434' APPENDIX. 



the southern half of the tropical seas had already 

 assumed an autumnal temperature, since we ob- 

 serve here, in 17° S., the same warmth (18J° R.) 

 as in 30° N. 



The difference of the seasons, of the places, and 

 of tlie depths in which these observations were made, 

 allow us neither an accurate comparison of them 

 among each other, nor with the small number of 

 facts collected by other observers so as to enable 

 us to deduce general and definite rules. Desirable 

 as it would be that these examinations of the tem- 

 perature of the air should be made in certain 

 parallels, for example, from 10° to 10° degrees, 

 circumstances cannot always be commanded, 

 though the excellent series of daily observations 

 which the preceding table contains, show what 

 may be done without waiting for calms. But the 

 observers might at least agree to make all their 

 experiments at the same depths, namely, at 25, 50, 

 100, 200 fathoms, in which the frequent deviation 

 of the plumb-line from the perpendicular direction 

 must be taken into account. Fiequent repetitions 

 of the observations, at every opportunity, is the 

 more advisable, because, in unexpected and ano- 

 malous results, it is the only means of guaran- 

 teeing their accuracy ; and we would particularly 

 recommend attention to a sufficient duration of the 

 experiment, and to frequent comparison of the 

 thermometrograph with an accurate thermometer, 

 at different temperatures. Lastly, the more detailed 



19 



