440 Mr. E. L. Garbett's Description of some Parhelia 



very clear, at Brighton there was a common halo, by far the 

 finest I ever saw, but no traces of any other arcs. On the 

 afternoon in London there was still a taint halo, but on the 

 next day nothing remarkable. 



II. Remarks. 



The extreme variety and complexity of phaenomena of this 

 kind render it difficnlt to determine the peculiar structure 

 requisite in the atmosphere to produce them. If the common 

 halo can only be satisfactorily accounted for by the supposition 

 of innumerable crystals of ice turned in all possible directions, 

 the present phsenomenon requires us to suppose all these 

 crystals turned in 07ie direction, or at least having all their 

 axes parallel. This is plain from the fixed relation which all 

 the appearances bore to the horizon and zenith, as well as to 

 the sun. 



Let us take, for instance, the case of the most conspicuous 

 feature, which happens also to be that which admits of the 

 simplest explanation, viz. the white horizontal circle. All the 

 rays by which this was seen had the same inclination to the 

 horizon as the direct solar rays (or some of them, viz. those 

 coming from some part of the sun's disc). Moreover, the absence 

 of colour showed that these rays had either suffered no refrac- 

 tion, or equal and opposite amounts of refraction, so that their 

 change of direction could be due only to reflexion. Now in 

 order that a reflected ray may retain the same inclination to 

 the horizon that it had before reflexion, it is plain that the 

 reflecting surface must be vertical^ though it may be turned 

 towards any azimuth. The appearance of this circle therefore 

 requires us to suppose innumerable reflecting surfaces turned 

 to all possible azimuths, but all of them truly vertical. This 

 would be the case if the air contained innumerable prisms of 

 ice, having all their axes vertical, but their sides turned in all 

 directions. All the light reflected from these sides, without 

 entering the prisms, would contribute to form the white circle 

 above described. And this would also be the case with that 

 light which entered the prism and suffered an internal re- 

 flexion from the opposite side, provided it passed out again 

 through a vertical face; for though the two refractions might 

 not always compensate each other horizontally, so as to pro- 

 duce no change of azimuth in the I'ay, yet they would always 

 compensate each other vertically, so that the ray would suffer 

 no change of inclination, and consequently no dispersion into 

 colours in a vertical plane : and its dispersion horizontally 

 would not impair the achromatism of the appearance ; for if 



