effect of surface curvature on the calibration, 

 the carriage was moved under the gage until either a crest or trough 

 was directly under the gage. Without any changes or adjustment of 

 the gage circuit, the gage elevation was reset to maintain a constant 

 distance above the formed sheet metal and a voltage reading was then 

 taken. Following this test, with no change in gage circuit and with , 

 the same conditions as above, the gage position was reset to main- 

 tain a constant voltage output from the gage, then a reading of the 

 gage distance above the formed sheet metal taken. The percent of 

 change corresponding to the reading over the flat plate was plotted 

 against ratio of the height of the formed wave to the wave length 

 and is shown in Figure 5; the error indicated on this plot representing 

 the errors introduced by the curvature of the surface of the wave 

 train. 



40 



