CHAP, v.] GENERAL CONCLUSIONS. 259 



nieaii of two observations ; a string of thermometers was then 

 sent down in detachments, to avoid the risk of too great a loss 

 in case of an accident, at intervals of 100 fathoms, to within 100 

 fathoms of the bottom, or more usually to a depth of 1500 fath- 

 oms — considerably beyond the uniform layer. 



Such observations gave us a very fair idea of the distribu- 

 tion of temperature along a section, and the general course of 

 groups of lines joining points of equal temperature along the 

 section gave very delicate indications of any general rise or fall. 

 The word " isotherm " having been hitherto so specially appro- 

 priated to lines passing through places of equal temperature on 

 the surface of the earth, I have found it convenient, in consid- 

 ering these questions of ocean temperature, to use the terms 

 " isothermobath " and " isobathytherm ;" the former to indicate 

 a line drawn through points of equal temperature in a vertical 

 section, and the latter a line drawn through points of equal 

 depth at which a given temperature occurs. Isothermobaths 

 are shown in schemes of a vertical section, such as those in 

 Plates v., IX,, XI., etc. ; isobathytherms are, of course, pro- 

 jected on the surface of the globe. All the temperature obser- 

 vations have been made with the modification of Six's register- 

 inff instrument known under the name of the Miller-Casella 

 thermometer ; and this instrument, although a great advance 

 upon any other hitherto constructed, is essentially uncertain 

 and liable to error from various causes ; thus even a slight jerk 

 causes the index to move slightly either up or down, and an 

 observation is in this way very frequently vitiated. In almost 

 every serial temperature sounding, one or two of the thermom- 

 eters were evidently adrift from some such cause. There was 

 an excellent proof that these eccentricities did not always de- 

 pend upon differences of temperature. Yery frequently, es- 

 pecially at considerable depths, where the differences were very 

 slight, thermometers sent to greater depths gave indications 

 higher than those above them. There may be no absolute rea- 



