December 16, 1891.] 



Garden and Forest. 



597 



pear early in June, and last nearly a month. The variety dif- 

 fers in having double flowers, and in blooming more freely. 

 The flowers of the double form continue in a perfect state 

 much the longest, and they are excellent for cutting, lasting 

 from two to three weeks in fresh water. The stems should 

 be cut away immediately after the flowers fade, and then the 

 foliage will retain all its luxuriant beauty throughout the sum- 

 mer and autumn. The variety can be propagated only by 

 division in spring and autumn ; but seeds of the species are 

 obtainable occasionally, and they should be sown under glass 

 early in spring. 



Thermopsis Caroliniana— As the specific name implies, this 

 excellent plant is a native of Carolina, the northern state of that 

 name, where it is said to occur plentifully in the mountainous 

 regions. The plant ranges in height from four to five feet, the 

 unbranched stems bearing large ternate leaves of bright green 

 color, and a terminal spike of showy yellow pea-shaped flow- 

 ers closely packed around the stalk in a mass from six to 

 nine inches in length. The flowers are very attractive in the 

 mixed border during the month of June, and the plant blooms 

 quite freely. The best position for it is the back line of a nar- 

 row mixed border or the centre of a broad bed or border in 

 which the plants decrease in stature to more sides than one. 

 I have seen a large mass of it planted in an isolated and fully 

 exposed position, but it is more striking than beautiful in this 

 way. It is more effective when planted in strong clumps 

 among other plants of similar size and contrasting colors, and 

 the flowers light up the dark green of dull shrubberies won- 

 derfully while they last. The various stems of each clump 

 should be loosely secured to a stout stake in the centre when 

 they become too long to stand against wind and heavy rain, 

 but established plants may otherwise be safely trusted to care 

 for themselves. It is not exacting in its requirements as to soil, 

 and is often found in luxuriant health where many plants of 

 similarly robust appearance would certainly succumb. The 

 large crop of seeds ripened annually afford a ready means of 

 propagation ; but they are, however, slow to germinate, and 

 should be raised in heat during early spring. 



Cambridge, Mass. __ M. Barker. 



The Forest. 



The Subjection of Torrents by Reforestation of 

 Mountains. — II. 



CEVENNES. 



THE range of the CeVennes is formed by a series of massive, 

 but less rugged mountains than the Alps or the Pyrenees. 

 These mountains are veritable fountains of water, and the 

 rains there fall in deluging showers, such as are seen no- 

 where else, and all the waterrlow there is eminently torrential. 

 It is usually in September that these terrific water-spouts are 

 formed, when the mountains (relatively slightly elevated) are 

 unable, as in the Alps and Pyrenees, to store a part of the pre- 

 cipitation in the form of snow. 



As a consequence of the orographic arrangement of the 

 ground, the valleys are continuous, and do not empty, as in the 

 Alps, directly and abruptly into an alluvial plain. Thus one 

 never finds there, even among the most powerful and erosive 

 streams, the great cones of eroded material so characteristic 

 of the mouth of torrents. The transportation of the materials 

 which they bear, after having torn them from the mountains, 

 continues in the torrential river, whose channel becomes 

 wider and wider as it fills up, more erratic and hence more 

 dangerous for the rich farm-lands or the great industrial 

 centres in the lower valleys. 



The absolute denudation of the summits and of the higher 

 basins is characteristic of this region. Forests there are still 

 more rare than in the Alps ; the slopes are there furrowed by 

 a series of torrents side by side, and by myriads of deep and 

 yawning ravines, even on the most gentle inclinations. 



THE PYRENEES. 



The chain of the Pyrenees is a geographic unit clearly de- 

 fined. It includes a variety of climates less complete than the 

 Alps, but having this inestimable advantage that limestone is 

 replaced in great part by igneous rocks, which are more solid 

 and less easily eroded. The general exposure is toward the 

 north, and rains are copious and frequent. The forests, as 

 well as the pastures, reach over vast areas, and are in a com- 

 paratively good state of preservation. 



