t^a Accounts of Books, 



ferent appearances, and reprobating the conftruftlon of frozen machineries for the purpofe, 

 he proceeds to ftate the circumftances under which the irides of the firft fort, the objefts 

 of immediate enquiry, arc exhibited. 



Thefe irides appear of one or more concentric orders and circles of colours, clofe around 

 and contiguous to the fun or moon, and fometimes around the brighteft of the planets and 

 fixed ftars. The colours from the luminary outwards are greyiih, black or blue, or faint 

 white, fucceeded by a broad denfe white, followed by yellow, then by red ; then by violet, 

 blue, green, yellow, red ; green, diluted yellow, red ; diluted green, diluted red. The 

 fecond order is frequently only green, yellow, red; and the third only diluted green, red. 

 The two firft orders are moft frequently feen, the third lefs frequently, the fourth moft 

 rarely. The firft may be feen every night, when the moon, with more than half a face, 

 ihines through thin white fleecy clouds, and through fimilar clouds by receiving the fun's 

 image through a fmall hole into a darkened chamber on a fheet of white paper. 



"Of thefe irides the diameters vary confiderably : that of the firft order from one degree 

 to 5-1:, and that of the fecond from 3!- to \o\ degrees. The general breadth of the firft 

 order of colours is rather more than 45 minutes, and thofe of the others arc fucceffively 

 lefs. 



In explaining thefe phaenomena, icy cryftalizations of all forts are reje£ted, and globules 

 of water, the only regular forms of concretion which the vapours of the atmofphere admit 

 of, are alone to be reforted to for the principles of exiftence of regular and conftant 

 phsenomena. 



For explaining thefe phxnomena as produced by globules of water, all the hitherto ac- 

 cepted principles are fhewn to be infufficient. The principle of parallel, or efficient rays, 

 employed in the fchools to account for the exhibition of the primary and fecondary rain- 

 bows, is exceptionable in its former ufe, and is not here tcrbe received, becaufe vifion of 

 images is not by parallel rays, becaufe the light after paflage, if fufficiently ftrong to be 

 feen, would only produce a broad circle round the luminary of decaying white light. 



Newton explains thefe phaenomena by the fits of eafy tranfmiffion and refle£tion, and by 

 comparing the refra£lion of the rays paffing through a drop of water, to the refleftion of a 

 flender beam of light from the back part of a glafs lens. The fits of eafy tranfmiffion and 

 reflexion have no exiftence. The circumftances of the refra£l;ed light in the phaenomena 

 are dltogether, and eflentially different from thofe of the light in the experiment by re- 

 flexion from a lens, and it is further fhewn to be impoffible by any refraftions through a 

 drop to produce appearances of the obferved orders of colours. It is a principle derived 

 from the doftrine of images, and ufed in explaining the primary and fecondary rainbows, 

 that the colours of the drop are inverted in the bow \ and according to this principle, thefe 

 irides, if formed by refraftions through the drops, would have their blues external, their 

 reds internal, contrary to obfervation. For this fame reafon, and from an eflential change 

 in the attendant circumftances, thefe phsenomena cannot be produced by the infleiSlions 

 obferved in . flender beam of light paffing by the edge of a fingle body. 



4 The 



