FLORA or "WESTERN ESKIMAUX-LAND. 25 
De Cand., Prodr. vol. i. p. 201. — Deless. Icon. Sel. vol. ii. t. 70. — TttraceUion ellipsoidcum, Turcz. 
Fort of St. Michael^ Norton Sound, in alluvial soil ; apparently introduced by the Russians from 
Siberia. 
No one comparing tbe figure in Delessert's * Iconcs SelectEe,' with that given on Plate II. of this work, 
would fancy that the two were intended for the same plant. In Delessert's figure the silioula* are nearly 
globose, iu mine they are pyrifonn. The difference arises from the blunder committed by the artist of the 
' Iconea Selectae,' who has represented an immature fruit magnified, for a full-grown one. The fruit, from 
being almost globose, becomes, in advancing, stretched at the base, until, when ripe, it has assumed the 8liai>e 
of a pear. Turczaninow enumerates two species of Tefrapoma, distinguishing them principally by the length 
of the style. I find that no importance can be attached to that character. In one and the same specimen, 
as Dr. Planchon has well remarked in Herb. Houk., the stigma is cither sessile or placed upon a style of 
more or less length, and as no other difference can be detected between T. larlar<Ffoliuni and T. Crusianum, I 
have united the two species, under the name T. pyriforme. The size of the plant is extremely variable. Some 
of the specimens before me measure only a few inches ; others, gathered in Norton Soimd, are from two to 
three feet high. The leaves also are subject to great variation. They are entire, lyrate, or deeply pinnatifid, 
and, like the whole plant, more or less covered with hair. The valves of the silicula are generally four in 
number, and the genus Tetrapoma has on that account been looked upon as the normal type of the Order 
Oruciferce. That number, however, is not constant. I observed some fruits with two, some ^\ith three, 
and even some with five and six valves. The native country of the pknt I believe to be Northern Asia. 
It has been found between Aldan and Ochotsk {Turczaninou^ , at Isinga, at Koragisk, and by me in 
Norton Sound, where I have reason to believe it has been brought by the Eussians, but where it is now 
perfectly wild. 
Plate II. T'ig. 1, flower ; 2, sepal ; 3, petal ; 4, stamen ; 5, G, ovary ; 7, ovary cut horizontally ; 8, ripe 
fruit ; 9, fruit with one of the valves removed ; 10, valve ; 11, fruit cut horizontally ; 12, seed ; 13, seed cut 
horizontally ; 14, embryo : — all mcujmfied. 
34. HuTCHiNsiA calycina, Desv., var. integrifolia, Hook., Fl. Bor. Amer. vol. i. p. 59.— Terr, ct 
Gray, Flor. of N. Amer. vol. i. p. 114. Cape Kruzenstem. 
HYPERICINE.^. 
35. Parnassia palustris, Linn., Hook. PI. Bor. Amer. vol. i. p. 82.— Ton. et Gray, Fl. of N. 
Amer. vol. i. p. 148. Kotzebuc Sound and Port Clarence; from Point Barrow to Bear Lake River 
(PuUen) . 
86. Pakxassia Kotzebuei, Cham, et Schlecht., Hook. Fl. Bor. Amer. vol. i. p. 83. t. 28.-Torr. 
et Gray, Fl. of N. Amer. vol. i. p. 149- From Port Clarence to Cape Lisbume. 
On' examining P. Kotzeluei I found that nearly one-half of the specimens collected in Western 
Eskimaux-land had five stigmas and a capsule with five valves ; and in analyzing P. palustru &:om that 
country the same result was obtained. It remains, therefore, to be seen whether the same formation exists 
in other species, and whether P. palustri. from other locaUties may not have five ™lves, Torrey and Gray, 
iu speaking of P. Kot.cluei, remark, that the figure in Hooker's ' Mora Boreah-Amencana exhibits views 
of pentacarpeUaiy capsules, and conclude that they represent monstrosities. But if, with Don and Ln^ey, 
we look upon the genus Farna^sia as a true member of the order J^pen^nne. we cannot be suj^nsed ^ 
find a ciuLry arrangement of the fructification, and mu.t rather be mchned to consider it as the normal 
formation than as a monstrosity. 
