172 Pfior. o. T. JONES ON [vol. Ixxvii, 



formation o£ the sediments. It m^j, in fact, be argued that a 

 thin band of graptoKtic shale has taken longer to form than the 

 intervening non-gra]:)tolitiferous strata. In the accompanying 

 table the zones comprise the shale-bands containing the zone-fossil, 

 as well as the other sediments which occur above, below, or between 

 these bands ; and for this reason it may be assumed with some 

 probability that, where the conditions of deposition have remained 

 tjlerably uniform, as in the Birkhill Period, that the thickness of 

 a zone bears some proportion to the length of time which elapsed 

 during its formation. 



It cannot be assumed, however, that the conditions of deposition 

 which prevailed during the Birkhill Period persisted into the Gala 

 Period. On the contrary, it is certain that a considerable change 

 of physical conditions occurred between the two periods. The 

 space allotted to the Upper Valentian is therefore made arbitrarily 

 equal to that of the Lower Valentian, and subdivided among its 

 five zones in the proportion of the thickness of these in the 

 Tarannon area. 



An important fact which emerges from the table is that in all 

 districts where the shelly facies prevails there is no evidence of a 

 transition from the Lower to tlie Upper Valentian ; but, on the 

 other hand, there is every indication of a physical break at this 

 horizon. 



IX. The Base oe the Valentian. 



Since it is now agreed to include the whole of the Valentian 

 Series in the Silurian, it becomes important to define as accurately 

 as possible the base of that series. As previously mentioned, 

 Aveline attempted to draw a line separating the Lower Llandovery 

 rocks from the Bala ; but he was unsuccessful, and for the shelly facies 

 this was first accomplished by J. E. Marr & T. Roberts ^ in the 

 Haverfordwest district, where the}^ recognized a group of conglo- 

 merates and sandstones underlying the fossiliferous Gras works Beds, 

 and probably representing their basal members. This was confirmed 

 by the officers of the Geological Survey during the re-examination 

 of the area around Narberth and Haverfordwest, and by the aid of 

 these beds the base of the Silurian in that area was mapped in 

 detail. 



No unconformity was detected between the Silurian and the 

 Ordovician. The fossiliferous Slade Beds (Ordovician) are suc- 

 ceeded in some sections by a group of dark shalj^ mudstones which 

 are overlain by a grey sandstone. In others a conglomerate of 

 varying thickness is associated with the mudstones, occurring in 

 places within them and in others at their base. It is only in the 

 latter case that the base of the Silurian can be accurately located. 

 In Girvan that horizon is defined by massive conglomerates 

 (Mulloch Hill, etc.) below which there is probably an unconformity 

 and overstep. In the Moffat area the base of the Birkhill was 

 drawn by Charles Lapworth at the bottom of a peculiar band 



1 Q. J. G. S. vol. xli (1885) p. 476. 



