SECT, lj WATER SAMPLING AND THERMOMETERS 127 



recorded by a distant reading thermometer or thermograph (Sverdrup et al., 

 loc. cit), but it must be remembered that the thermometer bulb, whether in an 

 intake pipe or in a pocket in the hull, may be 9 ft or more below the actual 

 surface, depending on the size of the ship. This is of little importance in the open 

 ocean where a homogeneous surface layer normally extends to a depth greater 

 than 5 m, but can give quite a fictitious reading in inshore waters. 



2. Thermometry 



Sverdrup, Johnson and Fleming (loc. cit) have given a very full account of 

 the measurement of temperature and depth by means of protected and un- 

 protected deep-sea thermometers. The thermometric method is still the 

 simplest, cheapest and most accurate way of determining the depth at which 

 a water-sampling bottle closes, even if the ship is drifting and there is con- 

 siderable stray on the wire. An unprotected depth thermometer on at least 

 alternate sampling bottles on a cast will give a more accurate figure for depth 

 of reversal than calculations based on the length of wire and the wire angle 

 between ship and surface. 



In the course of years the deep-sea reversing thermometer has become an 

 accepted instrument in oceanography. Nevertheless, it should still be con- 

 sidered a tribute to the glassblower's art rather than as a precision instrument 

 since it is at times liable to malfunction without apparent reason. Herdman 

 (1958) has analysed the performance of the reversing thermometers used by 

 the Discovery Investigation Committee between 1925 and 1939, and after- 

 wards by the National Institute of Oceanography. The percentage failure in 

 21,858 observations was 2.04, and there was little difference in this respect 

 between German and British instruments. Whitney (1952) has commented on 

 malfunctional behaviour in deep-sea thermometers and its correction and has 

 also remarked (1955) on non-linear behaviour of unprotected reversing 

 thermometers. 



References 



Fjarlie, R. L. I., 1953. A seawater sampling bottle. J. Mar. Res., 12, 21-30. 



Herdman, H. F. P., 1958. The reliability of deep-sea reversing thermometers. Discovery 



Reps., 29, 229-244. 

 Knudsen, M., 1929. A frameless reversing water-bottle. J. Cons. Explor. Mer, 4, 192-193. 

 Lumby, J. R., 1927. The surface sampler, an apparatus for the collection of samples from 



the sea surface from ships in motion. J. Cons. Explor. Mer, 2, 332-342. 

 Lumby, J. R., 1928. Modification of the surface sampler with a view to the improvement 



of temperature observation. J. Cons. Explor. Mer, 3, 340-350. 

 Sverdrup, H. IL, M. W. Johnson and R. H. Fleming, 1942. The Oceans: their Physics, 



Chemistry and General Biology. Prentice-Hall, New York. 

 Whitney, G. C., Jr., 1952. Notes on malfunctional behaviour and its correction in deep-sea 



reversing thermometers. Woods Hole Oceanographic Institution Rep. No. 52-29. 

 Whitney, G. C, Jr., 1955. A note on non-linear behaviour of unprotected reversing 



thermometers. Woods Hole Oceanographic Institution Rep. No. 55-62. 



