178 Mr. Ivory on Atmospheric Refraction 



or become much more erroneous, at the unusually low tem- 

 peratures in the Quarterly Journal. 



1. With respect to the first question, it is already well 

 known that all the tables entirely fail when the altitude is be- 

 low 2°. But in the Quarterly Journal it seems to be insinu- 

 ated that Dr. Brinkley's table has a great superiority in this 

 respect. The following comparison of five solar retractions 

 at altitudes less than 1°, with the quantities calculated by Dr. 

 Brinkley's table, will enable us to judge of this point. 



App. Alt. 



Thermom. 

 Fahr. 



Barom. 



Obs. 



Refr. 



Brinkley. 



0° 8' 40" 



— 28° 



30-11 



81' 



19" 



48' 42" 



18 40 



-30 



30-37 



56 



19 



44 45 



28 22 



— 28 



30-11 



54 



22 



40 49 



39 00 



— 30 



30-37 



41 



20 



38 38 



50 2 



— 28 



30-11 



39 



47 



35 11 



Here it can hardly be said that the calculated refractions ap- 

 proach in any degree to the observed quantities. I next take 

 all the sidereal refractions at very low altitudes. They are 

 seven in number, and are contained in the following table : 



App. Alt. 



Thermom. 

 Fahr. 



Barom. 



Obs. Refr. 



Brinkley. 



0° 29' 25" 



— 28° 



30-00 



40' 29" 



40' 21" 



30 55 



-20 



30-60 



42 8 



39 18 



43 14 



— 28 



30-00 



36 25 



36 40 



45 33 



— 20 



30-60 



37 23 



35 41 



50 53 



-22 



29-83 



38 8 



33 45 



58 22 



— 20 



30-60 



33 26 



32 51 



58 49 



—28 



30-00 



32 4 



33 7 



In three instances in the table the differences between the ob- 

 served and the calculated refractions cannot be considered as 

 excessive in the circumstances of the observations : but in the 

 other four instances the differences are 170", 102", 263", and 

 63". It does not appear, therefore, that Dr. Brinkley's table 

 can be followed as a very sure guide at such low altitudes. 



We may therefore conclude that no table of refractions 

 hitherto published can safely be trusted to at altitudes less 

 than 2°. And this conclusion is corroborated by the parallel 

 instance of the terrestrial refractions, which are found to vary 



from 



