5S ACCOUNT or TWO HALOS. 



will exhibit a bright parhelion immediately over the fun, with 

 an appearance of wings, or horns, diverging upwards from 

 the parheh'on. 

 Exp. with the For all fuch particles as are direcled nearly towards the 

 prifm. fpedlator, will confpire in tranfmitting the light much more 



copioufly than it can arrive from any other part of the circle; 

 but fuch as are turned more obliquely, will produce a greater 

 deviation in the light, and at the fame time a defieflion from 

 the original vertical plane. This may be eaftly underflood, by 

 looking at a long line through a prifm, held parallel to it : the 

 line appears, inftead of a right line, to become a curve, the 

 deviation being greater in thofe rays that pafs obliquely with 

 refpecl to the axis of the prifm ; which are alfo deflefted from 

 the plane in which they were puffing. 



The line viewed through the prifm has no point of contrary 

 flexure, but if its ordinates were referred to a centre, as in 

 the cafe of the halos, it would adUme a form fimilar to that 

 which Sir Henry Englefield has defcribed. 

 Tho'fnow flakes The form of" the flakes of fnow as they ufually fall, is in- 

 are complicated, jggj more complicated than we have been fuppofing, but 

 maybe fimple in their elements in the upper regions of the air are probably 

 the upper more fimple. The coincidence in the magnitude of the 



jegions, . obferved and calculated angles is fo ftriking, as to be nearly 

 decifive with refpecl to halos, and it is not difficult to imagine 

 that many circum fiances may exifl, whicii may caufe the axis 

 of the greater number of the prifms to affiime a pofition nearly 

 horizontal, which is all that is required for the explanation of 

 the parhelia with their curved appendages. Perhaps alfo, the 

 effeft may fometimes be facilitated by the partial -melting of 

 the fnow into conoidal drops : for it may be fliown, by the 

 light of a'candle tranfmitted through a wine glafs full of water, 

 that fuch a form is accommodated to the produdiion of an in- 

 verted arch of light, like that which js frequently obferyed to 

 ■accompany a parhelion. 



A Defcription 



