part 2] THE TALENTIA^S' SERIES. 171 



Valentian age by direct reference to the graptolitic 

 scale. 



At Haverfordwest, the only district outside Girvan where the 

 lowest-known Upper Llandovery rocks exist, the base of the 

 Uzmaston Group coincides with the base of the Grala. We have 

 seen, also, that in the southern areas comprising Woolhope, May 

 Hill, Malvern, Marloes, and Wooltack, the highest Upper Llan- 

 dovery rocks are succeeded conformably by the lowest Wenlock 

 rocks or their representative, the Woolhope Limestone. Elsewhere, 

 especially in the Caradoc and Llandovery districts, a thin band of 

 greenish shales intervenes between the Upper Llandovery and the 

 base of the Wenlock ; but the affinities of those shales link them 

 with the underlj'ing beds, and they are probabl}'' represented in the 

 districts previously mentioned by calcareous sandstones indistin- 

 guishable from the ordinary Upper Llandovery type. They ma}^ 

 therefore, be grouped with the underlying rocks as a special local 

 development. If this be done, then the upper limit of the Upper 

 Llandovery is defined by the base of the Wenlock, as, in fact, is 

 the upper limit of the Grala Stage. Thus both upper and lower 

 limits of the Upper Llandovery, where fully represented, coincide 

 with those of the Gala Group. 



Dame Ethel Shakespear's investigations showed that the Taran- 

 non rocks of Tarannon as redefined were also equivalent to the 

 Gala rocks, including the Hawick subdivision, and are therefore of 

 the same age as the Upper Llandovery shelly rocks, so that the 

 classification of the Upper Yalentian into Upper Llandovery + 

 Tarannon is without meaning. Until the Tarannon area was 

 revised, the chief purpose of the term was to denote a group of 

 purple and green shales which underlay the base of the Wenlock, 

 and it was coupled with Llandovery in order to make clear the 

 intention to include all the rocks up to the base of the Wenlock. 



Now that it has been proved to be equivalent to the Gala on 

 the one hand and to the Upper Llandovery on the other, a more 

 homogeneous classification of the graptolitic and the shelly facies 

 of the Valentian is obtained by dropping this term, which has been 

 the source of so much confusion, and reverting to the older- 

 established Birkhill-Gala, and Lower and Upper Llandovery re- 

 spectively. The Tarannon area will, however, always be known 

 by reason of the clear succession therein displayed, and of Dame 

 Ethel Shakespear's classic researches upon the rocks and their fauna. 



The foregoing conclusions are summarized in the accompanying' 

 table (p. 170). Here the space allocated to each graptolitic zone 

 in the Lower and the Upper Yalentian respectively is proportional 

 to its thickness, in a given district. It is found that, although 

 the actual thickness of the zones varies greatly from one locality to 

 another, their relative proportions remain approximately the same. 

 It is not known, however, whether the thickness of a given zone 

 bears any strict relation to the length of time involved in the 



Q. J. G. S. No. 306. o 



