220 



BULLARD 



[chap. 1 1 



or a vibrating reed may be used between the bridge and the ampHfier. A 

 system of relays or a rotary switch is necessary to connect the thermistors 

 successively to the bridge; calibrating resistors may be included in the cycle so 

 that the resistance of the thermistors is obtained by interpolation and not by 

 assumiiig the zero and sensitivity of the system to be known. A check on the 

 resistance of the thermistors when at the same temperature may be obtained 

 by noting the readings when the instrument is suspended a short distance 

 above the bottom. 



When a probe weighing 160 kg in air and 130 kg in water strikes the bottom 

 at a speed of 3 m/sec and penetrates 5 m, 1700 cal of heat are produced. This 

 heat appears partly in the probe and partly in the sediment ; if it all appeared 



20 

 Time (min) 



Fig. 2. Variation with time of the temperature difference between the top and the bottom 

 of a probe 2.7 cm in diameter. The full-line curve represents the theoretical ex- 

 pression (1). (After Bullard et al., 1956, fig. 2.) 



in the probe used in Discovery II it would raise its temperature by 0.8°C, which 

 is considerably larger than the temi^erature difference to be measured. In 

 practice part of the heat goes into the sediment. 



The time taken for heat to spread from the surface of the probe to the centre 

 is only a few minutes and is much shorter than the time needed for the probe 

 to reach temperature equilibrium with the sediment. If the lower end of 

 the probe is heated above the temperature of the sediment into which it has 

 penetrated the temperature will rise for about 3 min, pass through a maximum 

 and then fall towards its final value as shown in Fig. 2. 



It is not practicable to leave a probe in the bottom for more than about 

 40 min ; after this time the temperature gradient in the probe usually exceeds 

 that in the sediment by 10 to 15%. It is, therefore, desirable to extrapolate the 

 observed temperature differences to get the value which would be obtained 

 after a very long time. In making this correction it may be assumed that any 

 cross-section of the probe is at a uniform temperature and that conduction is 



