98 THE VOYAGE OF H.M.S. CHALLENGER. 



depths, protecting them from pressure by enclosing them in strong metal tubes with 

 a top firmly screwed on. This method was extremely uncertain and generally failed. 

 The tube generally came up quite full of water, indicating that it had afforded no 

 protection to the instrument inside it. In some instances Walferdin's thermometer, 

 which is a straight-tubed instrument, and not curved like Six's, was used entirely 

 enclosed in a glass tube hermetically sealed. In this way, of course, complete protection 

 was afforded so long as the glass tube did not collapse. 



The method of protection used in the case of the thermometers supplied to the 

 Challenger has been described above. It consists in encasing the true thermometer bulb in 

 another bulb partially filled with liquid to facilitate transmission of heat. The remainder 

 of the space is filled with the vapour of the liquid. Any compression therefore which 

 might be suffered by the outer bulb would produce no rise of pressure in the space 

 between the two bulbs, and would therefore not be transmitted to the inner bulb. 



The effect of pressure on a glass vessel is to produce compression and diminution of 

 internal volume while it lasts. When the bulb of a thermometer is compressed and its 

 capacity diminished, the liquid contained in it is squeezed up into the stem, and the 

 top of the column stands higher than it did before, so that the compression of the bulb 

 produces the same effect as a slight rise of temperature. 



If now the thermometer be a self-registering one, and it be sunk to a certain depth 

 in a sea of uniform temperature identical with that of the thermometer, the index or 

 recording mechanism will indicate the rise of the thermometric column in the tube due 

 to the compression of the instrument. If the same thermometer, at the same temperature 

 to begin with, be carefully warmed, exactly the same apparent effect will be produced, 

 namely, the thermometric column will rise, and when the temperature has risen to a 

 certain height, it will place the index in exactly the same position as was the case when 

 it was sunk in the sea of uniform temperature. If in the latter case the effect of pres- 

 sure be neglected, we shall ascribe to the water at the particular depth a temperature 

 higher than the true temperature by the thermometric equivalent of the shift of the index 

 produced by the pressure of the column of water. 



It does not require demonstration to show that the apparent effect of pressure on a ther- 

 mometer will be almost wholly due to its effect on the bulb. The stem suffers compression 

 also, but the apparent effect so produced is negligible compared with that due to the com- 

 pression of the bulb. Hence when Six's thermometers had to be protected from pressure, 

 it was held sufficient to protect the bulb. There seems to be considerable uncertainty 

 as to who first proposed and carried out the preparation of thermometers with a double 

 bulb, but they were certainly used on board H.M.S. " Cyclops " by Captain Pullen ' in 

 1858, and there seems to be good reason for believing that the thermometers used by 

 Sir John Ross in 1818 were protected by the same or some similar device. 



1 Phil. Tram., vol. clxv. pp. 608, 609, 1875. 



