108 SAPROl'HYTEH IX WATER, ON THE liAIiK UF TREES, AXU OX ROCKS. 



Schilleriana, a native of the Philippine IslaiKls; its roots ai-e rigid, compressed, 

 and about 1 cm. in breadth; the surface turned away from the trunk is slightly 

 convex, and has a granular structure and metallic glitter like a lizard's or chame- 

 leon's tail. The surface towards the trunk is flat and without metallic glitter, 

 and upon it, close behind the growing point, there is a whitish fur consisting of 

 >short, thickly packed, absorptive cells. When the tip of one of these roots comes 

 into contact with the bark it grows so tirmly to the substratum by means of the 

 absorption-cells, that it is easier to detach superficial bits of the bark itself than 

 the I'oot. The latter, once fixed, flattens out still more and becomes strap-shaped, 

 whilst creeping outgrowths proceed from it, forming strips which may ultimately 

 attain a length of 1^ metres. The sight of a trunk covered with these long 

 metallic bands is one that never fails to excite wonder even in the midst of the 

 world of orchids, wherein, as is well known, there is much to marvel at. 



In other species of tropical oi'chids, e.g. in Sarcanthus rostratus (fig. 15), the 

 roots are not flat from the beginning, but become so when they come into con- 

 tact with the bark. A root is often to be seen which arises as a cylindrical cord 

 from the axis, then lays itself upon the bark in the form of a band, and further on 

 lifts itself once more, resuming at the same time the rope form, as is shown in the 

 illustration. Here also complete coalescence takes place between the bands and the 

 bark, and the union is extremely close. Similar conditions have been observed to 

 hold in many Aroidece living on the bark of trees. The plants in question lie with 

 their stems, leaves, and roots flat against the trunks, so that they suggest a covering 

 of drapery. Taking, for instance, the Marcgravice (Marcgravia paradoxa, M. 

 umhellata), one might at first sight suppose that they adhere to the bark not only 

 by the roots, but also by the large discoid leaves, which are arranged in two rows. 

 A very remarkable fact also, in connection with these plants, is that they only 

 grow on very smooth and firm bark. When transferred to a soft sulistratum, such 

 as jnould or moss, they lauguish, because their roots are unable to enter into close 

 union with a support of such loose texture. This is also true of most tropical 

 orchids living on bark. When their seeds are transferred to loose earth devoid of 

 humus, they do indeed germinate, but then perish, whereas when sown on the bark 

 of a tree, they not only germinate, but grow up with ease into hardy plants. 



Where steep rocks occur near clumps of trees it is not uncommon for the same 

 species of plants to grow on both. Allusion is not here made to kinds which, like 

 ivy, have their roots in the earth at the foot of rocks and trees, and creep up the 

 one or the other inditterently, using both merelj»^ for support and not as sources of 

 nxitriinent, and clinging to them by means of special attachment-roots. The remark 

 is applicable also to plants which live on the products of the decay of organic 

 bodies, for example many tropical Orchidew, Dorstenice, Begonia, and Ferns: and in 

 cooler parts a number of Mosses and Liverworts. It is not diflicult to explain this 

 phenomenon in the case of species which derive their food from vegetable mould. 

 The crannied wall of rock is, in a certain way, analogous to the rugged bark of a 

 tree. The holes in the rock are filled in course of time with black vegetable mould. 



