THE TRANSATLANTIC LONGITUDE 1^ 



end of the axis being the larger. The value resulting from Pensacola levelings 

 was O'.Olo; and the mean of these has been applied to all level-readings as correc- 

 tion for inequality of pivots. 



On and after November 5, the transit-observations upon which the longitudes 

 depend were made by Mr. Mosman alone. On the 25th and 28th October, they were 

 made by myself; and on the other dates which enter in any way, however implicitly, 

 into the longitude-determinations transits were observed by both of us. This 

 circumstance, undesirable in itself, was, from the necessities of the case, not to be 

 avoided. I have, however, whenever possible, employed Mr. Mosman's observations 

 only for determining the clock-correction, and for those cases where this was not 

 feasible have applied to my own observations the constant correction of — 0'.08 for 

 personal equation to reduce them to Mr. Mosman's, as will be explained in a sub- 

 sequent chapter. 



With these few explanations, and the added remark that the observations for 

 time were almost without exception obtained with extreme difficulty in the inter- 

 vals of clouds and rain in one of the most unfavorable climates of the globe for an 

 astronomer, I give the crude observations, and their reduction for the groups imme- 

 diately preceding and following each series of longitude-signals, omitting the others 

 generally as needless. The notation and methods of observation and reduction are 

 those prepared by me for the longitude work of the Coast Survey some fifteen years 

 ago, and are described in detail by Mr. Dean in the appendix to the Coast Survey 

 Report for 1856. The conditional equations for clock-correction and azimuth are 

 solved by least squares, after correcting for level-error and clock-rate ; the normal 

 equations and resultant values being appended to each group. As already stated, 

 my own observations have been referred to Mr. Mosman in every case, by subtract 

 inff 0".08 from the observed times. 



May, 1869. 



