SOLAR E(3LIPSE, MAY C, 1S83. 



31 



eclipse, the latter being kept carefully as the staudard of reference. The observations with the 

 transit showed that on shore it had a gaining rate of 0'.47. a much smaller value than its sea rate. 

 No. 2786 was not removed from the vessel. 



We are now ready to determine the longitude from the chronometers, making use of the Green- 

 wich corrections given in Table I, and the Caroline Island corrections given in Table IV. Two 

 chronometers, 1340 and 1536, are available for the journey, New York to San Francisco, and the 

 whole nine between Callao and Honolulu. The following table contains the mean rates of each 

 chronometer and the resulting longitude. In obtaining the former the ditterences between the 

 chronometer corrections at New York, Callao, Honolulu, aiul San Francisco (Table I) were taken 

 and diminished by the changes which took place during the stay at Caroline Island (Table IV, 

 last column). Dividing these by the resjiective intervals of time, diminished by that from April 21 

 to Maj' 9, we have the mean sea rates of the chronometers. This assumes that the rates of those 

 chronometers, viz, 1340, 202, 1589, and 1530, which were removed to the island, and were observed 

 to have difl'erent rates on shore, returned to their former traveling rates after the resumj)tion of 

 the voyage. The daily chronometer comparisons show the same relative rates after leaving Caro- 

 line Island as before arrival, and hence justify the assumption, except in the case of 1589, which 

 had been used in the transit observations on shore and also had had its hands moved. A correc- 

 tion to the longitude was therefore obtained for this chronometer by summing up the departures 

 of the observed relative rates between 1340 and 1580 from their mean values, the resulting longi- 

 tude being given in the sixth column. By means of the rates given in columns two and tliree 

 the Greenwich corrections were obtained for the times April 21.101, and May 9.408, and the local 

 corrections from Table IV at those dates subtracted from them. Both of the computations give 

 the same longitude, and are simply a numerical check on each other. 



Table V. — Chronometer rafcs and longitude. 



The last column contains the seconds of the longitude used in deriving the mean result. Nos. 

 202 and 2788 were both excluded, the former because of its daily use as a hack, and its poor run- 

 ning, shown by the daily comparisons; the latter also for its poor ruiniing. (See Table of rates.) 



It remains to inquire if any corrections can be obtained to the resulting longitude from the 

 daily comparisons. This is unnecessary in the case of 1340 and 1536 between New York and San 

 Francisco, as but two chronometers were used, and the corrections, if obtained, would .simply reduce 



