602 CYCLES OF ORGANIC AND INORGANIC SUBSTANCES 



and one as large as 0.12° C. The paired determinations of depth 

 at 1800 and 1900 m nominal are \-ery concordant but the difference 

 between them is 105 m, 5 greater than the distance between them 

 (100 m) and 25 m greater than the depth intervals observed be- 

 tween the shallower bottles. There is no doubt that while the 

 messenger was running between the bottom pair of bottles, the 

 whole hoist sank by at least 5 m or, more likely, by 20 m. 



Wiist (1932) studied the mean errors of thermometric measure- 

 ments of depth during the Meteor expedition. At 1200 m he esti- 

 mated the mean error as about ±7 m. In terms of precision, 

 though not necessarily of accuracy, it is possible to improve on 

 this by using paired unprotected and paired protected thermome- 

 ters on each bottle. 



In Table II if a uniform wire angle between 1400 and 1800 m 

 nominal is considered to extend to 1900 m nominal, the depth of 

 the bottom bottle would fall short by 18 m of that measured by 

 paired thermometers. Since we have a number of other hoists 

 with large wire angles which yield a similar answer, the interpreta- 

 t'DU of the results as due to vertical yawing seems reasonable. 



Now that we have more accurate salinity measurements, the 

 error introduced by vertical yawing is more serious than appears 

 at first sight. There is not only error in depth but temperature and 



Table I. Hoist of Average Qualit\- Made with Small Wire Angle (10°) 

 Station Bedivere. 20 April, 1958 



Depth 

 Wire Observed Observed Assuming Error in 



Depth, Temperature, AT, Thermometric A/). X'ertical Depth, 



m °C °C Depth, m m Deep Wire, m 



m 



1900 4.13 0.00 1875.8 5.2 1871 -2 



4.13 1870.6 



2000 4.06 0.01 1969.0 — 1971 +2 



4.05 



2100 3.73 0.01 — — 2071 — 



3.72 



2200 3.66 0.01 2161.1 10.9 2171 -4 



3.67 2172.0 



2300 3.44 0.00 2273.8 1.3 2271 +3 



