the blood pooling than from either 18° or 25°C. Five samples were 

 taken at each time interval in each test condition except in some 

 extreme conditions where heavy mortalities precluded the use of more 

 than one or two samples per time period. The samples were centri- 

 fuged in a refrigerated centrifuge (-4°C) for 20 minutes at 3000 rpm 

 and the clear serum preserved for analysis. 



Blood Analyses 



Osmotic concentration 



The osmotic concentration of serum and water samples was mea- 

 sured on an Advanced DigiMatic Osmometer (Model 3D, Advanced Instru- 

 ments, Inc.). Each determination required 0.2 ml of sample fluid. 

 The apparatus reads the osmotic concentration in mOsm/kg. It is 

 accurate up to + 2 mOsm if the sample concentration is less than 

 500 mOsm but above 500 mOsm the accuracy increases to I'o of the 

 reading. Repeatability is within 1%. The instrument was periodi- 

 cally calibrated with sodium chloride standard within the 100 to 

 1000 mOsm/kg concentration range. 



Blood chloride 



Blood chloride concentration was estimated on a Buchler-Cotlove 

 Chloridometer (Model 4-2008, Buchler Instruments, Inc.). The instru- 

 ment operates on an amperometric end-point principle and shuts off 

 automatically at a preset increment of indicator current. Each 

 sample contains 0.1 ml of serum, 4 ml of nitric-acetic acid reagent 

 (O.IN) and four drops of gelatin reagent. Under normal conditions 

 the instrument is accurate up to 0.1% mEq/L. Estimations were made 

 on a 5-6400 mEq/L (high setting) range. 



Other electrolytes 



Potassium, magnesium and calcium ion concentrations were de- 

 termined on the Atomic Absorption Spectrophotometer (Model 305, 



41 



