1XX PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [May I9II, 



in the mass, and they must necessarily make an angle with the 

 upper and lower surfaces of the mass as a whole when its formation 

 is completed. 



Neither the upper nor the lower surface of the ' bed'' so formed 

 will he a true time-plane. Its lower surface will be in contact 

 with the floor on to which the deposit has been pushed outwards, 

 and, therefore, will be of decreasing antiquity when followed 

 outwards from the area of denudation. As the 'floor' may be 

 receiving fine-grained deposit (for instance, air-borne or laid down 

 in a lake), portions of this finer-grained deposit will be contem- 

 poraneous with parts of the pebble-deposit. Thus the ' time-planes ' 

 will pass down through the pebble-deposit into the finer-grained 

 'bed' beneath. On the other hand, the top of the ' bed' is con- 

 ditioned by slope and stream-velocity ; and, where pebble-deposit 

 ceases at any point, if may be succeeded there by the laying down 

 of fine-grained material (except along the actual stream-courses), 

 each part of which, in its turn, will be synchronous with some part 

 of the pebble-deposit on tho outer margins. The ' time-planes ' 

 will, therefore, here pass up into the overlying fine-grained deposit. 



Thus, while it may be convenient to map such masses of pebble- 

 1 beds ' as single units, and even to think of them as individual 

 i formations,' erroneous conclusions will inevitably follow unless it 

 is clearly realised that they are, in the main, units of structure, 

 texture, and condition, rather than of time. In converting the 

 record yielded by them into terms of time, it is essential in most 

 cases to ascertain their exact conditions of formation, and to be 

 guided by these in interpreting their structures and relationships. 



From the variable and intensely localised conditions which 

 determine the accumulation of ' terrestrial ' deposits, it must follow 

 that formations originating thus will be lenticular and irregular in 

 their distribution, while means of correlation by organisms will be 

 of the scantiest and the most unsatisfactory nature. Breccias formed 

 by weather-action, or by glaciers in one part of the area, may be 

 ■contemporaneous with pebble-beds, gravels, and alluvia in another, 

 with lacustrine sediments or even with wind-drifted sands or loess 

 elsewhere. Delicate correlation in the case of such deposits as 

 those to which I have just alluded will probably always be a matter 

 of no little difficulty. 



In connexion with these deposits, it is instructive to study 

 not only the work of W. T. Blanford and Sir Arthur McMahon 

 in Central Asia, but to consult also the more recent work of 

 Mr. W. R. Eickmers in that region. 



