34 The Sea-water ami its Physical and Chemical Properties 



Sampling bottles and thermometers are the most important of the instruments used 

 at an oceanographic station. To be suitable for series observations the sampling bottle 

 must be as light as possible; while still having sufficient capacity, it must allow free 

 circulation of water and it must function and close reliably. There are many differen 

 models of sampling bottles. They are all lowered open, allowing the water to pass 

 through freely as the bottle sinks and are closed automatically for hauling to the sur- 

 face. The most successful design is that of Nansen with two plug valves. The series 

 water bottle used by the "Meteor" Expedition 1925-27 was constructed on the same 

 principles but was a little larger and had a number of minor improvements. This water 

 bottle and its function is illustrated in Fig. 19 (WiJST, 1932). It had a capacity of 

 1250 cm^, weighed 44 kg (with thermometer frame approx. 5 kg) and had an over- 

 all length of 75 cm. Among the older designs may be mentioned that of Ekman 

 (1905) with improvements by Knudsen (1923) and a special 4 1. water bottle {'"Meteor" 

 Report, 4, No. 1, 1932). 



Small 100-200 cm^ bottles of ordinary green glass are suitable for storage of 

 water samples (for chlorine titration and analysis) since they have been found by the 

 investigations of Helland-Hansen and Nansen to have very slight solubility; they are 

 fitted with a patent stopper with a porcelain head carrying the sample number. 

 Before use the bottles must be boiled, cleaned with chromic acid-sulphuric mixture, 

 rinsed with distilled water and very carefully dried. 



A definitive programme has been worked out for the work required at each oceano- 

 graphic station and this has been found to be very successful as, for instance, during 

 the "Meteor" Expedition 1925-27, and has been described in Vol. 4, No. 1 of the 

 ''Meteor'' Report. It is worth mentioning particularly that a machine and an obser- 

 vations schedule containing everything of importance in the working programme for 

 the series should be kept for each oceanographic station. Very often the results of an 

 oceanographic series depend on the careful compilation of the machine and observa- 

 tions schedules. Apparently unimportant details may become important later during the 

 interpretation of the observations and can contribute to the uniformity and homo- 

 genity of the observations. 



2. Temperature Determination for all Layers of the Ocean 



The determination of the temperature of the surface layer of the sea offers little 

 difficulty. A sample taken from water collected in an ordinary bucket, lowered into 

 the sea for a short lime while the vessel is under way, is put immediately in a shady 

 place and its temperature is taken with a sensitive thermometer while at the same time 

 it is kept stirred. The water sample must be drawn from as far forward as possible (on 

 steam ships forward of the condenser exhaust). See Lumby (1927) on the measurement 

 of surface temperatures and the collection of suitable water samples. New surface 

 sampling bottles have been designed by Sund (1931) and improved by Schumacher 

 (1938). 



The determination of the temperature of the deeper layers of the sea is considerably 

 more difficult, and this also needed the work of almost a decade to reach an accuracy 

 suitable for scientific requirements. In the upper layers temperatures correct to 0-1 °C 

 are usually sufficient, but in the deep layers the variations both horizontally and ver- 

 tically are usually so small that an accuracy of 0-01 °C is needed to get some idea of 



