figure 23. Each occupation of the section is thus characterized by 

 a single T-S point, As the intermediate water is assumed to 

 undergo an annual change the time of year should have some 

 effect. In figure 23, however, no discrimination has been made 

 as to date of occupation, although the actual dates of the occupa- 

 tions vary from June to August. There appears to be but little 

 relationship between either temperature or salinity and the date 

 of occupation within this summertime range. The mean date of 

 the post war series of occupations is some 5 1 /? days later than 

 that of the pre war series. The mean T-S points for the periods 

 1934_41 and 1948-56 have been shown within squares and differ 

 but little in temperature with the salinity lower in the later series. 



A similar numerical analysis of the temperature and salinity 

 of the deep water was attempted. Whereas in the intermediate 

 water the remarkable uniformity permitted the grouping of all 

 observations without regard to depth, in the deep water there 

 is a sufficient vertical gradient, especially in temperature, to re- 

 quire a reduction to a common depth. All stations where the 

 observations extended to a depth of 2,000 meters or more were 

 considered, the individual station curves of vertical distribution 

 of temperature and salinity being scaled for values at 2,000, 2,500, 

 3,000 and 3,500 meters. For each occupation all scaled values for 

 each of these levels were averaged to get the mean temperatures 

 and salinities plotted in figure 24. Near bottom along the section 

 there is a characteristic tendency for the isotherms and isohalines 

 to parallel the bottom. No adjustment has been made for these 

 horizontal gradients. Because of the symmetry of the bottom and 

 the spacing of the stations the shallower levels are less affected 

 but the results in the deeper levels, especially the 3,500-meter 

 level are somewhat erratic from the small number of data and 

 the fortuitous spacing of stations with respect to the 3,500-meter 

 isobath. The average temperatures for the periods 1934-41 and 

 1948-56, plotted against the corresponding average salinities in 

 figure 25, show about the same salinity at 2,000 meters but for 

 deeper levels show a lower salinity for the later period. In tem- 

 perature the change is small with an increase at 2,000 meters, little 

 change at 2,500 meters and a decrease at 3,000 meters. The small 

 number of stations extending to 3,500 meters leaves this level 

 doubtful. 



The values of o- t corresponding to average temperature and 

 average salinity for individual occupations of the section are 

 plotted against the year of occupation in figure 26. Again it must 

 be kept in mind that the 3,500-meter level is doubtful because of 

 the small number of data. 



62 



