HYDROLOGICAL METHODS 213 



when it was reversed. This w^as remedied by the addition of a small weight, after which 

 no further difficulty was experienced. 



The Knudsen full-speed water bottle^ was tried on a number of occasions in the 

 'Discovery', but it was found that the manometers, which indicate the depth at which 

 the samples are taken, failed to give accurate records. If this defect, which possibly 

 does not exist in other instruments of the same pattern, can be overcome, this type of 

 bottle should give valuable results. 



The thermometers used with the water bottles were all made by the firm of Richter 

 and Wiese of Berlin, who also supplied the unprotected thermometers which were used, 

 as explained on p. 198, for determinations of depth. The simple thermometers used with 

 the Nansen-Pettersson bottle are of two patterns, one having a range of — 2° to 

 + 32° C, for use in tropical and sub-tropical waters, and one with a range of -^ 2° 

 to + 20° C. for use in colder areas. The reversing thermometers have a range of — 2° 

 to + 16° C. All the thermometers are scaled to o-i° C. and provided with certificates 

 from the Physikalisch-Teknische Reichsanstalt. 



In all work at stations the Nansen-Pettersson bottle was used for taking surface 

 samples. For observations between stations, when the ships were under way, a sample 

 was taken in a leather bucket and the temperature read with thermometers made by 

 Messrs S. and A. Calderara. These thermometers are scaled to 0-2" C. and have 

 certificates from the National Physical Laboratory. 



Transparency or Secchi Disc. This apparatus, which is used to determine the 

 transparency of the water, consists merely of a circular disc, painted white and 50 cm. 

 in diameter, which is lowered in a horizontal position to the depth at which it just 

 ceases to be visible. It is extremely easy to manipulate, but recent investigations have 

 unfortunately shown that the reliability of the method is open to doubt. In studying 

 the penetration of light into sea water, Poole and Atkins- have endeavoured without 

 success to correlate observations made with the Secchi disc with those obtained with 

 their photo-electric apparatus. 



Hydrological Methods.^ Water samples and temperatures in routine work were 

 usually taken at depths of o, 10, 20, 30, 40, 50, 60, 80, 100, 150, 200, 300, 400, 600, 800 

 and 1000 m. and thence at 1000 m. intervals to the bottom. Samples of water for deter- 

 mination of salinity, hydrogen-ion concentration and phosphate-content were taken in 

 6 oz. swing top milk bottles, and tie-on labels were used giving particulars of station 

 number, date and depth. Samples for oxygen work were preserved in numbered glass- 

 stoppered bottles, each with a certificate of its volume, placed in a corked outer flask. 



The analysis of water samples was for the most part carried out at the Marine 

 Biological Station, South Georgia. In the 'William Scoresby' with her scanty 

 accommodation it was not possible to do any of this work, but in the 'Discovery' 



^ Knudsen, Piih. Ciir., Cons. Explor. Mer, 50, pp. i-ii, 1909. 



2 Poole and Atkins, Joiirn. Marine Biol. Assoc, n.s., xv, p. 480, 1928. 



3 We are indebted to Mr H. F. P. Herdman for assistance in this section. 