The torrent is only found here as an exception, but an ex- 

 ception that will become general if no restraint is placed upon 

 it, and the Pyrenees would suffer what the Alps have under- 

 gone, notably in the upper basins of Gave de Paul, the Ga- 



ronne and the Ariege. The remembrance of the inundation 

 of 1875 i s st i u thrilling, and the activity of certain torrents 

 shows a disquieting tendency to increase. It is characteristic 

 of this region that the active torrents are all of recent forma- 

 tion, most of them having arisen within our own memory. 



A comparison of the three regions leads to these conclu- 

 sions : (1) That restorative work in the Alps will demand far 

 more energy and expense than in the two other regions com- 

 bined ; (2) That in the Pyrenees the important work of correc- 

 tion could be carried through in a relatively short time, and 

 that reforestation there would not be necessary over any great 

 areas ; (3) That in the Ce"vennes a long series of small correc- 

 tive operations should be carried on in connection with the 

 creation of great masses of forest at the sources of the innu- 

 merable streams which descend both toward the ocean and 

 the Mediterranean ; (4) Finally, the Alps require the greatest 

 as well as the most numerous corrective operations ; and in 

 most of the higher basins a reforestation so complete as'to 

 hold in constant subjection the formidable torrents which 

 threaten to make a very desert of the frontier so important to 

 south-eastern France. 



SUMMARY. 



These works for the public safety are needed in seventeen 

 departments where torrents develop full activity, but adjoining 

 districts also need the benefit of the law, and it would be well 

 if at least one-third of the entire area of France should be 

 made the theatre of " permissible treatment " by the govern- 

 ment. 



From the beginning of this century means of defense against 

 torrents have been studied. 



In articles on this subject engineers have usually confined 

 themselves to the question of the regulation of the deposit of 

 material loosened by the streams, and its retention within the 

 gorges or at the point where they empty into the valleys. 

 But these studies and plans aimed solely to diminish the rav- 

 ages of the torrents ; no one even dreamed of suppressing 

 them entirely. Surrell first introduced the idea of fighting 

 torrents to their utter extinction. 



After having discovered by a series of precise observations 

 the causes of the formation and the action of torrents, as well 

 as the causes of the quieting or the extinction of certain ones 

 among them, he set forth the following propositions : 



1. The presence of a forest on a surface prevents the for- 

 mation of a torrent. 



2. The clearing away of a forest renders the soil a prey to 

 torrents. 



3. The development of forests leads to the extinction of tor- 

 rents. 



4. The felling of forests increases the violence of torrents, 

 and may even cause them to be renewed. 



Experiments were carried on during thirty years by foresters 

 over all the mountainous regions of southern France. Innu- 

 merable observations, made in the mean time, on the unex- 

 pected appearance of new torrents and the influence of exist- 

 ing forests in the same regions, and the definite extinction of 

 a great number of torrents, even in this short period, have 

 fully demonstrated the immovable solidity of these funda- 

 mental principles in the great enterprise of the restoration of 

 mountain lands. 



As Surrell has said, " Nature, in clothing the mountains 

 with forests, placed the remedy by the side of the danger ; 

 she opposes the active forces of the waters with other active 

 forces from the kingdom of life. Ineffectual, indeed, are all 

 our efforts beside these great remedies which Nature dis- 

 penses and patiently uses through the centuries ! Our paltry 

 works are only defenses, as their name indicates ; they are 

 passive masses opposed to active forces, inert obstacles in 

 opposition to eager forces which are always aggressive. 

 Why, then, does not man utilize these living forces to suppress 

 torrents at his will ? The fundamental problem is simply to find 

 the best means of covering with a mantle of vegetation the 

 lands ravaged by torrents now or threatened by them in the 

 future. The defense must not begin in the lower reaches of 

 the streams ; these will defend themselves as the conditions 

 above are corrected. The real warfare must be carried into 

 the higher regions of the mountains. Every system of de- 

 fense which does not first check the erosions in the mountain 

 will always remain incomplete." 



These ideas were not shared by all. Eminent engineers 

 divided torrents into two classes : the curable ones, which 

 they thought might be subdued by reforestation ; and the incur- 

 able, whose basins did not seem susceptible of being cov- 

 ered with trees. The first class they would commit to the 

 foresters, the second to engineers. The Legislature did not 



