Vihich took place on September 7, 1820. 



27 



36', divided into seconds by means of a vernier. Thd advan- 

 tages attending an instrument of this kind, are its convenient 

 size, and the ease and expedition with wliich the observations 

 can be made. Nevertheless I think it right to remark that, in 

 the present instance, it is probable there is a constant error of 

 a few seconds affecting the results, arishig from the indistinct- 

 ness of the borders of the sun and moon, which prevented me 

 from determining the exact point of contact. But as the junc- 

 tion was always made under the same apparent circumstances, 

 the proportion between the results will not be affected thereby. 

 This inconvenience may proixibly be overcome in any new 

 telescope on this construction. The following are the obser- 

 vations which were made with tliis instrument. 



At 2^ 54' I left this mstrument in order to prepare for ob- 

 serving the termination of the eclipse as above stated: and I 

 cannot but consider it as extremely fortunate that the sun was 

 entirely fi'ee from clouds both at the beginning and end of tlie 

 eclipse. 



All these observations were made from a laroe windov/; near 

 which was suspended a barometer, with a thermometer at- 

 tached : and I had also suspended another thermometer in the 

 o)5en air, in the shade. At noon these instruments stood as 



follow: Barometer ... = 29*86 



Thermometer, within ... = 65° 



, widiout ... =67" 



During the progress of the eclip.se, I watched the state of them, 

 hut could not observe any altci-ation in either of tliem. As soon 

 ns the eclipse was ended, 1 again noticed them more particu- 

 larly, and thev stood as follow : 



Barometer ... = 29^87 



Tliermouieter, within ... = 66° 



without ... = 68" 



'I'he diminuliou ol" ligiit was very trifling, and would .scarcely 



l»ave been p< icinliblc, had not mv attention been called to it. 



U2 ' It 



