\ - ' 



» 



ft 





I 



h 



h 



I 



Sect. IX.] 



METEOROLOGY, 



•* 



83 



Much 



which the index may he restored to its place, 

 patience and many trials are often required for success. 

 An oil-lamp with a very small clear 



flam 



should De 



used 



Both the self-registering thermometers should be 



9 h. A.M. observation, as it is 



very improbable that the temperature at that hour should 

 be such as to obliterate either record of the preceding 





^-- 



24 hours. 



Double maxima and minima, when 



thev 



S' 



occur, if remarkable, should be recorded as supernumerary 

 and sepnrately in a diai-y, and their accompanying cir- 

 cumstances noted. 



The observer should he furnished with several other 



thermometers, all of sufficient delicacy to allow of es- 



timating tenth 



s 



of degrees, for observation of the 



temperature of the sea, or of the earth, ( 



on 



,^ 



rain, &c., and for a reserve in case of 

 accident. All should he compared with the standard. 

 That in habitual use for the sea temperature should be 

 defended from accident in the act of immersion by a wire 



/ 



guard. 



The tliermometcr for solar radiation should have its 

 bulb blackened with a coat of Indian ink. It should be 

 defended from currents of air by enclosure in a. glass tube : 

 and it would add infinitely to the value of a series of 

 observations made with it if this tube were exhausted and 

 hermetically sealed. Its exposure to the sun should be 

 perfectly free and full, and it should he suspended in 

 free air, quite out of reach of any support or object 

 heated by the sun's rays. 



J. 



Hyu 



are of very yarious constinictionSj and 



\\ 



