89 



HOW TO AVOID ERRORS IN OBSERVATIONS 



It is very eas}^ for one not thoroughly schooled in the art of collect- 

 ing observations of the temperature and salinity of a water column to 

 make all sorts of errors. Usually the mistake is not detected until 

 some later date when, alas, it is too late to repeat observations and 

 rectify the error. It behooves observers to exercise the greatest care 

 in order that the degree of accuracy be raised as high as possible, and 

 the reputation of the records correspondingly enhanced. The fol- 

 lowing hints may be found useful by future investigators: 



(1) The water bottles should be in the finest working condition, and should 

 be gone over and oiled frequently. 



(2) Guard carefully against a tendency for the bottle to close prematurely. 

 (,3) Each bottle should be equipped with two thermometers. 



(4) Thermometers should be functioning properly and kept under close ob- 

 servation. 



(5) If thermometers in the same bottle do not check, they should be examined. 



(6) The mercury column should be continuous from the bulb end when the 

 bottles are lowered over the side. 



(7) The meter wheel should be checked occasionally for accuracy of measure- 

 ment. 



(8) The wire should be guarded against kinks. 



(9) The wire should be oiled occasionallj-. 



(10) The wire should be vertical when the top messenger is released. 



(11) Never take station observations if wire has a slant of more than 35° with 

 the sea surface. 



(12) Allow five minutes after lowering the instruments before releasing the 

 first messenger. 



(13) Determine time interval for bottom bottle to be tripped at various depths 

 and do not start hoisting until this interval has expired. 



(14) Do not capsize bottles when removing them from wire. 



(15) Read thermometers with great care and note registration in record book. 



(16) Each bottle should then be returned to its properly marked stall in the 

 rack in order of sequence. 



(17) When last bottle is being hoisted on board, or before, plot the temperature 

 readings of the various depths of observation on cross-section paper. If the 

 values do not form a smooth curve characteristic for the time and place, repeat 

 suspicious observations immediately before leaving station. Abilitj' to detect 

 errors in temperature curves comes with experience. 



(18) Citrate bottles should be clearly marked to indicate the station number 

 and the particular depth from which filled. 



(19) Stoppers on citrate bottles should be absolutely air-tight. 



(20) Coach oceanographic party in teamwork. 



(21) Determine salinity of water samples by running them through the electric 

 salinity tester on board and in accordance with instructions for same. 



(22) Test salinity values on cross section for smooth curve. 



(23) Apply to stem temperature of deep-sea thermometers the proper correc- 

 tion for auxiliary thermometer reading. 



(24) Obtain density values by entering temperature and salinitj- graph. 



(25) Test densities for smooth curve on cross-section paper. (See fig. 27, p. 88.) 



