206 
PROCEEDINGS OF SECTION A. 
We have now the determinations Madras-Singapore of 1871 and 
1882. It is not unfair to assume the superiority of the latter value. 
The chief weakness of the earlier one arises, perhaps, in the carriage 
of the chronometers to considerable distances for the exchange of 
signals, and in the unknown personal equation of the observers. There 
is a difference of more than half a second of time between the two 
results, and it is not quite certain, though most probable, that the 
whole of this error is attributable to the observations of Dr. Oudeman 
and Mr. Pogson. 
The interval Singapore to Port Darwin depends solely on one set of 
operations — viz., those of 1888. I can only say that the observers felt 
satisfied about the quality of their work ; but still the receiving of 
galvanic signals by observing the sudden motion of a beam of light 
not always regular or well-defined involves greater uncertainty than 
transit observations, and may be subject to comparatively large varia- 
tions in its amount. The result is partly checked by the two separate 
intervals formed by the intervention of Banjoewangie, but is not 
corroborated by entirely independent operations. The difference 
between the direct and indirect result is 0T2 sec. 
There remain now the Australian operations. In the two 
intervals Port Darwin to Adelaide and Port Darwin to Melbourne 
the unknown error of the results rests almost entirely on the time 
determinations at Port Darwin, as the exchange of signals was entirely 
automatic, and transit observations at the fixed observatories involve 
very little uncertainty. 
The various measurements of the interval Sydney-Melbourne, as 
we have repeatedly observed in these pages, range from 24 min. 
55*10 sec. to 24 min. 55*81 sec., which may give reason to suspect some 
unknown disturbing cause interfering with this kind of work. 
Portunately, fresh determinations may be frequently repeated without 
inconvenience, and I believo it is the intention of the Government 
Astronomers of these colonies to make arrangements for that purpose. 
We have, finally, the boundary longitudes. 
Here an error of more than half a second of time was disposed 
of in what was thought the only possible way under the circum- 
stances ; but it does not by any means clear the doubts attached to 
the discrepancies produced by the operations of 186S. 
These are the principal facts upon which an opinion is to be 
formed as to the amount of uncertainty inherent to the adopted 
results. 
I think that the longitudes of the Australian Observatories may 
be accepted as true only within one second of time; and those of 
the boundary lines of South Australia with Victoria and New South 
Wales, within 1,500 feet. 
POSSIBLE IMPROVEMENTS OF THE ADOPTED VALUES. 
No doubt, even with the present means of astronomical science, 
the Australian longitudes could be strengthened by a new determi- 
nation of the longitude of Aden, as recommended by Dr. Gill, and of 
the interval Ispahan-Kurrachee. The importance of these operations 
could not be overrated, and it is to be hoped that they will be under- 
taken at the first opportunity. 
