171 



are founded, are indeed exact on this liead ; but the difficulty of 

 introducing the variation into the tables seems to have occasioned 

 its omission. M. Bessel's investigations are rendered more intricate, 

 by developing this change in a manner so as to be conveniently 

 introduced into tables. In my method the development and tables 

 appear equally simple : 



Inches ' " 



A,ft-i„7Tt S Bar. 28,001 the mean ref. is C by French Tables 1.19,2 

 ' ^ Therm. 700 j tlecreased j by my Tables 1.29,1 



' At 88» Z D i^"' 30,501 the mean ref. is C by French Tables 1.26,6 

 \ Therm. 25 j increased l by my Tables 2. 4,0 



At90»Z.D. |?ar.J9 60K^j.^^^^i^^ C by French Tables SVAf' 



( Inerm. 32" ) ^by mme 36.18 



Whereas the mean refraction is by each 33.23 



At present we have not sufficient observations to determine, whe- 

 ther the actual variations of refractions at low altitudes are most 

 conformable to the theory of M. Bessel, to that of Dr. Young, or 

 to that above given. It can only be ascertained by help of very 

 numerous observations, so numerous that they must require the co- 

 operation of many observers ; but the exactest instruments are not 

 required, and much assistance might be given by instruments of in- 

 ferior accuracy, and of which the errors may amount to some seconds. 



In the following Tables : 



Tab. I. = log i(rCT) tan^ 



// 



r 1,0375 , ,^ 57,72 T 



Tab. II. = log Li+;oo-i5^T?=3i) ^ (i->oooi(^-so))x ^^J 



(57,72 \ 

 (1-,0001 {<-50)x2^; 



A a2 



