UPON PLANT-HOUSES. 93 



places, but also the difficulty to procure perfect seeds, or to 

 preserve growing plants alive which oppose our endeavours, Not 

 to mention such tribes as Lorauthaceae, Cytineae, Balauophoreae, 

 EafflesieacefE, and the like, who has ever been so fortunate as to 

 study in a glass-house a Lardizabalea, Schizandracea, Lacistemea, 

 or Gyrocarpea? It is curious that many kinds of seeds will out- 

 live long voyages, while others similarly constituted soon perish. 



But something more than the mere existence of some rare 

 productions is required to compensate for the cost and trouble of 

 obtaining them. The Seedling Bertholletia, Caryocar, Lecythis 

 and the like, which are annually imported from the colonies, perish 

 soon, and without being of the least use, that could not be equally 

 well derived from Herbarium specimens. Such articles may 

 parade in garden catalogues as great rarities, but are in reality 

 very worthless items. The longe^dty of trees, such as those just 

 mentioned, is generally in a reversed ratio as regards their period 

 of development ; but there are exceptions ; f. e. the Bombaceae, 

 which will grow with rapidity when transported into our houses. 

 We ought further to consider well, whether certain forms, which 

 may be easily made to flower, are of sufficient importance to take 

 up room that might be more usefully occupied by others. Whether 

 we cultivate 20 or 100 sorts of Erica is, scientifically speaking, 

 indifFereut ; and the same may be said of many Cape and New- 

 Holland Thymelacese, Proteacese, Epacridefe, Leguminosae, and 

 Rutacese. A botanic institution should enable the student to 

 examine the characters of living genera; but of numerous species 

 only in such cases, where the Herbarium renders their study 

 difficult. The sorts of Gnidia, Pimelea, Erica, many Proteaceae, 

 Rutacefe, Grasses, Cyperacefe, &c., may be quite satisfactorily ex- 

 amined in their dried state ; but the beautiful and minute syste- 

 matic character's of Schott's Aroidese, Roscoe's Scitamineae and 

 Nees V. Esenbeck's Acanthacese, require to be examined in the 

 hot-house ; so also with respect to the fertile researches of Gaspar- 

 rini and Miquel on the diversity of the floral structure in the 

 fig tribes. 



Owing to some peculiar predilection in favour of certain forms, 

 many of them are admitted in exorbitant numbers ; such as 

 Cape and New-Holland, Mexican and South European Compositse, 

 which are represented to a very disproportionate extent in many 

 botanic gardens. Thus the Paris Garden contained in 1850, 350 

 genera of that family, according to H. Brongniart's catalogue. 

 Peculiar moi'phological forms give great eclat to the house ; f. e. the 



