On Refraction. 440 



for 1» 



M. Bonne 1-00 25777 



Bradley. 1 00 25000 



Dalton 1-00 20701 



De Luc 100 20388 



Fahrenheit , TOO 25777 



Gay Lussac 1 00 20868 



Gioombridge 100 21000 



Hawksbee 100 00*033 



La Caille POO 22222 



Mayer 1*00 20444 



Shuckburg , 1*00 22222 



Mean of all except Hawksbee's . . 100 2240,0 



The refraction deduced from Bradley's very neat and 

 simple formula was in a few years adopted by nearly all 

 the astronomers of eminence throughout Europe, The ex- 

 treme facility with which it might he computed, and the 

 corrections applied, whether from the, formula itself or from 

 tables ready calculated for that purpose, was a powerful re- 

 commendation in its favour ; but its near agreement with 

 observations soon established it. 



In 1805, the very ingenious and profound M. de la Place 

 in his Mecavique Celeste* favoured the world with a chap- 

 ter on this subject, wherein he has displayed as much saga- 

 cious penetration as deep mathematical learning and ability. 

 He begins with considering the trajectory of a ray of light 

 traversing the atmosphere; and by supposing all its layers 

 spheric, and of variable density, according to some function 

 of their height, he deduces a differential formula for the 

 refraction whose integral he then finds ; but, he observes, 

 this equation supposes that the refractive forces of the layers 

 of the atmosphere are directly proportional to their density, 

 which is the result of Hawksbee's experiments. Never- 

 theless, it is possible that this assumption may not be 

 strictly correct, and it would be useful if more experiments 

 were made on the subject. He then find! that the hypo- 

 thesis of an uniform temperature is erroneous, as well as 

 that of the density decreasing in arithmetic progression, 

 when the height increases in a similar progression ; and he 

 says, " the constitution of the atmosphere being com- 

 prised between the two limits of a density decreasing in 

 arithmetic progression, and of one decreasing in geometric 



* Vol ir, page 231. 

 Vol. 36. No. 152. Dec. 1S10. 2 F pr?- 



