May 22, 1902 | 
Airy, guided by the insufficiency of the Archdeacon’s 
results to explain the numerical discrepancies, was fully 
justified in asserting that the magnitudes of attractions 
computed on the theory of gravitation would be too great. 
He was less happy in the reason assigned for this con- 
clusion. Airy seems to have considered that the 
Himalaya Mountains were floating in a sea of dense lava, 
and that the bases of the mountains displaced a quantity 
of denser material, much in the same way that an iceberg 
displaces the water of the ocean on which it floats. The 
more legitimate explanation seems to be that the elevations 
are composed largely of an expansion of the matter in 
the immediately subjacent strata of the earth’s crust, the 
masses above and below being mutually interdependent ; 
where high elevations exist, therefore, the strata below 
are deficient in density, having parted with some of their 
contents. At low elevations the density is normal, since 
there is no appreciable upheaval of matter, while under 
the sea there is a contraction of matter and consequently 
an increase of density. These views have generally been 
supported by the pendulum experiments of Von Sterneck 
in the neighbourhood of the Alps and the still more 
recent measurements carried out near Kolberg in connec- 
tion with the German geodetic operations. 
tune, however, in all these inquiries is that it is impossible 
to detect the distance below the surface at which the 
excess or defect of matter may exist, and, therefore, the 
intimate connection between the unevenness in the earth’s 
crust and the unequal distribution of subjacent material 
is not clearly demonstrated. 
But in India, the very wealth of observation spread 
over a district so wide introduces new difficulties and 
taxes ingenuity to the utmost. The latest authority to 
struggle with the problem is Major S. G. Burrard, already 
well known to geodesists for the skill with which he 
unravelled the perplexities connected with the collima- 
tion of the transit instruments used in the longitude 
inquiries, and later for the very successful determination 
of the longitude of Madras, in which the circuit errors 
are reduced toa minimum. One therefore watches his 
attempt to deal with this old problem with a great deal of 
interest, and is inclined to treat his deductions with con- 
siderable respect. 
The particular point at issue may be stated thus : How 
is the astronomical zenith situated with regard to the 
geodetic zenith at the principal station for reference of 
latitude in India? This station is Kalianpur, the astro- 
nomical latitude of which, after complete discussion, had 
been settled at 24° 7’ 11"10, and from this quantity, by the 
aid of observed azimuths and the constants of Clarke’s 
spheroid, the geodetic latitudes of all the fundamental 
stations have been computed. An examination of the 
results shows that the mean excess of the astronomical 
over the geodetic values of latitude is —2’"0, or, put in 
another way, it is shown that of the 148 astronomical 
latitudes available for geodetic investigation, there are 90 
cases of negative excess to 58 of positive. It appeared, 
therefore,tothelate General Walker, and his conclusion has 
been generally accepted, that the astronomical latitude of 
Kalianpur was too great by 2”, and that the plumb-line was 
not deflected to the north, under the influence of the 
Himalaya Mountains, but was in reality deflected to the 
south. With the view of settling the question, he recom- 
mended that the latitude of a number of subsidiary 
stations within a moderate distance of the central station 
should be derived, and the mean latitude be used for the 
central station, since it might be assumed that such a final 
result would be more free from the effects of deflection 
than the latitude of any single point. Sucha view regards 
the deflection as arising from local causes operative over 
a small area, but Sir David Gill has since pointed out the 
very obvious objection that if local attraction is persistent 
in one direction over large continuous areas, group 
observations such as those recommended would be 
NO. 1699, VOL. 66 | 
NATURE 
The misfor- ; 
81 
insufficient to eliminate the effects, and it is not the least 
important part of Major Burrard’s investigation to show 
that the latent causes of disturbances must be sought 
over very extended areas. 
This work of latitude determination has now been 
completed under the superintendence of Captain L. 
Conyngham, and Kalianpur has been surrounded by a 
chain of stations, of which four are situated at an average 
distance of nine miles and four at an average distance of 
thirty-five, and the unexpected result of the discussion 
is to show that local attraction causes a northerly deflection 
of the plumb-line to the amount of o” 60, thus differing 
by 2’"60 from the value found by General Walker drawn 
from the whole of the Indian observations. The results in 
the prime vertical are not less contradictory, and in the 
following table is exhibited the amount of deflection in 
the two planes, at each of the group of latitude stations 
around Kalianpur, situated within the extreme parallels 
23° 30 to 24° 38’ :-— 
Deflection in the Prime 
| Deflection in the Meridian. Vertical: 
The Group pheleos | The Group Wholeak: 
SYSuST: System. | System. System. 
Daiadhari +ror S. +3°61 S. | +2°13 Ww. +5°35 W. 
Surantal +0'82 S. +3742S. | +3764 W. | +686 W. 
Siron} +169 S. +4'29 S. +2'54 W. +5°76 W. 
Bhaorasa +1°17 S. +3°77S. +o'22 W. +3744 W. 
Kalianpur —o'6o N. +2'00 S. —o'22 E. +3'00 W. 
Losalli —1o02N. +158 S. —6°38 E. —3°16 E. 
‘Jinsia +0798 S. +3758 S. _ _ 
Salot = — —4°49 E. —1'27, K 
Kamkhera —a'ts N. +o'45 S. +o'o4 W. +3'26 W. 
Abmadpur —2'49 N. +o'rr S. +2°27 W. +549 W. 
The stations in this table proceed regularly from the 
north towards the south, and, confining attention solely 
to the deflections in the plane of the meridian, it is clear 
that north of Kalianpur we get a southerly deflection, 
while on the southern side the plumb-line tends to the 
north. Clearly, then, the Himalaya chain,which has been so 
frequently invoked to explain inconvenient discrepancies, 
will not avail here. And the insufficiency of such a hy- 
pothesis is still more clearly shown if the deflections be 
examined at stations nearer to the mountains. At Dehra 
Dun, the most northerly, in latitude 30° 19’, the deflection 
is 38” ; in latitude 29° 31’, the deflection is reduced to 7”, 
and disappears entirely in latitude 27° 51 ; while south 
of Kalianpur we meet with northerly deflections diminish- 
ing in amount as Cape Comorin is approached. Major 
Burrard clearly puts the dilemma thus: ‘‘ If Himalayan 
attraction is capable of producing a deflection of 38” at 
Dehra Dun, its effects must be felt at Cape Comorin ; on 
the other hand, if Himalayan attraction exercises no 
effect on plumb-lines south of latitude 29° 31’, it cannot 
produce a deflection of 38” at Dehra Dun.” 
We cannot follow Major Burrard through the various 
steps by which he seeks to remove the anomalous re- 
sults, but his method is as exhaustive in theory as it is 
laborious in practice. He considers the effect of the 
surrounding ocean on the derived longitudes, and shows 
that these results stubbornly enforce the necessity of 
admitting the entire compensation of the ocean. He also 
asks whether it is possible to introduce any admissible 
alteration in the dimensions and ellipticity of Clarke’s 
spheroid, and the answer is not less certain. He finds 
that Clarke’s major axis is the most suitable for the Indian 
longitude arcs, and that, as concerns latitude, while one 
belt of negative maxima requires an ellipticity greater 
than 1/289, the large sub-Himalayan deflections demand 
an ellipticity smallér than 1/311. He therefore concludes 
that the accepted spheroid is not a source of serious 
error, and that the Indian observed latitudes favour the 
Clarke spheroid. 
