wire size)o The resistance is further reduced by lowering the elements 

 into the water until they are completely submergedj at which point only 

 the resistance of the recording circuit remains. Hence, it is seen 

 that the galvanometer (a current measuring instrument in this case) 

 will record from a zero current to some constant current of the circuit 

 with respect to the variance of the resistance as described o 



!Die type of electrodes used have been (1) silver plated constantine 

 wire, (2) silver plated music wire, and (3) iron welding rodo Sample 

 construction of elements are shown in Figures 3-As B, G. For conven- 

 ience of measurement the wire elements are attached to point gages which 

 have a least count of OoOOl footo Types (1) and (2) were 0,033-inch 

 diameter wire and 6 inches in length. Type (3) was I/L 6-inch diameter 

 wire and 15 inches in length. The 0,033-inch diameter wires and 6 inches 

 long were used to measure waves of Oo^5 feet maximum height and a 

 magnification of the actual wave was obtained when desired. This type 

 of element requires a smaller power supply (1 to 1^ volts) than does 

 the larger elements. It is more sensitive to amplitude changes and 

 less sensitive to power changes than the larger elements. 



The l/l6-inch diameter wires and 15 inches long were used to 

 measure waves of 1,3 foot maximum height and a reduction of the actual 

 wave was obtained. This type of element requires a larger power supply 

 (1^ - 2 volts) and is more sensitive to power changes than is the smaller 

 element. The l/l6 inch diameter elements with a high power input 

 (2„5 volts) were used satisfactorily for measurement of small model 

 waves , 



The power supply to the elements, of course, depends on the 

 resistance and the magnification desired. Thus, it will be different 

 for each experimental set up. 



The effect of the meniscus or surface tension on the wire element 

 compared to another type of wave measurement was found to be about 

 5 per cent vrhen compared with a point gage. The experiments showed 

 that there was 3 per cent variation in the wave heights as measured by 

 the point gage due to its meniscus effect and the variation of waves 

 generated. The corresponding measurements made with the wire elements 

 were found to differ from the point gage measurements ty a maximum 

 of 7 per cent in all cases. Thus, it is seen that the actual meniscus 

 effect is about 4 or 5 per cent. Furthermore, the difference was on 

 the high side part of the time and on the low side at other times. 

 The meniscus effect will be neglected in the futiire. 



Two or More Circuit Method 



The measurement of wave height at several stations simultaneously 

 is often desired so that not only the wave height can be determined 

 from the record but also the wave period, wave velocity and wave length. 

 The method proposed is best described by Figure 1-A, It is seen that 

 for any one power supply or transformer that one and only one circuit 

 can be traced out and, hence, the water path connecting pairs of 



' 17 



