900-foot bathythermograph and the salinity samples were 

 obtained from the Nansen bottle. The depths of the samples 

 are the length of wire payed out according to the counter on 

 the bathythermograph winch. 



Accuracy of Observations . The temperature cannot be 

 read from bathythermograms to a greater accuracy than 

 10.1 degree F.*, the instrumental error is estimated at ±0.1 de- 

 gree F. There may also be some difference between the actual 

 depth at which the water sample was taken and the depth at 

 which the bathythermogram was read. It is thus probable that 

 the temperatures corresponding to salinities given here are 

 accurate to about 0.1 degree C. The depths at which the water 

 samples were obtained from the counter of the bathythermo- 

 graph winch are fairly accurate at shallow depths. 



The water samples were titrated with silver nitrate by 

 the Knudsen method^ to determine chlorinity and salinity. 

 The salinities are expressed in parts per thousand (o/oo) 

 and are accurate to 0.02 o/oo. 



The density of the water was calculated from the tempera- 

 ture and the salinity by means of Knudsen's Tables. 3 in 

 discussing density in this report the common notation, <j , 

 will be used. This density is defined by the relation: 



cr t = 1000 (p 9iS>0 ~ 1) 



where/55 >s ,ois the density of a parcel of water of salinity s, 

 temperatures, and pressure (one atmosphere). Because of 

 the low temperature range encountered, the 0.1 -degree C. 

 error does not introduce a serious error in the computed 

 density. Within the temperature and salinity range encoun- 

 tered, the probable error in temperature corresponds to an 

 error of less than 0.01 ino^. 



Results . Earlier in this report, there is a discussion of 

 the results of temperature observations made with bathy- 

 thermograms. The general conclusions which apply to the 

 over-all temperature structure (as given in the bathythermo- 

 gram discussion) apply also to the temperature distributions 

 that are presented here. The temperature-depth structure 

 for the six hydrographic stations is shown in figure 21. In 

 general, these temperature traces show certain charac- 

 teristic trends. At all stations, except station C-l, a minimum 



40 



