The Distribution of Salinity (figs. 11-19, 21) 

 Salinities varied from maximum values 

 (>34.5%o) near the bottom of the southern 

 portions of northern Baffin Bay (sections E, F, 

 G and H) and in Lancaster Sound (section I) 

 to minimum values (<32.5%o) in near-surface 

 coastal areas (e.g. the western parts of section 

 F and H) and in Smith Sound (section A). In 

 the surface layers, salinity values were some- 

 what greater in the southern part of the study 

 area (>33.0%o) than in the northern part 

 (<32.0%o) (fig. 21). 



The lack of appreciable horizontal salinity 

 variations, relative to temperature, is similar 

 to that found during previous summers, and is 

 a major factor in the use of temperature as a 

 means to trace water masses in the region 

 (Muench, 1971a). 



The Circulation 



The total circulation is the sum of both baro- 

 clinic (internal) and barotropic (external) 

 modes. The baroclinic circulation is presented 

 both as a dynamic topography (fig. 22) and as 

 computed volume transports (Table I). Both 

 were obtained from the oceanographic data 

 using the dynamic method and are subject to 

 certain errors inherent in that method. These 

 errors include: 



a. Errors in computation of dynamic depth 

 anomaly due to inaccuracy in measurement of 

 temperature and salinity and its eff'ect on den- 

 sity determination. If all errors in temperature 

 and salinity measurement were maximum and 

 in the same sense, the resulting dynamic depth 

 errors would be greater than the differences 

 involved. It is assumed, since there is no reason 

 to believe otherwise, that the eri'ors occurred in 

 a random fashion and cancelled each other. 



b. Dynamic depth variations not connected 

 with the baroclinic circulation due to internal 

 waves on the pycnocline. The data were insuffi- 

 cient to determine presence or absence of inter- 

 nal waves and allow an estimate of the error 

 induced by them. 



c. Errors due to the neglect, inherent in the 

 dynamic method, of non-linear, frictional, and 

 time-dependent terms in the equations of 

 motion. The data were insufficient for evalua- 

 tion of these terms. 



Use of Defant's (1961) method revealed, as 

 it did in data from previous summers (Muench, 



1971a) , that a suitable level of no motion oc- 

 curred at 500 db. Distinguishable features of 

 the surface topography relative to 500 db (fig. 

 22) include a variable southward to westward 

 flow in the north central region west of Thule, 

 a southerly flow in the southwest region off 

 Baffin Island, and net eastward flows through 

 Jones and Lancaster Sounds. A dynamic topog- 

 raphy of the 200 db relative to the 500 db sur- 

 face (not shown) indicated a baroclinic circula- 

 tion at 200 db similar to that at the surface. 



Geostrophic transports through sections A, 

 C (southern channel only), I, and H were com- 

 puted. The remaining sections were too shallow 

 to allow computation relative to a reference 

 level determined by Defant's method. The data 

 from the southern channel of section C were 

 limited to the northern and central portions of 

 the channel and may not be representative. The 

 results (Table I) corroborate the net eastward 

 flows through Lancaster and Jones Sounds and 

 the net southward flow through the system sug- 

 gested by the dynamic topography. In addition, 

 a small but net southward baroclinic flow was 

 occurring through Smith Sound. Lack of a vol- 

 ume balance may have been due to any or all 

 of the errors noted above. These computations 

 neglected, moreover, any barotropic transports 

 which may have been occurring and which were 

 of unknown magnitude. 



Both the dynamic topography and the net 

 southward flow through the system, compared 

 to a previous mean of 2.0 x 10« m.^/sec (Muench, 

 1971a), suggest that the baroclinic flow in sum- 

 mer 1970 was weak throughout most of the 

 region relative to that during preceding sum- 

 mers. On the other hand, the transport through 

 Lancaster Sound appeared to be large relative 

 to that for preceding summers. 



Dynamic computations were inadequate for 

 determination of baroclinic currents at 300- 

 500 m. depth and below. However, the 27.5 



3 



