PROCEEDINGS OF SECTION A. 



53 



The table below shows two sets of values for the Longitudes of the 

 Australian Stations, The values in the first column are those depending 

 on Port Darwin as the initial meridian, after correcting its longitude 

 according to the new determinations and adjustments made along the 

 route Greenwich-Potsdam-Persia-India-Singapore since 1894 to the 

 present time. 



The values in the second column are those depending on the 

 Meridian of Southport as the Terminal Station of the Transpacific 

 Longitudes. 



These results seem so satisfactory that we might be tempted to 

 allow the Longitudes of the Australian Observatories to rest in their 

 present state for a long time to come in the belief that we could hardly 

 expect to still reduce their margin of uncertainty, which is already 

 less than 100 feet according to the closing error of the measured arcs 

 which together form a whole longitude circuit round the world. 



It is difficult, however, to accept, with full confidence, the reality 

 of so close an agreement between two results built up by many steps, 

 each step involving some unknown error, and depending upon a great 

 number of exceedingly delicate operations carried out by many 

 observers through many countries under different climates and con- 

 ditions at long intervals of time covering a period of twenty years. 



This pessimistic view is unfortunately borne out by experience, 

 for there are instances of disagreement amounting to much more than 

 0*079s. (the closing error of the entire circuit) in the independent values 

 even of a single arc. 



Undoubtedly the evidence of the Transpacific Longitudes and of 

 the later determination of the interval Greenwich-Madras would 

 justify a greater confidence in the smallness of the error affecting 

 the Longitudes of the Australian Stations, if the gap between Madras 



• Thi- value for Brisbane was obtained by telegraphic exchange of time signals with Sydney 

 in 1884, 1891, and 1892. 



