INTEODUCTION. Ill 



a very common one among leaves, prevailed upon me ultimately to call the first 

 an oval and the last an ellipsis. LinnaBus describes elliptical or oval as a form 

 whose length exceeds the breadth, " superiore et inferiore extremitate angus- 

 tiore," a phrase which would at least include my elliptic, if it does not limit the 

 sense to that form. Lindley says that oval or elliptic differs from oblong in being 

 acute at each end. Lanceolate is only a narrower form of elliptic. 



Thus we have, in the blunt series, subrotimd, where the length is less than \\ 

 times the breadth ; oval, where it is from 1^ to 4 times ; ohlong, between 4 

 and 8 times ; linear, 8 or more times. I should, however, call a leaf ohlong if the 

 sides were distinctly parallel, even though the length were hardly twice the 

 breadth. Linnaeus, in the ' Philosophia,' has introduced the word Ugulate or 

 strap-shaped, but has hardly made any use of it. Perjiaps it would be advan- 

 tageously employed for a form between oblong and linear, restricting the former 

 within narrower limits. In the acute series, we have only elliptic, from 1 j to 3 

 times the breadth, and lanceolate, where the length is more than three times the 

 breadth. 



AH these forms are further distinguished, when needful, by the addition of the 

 terms broad, exact (exquisite of Bertoloni), and narrow. De CandoUe seems to 

 have taken a fixed type for each form ; and he designates as compound all 

 deviations from it. It is, however, more convenient to admit a wide variety of 

 proportion under each term, and to use the compound only when the form is so 

 intermediate that we can hardly teU which word ought to be used. I will add 

 here, that a compound epithet indicates an intermediate, and not a compound 

 form. Thus ovato-lanceolate is not ovate at the bottom and lanceolate at the 

 top, but in all parts intermediate between ovate and lanceolate. Bpatulato- is an 

 exception to this rule ; a spatulato-ovate leaf being a spatulate leaf of which the 

 blade is ovate. In this, I believe, I foUow the usual practice, but I do not know 

 that it is anywhere distinctly explained. 



Cordate I consider as ovate with an indentation at the base. Cordato-oblong 

 would therefore indicate a leaf between ovate and oblong, indented at the base ; 

 and oblong with a cordate base, a strictly oblong leaf with such an indentation. 



Triangular has been used ambiguously, because it is not explained whether the 



angle at the base is included in the number. I understand by it a form nearly 



triangular, where the stalk is attached to one of the sides (fig. 16). Where the 



insertion of the stalk also forms an angle, but where the upper and lower parts of 



the leaf are very unequal, it becomes deltoid, and this is the way in which I think 



Linnaeus used the word ; though his definition will not, perhaps, bear a critical 



examination, and his figure in the ' Philosophia ' does not represent a leaf at all 



corresponding with the general use of the word. Rhomboid, or rather rhombic, 



is where the lateral angles are nearly equally distant from the summit and base. 



5 2' 



