Account of Halos and Parhelia in America. 113 



tfr=20° 



In these experiments the instrument could not be used as a 

 repeating one on the first angles. They are therefore, in re- 

 spect to the seconds, somewhat less accurate than those ob- 

 tained by light received vertically ; but as the differences on 

 one or the other side of the axis already amount to several de- 

 grees, a few seconds are no longer of consequence. 



Thus, in the aggregate experiments of this kind, as D— » 

 has been obtained differently from D+i, so likewise F— i from 

 F + i ; and the same was the case for the remaining coloured 

 rays, of which I here omit the experiments. When the light 

 is not received vertically, the spectra produced by the system 

 of lines are not at all symmetrical on both sides of the axis ; 

 nay, the difference of their position, with large angles of in- 

 cidence, is so considerable, that, even if it amounted only to the 

 hundredth part, it could be easily observed. 



( To be concluded in next Number.) 



Art. XXII. — Account of Halos and Parhelia observed in 



America, * , 



On the 8th September 1816, between two and three o'*clock, 

 there was observed about the sun, at New Port, Rhode Island, 

 a very curious halo, which lasted between 40' and an hour. It was 

 drawn b}'^ Mr D. Melville, and is represented in Plate I. Fig. 

 3. The halo encircling the sun S was of the ordinary size, 

 but very bright, having the whole of its circumference tinged 

 with the prismatic colours. On its upper and north-east limb 

 there was a bright mock sun, the rays of which formed a se- 

 cond halo of a smoky white colour, well-defined in its whole 

 circumference, but more faintly as it approached the primary 



* Collected and abridged from Dr Sillirinan's Journal, vol. x. p. 368, and 

 vol. xi. p. 325. 1^ S>i|} 



VOL. VII. NO. I. JULY 1827. H 



