114 CARRITT [CHAP. 5 



salinities. In the conventional Wheatstone Bridge, temperature control to 

 ± 0.003°C would be necessary to achieve the same precision in salinity. 



The basic Wenner Bridge design has undergone at least three recent revisions 

 and modifications. Each has been an attempt to improve the accuracy and 

 precision of chlorinity measurements by incorporating new components, circuit 

 features and controls into the original conductivity comparison bridge. Brad- 

 shaw and Schleicher (1956), Cox (1958) and Paquette (1958) have each con- 

 structed several conductivity bridges that have been used in their own and 

 other laboratories. The precision (standard deviation of an individual measure- 

 ment from the mean) of each of these instruments is in the range of 0.001 to 

 0.003 % in salinity. 



The conductance-bridge type salinometers mentioned above are laboratory- 

 type instruments which cannot be readily modified to make in situ measure- 

 ments. They all use classical kinds of conductivity cells in which the primary 

 measurement is of the resistance between two metallic electrodes in contact 

 with the samples being measured. Several in situ conductivity devices have 

 been constructed using the classical -type cell but without the double cell 

 features of the laboratory instruments. 



The Woods Hole Oceanographic Institution salinity, temperature and depth 

 recorder (S.T.D.) (Jacobson, 1948), the Johns Hopkins University Conductivity- 

 Temperature Indicator (C.T.I.) (Carritt, 1952) and the portable temperature- 

 chlorinity bridge described by Hamon (1956) are recent examples of devices for 

 in situ measurements. Apparently all of these devices suffer from a common 

 fault which limits their usefulness in open ocean studies. The difficulty comes 

 from changes in cell characteristics for which there is no compensation as in the 

 double-celled laboratory-type instrument. Even with frequent standardization 

 and recalibration, which is difficult to do routinely at sea, the precision appears 

 to be limited to 0.05 % to 0.10% o in chlorinity. 



Many of the objectionable features of the classical conductivity cell have 

 been overcome by a design first described by Esterson (1957). This cell is a 

 toroidal transformer in which sea-water forms the equivalent of a one-turn 

 winding linking two co-axially mounted inductors. A potential impressed on 

 the primary element of the transformer induces a current flow in the sea-water 

 link which, through inductive coupling, causes current to flow through the 

 second core and associated components. The current in the sea- water path is 

 proportional to the conductivity of the sea-water. Thus, the output signal from 

 the second core is a measure of the conductivity of the sea-water that forms 

 the connecting link between the two transformer elements. The transformer 

 elements are contained in an insulating housing which maintains the relative 

 geometry of the two elements and the sea-water path that passes co-axially 

 through them. Since there is no direct contact of water and electrodes, polariza- 

 tion, fouling and poisoning of the kind encountered in classical cells are 

 eliminated. 



Field instruments constructed on the Esterson design have been used for 

 several years by the Chesapeake Bay Institute of the Johns Hopkins University 



