nf Injlrument fir exhibiting Jupiter and his Moons. 



Though tKc dimenfions of the wheelwork here given are convenient for the conftru^lion- 

 of the fatellitiau, yet the box may be in the form of a fquare, an oftagon, or a circle ; 

 though a parallelogram feems to be the moft portable; but of whatever fliape, there mud 

 bfe a contrivance for cramping or otherwife filing it to a table when ufed. — The fize of tlie 

 bail, which reprefents Jupiter in the plate, is proportioned to the length of the arms which 

 carry the fatellites, and ought not to be larger ; for an increafed length of the arms would 

 not admit the fcreen to approach near enough to prevent the divergence of the fliadows.— 

 The balls which reprefent the fatcllites fhould, however, be as diminutive as poflible, fo 

 that they give adiflin£l fliadow by candlelight. According to Mr. Herfchell's obfervations, 

 the third is the largeft, the firft and fourth nearly equal, and the fecond the fmalleft : their 

 colours are, in his own words, as follows : the firft is " white, more fo fometimes than 

 others-," the fecond " white, blueifli, and afh-coloured ;" the third " white, differently in 

 different fituationsj" and the fourth " dufky, dingey inclining to orange, reddifli, and ruddy 

 at different times." 



A motion is communicated to the machinery by means of a fingle endlefs fcrcw, on an 

 horizontal axle, lying acrofs the box at the fmall black circle near 73, and working with the 

 teeth of this wheel : a handle is put upon the projeding end of this axle, which cannot ap- 

 pear in the plate, at fuch a dlftance from the fide of the box, as may prevent the hand of 

 the perfon who turns it from touching any of the fmall balls, or intercepting the view of a 

 fmall company of fpe£l:ators. 



As appendages to the fatellitian, a fmall Aiding fcreen of thin paper and an appropriate 

 candleftick are necefTary. — The mode of adjufting the fcreen is optional, and therefore re- 

 quires no defcription. — The bafe of the candleftick, of which a fketchis given in miniature 

 ^' fig- 3> confifts of a circular plate of brafs 4 inches diameter, divided into the 12 figns, 

 and graduated in concentric circles near the extremity : in the centre of this ftands a fliort 

 perpendicular flem of a cylindrical form, which fcrews into a folid piece foldered on the 

 furface of the plate, and has a focket proper for containing a candle or lamp at the upper 

 end : on the low end of this ftem turns a broad ring, into which an arm is riveted 

 3 inches long, which has a fimilar focket on its upper end, and oppofite thereto a hand alfo 

 riveted, which points to the giraduated ecliptic oppofite to the part on which the arm refts^ 

 at all times. 



In order to eftimate the accuracy of the mean motions of the little balls in the fatellitian, 

 we muft confider each pair of correfponding wheels as an improper fra£tion of a day ; 

 thus 11 of 24 hours is equal to i d. i8 h. 27 min. 41 fee. ; |^ is equal to 3 d. I3h. aomin.j 

 ^^ is equal to 7 d. 4 h, o min. ; and '-j-* equal to 16 A. 1 8 h. o min. ; whence it appears 

 that the error in one day's mean motion of the firft fatellite is not quite + 37''; of the 

 fecond fomewhat more than + 3.7''; of the third not quite + 3''4: » *««! of the fourth 



about— 18". The firft and fecond of thefe errors will amount to one hour's motion: 



of each fatellite in about 97! days' ufe with one redification ; but the third and fourth will: 

 not be perceptible by the eye in fcveral years, by reafon of the flownefs of the fatellites'' 

 motions, as well as fmallnefs of their daily errors. With regard to the calculation of the. 

 other wheels, the fradion '/ is equal to 7 days for the week-hand ; and the compound 

 one 'y ^ V is equal to 73,5 but the foiral face on the cover has five lines divided into 



365 equal 



