ferred. The work of Fjeldstad, Defant, and others ^ indicates that 

 an unlimited number of internal waves of differing phase, period and 

 order may be present in the usual sea condition of continuous vertical 

 density gradient. Such analyses as have been made emphasize the 

 waves of longer period, of the order of magnitude of several hours. 

 This may be because of the sampling methods which so far have been 

 available. Figures 22 to 35 indicate that much shorter period waves, 

 of the order of magnitude of several minutes, also are present. For 

 the study of such waves a rapid response water sampler or salinity 

 measuring device is necessar3^ The Nansen type water bottle is 

 not a rapid flushing bottle but instead requires somewhat more than 

 6 minutes to trap a sample representative of its surroundings.^ Evi- 

 dently then it cannot be used with a rapid response thermometer such 

 as the TD. However, just as the reversing thermometer averages 

 the temperature over the several minutes prior to its reversal, so also 

 does the Nansen type water bottle and the two together eliminate 

 many of the uncertainties in dynamic topography which might other- 

 wise arise from short period internal waves. The undulations with 

 periods of several hours, of course, remain in the topography developed 

 from measurements made with the reversing thermometer-Nansen type 

 bottle methods and may be the cause of some of the peculiarities in 

 our topographic charts which have been difficult to explain. While 

 remembering that the dynamic topographic charts have been in 

 sufficiently good accord with observed drifts of bergs to be of great 

 practical use, the degree of reliance which may be placed on such 

 computations as volume and heat transport and mean temperature 

 derived from such observations is in doubt. 



It has been found that in the Grand Banks region Labrador Current 

 water and Atlantic Current water are present as water masses having 

 characteristic temperature-salinity relationships and in addition the 

 mixed water formed from these parent water masses is usually present 

 in sufficiently uniform proportions of each of its components so that 

 it may be regarded as a virtual water mass. Of the data collected 

 during 1953, those from the first and fourth surveys were from the 

 same area which has been the source of the T-S data in the past 

 and so provide a good basis for comparison. Except for its southern 

 end, the area covered by the second survey, which included the 

 northeastern slope of the Grand Banks, is north of the area from 

 which earlier data have been considered. The third survey was 

 comprised of the Bonavista triangle only and was not analyzed for 



' As summarized by Sverdrup, Johnson, and Flemine in "The Oceans" pp. 685-602 (1912), Prentice-Hall, 

 New York. 



* See R. L. I. 'Fjarlie "A seawater sampling bottle" Pacific Oceanographic Group, Joint Committee on 

 Oceanography P. O. G. fileN7-7 (15 Aug 1952') Nanaimo, B.C., wherein the author describes the performance 

 of a new bottle desiened to flush rapidly. He compared the new bottle with other types of bottles at 0, 2, 

 4, and 6 meters. The .salinity at 2 meters was essentially the same as at the surface. A replot of his data 

 for only the 4 and 6 meter levels shows the Nansen type bottle to require about 6 minutes to trap samples 

 Of the same salinity trapped by the new bottle under the existing sea conditions (not stated). 



76 



