Step 11. Repeat the process beginning with 

 drawing the sample with the automatic pipette. 

 Wlien running unknown sea water sample titra- 

 tions, a minimum of two titrations of each 

 sample must be made, and two burette readings 

 must be within 0.01 to be acceptable. Whenever 

 the readings have a difference greater than 0.01, 

 a third or fourth titration, if necessary, must be 

 run until two readings are obtained within the 

 prescribed limits. When making the standardi- 

 zation titrations of standard sea water, a mini- 

 mum of three titrations must be made. 



1-13 Standardizing and Adjusting the Sil- 

 ver Nitrate Solution. — As the titrations are 

 performed to standardize the silver nitrate so- 

 lution, record the burette readings (A) in the 

 Standardization column of the D-Sheet and 

 compute the difference alpha (a) between the 

 chlorinity of the standard water sample (N) 

 and the burette reading (A) for each titration. 

 For example : 



An average alpha is obtained from the three 

 best titrations of the standard sea water. This 

 average alpha is the value used to obtain the 

 correction k from Knudsen's tables, which in 

 turn is a^Dplied to tlie burette readings of the 

 sea water samples to obtain chlorinity. For 

 example : 



(N)- (A)= (a) 



19.395 19.375 +0.020 

 19. 395 19. 365 . 030 



19. 395 19. 370 . 025 



— . 025 (average) 



In the event the standardization reveals that 

 alpha is not within —0.150 to +0.145, the solu- 

 tion must be adjusted. This adjustment is made 

 by adding more silver nitrate crystals if the 

 solution is too weak, i.e., shows a negative alpha 

 lower than —0.150 or by adding distilled water 

 if too strong, i.e., shows a positive alpha of more 

 than 0.145. 



To assist in making these adjustments, two 

 linear graphs are shown in figures I-S and 1-9. 

 The first graph shows the number of grams of 

 silver nitrate to be added to each liter of solu- 

 tion when the titration reveals a negati^-e alpha 

 value. The second graph shows the number of 

 milliliters of distilled water to be added to each 

 liter of solution when titration reveals a posi- 

 tive alpha. 



Figure I-t. The true meniscus and the observer's eye. 



After the silver nitrate solution is adjusted, 

 repeat the titration process with a standard sea 

 water sample. 



1-14 Computing the D-Sheet. — After the 

 standardizations and titrations have been made, 

 compute the salinity of each sample. 



Step. 1. Compute the average burette reading 

 for each sample, and enter in Average Burette 

 Reading column. 



Step. 2. Average the beginning and the ending 

 alpha (a), and turn to the appropriate alpha 

 (a) column of table I-l, Table of the correction 

 k. 



Step. 3. Read down the column until the av- 

 erage burette reading (a) is less than the upper 

 number and greater than the lower number, 

 and read the value of k in the right hand col- 

 umn of table. Enter this value in the h column 

 of the D-Sheet. 



Step 4. Algebraically add k to the average 

 burette reading and enter the value in the 

 Chlorinity %o column. 



Step. 5. Compute salinity with Knudsen's 

 tables or by the following formula: S%o = 0.03 + 



Table I-l. Table of the correction k (taken from 

 Hydrographical tables, edited by Martin Knudsen) 



a =0.025 



20.15 

 19.76 

 19.34 

 18.93 



k = 



+ 0.01 

 + 0.02 

 + 0.03 



a is average burette reading 



1-7 



