COMMENTATIO ASTRONOM IC Ä. S5 



(f rallel to the «cliplic , and the minor in the direction of a aecondary to that greal cir- 

 « ele , ihe true place of Ihe slars being in the centre. Were these ellipses of the same 

 « magnitude, for eaeh of Iwo conliguous slars, ihe line joining their apparent places 

 « (which are necessarily homologous points in the circumferenses of cach ) would pre- 

 « serve its parallelism at all times; but as the axes of the ellipses are reciprocally as Ihe 

 •« distances of the stars , that parallelism cannot obtain when the stars are situaled at 

 « v^ry unequal distances from the carth ,. and an alternate increase and decrease of the 

 « angle of position made by this line Tvith any fixed direction must be the neeessary 

 « consequence. " 



Indagat extensionem harum varialionum „ et eventum mensurae. Concludit, distanlia 

 si fuerit 3" , angulum positionis ad 30' cxacle observatum iri. 



<c Now, ll»e tangent of 30' to a radius of 3" corresponds lo a subtense of 0",02(V 

 « OT -%Üi of a second , so that a difference of parallaxes to the amount of a 40lh of a 

 « second, exisling in ihe two stars of a double slar so cireumslanced , could scarsely 

 « escape detection; and that even much less quantities than this, under favourable cir- 

 « cumstances , miglh be rendered sensible , I think n- ay fairly be concluded , v/hen we 

 « consider that in this estimate , the data are ccrtainly assumed within bounds. No ac- 

 « counl is here taken of the improvements in the position micromeler , which may rea- 

 « sonably be expeeted, because the object at präsent is only to apprcciate the dcgree of 

 « delicacy which the method now proposed may lay claim tö , with our present instru- 

 « ments and habits of observing, and to compare it with those which hav« hilherto 

 « been resorted to in the iiivestigation of parallax. 



« In selecting stars for examination , it appears to me lliat we ouglh by no means to 

 « confine oursclves, by assuming it as a universal law, that Ihe brightest stars are the 

 « nearest to us» From what we know of the variety of nalure and the enornjous diffe- 

 « rences in point of magnitude belween the bodies- of our own System, iJ seeras impro- 

 « bable that the real magnitude and brigthness of jlhe slars should be confined withi» 

 « narrow limits. Their distances are equally undetermined ; nor have we any reason 

 « whatever to conceive these two- Clements related tö each other. There is not tterefore 

 « the slighlest a priori improbabilily, in- supposing that among stars of apparenlly 

 « equal lustre , the greatest diversity of dislance may exist , or that inniimerable of Ihe- 

 « minulest stars visible in tclescopes may be nearer to us- than any of those of the 

 « firsl magnitude ; and eonsequenlly , that that delicate dement in seareh of which as- 

 «r tronomers have exhausted refinemcnt , may with nearly or quite equal probability of 

 « sticcess be-sougtb among stars of far inferior magnitudcs. The proper motions of the- 

 «■stars afford- anargument from analngy- — ihey bear no rclation lo' their apparenf luslre, 



« and 



