198 FORMATIONS AND GUILDS [Part I 



impossible are confined to the tropics. The rain-forests of tlie tropics an 

 always moist. This is much less true of rain-forests of the warme 

 temperate zones and not at all true of the summer-forests of highe 

 latitudes, for the cold of winter there constitutes a period of physiologica 

 drought, which, even with the heaviest atmospheric precipitation, is mor 

 opposed to the suppl}^ of water than great dryness when united with hea 

 Under heat and dryness transpiration is indeed much greater, but thf 

 absorption of water is not hindered and the nightly dew is of direi 

 advantage to the superficial roots of the epiphjtes, whereas under temperati 

 conditions there is no supply of water to be set against its loss by epiph}-te! 

 for the frozen or at any rate very cold exposed roots transpire, but absoi 

 nothing. 



Corresponding to these conditions of life, the vast majority of epiphyti 

 belong to tropical rain-forests. Only there do they luxuriantly cov 

 stems, branches, and frequently even the leaves of trees, and often ther 

 selves attain the dimensions of trees. In districts with markedl}- d)' 

 seasons, and on the isolated trees of savannahs, epiphytes are eith 

 completely wanting, or rare and represented by relatively few forms. Sut 

 forms as are found are emigrants from the rain-forests, and their presence 

 always a sign that the dry season is not long, or, as in the monsoon-fores 

 is accompanied by copious dew. 



The origin of the guild of epiphytes in tropical forests may have cor 

 about in the following way. Many teirestrial plants living in the forest a; 

 able to settle and grow on rough fissured stems, in the forks of boughs, a 

 on other spots where humus collects. This happens in the tropics in t 

 case of many Solanaceae, Melastomaceae, and ferns. From such accident 

 epiphytes true epiphytes were derived, since many of these plants ovvfl 

 their existence to this faculty, which secured for them a safe retreat outsife 

 the seat of conflict. The competition on the trees was limited to ii 

 species, because the faculty of existing as an epiphyte demands cerfci 

 definite and by no means common characters. Obviously, for instan, 

 only such plants germinate on trees as are provided with seeds capableJf 

 dispersal not only in a horizontal, but also in a vertical direction, and 1 

 latter demands adaptations to arboreal animals and to the wind. Moreo\[, 

 the seeds must be very small, so that they can enter narrow crevices, a i 

 in the case of dispersal by the wind they must be extremely light, beca 

 vertical wind-currents are weak in the forest. The seeds of epiphy 

 actually fulfil all these conditions ; they are always small, and eit 

 surrounded by succulent envelopes, as in Aroideae, many Bromeliacei; 

 Rubiaceae, Melastomaceae, Ficus, Cactaceae, and Gesneraceae, or they e 

 extremely light, even like powder, as for instance the spores of ferns, e 

 seeds of orchids, or they are provided, in spite of their ver\- small dimensics, 

 with a most suitable parachute, as in Rhododendron, many BromeliaceS, 



