The Sea-water and its Physical and Chemical Properties 35 



the spatial variations in temperature. This accuracy is also necessary for the calculation 

 of densities accurate to the fifth decimal place. Deep-sea thermometers are thus 

 extremely accurate and sensitive instruments which cannot be handled skilfully just 

 by anyone. 



An ordinary thermometer suspended freely in the water will not show the correct 

 temperature since the pressure of the water will compress the thermometer bulb and 

 force the mercury to a higher level. It is therefore necessary to protect the thermometer 

 against the water pressure by enclosing it in a thick-walled glass tube. The part of the 

 tube surrounding the thermometer bulb is filled with mercury to improve the heat 

 transfer between the water and the bulb. Since the temperature usually decreases with 

 depth the instrument first used was a maximum and minimum thermometer con- 

 structed by Six and adapted for deep-sea use, and this was the classical instrument 

 used on the "Challenger" and the "Gazelle" Expeditions. Since 1874 the reversing 

 thermometer, first produced commercially by the firm Negretti and Zambra, has been 

 used instead, and with numerous modifications is still used at the present time as the 

 standard instrument for oceanographic temperature recording. This is a thermo- 

 meter with the capillary considerably constricted a little above the mercury bulb, 

 so that the mercury thread will break at this point when the thermometer is turned 

 through 180° and slide down to the other end of the capillary. The higher the tem- 

 perature when the thermometer is reversed the longer the mercury thread that is 

 broken off. This thread gives a direct reading of the temperature at that time when 

 read against a scale running in the reverse direction with appropriate corrections. The 

 accuracy of the thermometer is very dependent on the shape of the constriction. It 

 must, of course, be made so that the mercury thread always breaks at the same point 

 and it must be designed so that further mercury cannot follow the thread if the ther- 

 mometer passes subsequently through a warmer layer of water. All the initial diffi- 

 culties were overcome by the work of Richter (of the firm Richter & Wiese, BerHn) so 

 that the reversing thermometer is now a true precision instrument. The shape of the 

 constricted part of the capillary is shown in Fig. 20. Further details are given in the 

 ''Meteor'' Report, 4, No I, by Bohnecke (1932), and in "Oceanographic Instrumenta- 

 tion" (Rep. Conf. Rancho Santa Fe, Calif. 21-23 June 1952, p. 55). 



In use the reversing thermometer is enclosed in a suitable holder (a brass tube) 

 which is attached directly to a reversing sampling bottle or to a frame which can be 

 reversed at the desired depth (reversing frame, propeller frame). 



The reversing thermometer does not show the true temperature {in situ) directly 

 since it will have been brought back to the surface through layers of water at different 

 temperatures. After removal from the sampling bottle on deck it is placed immediately 

 in a water bath and allowed to adjust to the water temperature before it is read. To 

 show the temperature of the water bath every reversing thermometer is fitted with a 

 normal auxiliary thermometer. To correct the reading to the temperature in situ a 

 small correction given by the formula 



{T -t){r+ Kq) 

 6100 



J ^ {r~t){r+ V,) 



6100 



must be applied. In this equation T' is the uncorrected reading of the reversing ther- 

 mometer, t the reading of the auxiliary thermometer (bath temperature), Vq is the 



