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



difficulties have been discovered and met, and I have recently repeated the investigation 

 in a tenth of the time it originally cost me. The data given in columns 7, 8, 9, 12, and 

 1 3 of the Table in Appendix E below, are (with the exception of those for one thermo- 

 meter in which the mercury column had been accidentally broken) those obtained in this 

 repetition of the inquiry. They were found to agree so well with the earlier data that it 

 was considered unnecessary to print these. 



I found myself at the beginning of the inquiry very much in the position of a chemist 

 who has given to him a mixture containing half a dozen absolutely unknown elements, all 

 in very small and in nearly equal quantities, and who is required to determine the nature 

 and properties of each, and also the proportions in which they occur in the mixture. 



A great many very curious offshoots have sprung from the inquiry, some of which 

 are of real scientific importance. For instance, the determination of the amount of heat 

 developed by exposing to very high pressures, under different circumstances, various 

 kinds of substances. This question, so far as I am aware, has as yet been treated (even 

 theoretically) only for moderate pressures. Again, there is the very curious question, 

 What is the cause of the breaking of a piece of glass or other fragile body, under 

 hydrostatic pressure ? Does it break in consequence of uniform compression, or of 

 shearing, or of extension only ; and at what amount of compression, or shear, or exten- 

 sion, does it give way ? And there is the very important practical question of the 

 accurate measurement of pressures greater than can readily be compared with the weight 

 of a tall column of mercury. Amagat has successfully worked with a column of mercury 

 of more than 1000 feet in height, corresponding to a pressure of about 3 tons weight per 

 square inch. But there is a limit, to experiment in this direction, which he has nearly 

 reached. The simple and easily manageable apparatus described below has been found 

 capable of giving results of considerable accuracy up to pressures of 12 tons weight on 

 the square inch, and will probably be applicable much farther. 



After some consideration I have decided to give, first, a general account of the whole 

 work in terms which will be easily intelligible to all readers ; and then to develop at 

 length special parts of the inquiry which have scientific interest, either pure or practical, 

 but which are not of a nature to be easily comprehended except by specialists. Of 

 course I reserve for the latter part the proofs (experimental or mathematical) of the 

 statements now to be made. 



For convenience, this subject is arranged as follows : 



The Pressure- Corrections supplied to the Challenger along imth the Thermometers. 



Construction of the Thermometers. 



Wholly protected Instruments. Their Defect. 



Individual Peculiarities of some of the Challenger Thermometers. 



