VALUE OF AUSTRALIAN LONGITUDES. 
205 
systematic errors is revealed, if the results do not agree. But even 
then it is difficult, if not sometimes impossible, to locate them. There 
are, besides, inaccuracies the causes of which are traceable, such as 
unsteadiness of stations, imperfect adjustment of electric instruments, 
changeable strength of circuits, level imperfections, unfavourable 
conditions such as having to carry time pieces to a distance, and 
others ; but their effect can only be made evident by new 
measurements. 
Every determination of differential longitude, however short the 
interval may be, is weakened by at least some of the causes here 
enumerated. 
Admitting consummate skill in the great majority of the observers 
concerned, we may then look at the conditions under which this long 
longitude chain Greenwich- Australia was developed, in order to see 
where its deficiency in strength is more especially to be feared. 
There appears to be at first a natural division at Aden. The 
three intervals on the western portion were all measured twice, the 
results giving, as we have seen, the following discordances: — 
see. 
Greenwich- Alexandria ... ... ... 0T 56 
Alexandria-Suez ... ... ... ... 0*456 
Suez-Aden 0177 
Indeed, remembering the circumstances, these differences seem very 
small. Tet, although the aggregate error in the Aden longitude may 
not be more than one-fifth of their sum, it would not be unreasonabfe 
to suspect that it may amount to half a second of time or even more, 
for the unsteadiness of the stations at Alexandria and Suez and the 
great variations in the personal equation of the observer at Mokattam 
are serious matters. 
The operations east of Aden all along to Australia were decidedly 
made under better conditions and with more completo equipments, 
and, unlike the others (which were only chiefly made for the purposes 
of the observations of the Transit of Yenus), they were intended for 
the establishment of fundamental longitudes. 
The portion from Aden to Madras depends on the elaborate and 
refined operations of the officers of the Great Trigonometrical Survey 
of India, of which the interval Aden-Bombay, with its two independent 
and extremely accordant values, obtained under such uneven share of 
advantages, offers a remarkable instance of how a good result is 
sometimes found where we might be justified by the nature of the 
case in giving it but little weight. 
Erom Bombay to Madras the telegraphic results, though in every 
respect highly trustworthy, are not corroborated by any other entirely 
independent telegraphic determination. It appears also that the 
geodetic value of this interval, derived from the principal triangulation, 
is 12 //, 29 = 0 819 sec. in excess of the telegraphic value, the 
difference being partly attributed to local attractions — (11) Preface, 
page xviii. 
Up to this point we have another test for the whole of the opera- 
tions in the longitude chain via Berlin-Ispahan-Kurrachee, and but 
for the doubts attached to the Kurrachee station this test would be 
invaluable. 
