THE PACIFIC OCEAN 



89 



a salinity which is lower than the salinity of both the 

 surface and the deep water. 



The deep water which fills the Pacific Basin below 

 a depth of 2000 meters, is, according to these sections, 

 of a very uniform character, with a temperature slightly 

 below 2° and a salinity somewhat above 34.6 per mille. 



Defant (1928) has pointed out the adyantage of dis- 

 tinguishing between two different horizontal strata in the 

 sea, namely: an upper stratum, the troposphere, within 

 which great variations of temperature and salinity in 

 horizontal and vertical directions are found, and to which 

 the most important currents are confined; and a deeper 

 stratum, the stratosphere, which is characterized by 

 small variations in temperature and salinity both in hor- 

 izontal and vertical directions, and consequently, by 

 very slow currents. In the Pacific, in accordance with 

 WQst, we may regard the isothermal surface of 10° as 

 separating the troposphere from the stratosphere. We 

 find then that the troposphere has a maximum vertical 

 extension of about 650 meters in the western Pacific and 

 of less than 500 meters in the central Pacific, and that 

 the stratosphere reaches to the curface of the sea north 

 and south of latitudes about 50° north and south, respec- 

 tively. The circulation within the troposphere, also in 

 accordance with Wust, will be called the warm-water 

 circulation, and the circulation within the stratosphere 

 will be called the cold-water circulation. 



The Available Data 



The observations of the Carnegie are not so numer- 

 ous that by means of these we can undertake a complete 

 discussion of the physical oceanography of the Pacific. 

 Most of the observations were made north of latitude 20° 

 south. Observations south of this latitude are available 

 only from the South American coast to longitude 120° 

 west. Thus no stations were occupied in the greater 

 part of the South Pacific south of latitude 20° south, and 

 in the North Pacific great regions have not been visited. 

 Considering these wide gaps, it might appear desirable 

 to amplify the data of the Carnegie by means of data 

 from earlier expeditions, in order to make the best pos- 

 sible use of the existing material in the following dis- 

 cussion. None of the earlier observations from the Pa- 

 cific, however, are of the same quality as to accuracy 

 as the Carnegie data except later observations which 

 have been taken off the coast of California, in the Gulf of 

 Alaska, in the Japanese waters, and the observations of 

 the Planet between New Guinea and Japan. Most of these 

 observations do not reach to as great depths as those at 

 the Carnegie stations. They are well suited for the study 

 of details in the different regions but they contribute lit- 

 tle to the knowledge of the major features which enter in 

 the foreground when dealing with the work of the Carne - 

 gie . We shall, therefore, make use only of some stations 

 in the Gulf of Alaska which directly form a supplement to 

 the Carnegie stations in this region. 



As to the older observations, it has already been 

 emphasized that these are less accurate than the later 

 ones. In order to illustrate this the vertical tempera- 

 ture distribution at three Challenger stations and three 

 Carnegie stations (which were taken in approximately 

 the same localties) is represented in figure 12. It is 

 seen that the general features of the temperature dis- 

 tribution agree well, but the results differ considerably 

 in details. Comparing the stations Challenger 254 and 



Carnegie 143 we find that the Carnegie observations 

 show a decrease of the temperature at all levels, where- 

 as the Challenger observations show three inversions. 

 The two lower inversions, however, fall between levels 

 at which the Carnegie observed, and at the latter levels 

 agreement exists between the Challenger and Carnegie 

 temperatures. It is possible, therefore, that the inver- 

 sions existed when the Carnegie station was occupied 

 but escaped detection because the observations were 

 made at too great intervals. A comparison between sta- 

 tions Challenger 262 and Carnegie 139 shows that this 

 conception can hardly be upheld. The Carnegie station 

 again shows a decrease of temperature at all levels, 

 whereas the Challenger station indicates a succession of 

 intervals with small decrease or inversions. At this 

 station the intervals of the Carnegie observations are 

 again much greater than the intervals of the Challenger 

 observations, but a Carnegie observation was made at 

 the level at which a Challenger observation indicates an 

 intermediate minimum of l.°56. The Carnegie observa- 

 tion shows no such minimum but the value lies pricti- 

 cally on the straight line joining the two adjacent obser- 

 vations. The irregularities of the Challenger values 

 which occur above a level of 500 meters are actually of 

 the same order of magnitude as the irregularities at a 

 greater depth but they are less conspicuous because of 

 the rapid change of temperature with depth. The obser- 

 vations at the two stations Challenger 280 and Carnegie 

 87 agree, on the whole, very well but below a level of 

 800 meters the Challenger temperatures appear to be as 

 much as 0?3 too high. 



These three examples show that the reality of the 

 small inversions which were observed at great depths on 

 the Challenger must be doubted and the same also ap- 

 plies to the intervals with small temperature gradients 

 in the upper layers. Such irregularities are never found 

 at the Carnegie stations, as is evident from the temper - 

 ture curves of the following figures reproduced in I-B: 

 94, 100, 106, 112, 118, 127, 133, 142, 148, 154, 160, 166, 

 172, 178, 187, and 196. 



It is true, as already mentioned, that the Carnegie 

 observations have been made at greater intervals than 

 the Challenger observations, but if inversions at great 

 depths were as frequent as indicated by the Challenger 

 data, they would undoubtedly have been observed at 

 some stations and have changed the smooth course of the 

 curve. Considering this circumstance we cannot agree 

 with Wust in accepting the small inversions as repre- 

 senting actual conditions. The Carnegie observations 

 strongly indicate that, although the Challenger data give 

 a true representation of the major features of the tem- 

 perature distribution, the details cannot be relied on. 

 This result is in agreement with the conception of the 

 officers of the Challenger because, in the report on the 

 deep-sea temperature observations, smooth curves have 

 been drawn by means of the observed data and the values 

 scaled from curves have been given beside the observed 

 values. The smooth curves agree, on the whole, with 

 the Carnegie curves, and the results from the upper lay- 

 ers could be used for amplification of the Carnegie data, 

 but the deep-sea temperatures from the two expeditions 

 are not comparable. 



With the exception of the later expeditions referred 

 to previously, the observations of temperature by the 

 other expeditions which have cruised in the Pacific have 

 not been made by means of more superior methods. It 

 is not advisable, therefore, to combine the results of 



