THE UNDERGROUND FORM OF FLOOR OF GANGETTO TROUGH. 89 



This, however, is not the only possible explanation, or the only 

 supposition which will fit in with the facts. If we supposed the 

 depth of the trough under the Siwaliks to be about 10,000 feet the 

 effect of the defect of density through this depth would be approxi- 

 mately equal in amount to that of the Siwalik plateau, but opposite 

 in sign. The two would, in these circumstances, neutralise each 

 other, and if we then supposed the trough to be deepened outside 

 the Siwaliks, or, in other words, the floor of the trough to form a 

 step upwards under the outer edge of the Sub-Himalayan region, 

 we would have much the same effect produced as in the supposi- 

 tion just examined. In reckoning the effect of such an hypothesis 

 as has been outlined it will be necessary to make some modification 

 in the distance from the outer edge of the hills, as given in table 

 No. 26, for the two stations Lachkua and Bullawala. The dis- 

 tances given in the table are measured from the outer edge of the 

 visible hills, as marked on the one 1-inch map, but it is not reasonable 

 to suppose that a rise in the floor of the trough, if it exists, would 

 follow all the sinuosities of the boundary between the hills and the 

 gravel slope at their base, and these two stations are situated where 

 the outer edge of the hills takes a distinct curve inwards from its 

 general course. If we suppose that the rise in the floor of the trough 

 spans this inward bend, these two stations would lie at some three 

 or four miles from the course of the rise, or from the edge of 

 the deeper trough. Making this allowance, and assuming a depth 

 of 10,000 feet under the Siwalik plateau and of 15,000 feet outside 

 it, we get the figures given in column II of the table, which will be 

 seen to agree almost equally well with the results of observation as 

 those in column I. A slightly closer agreement might be obtained 

 by varying the assumed depth of the trough outside and within the 

 Siwalik area, but no real advantage would be obtained by trying 

 to attain a greater degree of precision than the method permits. 



As has already been pointed out, the calculations, both of the 

 observed deflections and of the deflections which are to be expected 

 on any given hypothesis, involve the adoption of certain assump- 

 tions, which in no case exactly agree with what is found in nature, 

 but are approximations to the conditions which are either known, 

 or may be expected, to exist. A variation in these assumptions 

 would produce a change in the absolute value of the deflections 

 given in the table, but any such variation, if applied to every station, 

 would produce a similar change in all, and the differences would be 



[ 237 ] 



