October 6, 1911] 



SCIENCE 



441 



to the Pottsville group, which covers the 

 " Lower " and " Middle Coal Measures " of 

 the British Isles. As to their Carboniferous 

 age neither of us had any doubt; and I think 

 I speak correctly for Mr. Kidston when I add 

 that the extensive discoveries of the past ten 

 years, though without exception confirmative 

 of our correlations, can have made us but 

 little more certain of our ground. In reply 

 we have heard repeated the arguments of the 

 " sixties," that the flora differs from all other 

 Devonian floras because it is estuarine or 

 marsh, and that the relative metamorphism 

 and the stratigraphy of the beds unmistakably 

 prove their Devonian age. 



It is impossible here to give particulars or 

 even the substance of the paleobotanical evi- 

 dence. Briefly, it is clear that the flora com- 

 prises an association of genera characteristic 

 of the Upper Carboniferous; that many of 

 the species are identical with plants in the 

 Pottsville of the Appalachian trough, while 

 other forms differ no more than may naturally 

 be expected in view of the remoteness and 

 isolation of the basin ; that all types character- 

 istic of the Devonian, including estuarine and 

 delta beds in other parts of the world, are ab- 

 sent; that the evidence of the associated ani- 

 mal fossils is in agreement with that of the 

 plants; that the metamorphism is not greater 

 than in the Rhode Island Coal Measures ; and 

 that, in this region of extensive Pleistocene 

 and sea concealment, and of folding, faulting 

 and metamorphism, the stratigraphic evidence 

 presented is neither clear nor conclusive. 



We are now told that the floras (" faunas ") 

 of the " fern ledges " are Silurian ! They are 

 said to differ from all other pre-Carboniferous 

 floras because they are " delta " floras ! To be 

 more explicit, the plant-bearing delta deposits 

 are correlated by him with other beds in dif- 

 ferent regions shown by their marine remains 

 to be Silurian. The " Dadoxylon sandstones " 

 are accordingly referred to No. 2 of the Mas- 

 carene Silurian series, while the " Cordaites 

 shales " are said to belong to No. 3 of the same 

 series. In other words. Dr. Matthews now 

 concludes that the " fern ledges " are of Clin- 

 ton and Niagara ages. By no process can he 



possibly be interpreted as permitting the 

 youngest plant beds to be above the Helder- 

 berg. Hence, if any paleobotanist has at any 

 time entertained sufficient confidence in the 

 stratigraphic arguments to cause real anxiety 

 lest the " fern ledges " might possibly be 

 Devonian, the new stratigraphic " correla- 

 tion " must certainly put him completely at 

 ease. 



The almost astounding faunal discoveries 

 brought to light by Dr. Walcott in the Ca- 

 nadian Rockies should deeply impress on every 

 paleontologist the virtue of conservatism; but 

 the possible analogies with the " fern ledges " 

 floras are very limited. The wonderfully pre- 

 served fauna exhibiting so wide a systematic 

 range and such singular biologic relations in 

 the Cambrian of Canada nevertheless com- 

 prises characteristic Cambrian fossils. On 

 the other hand, to assume that under local 

 environmental conditions (which there is no 

 reason for regarding as unique) there were 

 developed at one known spot in the world not 

 only a group of identical genera in character- 

 istic association, but also species in part iden- 

 tical with those later reproduced in the 

 " Upper Carboniferous," the flora being largely 

 composed of fern genera nowhere else known 

 in pre-Carboniferous beds and devoid of all 

 the Devonian and Silurian types supposed to 

 be contemporaneous, is certainly going to the 

 extreme in the doctrine of parallelism in de- 

 velopment. 



To return to Dr. Matthews's paper: The 

 discussion of the " fern ledges " floras and 

 their ages is supplementary to' the announce- 

 ment of the discovery, in beds correlated by 

 Dr. Matthews with the No. 1 (Medina) divi- 

 sion of the Mascarene series, at Beaver Har- 

 bor, New Brunswick, of an Arthrostigma flora. 

 Arthrostigma has been regarded as character- 

 istic of the Devonian. We shall therefore 

 look forward with keen interest to the full 

 publication with, let us hope, adequate illus- 

 tration of this older flora. The new flora 

 which is from a different region is said to have 

 nothing in common with the " fern-ledges " 

 floras, which include such common Carbonif- 

 erous genera as Galamites, Annularia. Astero- 



