DEVONIAN. 327 



The outlines of this formation were carefully traced along the shores from the St. Croix 

 River to Point Lepreau, and in some places, as at Pea Point, and L'Etang Head, the basal 

 conglomerate was found to have a great thickness, aggregating not far from 4,000 feet. At 

 Lepreau Harbor, where these rocks are well exposed, they apparently rest directly and con- 

 formably upon the Devonian shales and sandstones of the Mispec and Little River groups of 

 the St. John Devonian basin. As they elsewhere underhe the basal beds of the lower Carbon- 

 iferous, including the marine limestones of that series, it would now appear that the rocks of 

 the Perry River group, as a whole, represent the upper portion of the Devonian system of 

 southern New Brunswick, as was early suggested by Sir Wilham Dawson and others from the 

 evidence of the contained plants. 



The above statement by Ells that the rocks of the so-called " Perry" of Lepreau 

 Harbor rest conformably upon Devonian rocks of the Mispec group raises a question 

 as to the relation of the " Perry" at Lepreau with the true Perry formation of Maine. 

 If the Mispec rocks are meso-Carboniferous (see p. 331) strata which rest conformably 

 upon them must be middle Carboniferous or later; whereas the true Perry is Devo- 

 nian, probably Upper Devonian. Furthermore, plant remains which, like those of 

 the Mispec, are now referred to the middle Carboniferous have been found near 

 Lepreau Harbor. In view of the great thickness and highly disturbed condition of 

 the strata it is probable that more than one terrane is represented in that vicinity. 



The Devonian age of the Perry formation in the type locality in Maine has 

 recently been placed beyond question by the investigation made by Smith and 

 White/^* who have assembled in the field and from the literature the essential facts 

 of stratigraphy and paleontology. They thus describe the Perry formation: 



The Perry formation as exposed in the towns of Robbinston and Perry is divisible into four 

 members. Two of these members consist of sedimentary strata, while the other two are inter- 

 bedded lavas and associated volcanic breccia. The section may be described thus: 



Upper lava: Green lava with columnar parting and amygdaloidal texture. Includes one bed of conglomerate 

 and sandstone. 



Upper sandstone: Coarse red and brown sandstone and conglomerate, with small amounts of shale. Includes 

 one flow of lava near base. 



Lower lava: Basaltic lava and breccia. Includes one thin bed of conglomerate near top. 



Lower conglomerate: Coarse conglomerate and red sandstone, with thin beds of sandy shale. 



The thickness of these members can not be definitely given. The lowest member is evi- 

 dently the thickest, but probably its thickness varies considerably in different parts of the area. 

 As will be noted later, the Perry sediments were largely shore deposits, so that the distance 

 from the source of the material would control largely the thickness of the beds. In case of the 

 volcanic members the different lava flows give the impression of beuig rather more persistent 

 in thickness than might be expected. « 



White made a thorough study of the plant remains originally described by Sir 

 William Dawson, and of others including new collections from the Perry. He 

 describes the species, discusses their distribution, and concludes : ''^ 



Of the entire flora described " from Perry, Sphenopteris filicula and Bepidocystis siliqua 

 are types which appear to belong to groups more or less distinctly characteristic of the lower 

 portion of the lower Carboniferous. The fragment described as Palseostachya? sp., and the 

 forms earlier pubUshed as Lycopodites comosus and Carpoliihes lunatus are too Httle known or 



" Lepidodendron gaspianum, a characteristic Devonian plant, unknown in the Carboniferous, and reported from 

 Perry by Dawson, is omitted from the list as insufficiently represented for identification in the Perry material which 

 I have seen. 



