PRESIDENTIAL ADDRESS, 757 



its method of construction and its ultimate use, whether for military, adminis- 

 trative, engineering, or purely scientific purposes. This enlargement of our scope 

 does not, I think, modify our previous conclusions, and were I now called upon 

 to devise a scheme for the mapping of British Africa, I should hase it upon the 

 principle of a central Imperial body for executing the triangulation and topo- 

 graphy, leaving the land survey to local organisations. 



The arguments in favour of this policy are manifold. As regards the trian- 

 gulation they hardly require stating. It will be obvious to all that such work 

 must be closely co-ordinated, and that some central, directing head is imperatively 

 called for. The enormous waste of money that is ultimately involved by tolerat- 

 ing imperfect work, of which many examples could be cited, is alone a sufficient 

 justification for holding this view. We may, however, pause to examine a little 

 more closely into the advantages of centralisation as regards one particular opera- 

 tion in a survey. That is the measurement of the initial base line upon which the 

 accuracy of the whole framework depends. This task used to be one of the most 

 laborious and difficult with which the surveyor is confronted. The apparatus 

 employed, some form of compensation bar, was cumbrous and difficult to use, the 

 site selected had to be levelled, and the preparatory alignment carried out with the 

 mos't scrupulous care. Thus the Loch Foyle base for the triangulation of Great 

 Britain and Ireland was about six miles long, and the actual measurement, quite 

 apart from the time spent on the preparation of the ground, took sixty days, an 

 average rate of work of just over 500 feet per working day. 



A few years ago the discovery was made of the nickel steel alloy with a very 

 small or zero coefficient of expansion, the so-called invar. This valuable metal, by 

 abolishing the necessity for any temperature correction, has enormously simplified 

 all physical measurements of length, and, a fortiori, those measurements, such as 

 base lines, which are perforce done in the open air and over a large range of tem- 

 perature. Survey bases are now measured with an invar wire stretched to care- 

 fully regulated tension, and either laid along a flat trough, or what appears to give 

 equally good results, hung freely between supports. The gain in precision due to 

 the avoidance of errors of expansion or contraction in the measuring apparatus is 

 substantial, while the gain in rapidity is very great. Thus, as a contrast to the 

 Loch Foyle base, let me give a short account of the measurement of a base in Spitz- 

 bergen by the Russian party of the joint Swedish and Russian missions in 1900, 

 extracted from a review already written for the ' Geographical Journal.' 



The conditions for accurate work were very unfavourable : no site even ap- 

 proximately flat could be found, and the base was therefore irregular in contour 

 and traversed rough and in some parts marshy ground. The weather conditions 

 were far from ideal. The cycle of operations was as follows : An auxiliary base 

 175 metres long was measured with Struve's apparatus, twice before the main base 

 measurement and twice afterwards. The two wires used for the main base were 

 standardised on this subsidiary base four times, twice before and twice after use. 

 The main base, 6'2 kilometres long, was measured twice in each direction by each 

 of two wires, eight measures in all. The limit of error in the final value was 

 17 millimetres — say, one part in 360,000. 



The whole of these operations, including the laying out of the standard and the 

 comparison of the wires, were completed in a period of three weeks ; Monsieur 

 Backlund, who superintended the actual measurement, left the observatory at 

 Pulkowa on June 11 and returned to it on July 24. It was therefore possible 

 to standardise the wires not only by the check base upon the spot, but also by the 

 permanent standards of the observatory within three weeks of their use for the 

 actual measurement. It need hardly be pointed out that this was eminently 

 favourable to the attainment of the highest exactitude, and we have here a marked 

 example of the value of centralisation. The proposed trigonometrical survey 

 department of Africa would probably find it advantageous to adopt similar proce- 

 dure, and, instead of trusting a base measurement to a local staff unacquainted 

 ■with the work, it would send out one or two men of highly trained technical skill 

 equipped with the best apparatus. The money spent in journeys would be more 

 than saved— first, by the unquestionable gain in accuracy and the consequent 



