167 



anchor, an analyst can often make a single adjustment satisfactorily. 



(b) Electrical power on most ships has regulation of voltage, frequency, 

 and wave form that is inferior to the regulation of most commercial power. The 

 efficiency of instruments requiring well regulated supplies is thus often reduced 

 on shipboard unless supported by costly secondary supplies or regulators. Bat- 

 tery-operated instruments obviously overcome a part of this difficulty, and ex- 

 cept when working immediately offshore have to be used in recording or tele- 

 metering equipment. 



'Sea water batteries", in which sea water is the electrolyte and a rela- 

 tively reactive metal such as magnesium serves as one of the electrodes, con- 

 ceivably could find application in situations where the battery or the entire in- 

 strument is to be jettisoned. 



(c) Maintenance and repair facilities are much more limited at sea than 

 ashore. 



The efficiency of a seagoing instrument for chemical measurements, 

 then, can be judged by the following criteria: (1) it should be unaffected by mo- 

 tion and vibration; (2) it should be simple to operate (especially if put in the 

 hands of semi-skilled technicians); (3) it should be able to operate efficiently 

 with poorly regulated power supplies or with batteries; and (4) it should require 

 a minimum of maintenance while at sea. 



CHLORINITY-SALINITY-DENSITY MEASUREMENTS 



One of the persistent problems in chemical oceanography is that of de- 

 vising a simple, rapid, accurate, and precise method for the determination of 

 sea water density. A great deal of work has been done to avoid the direct de- 

 termination of density which is considered too time-consuming and delicate for 

 routine operations. It should be noted, however, that modern, direct-reading 

 analytical balances, precision thermostats, and plastic sample bottles make the 

 direct determination now much more feasible than in the past. 



The classical works of Dittmar and his contemporaries on the- composi- 

 tion of sea water were expanded by Knudsen and Patterson to establish a stand- 

 ardized, indirect density measurement now known as the Knudsen Method (Ox- 

 ner, 1946). This method uses a relatively simple titrimetric procedure which 

 gives results that can be directly converted to density with the aid of Knadsen's 

 Hydrographical Tables. Despite the simplicity of the Knudsen Method, which 

 can be carried out on shipboard as well as ashore, it has undesirable features 

 when a large number of samples must be processed. The undesirable features 

 appear to be associated with hunnan error rather than with any inherent technical 

 inconsistency in the method. Under controlled conditions with competent tech- 

 nicians, the titration will give salinity values with an error not over jf 0.02 °/oo. 



To maintain this accuracy when several technicians are involved appar- 

 ently is difficult. It is possible to obtain a measure of the accuracy and pre- 

 cision of the results from each of several operators by interspersing samples 

 of known or standard solutions with field samples in such a way that the opera- 

 tors have no way of distinguishing between the two. The results of two such 

 tests, the only ones known to the author, have not been published. They did in- 

 dicate, however, that improvement was needed in order that the resulting densi- 

 ty values be useful in computations in physical oceanography. Continued checks 

 of this sOrt are time-consuming and by themseWes give no indication of the real 

 source of errors, except of course to point out grossly incompetent technicians. 



