the similarity of the leaves is rather with some Mexican 

 Oak. 



It differs from C. mexicajia, chiefly in its acute leaves, 

 alwavs more or less toothed at the margin, circumstances not 

 occurring in the specimens from Hartweg numbered 341, re- 

 ferred by Mr. Bentham to that species, but which I regard as 

 entirely different from the No. 488 above alluded to ; and at 

 variance with the specific character of DeCandolle, with 

 which the No. oil perfectly agrees. It may however be a 

 mere variety of that species, which was found in the cold dis- 

 tricts about Bolanos, while CI. quercifolia comes from the 

 warmer climate of Jalapa. 



It is in some respects very like CL Brasiliensis, or at 

 least the supposed variety which Loiseleur has called C. 

 scabra, but the leaves of that species are much more lanceo- 

 late, and the flowers are not a quarter the size, and in dense 

 racemes. 



It is not hardier than C. arborea, and requires the same 

 treatment, growing freely if planted in any good garden soil 

 w4iich is rather light. It is easily increased either by layers 

 or seeds. 



