THE ANCIENT AND MODERN FLOEAS OF MONTPELLTEE. 223 



country which probably included fifty or sixty times that number, 

 or to conclude that a plant was not in the neighbourhood or even was 

 not abundant because its remains have not been there found, is what 

 we can by no means admit. Supposing, as is most likely to have 

 been the ease, from the scanty evidence at hand, that the spot in ques- 

 tion was broken ground mostly covered by a dense mass of wood and 

 thicket in the midst of which arose these springs, with perhaps here 

 and there some bold rocks around them ; what would be the leaves 

 and fragments most likely to fall into the streams thus formed before 

 leaving the woods ? Surely those only from the immediately sur- 

 rounding trees or bushes, or the very few herbaceous plants which 

 would grow under them. The droppings of those which covered the 

 country even at a hundred yards from the encrusting fluid would 

 have very little chance of finding their way to it. Winds would not 

 carry them far into a wood; rains would wash them into the beds of 

 torrents which almost always eat their way into a lower level than 

 the springs of a rocky valley ; and the fragments that might be 

 brought by birds or other animals would be very few and far between. 

 Let us imagine one of our own wooded Welsh valleys with a copious 

 spring issuing from the side of the hill, let us follow this from its rise 

 for a few hundred yards, picking up here and there along its course a 

 basket of dead leaves, sticks and other rubbish caught in its eddies. 

 Suppose that in sorting out our booty, we can detect representations of 

 twenty different species (taking this number as proportional to the 

 general vegetation of the country) what would they most probably 

 be — leaves of oak, ash, wych-elm, alder, sallow or osier, ivy, bramble 

 — perhaps a wild rose or hawthorn leaf— a fragment of dead rush or 

 of brake fern, and a few other such. Should we be likely to find any 

 traces of heath or furze, any wild thyme, bilberry, cranberry or club- 

 moss, and if not, should we at once conclude that there were no such 

 plants in Wales ? and yet the chances that these should be among our 

 twenty species are quite as great, as that Cistuses, Genista scorpms, 

 thyme, rosemary, and lavender, should be among Planchon's thirty, 

 although they may have been as common as they are now. He has 

 fomid one single leaf of each of two oaks Q, robur and Q. ilex, he 

 has found none of Q. coccifera, from that he concludes that the two 

 former were then in the country, and the latter a more recent inva- 

 sion. But supposing them all three to be equally abundant and at no 

 great distance from the spot, among the millions and millions of 

 leaves they must annually shed over a few acres of land, the diffe- 



