VERTICAL AND HORIZONTAL MOVEMENTS IN THE OCEAN 603 



salinity also do not correspond. The flushing of our bottles is 

 probably efficient so that the salinity represents the depth at 

 which the bottle closed. Thermometers, however, require a few 

 minutes to equilibrate. During gusty weather the thermometers 

 can do no more than indicate some sort of mean value of the 

 water through which the bottles are rising and falling. Moreover, 

 each of a pair of thermometers is not likely to have the same rate 

 of approach to equilibrium; nor are unprotected thermometers 

 likely to respond at the same rate as protected thermometers. 

 Consequently, during vertical yawing anomalous and discordant 

 temperatures and thermometric depths are much more likely than 

 when work is done with a vertical wire. 



Since temperature and salinity may not quite correspond, a 

 temperature-salinity plot will show excessive scatter about the 

 best curve drawn through them. A measure of standard deviation 

 will be larger than with samples drawn on a vertical wire. 



At New York the very careful study of the yawing problem by 

 Wust (1932, pp. 140-149) was overlooked. Nevertheless no excuse 

 is needed for raising the issue again. It will be essential not only to 



Table II. Hoist with Large Wire Angle at Surface (45°) to Illustrate \'ertical 



Yaw ing 

 Station Bedi\"ere. 28 September, 1958 



Distance 



Wire Obser^•ed Obser\ed Mean between 



Depth, Temperature, AT. Thermometric AD, Observed x-^djacent 



m °C °C Depth, m m Depth, Bottles, 



