478 VIGOUREUX AND HERSEY [CHAP. 12 



Wilson made 581 measurements of the velocity of sea-water at temperatures 

 from —3° to 30°C and for five salinities between 33 %o and 37 %o. These were 

 fitted to MacKenzie's (1960) formula, which is based on Kuwahara's tables. 

 Of this work Wilson reports : "this equation resulted in a standard deviation 

 from the mean of the differences between the computed data and the measured 

 data of 0.29 m/sec. The equation was consequently changed and only the terms 

 resulting from the differential analysis of fourth order polynomials were re- 

 tained." The resulting terms from equation (7) are given below. 



co.35,0 = 1449.22 



Act = 4.6233^- 5.4585 X 10-2^2 



-1- 2.822 X 10-4^3 - 5.07 x lO-^i^ 

 Acp = 1.60518x10-^+1-0279x10-5^2 



+ 3.451 X 10-9p3 _ 3.503 x 10-12^4^ 

 Acs = 1.391(>S- 35) -7.8x10-2(^-35)2, and 

 Acstp = (^-35)(- 1.197x10-2^ + 2.61x10-4^ 

 -1.96X 10-7^2 _ 209 X 10-6^0 

 +p{ - 2.796 X 10-4^ + 1.3302 x 10-5^2 

 - 6.644 X 10-8^3) +p2( _ 2.391 X 10-'?« 

 + 9.286 X 10-10^2) _ 1.745 x lO-i^^^. 



where p is in kg/cm2, S in parts per thousand (%o), and c in m/sec. 



Hays (1961) made seven sound- velocity profiles in the Mediterranean Sea in 

 conjunction with seven hydrographic stations, during which many determinations 

 of temperature and salinity were made. He used an instrument designed by 

 Tschiegg (see Tschiegg and Hays, 1959). He has compared his measured 

 velocities at the same depths as the salinity /temperature measurements with 

 velocity values computed by Wilson's formula. The differences between the two 

 show a nearly linear increase with depth, with the computed velocities less 

 than the measured by about 0.2 m/sec at 700 m and by about 0.5 m/sec at 

 2200 m. At shallower depths than 700 m the differences are much larger and 

 less systematic, probably owing to horizontal gradients in the water through 

 which the ship was drifting during the observations. At greater depths than 

 2200 m the difference decreases to about 0.4 m/sec at 2800 m. 



The systematic difference between Wilson's formula and Hays's measure- 

 ments suggests that further investigation is needed to assure an increase by a 

 factor of ten in accuracy from the work of Matthews and Kuwahara. Neverthe- 

 less, the discrepancy is small ; of the same order as the differences between the 

 laboratory results for distilled water of Del Grosso, Greenspan, and Wilson. 



At the time of writing, the velocimeters of Greenspan and Tschiegg (1955) 

 have been used under severe conditions at sea as well as in the laboratory. 

 They have proven to give reproducible results and trouble-free service. By now 

 several tens of velocity soundings have been made in the deep ocean. Never- 

 theless, for some time to come acousticians will be dependent on measurements 

 of salinity and temperature as a function of depth. 



