VARIABLE STAR*. 27j 



tions of the increafe and decrease of brightnefs, as thereby the 

 ciaanges and times elapfed are pointed out. (See Table V. 

 page 136; and Phil. Tranf. for 1797.) 



I have tried pradlically the effedt of the above fuppofitions, Experiment with 

 by placing fmall white fpots on a dark fphere, which being ^ ^ ^^^* 

 revolved round reprefented the various changes as nearly as 

 could be exped«d : proceeding therefore with thefe and other 

 eonfiderations, I iliall make ideal drawings of the.%r with the 

 fraall illuminated parts in its almofphere, and apply to them 

 fbme of the aSual obfervations from both the preceding tables, 

 having always in view that each period may, more or lefs, 

 require a different difpofition of fpots, ifl jpojifequence of their 

 conftant changeability. 



Plate XIII. Fig. 1, A B, the ftar's polar axis, round which View of the ftar 

 its rotation takes place in 6? days from C to D. at its greateft 



C D, its equator, the 360 degrees of which being revolved *' *' 



in 62 days, gives nearly 5| degrees for each day's motion ; the 

 brighteft part or fpot is reprefented as centrally facing us, 

 and accordingly (lie wing the fiar in its greateft luftre. Were 

 this bright fpot and the other parts to remain unchangeabfe, 

 they would after having completed the revolution of 360 de- 

 grees, or 62 days, (the ftar's rotation on its axis,) appear 

 again as at firft, and at every return continue to give exa^ 

 periodical times, as was nearly the cafe in 1799, between 

 Auguft and Oaober, (fee Table IX. p, 142) ; but if the fpot 

 becomes obfcure, and another brightens up in a different 

 place, this latter wili make the (lar appear a^ its next full 

 fplendour either fooner or ialer than tl^e real rotation accordr 

 jng to its pofition, thus, 



2d View, 



Fig, 1. A full brightnefs having been fliewn by (be fame Anomaly from 



'fpot, it afterwards lofes its light, and another as bright is pro- ^""^^^"^ ^"S^*^ 

 , , , . , , ^ ,. . r i- fpotcaufingan 



duced D days motion (or 29 degrees) precedmg it at Ji., lee apparently fliort 



Fig. 2. This latter, when turned centrally to ihe earth, will period. 

 appear 5 days fooner than the former one, now obfcured, (here 

 marked P,) and fliow the ftar at its full luftre, making the 

 ,rotation 57 days inftead of 62, which was tiie cafe in 1796, 

 the obferved revolution between September 17 and Novem- 

 ber 13. {See Table IX.) 



T2 H 



I 



