THE 



VOYAGE OF H.M.S. CHALLENGEE. 



THE PRESSURE ERRORS OF THE CHALLENGER THERMOMETERS. 



By Professor TAIT. 



THOUGH the contents of the following paper have been, with the sanction of Sir 

 "\Vyville Thomson, communicated at intervals during the last two sessions, and in par- 

 ticular on April 4th, 1881, to the Royal Society of Edinburgh, they are now published 

 for the first time. The brief abstracts which have appeared in some scientific journals 

 have given an inadequate, and by no means accurate, account of my method and results. 



The subject is the reduction of the deep-sea observations which were made on the 

 Challenger, in so far as these are affected by pressure. The thermometers employed 

 had protected bulbs ; but the stems, in which there were certain aneurisms, 1 were 

 wholly unprotected. The determination of the necessary pressure corrections is of great 

 importance, especially in the bearing of the results upon ocean circulation and other 

 grand points of physical geography : and when, at Sir Wyville Thomson's request, I 

 undertook the inquiry, I resolved to carry it out with a degree of accuracy suitable at 

 once to the capabilities of the thermometers employed and to the magnitude of the issues 

 involved. 



In the course of my work several improvements, which may be useful in future in- 

 vestigations of a similar kind, have suggested themselves ; but my primary object was 

 simply to find how to obtain the most trustworthy results from a set of observations 

 already made, the instruments with which they were made having been put in my pos- 

 session. 



The work has extended through a very considerable time, having occupied my leisure 

 moments for a large part of each of the last three years. The nature of the difficulties 

 which were successively met with and overcome will be easily seen from what follows 

 without farther preface. 



The whole matter looks uncommonly simple now that the instrumental and other 



1 From Hvffji*, a widening or swelling ( and rfr ? i/ f ) ; not, as is sometimes stated, A-m/yn (without sinews). 

 Hence the word is correctly used for the peculiarity in the thermometers. 



