16 TRANSIT OF VENUS, 1874. HONOLULU. 



39. For the convenient application of the corrections for level, collimation, 

 and azimuth, a table of factors was computed for every degree of N.P.D., 

 similar to that contained in the Introduction to the Greenwich Observations. 

 Similar factors were computed for the four azimuth-stars for every 15 days. 

 Tor the more accurate reduction of circumpolar stars observed rather far 

 from the meridian, and of imperfect transits of clock stars, the collimation 

 factors were computed to four places of decimals. The observations of 

 circumpolar stars more than eight minutes of time from the meridian have 

 been reduced by the rigorous formula. 



40. The clock-correction was required for (1) the transit of the Moon ; 

 (2) the Altazimuth observation of the Moon ; (3) the comparisons of travel- 

 ling chronometers ; and (4) for latitude observations. No night was allowed 

 to pass without the clock and instrumental errors being determined, if 

 possible ; and a sufficient number of observations were made to obtain the 

 relative personal equations of the observers ; which was necessary, because 

 the observation of the Moon with the Altazimuth was too fatiguing to admit 

 of many stars being taken by the same observer. 



41. The observed Transits of the Moon and Stars, as arranged in Table IV., 

 require little explanation. The day commences at noon, 10 h . 31 m later than 

 noon at Greenwich. The reading and position of the micrometer are given 

 (in the third column) for every star, it being understood that a given reading 

 is to be carried down the column until a change occurs'. In this column the 

 adopted reading for zero of collimation for the day is also given in a bracket, 

 thus [20'070]. It generally applies to all the observations made in one night, 

 but occasionally different values have been used for different observers. When a 

 circumpolar star has been observed with reversed positions of the transit-axis, 

 the two observations are separately reduced. The clock correction proper to 

 apply to the true transit of the Moon's limb is entered in the eighth column 

 in a bracket, thus [2O86], on October 17, Table IV. 



42. It has been considered, sufficient to print in detail the transits of the 

 stars observed with the Moon, and those observed for azimuth-error on the 

 days when the Moon was observed with the Altazimuth. All other transits 

 are omitted. Each observer's mean result for clock error is given in Table V. 



43. Determination of tlie relative personal equation in observing transits of stars 

 of the observers at Honolulu. 



In the following table T R = +0 S '22 signifies that T makes the clock 

 O s- 22 more slow than does R. The numbers are obtained from Table V. by 

 interpolation. 



