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



those which lay above, should be equal to that of those lying below. This would have 

 shown a rise of temperature of the water in proportion to the increase of pressure, but 

 its amount would have been considerably under the truth. But the experimenters used 

 only the white spots ; and by the help of these drew the full line in the figure as 

 indicating their average ; thus obtaining a very slow increase of heating by pressure. 



By the help of Sir William Thomson's 1 formula for the heat developed by compression, 

 I have calculated what amount of heating should have been obtained in these experi- 

 ments, on the supposition that the pressure is applied with sufficient suddenness to let 

 the full effect be produced on the thermometers before there is any sensible absorption 

 of heat by the walls of the pressure vessel : and on the farther assumption (not, as will be 

 seen, justified by experiment) that there is no heating of the glass protecting sheath by 

 pressure. It is represented by the dashed line in the figure, and it is certainly very 

 remarkable that this line (the true one) runs through the group of rejected observations, 

 paying as it were no attention whatever to those which were retained ! The formula re- 

 ferred to is discussed in Appendix C below ; but, as will be seen at a later stage, the effect 

 on the protected thermometer is not due solely to the heating of water by compression. 



Captain Davis concluded from two sets of observations, one at 55F. and the other 

 about 39 F., that little attention need be paid to the heating of water by compression, 

 (obtaining, in fact, the dotted line), and thus that the effect observed in the hydraulic 

 press was due mainly to direct pressure, and would, of course, be experienced by the 

 thermometers when they were let down into the sea. 



The officers who managed the thermometers of the expedition, were, in conse- 

 quence, furnished with corrections for each thermometer, all of the order already 

 indicated, i.e., about half a degree for each mile under the surface of the sea. These 

 corrections were, of course, for the maximum side of each instrument. 



Consequent Correction for the Minimum Side. 



Looking at the thermometers, it seemed to me perfectly evident that this correction, 

 if it was to be applied at all, must be applied in very nearly the same amount both to the 

 maximum index, for which it was determined, and also to the minimum. Any difference 

 between these two must be due solely to the effects of temperature upon the column of 

 mercury which lies between the two indices, and of pressure on the tube containing that 

 mercury. Unless the heating effect were confined to the space between the indices, the 

 former is provided for by the graduation of the instrument itself; and it was quite 

 certain that the two together could not produce an effect amounting to more than a 

 small fraction of the degree and a half for three tons pressure. 



Therefore, as all the readings of the Challenger thermometers were taken from the 



1 " On the Alterations of Temperature accompanying Changes of Pressure in Fluids," by Professor W. Thomson 

 (Proc. R.S., June 1857). 



