CHAP, xcv THE PEOPLING OF NEW SOIL 353 



Hclianthemum annuum, Ononis spinosa, O. repens, Origanum vulgare, 

 and Anthyllis Vulneraria, also such shrubs as Crataegus, Juniperus, and 

 Viburnum Lantana. Juniperus itself gives rise to a dense bushy growth. 

 When the vegetation had reached this stage several other kinds of shrubs 

 with fleshy fruits established themselves, and in twelve years gave rise 

 to an impenetrable bushland. Finally, Sorbus, Fagus, and other trees 

 appeared, and forest arose. The soil was constantly changed and improved 

 by the death of the previous occupants ; each kind of vegetation sup- 

 [jressed another, until finally forest vanquished busliland, which could 

 maintain itself solely at the margin of the forest as a mere fringe. 



The slopes of marly diluvial hills laid bare by water usually clothe 

 themselves at first with a community consisting mainly of members of 

 the segetal and ruderal flora, and especially with annuals ; and not until 

 later is there found the flora that is characteristic of these sunny slopes 

 and is formed by longer-lived plants. 



Fires in Forest and Grassland. 



New soil is not always entirely free from propagative bodies. Its mode 

 of origin determines this question. For instance, soil whose vegetation 

 has been devastated by fire is seldom entirely sterilized thereby ; it 

 preserves seeds, living roots, and rhizomes in great numbers, and new 

 plants can sprout forth from these. Yet the vegetation may be interfered 

 with to such an extent that a new kind of vegetation can make its 

 entrance.^ 



Tropical and subtropical grasslands, including steppes and savannahs, 

 in various parts of the world are designedly fired by the inhabitants : in 

 some places for hunting pursuits ; in others for grazing purposes, because 

 the burning of the old, dry carpet of grass and perennial herbage causes 

 a rapid upgrowth of new grass. Several of these kinds of communities, 

 and savannahs in particular, entertain scattered trees.^ It naturally 

 suggests itself that where one tree can grow many trees could flourish and 

 give rise to forest. And when, as a matter of fact, no forest exists this 

 lack has been ascribed to fires. M. Christie, Mayr, and Redway^ suggest 

 that the prairies may be treeless because fires prevent trees from growing 

 up ; fires may also be the cause of the absence of earthworms and snails. 

 Asa Gray expresses the opinion that, between soil receiving rain sufficient 

 to bring forth forest, and soil receiving rain inadequate for this, there is 

 a debatable tract where relatively slight causes decide whether forest or 

 prairie shall prevail. 



The Brazilian campos are regarded by P. V. Lund* as derived from 

 forest that has been metamorphosed into savannahs (campos) by fire. 

 Reinhardt ^ and Warming ^ take another view, although neither of them 

 denies the important modifying influence of fires ; moreover Volkens ' 

 adopts this opinion. 



' There is a rich literature dealing with fires in prairies, savannahs, and forests. 

 [See Warming, 1892; L. S. Gibbs, 1906; Pearson, 1899; also see p. 298. 

 ' See pp. 296 and 299. 

 ' M. Christie, 1892 ; Mayr, 1890 ; Redway, 1894, 



* P. V. Lund, 1835. * Reinhardt, 1856. 



* Warming, 1892. ' Volkens, 1897. 



WARMING /^ a 



