THUYA, ELODEA, VALLISNERIA, AND CARKX. 325 



limit crosses the Albany at some distance from the sea, extending westward 

 to a point about seventy-five miles southwest of Trout lake, thence south- 

 ward to Lake Winnipeg and the United States boundary. It is one of the 

 trees most likely to be found in this formation. This species has been rec- 

 ognized by Sir William Dawson in the drift of the Roseau river, Manitoba, 

 and of Montreal (Leda clay) and the Ottawa river.* 



ELODEA CANADENSIS (?), MICHX. 



A specimen of soft stone bearing the impress of a small branching plant 

 and the carbonized remains of another of the same kind. This was from 

 the collection of Mr. Tyrrell, made in 1887, and obtained from Rolling 

 river, Manitoba, two miles above Heart hill. A slide of the same plant and 

 from the same locality, from Dr. G. M. Dawson, shows the plant to have 

 been herbaceous, but with a distinctly vascular axis, the wood cells of which 

 are thin walled and with rather blunt terminations. This vascular structure 

 is surrounded on all sides by a distinctly parenchymatous structure. Asso- 

 ciated with this plant are many diatomaceous remains belonging to fresh- 

 water species, among which I have recognized Navicida lata, N. legumen, 

 Encyonema prostratum, Denticida latda, and various species of Licmophora (?) 

 and Cocconeis. It is therefore clear that the plant is not a seaweed. The 

 distinctly branching habit and the structure suggest Elodea, although the 

 state of preservation is not such as to render exact comparison possible. I 

 therefore refer it provisionally to our common Canadian species, E. cana- 

 densis, which is everywhere found in fresh water. 



VALLISNERIA (?). 



Several fragments of the same earthy material as above, bearing each a 

 small fragment of a leaf. This is in each case linear, with a well-rounded 

 apex, and usually about 2.5 mm. wide. The epidermis is apparent under a 

 pocket lens. In fact the remains appear to consist wholly of the two epi- 

 dermal layers, which may be separated readily. Under the microscope the 

 epidermal cells are found to be well preserved. No stomata have been found, 

 and this, together with the presence of fresh-water diatoms, would indicate 

 that it must have been a submerged, aquatic plant. The structure strongly 

 reminds one of Vallisneria, to which I shall provisionally refer it. This 

 plant is everywhere common in fresh water, and is very likely to have 

 occurred in such a locality as that from which the fossil was obtained. 



CAREX MAGELLANK'A, LAMARCK. 



The Green's creek nodules contain an abundance of leaves, evidently of 

 grasses and sedges. In one nodule from the Miller collection and in two 



* Can. Nat., New Ser., Vol. Ill, 1808, p. Tl\ Report on 19th Parallel, 1875, p. l\\ ; Notes on IVst-I'lio- 

 cene, op. cit., 1871, p. 40L 



