Mr. U. 15. Iliiuls Oil UcDi/raji/iic llulamj. 99 



imxiii|j: together China ink uiul gamboge, and that by varying 

 the proportions, all the shades of green are ])rodnccd whieh are 

 found in the foliaceous organs of plants. DeCandolle adds, that 

 though this explanation is somewhat mechanieal, it is very likely 

 to be eorreet. 



The Chinese, with that praetieal a|)plication of faets to pur- 

 poses of utility whieh so ])re-eminently distinguishe.s them as a 

 nation, have availed themselves of some of these features in their 

 landscape gardening. To convey the appearance of distance, 

 trees of the loftiest and largest growth with foliage of the deepest 

 green are selected for the foreground. Others of smaller stature 

 and more subdued shades are jjlaced in the distance ; whilst to 

 vary the surface and increase the a])parent extent, groups of 

 suitable trees, selected also with a due regard to the influence of 

 the seasons, are judiciously scattered about. To aid the effects 

 produced by vegetation, representations of old nuns, receding 

 w^alls, and time-worn rock-work, are all made available. 



All the species whieh enter into the flora of a country do not 

 equally assist to give a certain jjliysiognomy ; and they will vary 

 considerably among themselves as to their })ower of doing so. 

 Trees and shrubs from their size will surpass herbaceous plants, 

 whilst these also will excel each other according to any pecu- 

 liarities of organization they may possess. The prevalence of 

 certain families in particular regions will often be highly charac- 

 teristic : none would feel at a loss when surrounded by Cruciferce 

 and Umhellifera; ; or when traversing a plain covered with Eri- 

 caceae, a grove of ChaDiarops pa/meito, or a thicket of Melastoma. 

 When plants become gregarious, especially those of larger grow^th, 

 they deeply impress features on the scenery around them. An 

 impression of this kind is also liable to occur from an opposite 

 cause. Supposing in a forest where the trees possess a similar 

 aspect and maiuier of growth, there occur one or two, or more 

 indi\ iduals of a totally different character ; these latter will 

 obtain a prominence which is not due to them from their num- 

 bers, but from their })eculiarities. A circumstance of this kind 

 strongly impressed me in the Brazilian forest, when encountering 

 a few individuals of Araucaria braziliensis in the midst of trees 

 not distinguished for their ])hysiognomy. 



To obtain some numerical value for those plants which appeared 

 to give a character, I assembled a number of different species in 

 Devonshire, and after carefully examining them, and making the 

 freest allowances, I came to the conclusion that in 100 species 

 forty-eight might be considered as contributing to the physio- 

 gnomy of the flora, and fifty-two were too insignificant to assist in 

 this. It occurred to me to repeat the observation within the 

 tropics, where I found the numbers very similar, though the pro- 



