656 J A M A I C A. 



that we were turiiinied with more regideis, and thole formed with 

 more regularity. To be corre(5lly accurate is Icarcely to be expeded, 

 iiiice thermometers, graduated alike, often vary much from each 

 other in their indications. But what chiefly impedes the enquiry is, 

 that experiments, made in ditlerent parts of the world, are by in- 

 flruments of very different graduation. It one llandard thermometer 

 was univerlally in ufe, a comparifon of remarks or regifters, formed 

 in different parts of the globe, or in the hme country, would be very 

 pra'flicable, and lead to a knowledge of their refpective temperatures 

 with far mure precifion. 



The tliermometers I made ufe of were a ipirit and a mercurial, 

 both of them graduated according to Fahrenheit. Yet, upon com- 

 parilbn, the former always ro'e two degrees iiigher, excepting in one 

 inftance, in x^pril 1762. 



The weather, havinj?' been fultrv and (howerv for fomc time, be- 

 came on a fudden very cool, and cleared up; when, to my great 

 furprize, I oblerved the mercurial tliermometer had rifen two degrees 

 higher than tiie other, I could only account for this, by fuppoling, 

 tliat, as the parts of mercury are more iulceptible of immediate im- 

 preffion from changes of the atmolphere than thofe of fpirir, they 

 are likewii'e fooner affe£led with any fudden rarefaction or denfity. 

 The elalficity of the air in the upper part of tlie tube has been found 

 fometimes to overcome the expanfion of the fpirit; and often the air, 

 when very hot, is faid to make the Ipirit rife difproportionately high 

 in the tube, v.'hen what is contained in the ball is not equally affedted. 

 The ball of my fpirit thermometer was covered with a mahogany 

 cap ; upon taking off which, in the month of February, the liquor 

 funk immediately two degrees, I left it uncovered a whole night ; 

 and, early in the morning, after fuffering the niglit air to blow upon 

 it freely, through a window left open for that purpofe, I perceived it 

 fallen fix degrees lower than had been ufual when the cap was on. 



I obferved frequently, that, the houfe being fliut up all night, 

 upon viewing the thermometer in the morning before the fun was any 

 height above the horizon, the air within doors was warmer by five 

 or fix degrees than the air abroad. 



Similar variations have been remarked in London. 



1748, june II, the thermometer out of doors, in a fhaded air, 



was at ■ • 83 1 



Within doors, at 68 



a difference oi fifteen degrees and an half; which is almofl: 

 incredible. 



At nine in the evening, without doors, 74 



5 Within, 



