part 3] OF THE INDUS, BEAHMAPUTEA, AND (iAXGES. . 141 



the trough deep ; but in the central Punjab they were long, and 

 the trough-area wide, shallow, and split up into a number of sub- 

 troughs, the total breadth from Cherat to the Grhaggar channel 

 being 300 miles. The Murree zone, the Siwalik /one of the Soan 

 valley, and perhaps the Jhilam, Chinab, Ravi, Sutlej, and Grhaggar 

 valleys are all sub-troughs or elements of this attenuated compound 

 trough. 



It has been established by Medlicott, Middlemiss, and others 

 that the Tertiary foothills of the Himalaya are divisible into 

 parallel zones separated one from the other by successive 'boundary- 

 Faults " — dislocations of a reversed nature — each of which forms 

 a close approximate boundary of deposition as regards the zone 

 whose northern boundary it constitutes. In other words, these 

 4 boundary-faults' mark the northern shore of the Tertiary sea or 

 the northern permanent bank of any longitudinal Tertiary river ; 

 north of each 'fault,' at the time of its formation, was land, 

 drained by mountain-streams which have left little or no deposit 

 -within the mountain-area, but brought down large quantities of silt 

 on to the plain below. In the Hazara mountains of the Northern 

 Punjab Mr. Middlemiss showed that there were three of these 

 faults dividing the area into four more or less defined zones, and 

 another farther south shutting off a fifth somewhat less defined 

 sub-zone, the Murree sub-zone. 1 



The southernmost of his Hazara zones Mr. Middlemiss called 

 4 the Nummulitic zone,' since it consisted mostly of these rocks ; 

 its northern boundary-fault is the limiting line of their deposition, 

 with the exception of a considerable patch of Nummulitic Lime- 

 stone east of Abbottabad. which, however, is itself cut off by 

 another • boundary-fault ' on the north-west. This Nummulitic 

 ;zone it is now proposed to trace, before we attempt to reconstruct 

 the secpience of events attending the birth and life of the lndo- 

 brahm. From a glance at a geological map (PI. X) it will be 

 seen that there is a continuous outcrop of Nummulitic beds ex- 

 tending from the Hazara hills through Kohat, Waziristan, and 

 Baluchistan to the Sind coast. East of Hazara in the Jammu hills 

 Nummulitic rocks again appear. Some unfossiliferous strata 

 marked as doubtful Trias or Carboniferous by Lydekker (the Kiol 

 beds), along the main boundary-fault of this area, 1 would also 

 include with the Nummulitics. Farther south-east there is a gap 

 until the Nummulitics of Simla are reached, and another until we 

 eome to a very thin interrupted band between Dehra Dun and 

 Naini Tal. 2 The Simla and Dehra, Dun Nummulitics consist of 

 marine beds, and what is probably a mixture of iluviatile and salt- 

 lake deposits. Similar alternations of lagoon and freshwater beds 

 compose most of the upper stage of the Xunimulitics of the 

 Northern Punjab, and to a varied extent are seen among the rest 

 of the Nummulitic belt from Kohat to Sind. But, since in every 



1 Mem. Geol. Surv. Ind. vol. xxvi (189G) p. 260. 



2 Further work may show these gaps to be shorter or non-existent. 



