884 



ness to the foliage as contrasted with the neater appearance in this 

 respect of Q. sessiliflora, in which the leaves are remarkably flat, ra- 

 diating, and lying over each other in a horizontal position and in 

 parallel planes, by which it may be distinguished at a great distance 

 from its congener. Whether Q. Robur and Q. sessiliflora be accounted 

 distinct or not, I can by no means agree with Fries (Corp. Flor. Pro- 

 vin. Suec. p. 144) in thinking that the differences between the two 

 are owing to poverty of soil, as he confidently affirms.* From what 

 we know of the mutability of character in Tilia and Ulmus, as also 

 in Salix, Populus and other Amentiferae,f we should be cautious in 

 admitting to specific distinction a tree which in many of its states 

 does not always present very definite limits to the marks separating it 

 from Q. Robur. 



* Q. sessiliflora seems to abound most in the hilly districts of the north of England 

 and in Wales, and there to reach its fullest dimensions. At Boultibrooke, near Pres- 

 teign, Radnorshire, the seat of Sir H. J. Brydges, Bart., are the finest, most charac- 

 teristic trees of the sessile oak I have ever seen, almost rivalling the Spanish Chestnut 

 in beauty of foliage, and of noble magnitude and height. 



•j- The word Amentaceae employed to designate the natural order of ament-bearing 

 plants is manifestly incorrect, and nearly as bad as Graminaceae for Gramina. The 

 termination in acece (which it would be highly desirable, for the sake of that neatness, 

 precision and uniformity, at once useful and ornamental in scientific nomenclature, 

 should be unanimously adopted instead of the vague, arbitrary and often ungramma- 

 tically-constructed and mixed terminations of De Candolle, still in partial use amongst 

 botanists) does, or ought to express, the relations which certain groups of plants bear 

 to some typical genus concentrating in itself the leading or most prominent characters 

 distinguishing such groups in the aggregate. Thus, Violaceae implies a group, family 

 or order of violaceous or violet-like plants, not violets themselves, but allied to or 

 closely resembling them, having most of the peculiarities of form, structure, appear- 

 ance, and properties of the genus Viola, which serves as a type or standard of compa- 

 rison to judge all its allies by. But to apply the term Graminaceae to designate the 

 order or family of Grasses, or Palmaceae that of Palms, as has been inconsiderately 

 done in some instances, at once destroys the entire force and meaning of the Latin 

 adjective termination in acece, which has a comparing efficacy, that an incorrect ap- 

 plication of it wholly stultifies. For as there is no such genus as Gramen or Palma, 

 these two natural orders represent themselves, and are not typified by any one genus 

 of Grasses or Palms, to which all the remaining genera composing these orders may 

 be referred as a standard of comparison. To compare a thing to itself is obviously 

 absurd, because every comparison necessitates the existence of two objects to be com- 

 pared ; to call the Grasses and Palms, therefore, Graminaceae and Palmaceae, is only 

 to declare them to be grass-like and palm-like plants, or in other words, like them- 

 selves, that is, Grasses and Palms, which is in fact a comparison only in sound. The 

 proper designation of these orders should be Gramina and Palmae ; for Gramineae 

 (plantse gramineae) and Palmeae (plantae palmeae) are only preferable to Graminaceae 

 and Palmaceae as being belter Latinity. In like manner Amentaceae is highly objec- 



