ogy if the Riversdale plants, whose correlation by Sir William Dawson with the Millstone Grit is not likely to prove far from correct, were now to be pronounced Devonian on account of their paleontological identity with the fern ledges whose erroneous reference to the Devonian was earlier forced upon Sir William by the findings of the stratigraphers. The study of the plants collected at several hundred localities in the Pottsville formation along the Appalachian trough proves conclusively that the St. John flora is from nearly the same stage, while, as I have elsewhere pointed out,' it is probable that a portion of the section at the "fern ledges" is contemporaneous with the upper portion of the Pottsville.

2

THE MACKAY'S HEAD PLANT BEDS.

MacKay's Head, another of the sections now placed in the Hamilton, furnished a number of the species described in 1873 by Sir William Dawson and referred by him to the Millstone Grit. Dawson's list includes alternateribbed species of Calamites, a large Carboniferous type of Lepidodendron identified by him as L. aculeatum, a fern figured as "Odontopteris?", and other ferns identified as Alethopteris lonchitica and Sphenopteris obtusiloba. These species, though few in number, are recognized by palæobotanists as distinctly Carboniferous. The material is hardly sufficient for a definite correlation, but it seems probable that the flora, especially the Lepidodendron and the Alethopteris, will eventually be found to lie above the Carboniferous Limestone. The small amount of published plant evidence points towards the Pottsville or Millstone Grit, to which the MacKay's Head beds were referred by Sir William Dawson.

1 20th Ann. Rept. U.S. Geol. Surv., pt. 2, 1900, pp. 913, 917.

2 Dr. G. F. Matthew suggests the reference of a portion of the St. John fern ledges to the Silurian.

DISCUSSION OF THE PROBLEM.

From the foregoing it will be seen that the reference by Mr. Fletcher and Dr. Ells of the Riversdale-Harrington River and MacKay's Head plant beds to the Hamilton would necessitate the existence of essentially Middle Carboniferous floras in the Middle Devonian. This could be admitted only on absolutely indisputable stratigraphical evidence, such as their occurrence in a continuous normal section, with abundant characteristic Middle Devonian marine fossils in close association or appropriate sequence. Evidence of this character appears to be wholly absent or quite inconclusive. It does not appear that the testimony of the animal fossils from the plant beds differs widely from that of the plants. The contradictions between the correlations by stratigraphy and those of palæozoologists1 are perhaps less striking only because the evidence of invertebrate fossils is less abundant or is of a class of animals (e.g. Phyllopoda and Anthracomyæ) of somewhat uncertain stratigraphical value.

The grounds for confidence in the palæobotanical evidence are strong. The plant remains on either side of the Lower Carboniferous-i. e. in the Devonian and in the Upper Carboniferous-exhibit the same succession of essentially identical contemporaneous floras in New Brunswick and in other regions of Nova Scotia as well as in the United States. This presumptive evidence in favor of the contemporaneity or approximate contemporaneity of the floras of the Lower Carboniferous and Pottsville (Millstone Grit) of the United States and Europe on the one hand with the corresponding floras of the Horton and Riversdale beds of Nova Scotia on the other hand is supported by the geographical relations as well as the conditions of migration. A land stretch of no very great extent, and at times reduced to a lowland, between the Nova Scotia and

1 Dr. Ami places the Riversdale beds within the Carboniferous period, but at the base of the Lower Carboniferous. (See Trans. Nova Scotia Inst. Science, Vol. X., p. 171.)

the Appalachian basins, must have been clothed throughout by the general flora characteristic of the particular geological time; and on the occasions of the great floral changes of the earth it should, with the important aid of air and water currents, have carried migrating species so rapidly from shore to shore that the period of each successive flora would be essentially contemporaneous in either basin. In view of these relations only undeniable stratigraphic evidence can prove that the identical floras do not belong to the same great period, and that we have in Nova Scotia the unique phenomenon of a typical middle Carboniferous flora in rocks of the middle Devonian. Such an occurrence, as yet unknown elsewhere in North America or Europe, and in direct contradiction of all biopalaeobotanical data, would scarcely be harder to explain for one district than would it be to account for its absence in other regions of the Northern hemisphere.

