were used throughout. Early in the series of runs the a,c bridge was 

 a,t)andoned in favor of a dc bridge. This bridge consisted of a Wheatstone 

 bridge (Leeds & Northrup, Model No. ^735 )? a guarded dc null detector 

 (Leeds & Northrup, Model No. 983^) > and a regulated power supply (Lambda, 

 Model No. LH 121 FM) . The bridges could be read to 0.1 ohm and estimated 

 to a few hundredths of an ohm. 



In a typical rim, approximately 300 ml of a high and a low salinity 

 seawater were pipetted into the inner and outer cylinders, respectively. 

 The high salinity seawater was covered with a thin layer of mineral oil 

 to prevent water vapor transport with its concomitant salinity change 

 and temperature gradient. The calorimeter was closed and cooled to a 

 temperature slightly lower than the desired temperature of the run, and 

 then placed into a dry, precooled, wide mouth Dewar flask. Styrofoam m 

 inserts were used in the Dewar flask to fill up dead-air space and to 

 prevent direct thermal contact between the calorimeter and the flask. 

 The Dewar flask was sealed with a, Styrofoam insulated plastic cover, and 

 the whole assembly was clamped into a constant temperature water bath 

 (Aminco, Model No. U-8605) with a temperature control of ±0.02°C. 



The temperature of the calorimeter was monitored by means of the ac 

 or dc bridge described above. The temperature in the water bath was 

 adjusted until the temperature in the calorimeter was relatively constant 

 with time. The criterion used for equilibration was a rate-of -resistance 

 change less than ±0.05 ohms/min. Equilibration time ranged from 1 to 2k 

 hours . The two seawaters were mixed by inverting and rocking the Dewar 

 flask/ calorimeter assembly. The possibility of introducing heat during 

 the mixing procedure was checked with 6OO ml of distilled water in the 

 calorimeter. No detectable change in temperature was observed. 



The heat capacity of the Dewar flask/ calorimeter assembly was determined 

 by several methods . Initially, a known weight of water at a known 

 temperature was introduced into the calorimeter. Since the initial and 

 final temperatures as well as the heat ca,pacity of the water were known, 

 the heat capacity of the assembly could be calculated. After extensive 

 modifications to the Dewar flask/ calorimeter assembly, a second set of 

 calibrations was carried out tha,t more nearly reflected the experimental 

 method used in the heat of mixing runs. Two sodium chloride solutions, 

 80.60^tf and 17.23'^o , were mixed at 25°C. From available data on the 

 heat of dilution of sodiiim chloride solutions at 25 °C (Harned and Owen, 

 1958) the theoretical temperature change was calculated. From this 

 theoretical temperature change, the observed temperature change, and the 

 weight and heat capacities of the solutions involved, the heat capacity 

 of the Dewar flask/ calorimeter was determined. This was assumed to be 

 constant over the temperature range of the mixing experiments; i.e., from 

 1.93° to 25.3°C. 



In an isothermal mixing experiment we would like to maximize the 

 observed temperature change. This requires a maximum difference in the 

 concentration of the waters to be mixed. The final concentra,tion should 

 be near c = 0.035 where c equals the grams of solute per gram of solution. 



429 



