His'roRr of the sociErr. 9 



•This arch did not exceed 7° or 8° in length' ;'iie Was of the 

 fame breadth with the principal bow ; it had the Colours in the 

 fame order, and nearly of the fame brightnefs ; or, if any differ- 

 ■ence was difcernible, it was, thap the tranficion from one co-- 

 lour to another was'' Wot made' with fo much delicacy iti the laft-» 

 mentioned rainbow as in the former. ' >'■ • i''^ ■; 



We recoUedled that a phenomenon fimilar to thij'is'defcri- 

 bed in the Philofopbical Tranfaciions, as having been feen at 

 Spithead,' and that it is aftribed by the gentleman who ob(erved 

 it to the refledlion of the fun's rays from the furface of ' the fea, 

 fo as X.6 fall oh the cloud where the rainbow wasformed. This 

 hypothefis feemed to agree exactly with the phenomenon now 

 before us. 



The accidental rainbow, for fo it may be called, was feen 

 only at the extremity where the principal arch rofe from the 

 fea, and where,- of confequence, the fun's rays, relieved from 

 the furface of the water, at that moment very fmooth, might 

 fall on the drops of rain. The other parts of the cloud could 

 not receive rays fo refledled, as the land intervened, and there, 

 accordingly, no.veftig-e of the accidental rainbow was obfer- 

 ved. . • , , ., 



. The accidental .rainbow lay, as was already faid, on, the ffde 

 toward the fun, and, this is agreeaBre to the hypothefis ;' for. the' 

 rays that after reflection from the furface of the water fell oii' , 

 the drops of rain, muft have come as from a poii:it asrnuch de- 

 preffed below .the horizon, as the fun was at that inftant eleva- 

 ted above-, it. ,Th^ axis of the accidental rainbow muft there-' 

 fore have made with the axis of the principal, an angle equal' 

 to twice the fun's elevation, and its centre muft have been ele- 

 vated by that fame quantity above the centre of the other, fo 

 that if it had been complete, it would have been wholly between 

 the principal rainbow and the fun. 



B The 