The existence of the wide gulf between the correlations by stratigraphy and those by palæontology does not admit of the conclusion that the verdict of the stratigraphers is final. With all respect for their ability and the conscientious character of their work in a field offering great difficulties of surface concealment, it is still proper to enquire of the Nova Scotia geologists if there is not a further stratigraphical explanation; whether, for instance, certain plant beds at different points may not have been wrongly correlated in carrying the stratigraphical identifications across to the type (reference) sections; or whether the Carboniferous Limestone of their sections is not much younger than is generally supposed; or whether certain of the superpositions are not possibly due to overthrust faulting.

NOTE OF EXPLANATION.

A seeming discrepancy between Mr. Kidston's and my own correlations of the Riversdale-Harrington River plants

appears in Mr. Fletcher's table,' where Mr. Kidston is recorded as referring these beds to the Coal Measures, while I am quoted as correlating them with the Pottsville (Millstone Grit). This difference represents only a variation in the nomenclature of the Coal Measures in Great Britain and in America. The upper portion of the Pennsylvanian Pottsville appears, as I have elsewhere indicated, to be contemporaneous with the Lower Coal Measures of Europe, while the succeeding beds in the northern Appalachian area of this country represent the Middle Coal Measures, etc., of the Old World. Accordingly my correlation of the fern beds at St. John with the upper Pottsville nearly corresponds to Mr. Kidston's correlation with the Lower Coal Measures. I do not insist that the Riversdale plants are necessarily so late as the Pottsville, though I believe they belong close to, if not in, the Pottsville stage. It is interesting to note that the conclusions reached by Mr. Kidston and myself, each independently and without the knowledge of the other, as to the age of the plant beds under discussion, are nearly identical. Not less interesting is it that our conclusions, with the exception of those relating to the St. John fern ledges, tend in general to sustain the correlations made by Sir William Dawson thirty years or more ago.

So far as I can recollect, I have never seen a fossil from the Union, or expressed an opinion as to the age of that formation, which in Mr. Fletcher's table I am credited as doubtfully referring to the Coal Measures. The terranes referred by Mr. Fletcher to this formation, which from the descriptions seems to have some features in common with the Catskill, do not appear to have yet yielded plant remains of stratigraphical value.

1 Op. cit., p. 243. See p. 3 of this paper.

2 Bull. Geol. Soc. America, Vol. XI., pp. 166, 173; 20th Ann. Rept. U.S. Geol. Surv., pt. 2, pp. 912, 917.

ADDITIONAL NOTES ON THE FLORA OF CAP-À-L'AIGLE.

By REV. ROBERT CAMPBELL, M.A., D.D.,

In the RECORD OF SCIENCE, Vol. IV., No. 1, pp. 54-68, and Vol. V., No. 1, pp. 38-40, appeared lists of plants collected at Cap-à-l'Aigle, County of Charlevoix, up to that date. The work of noting the plants of that region has gone on from year to year, during the latter part of July and the first three weeks of August. The Laurentian Mountains form not only the background of the landscape but also the backbone of the Flora, if I may be allowed a free metaphor. If the rocks that face those mountains. are the oldest on the earth's crust, as is generally held, then the plant-life developed on them may be regarded as also the oldest that exists in the world, since the geological evidence indicates that the north-eastern portion of the continent was the earliest to emerge from the primeval waters. On this ground a special interest attaches to the flora of the Saguenay basin and the heights of Labrador beyond. It will be observed that the district is specially fruitful in sedges, rushes, grasses and other endogenous plants. The following catalogue embraces all the additional species that I have noted up to date:

OSMUNDACEE R. BR.

OSMUNDA REGALIS L.-Royal Fern.-Loutre Marsh. July.

POLYPODIACEE R. BR.

WOODSIA ILVENSIS (L.) R. BR.-Rusty Woodsia.-Capà-l'Aigle rocks. August.

CYSTOPTERIS BULBIFERA (L.) BERNH.-Bulblet Cystopteris. -Banks of St. Lawrence. August.

CYSTOPTERIS FRAGILIS (L.) BERNH.-Brittle Fern.-In many places. July.