extended from the surface to as near tlie ])ottom as was practicable. 

 The 31 stations that comprised a part of the Gulf Stream, 1960, survey 

 were similarly taken except that a maximum depth of 4,000 meters 

 was imposed. At the remaining- stations, observations were limited 

 to the upper 1,500 meters. The intended depths of observations were 

 0, 25, 50, 75, 100, 150, 200, 300, 400, 600, 800, 1,000 and thence by 

 500-meter intervals except for the Gidf Stream, 1960, stations which, 

 below the 1,000-meter leve', were 1,250, 1,500, 1,750, 2,000, 2,250, 

 2,500, 3,000, 3,500 and 4,000 meters. Temperatures were measured 

 with protected deep sea reversing thermometers. Most of the ther- 

 miometers used were of Richter and Wiese manufacture, with a 

 scattering of Negretti and Zambra, Kahl vScientific Instrument G!orp. 

 and G. M. Manufacturing C^o., instruments. Depths of observation 

 were based on unprotected reversing thermometers made by Richter 

 and Wiese and bv Kahl. As in the past, a program of intercomparison 

 of the protected thermometers was carried out in the field to permit a 

 close control to be kept on the functioning of individual thermometers. 

 The thermometers were used in pairs and one of each pair was shifted 

 periodically so that any one thermometer was eventually jjaired with 

 several other thermometers. Mean differences between the corrected 

 readings of pairs of thermometers were adjusted to zero by the applica- 

 tion of consistent corrections. In all, 2,150 comparisons were made. 

 Of these, 683 involved thermometers having a range of —2° to +20° 

 or more and had a standard deviation (difference between the corrected 

 readings of a pair of thermometers) of ±0.013°. For the ranges of -f3 

 to +13 (311 comparisons) and —2° to +8° (1,156 comparisons), 

 where the scale is more open, the standard deviations were +0.007° 

 and ±0.008° respectively. Inasmuch as most of the observed 

 temperatures listed in the Table of Oceanographic Data are based on 

 the corrected readings of a pair of thermometers and since many of the 

 thermometers used had recent laboratory comparisons with thermom- 

 eters tested by the National Bureau of Standards, it is considered that 

 the tabulated observed temperatures have an accuracy of 0.01° (\ 



As in past years, the routine salinity measin-ements were made with a 

 Wenner salinity bridge with a precision of ±0.005°/oo- This year the 

 observed salinities listed in the Table of Oceanographic Data are con- 

 sidered to have an accuracy equal to this precision. As mentioned in 

 Bulletins Nos. 44 and 45 of this series, efforts to arrive at a better 

 standardization of salinities have been in progress for some time. 

 During the 1959 ice season, 11 large samples (ca. 12 gallons each) of 

 actual surface sea water, distributed over the salinity range from 31 to 

 36.4°/oo, were collected from the Ice Patrol operating area and stored 

 in polyethylene drimis. In February 1960, these samples were di- 

 vided, some portions of each being transferred to polyethylene and 

 glass containers for further storage, and other portions being measured 

 after a few days storage in glass bottles. The measurements consisted 



33 



