86 OLDHAM: THE STRUCTURE OF THE HIMALAYAS, ETC. 



outer edge towards the hills. The section is not complete but 

 it is probable that the maximum depth is not at the northern 

 boundary of the trough though nearer to it than to the southern, 

 and probably close to the outer edge of the hills, as is found to be 

 the case on the section through the Dehra Dun. The westerly 

 extension of this trough cannot be traced for want of observation, 

 but it is natural to expect that it dies out as the point is reached 

 where the Salt Range impinges on the Himalayas, just as it seems 

 to die out in the east where the Assam range, in a similar manner, 

 bridges the angle between the Himalayas and the ranges separating 

 India from Burma. 



The geodetic observations also show, in confirmation of the 

 deduction which was drawn from geological evidence, that the 

 great spread of alluvium in the Punjab differs from that of the 

 Gangetic plains, in that it is formed by a comparatively thin cover- 

 ing over the rocky floor, and only when the hills of the western 

 frontier are approached do we find indications of a trough com- 

 parable with that which borders the Himalayas ; but this is a 

 matter which cannot be dealt with here. 



The observations, dealt with so far, are confined to that portion 

 of the Gangetic trough which lies south of the limit of the hills, and 

 only incidental reference could be made to the form of that portion 

 of the trough which lies within the Siwalik area, between the outer 

 edge of the hills and the main boundary. There are only two series 

 of geodetic observations which cross this boundary, one near the 

 88° of longitude, where there is only a narrow fringe of hills between 

 the main boundary and the edge of the plain, and the other near 

 the 78° of longitude, where there is an exceptionally complete 

 series of latitude and gravity observations in the Dehra Dun and 

 in the Himalayas on the one hand, and the plains on the other. 



Taking the latitude observations first, these are included in 

 table No. 26 (page 87), to which the two northernmost stations 

 of the series in table No. 20 are added, in order to bring the two 

 series into relation with each other. In the table No. 26 the dis- 

 tance of each station from the main boundary, and from the south- 

 ern edge of the Siwalik hills is given in the second column, these 

 distances being in every case measured in a direct line, normal to the 

 course of the boundaries, and expressed in the nearest whole mile. 

 In the case of the first three stations two values are given for the 



[ 234 ] 



