189 



"... tailoring the instrument or method to fit the particular problem at hand. ..." 

 In this case it might be desirable to use two instruments, the first a slow re- 

 sponse device to obtain the mean value of salinity with good accuracy; the sec- 

 ond, a fast response instrument to obtain the random terms. The second instru- 

 ment need not have the accuracy demanded of the first, as the mean of its infor- 

 mation could be adjusted to fit the mean obtained with the first. By using two in- 

 struments it might be possible to eliminate the troubles encountered when one at- 

 tempts to combine "high accuracy" and "high sensitivity" in a single instrument. 



DISCUSSION: Norris W. Rakes traw 



I might add, in addition to what Dr. Carritt has said about the automatic 

 chlorinity titrator (ACT) that this instrument has operated satisfactorily on ship- 

 board under actual conditions. It is not affected by the vibration or the roll of 

 the ship and its effectiveness seems to be limited only by the ability of a man to 

 stand up and operate it. Nevertheless, it is not in its final stage of development 

 for some work should still be done to improve the design of electrodes and to 

 make their placement in the slipstream of the beaker less critical. Neither on 

 shore nor shipboard can this instrument be entirely relied upon when operated by 

 inexperienced technicians. It is not yet automatic to that degree. 



Modern instrumentation has also contributed to the measurement of pH, 

 which is important in determining the distribution of carbon dioxide and for many 

 other purposes in chemical oceanography. Until rather recently, this measure- 

 ment was almost always made by the use of indicators, sometimes with very 

 simple colorimetric apparatus. The use of a modern spectrophotometer makes 

 such a measurement very much more precise, but in any indicator method there 

 is always considerable uncertainty as to purity and properties of the indicator. 

 In recent years, the improvement of the glass electrode and of the electronic 

 amplifier circuits with which it is used, has entirely revolutionized the measure- 

 ment of pH in all fields of application. Nevertheless, pH meters and glass elec- 

 trodes must be pushed to their highest degree of sensitivity in order to yield re- 

 sults sufficiently precise for the chemical purposes of oceanography. Unless 

 one's results are reliable to one or two in the second decimal place, it is scarce- 

 ly worthwhile to make any measurement at all. The great difference in temper- 

 atures of the samples and the necessity for excluding air in order to prevent loss 

 or gain of carbon dioxide makes such precision rather difficult. But it is possi- 

 ble to overcome this difficulty by constructing long, narrow electrodes and keep- 

 ing the samples in polyethylene bottles without presence of air until they have 

 reached room temperature. Measurements made under such conditions seem to 

 be about the most satisfactory which we can accomplish at the present time. 



Dr. Carritt mentions the reagent dispensers which we have constructed 

 for use in several field methods. These represent a trend which I feel sure is 

 bound to continue, until the time comes when chemical apparatus for use at sea 

 will be totally different from anything that we use in the shore laboratories. 

 Conventional volumetric apparatus, such as flasks, pipettes, and burettes, were 

 never designed to go to sea, We are gradually learning how to eliminate them. 

 Beakers, bottles, funnels, and other measuring apparatus are all now available 

 in plastic material, generally polyethylene, and these are admirably designed 

 for seagoing use. Large carboys holding as much as 50 liters are now available 

 in such material. They are light, do not break, and will keep water silica- 

 free. 



The newer types of ultra-filters (such as the ones known as "Millipore" 

 filters) are becoming a useful tool in chenmical operations which require filtra- 



