306 



CATALOGUE. — ^PHYSICAL OPTICS. 



Supposes them produced by vesicles of which the thick- 

 ness is yJ, of the diameter. 



Brandes on parhelia. Gilb. XI. 414. 



Supposes vesicles filled with a medium of a certain den- 

 sity, producing the halos as the drops ofvrater produce the 

 rainbow. 



MuHotte Phenomenon \i. The great Coronae. 



" Sometimes when the air is pretty serene, a circle of about 

 4i° diameter is seen round the sun or moon : the colours are 

 not in general very lively, the blue is without and the red 

 within, their breadth is nearly as in the common external 

 rainbow. Explanation. I take for the cause of this ap- 

 pearance small filaments of snow, moderately transparent, 

 having the form of an equilateral triangular prism. I con- 

 jecture that tl^ smair flat flakes of snow, which fall during 

 a hard frost, and which have the figure of stars, are com- 

 posed of little filaments like equilateral prisms, particularly 

 those which are like fern leaves, as is easily seen by the 

 microscope. I have often looked at the filaments which 

 compose the hoar frost, that appears like little trees or plants 

 in the cold mornings of spring and autumn ': and I have 

 found them cutinto three equal facets ; and when viewed in 

 the sunshine they exhibited rainbow colours. Now it is 

 very probable, that before these little figures of trees or 

 stars are formed, there are floating among the thin vapours 

 in the air, some Of these separate prisms, which when they 

 unite form the compound figures. These little stars arc 

 very thin, and very light, and the little filaments, which 

 compose them, are still more so, and may often be sup- 

 ported a long time ih the air by the winds : hence when the 

 air is rtvoderately filled with them, so as not to be much 

 darkened, many of them, whether separate or united, will 

 turn in ever.-' direction as the air impels them, and will 

 be disposed to transmit to the eye for some time, a coloured 

 light nearly like to that which would be produced by equi- 

 lateral prisms of glass." 



The angles are then calculated, and 16' being deducted 

 for the semidiameter of the sun, and 3o' for the deviation 

 of the red rays, there remnins 22° 5o' for the ultimate an- 

 gular distance of the halo, 



■P. 2?6. Phenomenon 13. Parhelia or moek suns. 



" The most usual are at the same altitude as the sun. 

 Among the prisms of snow there are often many heavier 

 at one end than at the other, and consequently situated in 

 a vertical direction : these cause a bright parhelion, with a 

 tail, which cannot be above 70° long. I havejcad an ac- 



count of a halo seen in May, soon after sunrise, with par- 

 helia in its circumference, which after two or three hours 

 were more than a degree distant from it. This appearance 

 arises from the coincidence of the sun's rays with the trans- 

 verse section of the prism when they are nearly horizontal, 

 and from their obliquity when the sun is elevated, causing a 

 greater deviation, and throwing the parhelia outwards, as 

 may be shown by an experiment on t^vo prisms. There are 

 also accounts of parhelia above and below the sun, of an- 

 thelia, and of a white horizontal circle. I do not undertake 

 to explain these appearances, because I have never seen any 

 of them, and I have not certain information of the circum- 

 stances attending them." 



Rem<ir/cs on helos. See Journ. R. I. II. 4. 



The explanation of the primary and secondary rainbow 

 begun by De Dominis, and completed by Descartes and 

 Newton, derives an entire and satisfactory confirmation, from 

 the perfect coincidence of the observed angular magnitudes, 

 with the result of calculations of the etfect of spherical 

 drops. We know that drops of water, either accurately, or 

 very nearly spherical, exist in great abundance in every 

 cloud, and in every shower of rain ; and whatever their di- 

 mensions may be, they must necessarily conspire in the 

 same general effect, of producing the same rainbow, when- 

 ever a spectator is placed in a proper situation for observing 

 it ; conssquently such rainbows are of very frequent occur- 

 rence. 



I have attempted to show, that for producing the phe- 

 nomena of variable halos or coronae, often observable in hot 

 climates, it is only necessary that a considerable part of the 

 spherules of a cloud or mist, be either accurately, or very 

 nearly, of equal magnitude, a condition, of which the pos- 

 sibility is easily admitted from analogy, and the probability is 

 favoured by the apparent uniformity of the different parts 

 of such mists as we can examine. 



The hypotheses, by which Huygens attempted to explain 

 the production of halos and parhelia, are both arbitrary and 

 improbable. He imagined the existence of particles of hail, 

 some globular, others cylindrical, with an opaque part in 

 the middle of each, bearing a certain ratio to the whole ; 

 and he supposed the position of the cylinders to be some- 

 times vertical, and sometimes inclined to the horizon in a 

 given angle. 



!t has already been objected, that no such particles have 

 ever been observed to accompany halos ; and it is, besides, 

 highly improbable, that such an opaque part should bear 

 the same projwrtion in all the hailstones, and that the cy- 

 linders should have terminations so peculiar as is supposed ; 

 and the most incredible circumstance of all is, that all 



