542 REPORT— 1902. 



the generalisation wliicli followed the Moscow discovery v/as too hasty ; to the 

 best of my belief, no second plain has been discovered where deflections vary 

 sixteen seconds in 18 miles. On the other hand we have discovered ■within 

 the last two years the existence of an underground and invisible chain of extreme 

 density stretching across India from east to west for a thousand miles, and 

 which may possibly be disturbing the direction of gravity throughout the country. 

 The discovery of this underground chain brings home to us how little we really 

 know of mountain attraction. The underground chain of density underlies the 

 Vindhyan range of mountains: whilst then we have, ia obedience to theory, been 

 endeavouring to prove that the Himalaya Mountains overlie great deficiencies 

 of matter, we have stumbled on the fact, unforeseen by theorists, that the Vindhya 

 Mountains overlie great excesses. 



It has been said that the arcs of India are Great Britain's largest contribution 

 to the science of geodesy. It will not be out of place if I refer to their critics. 

 It has been repeatedly stated that the presence of the Himalayan Mountains 

 detracts from the value of the Indian arcs. This statement originated in the 

 account of the triangulation of Great Britain. The author of a popular handbook 

 describes the Himalayas as a reproach to the ares of India. Now if mountains 

 lead us to anticipate systematic error, do not oceans do so also ? Where then is 

 the country, and where the arc, where systematic error is not to be anticipated 1* 

 A theory has been devised that the visible deficiency of oceans is compensated by 

 the condensation and contraction of submarine strata, but it rests on insufficient 

 data. Every survey has discovered coast-stations where plumb-lines are deflected 

 seawards, and the theory of condensation has been built on them. As an example 

 we may suppose a seaward deflection of five seconds observed on the coast of 

 Mayo, whilst no deflections are found in Roscommon. Would this be evidence 

 that the strata underlying the Atlantic are condensed ? Would not a wide and 

 distant ocean afl'ect Mayo and Roscommon similarly ? If the Atlantic Ocean were 

 causing the deflection in Mayo, it would, I submit, affect plumb-lines in Germany 

 and Russia. The systematic error caused by the Atlantic Ocean cannot be gauged 

 from an observation on the Irish coast. Its determination requires masses of 

 observations spread over Europe. AVhilst we can observe at the foot and heart of 

 the Himalayas, no observations can be taken over Atlantic depths nor even on the 

 brink of its submerged cliffs. The real edge of the Atlantic is not at the visible 

 Irish coast-line, but far out to sea. If we had never approached nearer to the 

 Himalayas than the Irish coast-line is to the Atlantic, we should have been unable 

 even todiscuss the subject of mountain attraction, and oceanic effects can only 

 be studied by the aid of lessons from mountains. 



Instead, then, of reproaching the arcs of India, we must discover the systematic 

 errors affecting them. The aim of our geodesists must not be confined to the 

 determination of an imaginary mean figure for the earth ; and though it would be 

 fantastic to think of investigating all the many elevations and depressions that dis- 

 figure the level surface of the sea, we must at any rate endeavour to obtain a definite 

 idea of the order of their magnitude, whether their heights and depths, above and 

 below the mean surface, are to be estimated in feet or thousands of feet. If this 

 endeavour be regarded as a desirable and legitimate aim, the scientific value of 

 arcs affected by mountain attraction will be recognised. 



4. Report on Seismological Investigations. — See Reports, p, 59. 



5. Report on the Investigations of the Upper Atmosphere by Means of 

 Kites. — See Reports, p. 77. 



6, Report on Magnetic Observations at Falmouth Observatory^ 

 See Reports, p. 75. 



I 



