DEEP-SEA REVERSING THERMOMETERS 239 



Fo differs from thermometer to thermometer. Some years ago its value varied widely (reference to 

 Table i shows a variation between 49^ and 265^) but it has now been stabilized— at any rate in British 

 instruments— within what appear to be reasonable limits, though the size of the expansion chamber is 

 still determined to some extent by convenience of construction, and no two manufacturers are likely 

 to agree on an ideal size. It is, however, generally agreed that high volumes should be avoided. It 

 might perhaps be thought that the different weights of mercury below the constriction with different 

 values of Fq would be reflected in the performance of individual thermometers; this, as mentioned 

 above, does not appear to be borne out in practice. So far as we can see there is no connection between 

 Fo and the performance of a thermometer, other than the tendency of those with a high I'o for the 

 mercury to stick in the expansion chamber. 



Incidence of Failures 

 In Table 2 the number of successful temperature observations is related to the number of failures. It is 

 based on {a) the scientific log-books kept in the ship for each voyage since 1925 which record all 

 deep-sea temperatures taken, together with the thermometer numbers; (b) the temperature correction 



Table 2. Number of observations and incidence of failures in German and British thermometers 



- - • ' • Failures 



Make of thermometer 



and number of 



observations 



German 



British 



Ship 

 Discovery 

 William Scoresby 

 William Scoresby 

 Discovery II 

 William Scoresby 

 William Scoresby 

 Discovery II 

 Discovery II 

 Discovery II 

 Discovery II 

 William Scoresby 

 Discovery II 



Commission 



1925-7 



1926-7 



1928-9 



1929-31 



1929-31 



1931 



1931-3 



1933-5 



1935-7 



1937-9 



1950 



1950-1 



German 



678 



150 

 926 



1777 



363 

 941 



2275 

 2486 

 2921 



762 

 84 



242 



British 



137 



1780 

 901 

 129 



3334 



III 



1861 



No. 

 24 

 18* 

 26 



38 

 II 



13 



55 

 53 

 45 

 25 

 2 



5 



Age (%) 



3-5 

 12 

 2-8 



2-1 



3-0 



1-4 

 2-4 



2-1 



1-5 



3-3 



2-4 



2-1 



No. Age (%) 



2-9 



0-8 

 4-4 

 77 

 i-i 



5-4 

 i-i 



14 



4ot 



lot 



37 



6 



21 



* Mostly one erratic thermometer. 



t Mainly one poor batch of thermometers. 

 . . . . .• .1 » r 



J iviainiy one poui uaiLn ui iijv^iiiiuiiiv,..v,i^. 



X Mostly due to one erratic thermometer from 1933/5 batch. 



Table 3 summarizes the figures in Table 2 for the whole period 1925-5] 



Table 3. Summary of Table 2 



Failures 



Make of thermometer 

 German (R. and W.) 

 British (N. and Z.) 

 Totals 



No. of 

 observations 



13.605 



8,253 



21,858 



No. 



315 

 132 



447 



Age (%) 



2-31 

 I -60 



2-04 



books kept from 1933 onwards ; and (c) a thermometer log kept personally by one of the writers of this 

 paper Separate figures are shown for British and German thermometers and failures have only been 

 classified as such when due to one or other of the following: mercury failing to break at the constric- 

 tion, mercury sticking in the expansion chamber when the thermometer is reset, and uncertain causes 

 of anomalous temperatures. No account has been taken of the erratic behaviour of thermometers after 



