GENERAL STATEMENT OF THE PROBLEM. 



possible number of observations of every kind in each hemisphere, and to trust to the accumu- 

 lation of observations for eliminating the inaccuracies due to erroneous assumptions in the 

 adopted positions of the comparison stars, and to the unavoidably large mean error of the 

 meridian declinations. This course has been pursued, and, without any attempt to refute the 

 objections which may fairly be urged, the present discussion aims simply at deducing the best 

 result from the materials at hand, in the hope that whatever be the degree of trustworthiness 

 attributed to the final determination, this discussion may at least claim to have been based upon 

 the best available data, and to afford, for the result of previous research, the always welcome 

 corroboration which is furnished by different observations and different methods of investi- 

 gation. 



The observations available for this purpose are as follows : 



At Santiago 55 micrometric comparisons of Mare, I series 



89 do. do. II series 



79 meridian observations do. II series 



56 micrometric comparisons of Venus, I series 

 72 meridian observations do. I series 



37 micrometric comparisons do. II series 



46 meridian observations do. II series 

 At Washington 32 micrometric comparisons of Mars, I series 



5 do. do. ' II series 



16 do. of Venus, I series 



6 do. do. II series 

 At Greenwich .... . ... 8 micrometric comparisons of Mars, I series 



34 meridian observations do. I series 



36 do. do. II series 



17 do. of Venus, I scries 

 17 do. do. II series 



At the Cape of Good Hope .... 47 micrometric comparisons of Mars, I series 



47 meridian observations do. 1 series 

 46 do. do. II series 



At Cambridge 19 micrometric comparisons of Mars, I series 



At Minus 38 meridian observations do. I series 



At Crncow 16 do. of Venus, II series 



At KremsmiinJiter 13 do. of Mars, II series 



At Altona 3 do. of Venus, II series 



A few of these meridian series have been included up to dates somewhat beyond the proper 

 limits, especially when their reduction showed a mean error sufficiently small to warrant the 

 hope of attaining increased accuracy by their employment. A series consisting of five obser- 

 vations of Mars II. at Cracow, only one of which was within the limits of the Santiago obser- 

 vations, were omitted in consequence of the mean error coming out as over 3". A large part of 

 the Cambridge observations were inapplicable for combination with those at Santiago on 

 account of the observations being confined to a single limb, and still others are incapable of 

 employment on account of the smallness of the comparison star employed, entailing, naturally, 

 an impossibility of identification in any of the catalogues. 



Observations of right-ascension only are utterly without avail for our purpose. The 

 Greenwich observations were obtained from the annual volumes of the Eoyal Observatory, and 

 are, with few exceptions, solely meridional, conformably to the established usage of the 

 observatory. 



The only method which remains for making use of these observations to deduce a value of 

 the parallax, is to obtain the declinations of the several comparison stars as well as possible, 

 and with these to endeavor to elicit from the series of differential observations, with equatorial 

 instruments, sufficiently numerous determinations of the planet's declination, at both northern 

 and southern observatories, to authorize the hope that the inaccuracies of the individual star 

 positions may possibly disappear from their combined result. The results of these differential 

 measurements would then be used like determinations of absolute position. The meridian 

 observations at Santiago and Greenwich could be incorporated, with their appropriate weights, 

 so that every observation would be made to contribute to the final result. 



