ATLANTIC KEGION 581 



(now slates) were deposited. The hiatus between the Armorican sand- 

 stone and Angers mudstones corresponds to the hiatus between the Lower 

 Arenig and the Glenkiln of Scotland, and this indicates that the retreat 

 of the sea at the end of Arenig time resulted in a narrowing of the English 

 Channel. Whether the British area emerged completely, so that a con- 

 tinuous land-mass extended from the old land of Scotland ( Caledonia') 

 to France (Armoriea) must be determined by a further study of the 

 Welsh sections. If these, too, show a hiatus between the Arenig and 

 Llandeilo, such complete withdrawal of the sea would be indicated. The 

 fact that in South Wales the Upper Arenig and Lower Llandeilo form 

 apparently a unit, the Llanvirn group of Hicks, 2,000 feet thick, near 

 Saint Davids, indicates that here at any rate deposition was continuous. 

 This does not argue, however, for continuity to the east of this, for there 

 dry land probably existed, as is indicated by the conditions in the Baltic 

 region, to be discussed subsequently in this paper. The southern Wales 

 district may have constituted an embayment from the Atlantic Ocean of 

 Llanvirn time. 



Mediteeeanean Region 



iberian peninsula 



On the southern border of the ancient Armorican land -mass, or the 

 Iberian Peninsula of early Ordovicic time, the Tremadoc beds, as already 

 noted, are found at Montaigne Noire in southern France and at Barce- 

 lona in Spain; but inland (northward and westward) they are overlapped 

 by the transgressing Arenig, here also represented by rocks of the Armor- 

 ican sandstone type. Much, if not all, of the southern part of the pen- 

 insula (SjDain and Portugal) seems to have been covered by the Arenigian 

 during its greatest advance ; but central France probably remained above 

 water. With the succeeding retreat of the sea, however, the whole of this 

 region seems to have been uncovered again, for even at Barcelona the 

 Armorican sandstone is only slightly developed, having probably been in 

 part removed by erosion during the retreat. More evident, however, of 

 the presence of a hiatus between it and the succeeding beds is the fact 

 that it is followed by beds with Ortliis actonice Sow., 0. vespertilio Sow., 

 0. calligramma Dalm., 0. (Dalmanites) testudinaria Dalm., Plectam- 

 bonites sericens Sow., Echino pi ha ■rites cf. balticus, etcetera, these clearly 

 proving Caradoc age for these beds. Thus the Llandeilo is absent alto- 

 gether, being overlapped by the Caradoc, which probably represents a 

 continuation of the transgressive movement begun in Llandeilan time. 

 Speaking in terms of the American series, beds of early Beekmantowu 

 age are here followed by Middle Trenton, a condition not infrequently 



