Hydrography, arid the AH of Navigation. 39 



cribed to the precipitation of humidity on their surface, as they 

 descend from the cold region whence they originated, and tra- 

 verse atmospheric strata, gradually becoming warmer as they 

 come nearer the earth. It is then almost certain that, if sup 

 plementary rainbows are formed in equinoctial regions, they will 

 never reach the horizon, as is likewise the case in Europe ; but 

 the comparison of the angle of the height at which they cease to 

 be seen, with the angle of disappearance as noticed in our climates^ 

 seems to afford the means of obtaining meteorological results^ 

 which no other method at present known can supply. 



Halos. — In high latitudes, the vicinity of Cape Horn for ex-* 

 ample, the sun and the moon often appear surrounded by two 

 luminous circles, which meteorologists name halos. The ray of 

 the smallest of these circles is about 22'', the ray of the larger 

 is almost pxactly 46°. The first of these angular diraensiona 

 approaches near the minimum deviation which light under- 

 goes while traversing a glass prism of 60° ; the other would be 

 given by two prisms of 60°, or by a single prism of 90°. 



It seems, therefore, natural to seek for the causes of halos, as 

 Mariotte has done, in the rays refracted by the floating crystals 

 of snow, which, as every one knows, usually present angles of 

 60° or 90°. 



This theory, besides, has received a new confirmation, since 

 the power has been acquired of distinguishing between refracted 

 and reflected light by means of chromatic polarization. It is in 

 fact the colours of the first of these lights (that which is refracted) 

 which produces the polarized rays of the halo. What, then, still re- 

 mains to be known regarding this phenomenon ? It is the follow- 

 ing :^- According to theory, the horizontal diameter of a halo and 

 the vertical diameter ought to have the same angular dimensions ; 

 but we are assured, that these diameters are sometimes strikingly 

 unequal. Measurement alone can establish this fact ; for, if a 

 judgment has been formed by the naked eye, the causes of illusion 

 will not fail to explain how the most experienced natural philoso- 

 pher might be deceived. The reflecting circles of Borda are ad- 

 mirably adapted to the measurement of the angular distances at 

 sea. We may, therefore, without hesitation, recommend to the 

 officers of the Bonite, to apply the excellent instruments with 



