place names as an aid in the reconstruction of the original

PLACE NAMES AS AN AID IN THE RECONSTRUCTION OF THE ORIGINAL VEGETATION OF CUBA

LEO WAIBEL

[With separate map, Plate I, facing p. 3941

THE island of Cuba," said Las Casas,' "is so full of woods that one can walk its entire length beneath trees." Three centuries later Humboldt2 voiced the same thought: "It would seem as if the whole island had

been originally a forest." And down to our own day this has been the com- monly accepted belief3 But was this, in fact, the condition of the island at the time of the discovery? The question was posed a good many years ago by the eminent Spanish botanist Ram6n de la Sagra.4 In what state was this beautiful part of the world some three and a half centuries ago when the Lord entrusted it to the colonizing European race?" His answer was: "Cov- ered over the greater part with thick and dense woods." Again, more spe- cifically, he asked: "What relation existed . . . between the absolute extent of the forest vegetation and the herbaceous vegetation of the sabanas or lianuras?" And the reply: "We do not know; but historical tradition and the aspect of the yet uncultivated areas lead us to believe that the former was much more extensive than the latter." And there the matter has rested.

I became interested in the subject of the original nature of the Cuban landscape when I learned that the term savana or sabana is of aboriginal origin and means "a treeless plain" (see below). This immediately suggested a toponymic approach to the study. The problem is fundamentally similar to one that has long engaged the attention of European geographers, especially in Germany, where the relations between the ancient forests and natural clearings are of extreme interest.

IBartolome de las Casas: Historia de las Indias, 5 vols., Madrid, I875-I876, Vol. 3: "Es cuasi toda tierra llana y llena toda de montes o florestas" (p. 467); "La dicha isla de Cuba es, como dije, muy montuosa, que cuasi se pueden andar 300 leguas por debajo de arboles" (p. 468); "la isla tiene de luengo cerca de 300 leguas, y se puede andar toda por debajo de los arboles" (p. 469). Las Casas wrote from his two years' experience (I5II-I5I3) on the island as "clerigo" of the party of occupation under

Velazquez. 2 Alexander de Humboldt: Personal Narrative of Travels to the Equinoctial Regions of the New

Continent, during the Years I799-I804, by Alexander de Humboldt, and Aime Bonpland, written in French by Alexander de Humboldt, and translated into English by Helen Maria Williams, 7 vols.

(in 8), London, I8I8-I829, Vol. 7, p. 6i. 3 Censo de la Republica de Cuba, aiio de I919, p. 65. "The whole island was an immense forest

in the epoch of the discovery." 4 Historia fisica, politica y natural de la isla de Cuba, Vol. I3, Suplemento a la secci6n econ6mico-

politica, Paris, i86i, pp. 5o5-I.

VEGETATION OF CUBA 377

In Cuba the process of settlement is relatively recent and tracts of the original vegetation have been preserved; hence the method of field obser- vation can be used. In Europe, on the other hand, one has to rely on indirect evidence, such as place names, size and site of settlement, archeological data, and descriptions of ancient travelers. However, if the German method of Urlandschaftsforschung is applicable at all in a Hispanic-American country, it should work even without field study. Thus I am encouraged to present the results of a preliminary library investigation. My personal knowledge of Cuba is limited to the western part of Habana Province, which I visited briefly in I938; I plan to check the present conclusions by work in the field as soon as opportunity permits.

THE SOURCES USED

Fortunately for my purpose, there exists a large-scale map of Cuba that shows not only many place names but also such types of vegetation as woods and open country. This is the "Military Map of Cuba I906-I908" (referred to hereinafter as the Military Map), on the scale of i :62,500, published by the United States War Department in 70 sheets, each comprising a quarter of a square degree. It was desired "to obtain as much information as possible which will be permanently of value from a military standpoint";5 thus the vegetation cover-thick woods, thin woods, brush, open country, swamps, and cultivated land-was mapped in detail, with results of incom- parable value.

In I9II the War Department published a map of Cuba on the scale of I : 6oo,ooo with quadrilaterals corresponding to the sheets of the Military Map. Also on the basis of the Military Map, and on the scale of i :6oo,ooo, is the map "The Forest Regions of Cuba" which accompanies a report on the forest resources by Georgia E. Wharton (U. S. Department of Agri- culture, Forest Service, I919). Although it makes no distinction between types of forest, it nevertheless gives a clear picture of the forested area for the period. These two maps, adjusted to the Santiago de Cuba and Habana sheets of the American Geographical Society's Millionth Map of Hispanic America, have provided the base for my vegetation map, Plate I, facing page 394.

5 An account of the construction of the map is given in J. L. Schley: The Reconnaissance Map of Cuba 1906-7, U. S. Army Engineer School Occasional Paper No. 30, I908.

The Military Map I906-I908 is the basis of the "Carta militar de la Republica de Cuba," on the same scale, which, since I925, has been currently brought up to date.

378 THE GEOGRAPHICAL REVIEW

Especially important for my investigation were the works of two Cuban geographers. Esteban Pichardo, to whom we are indebted for the first large- scale map of Cuba (22 sheets were published between I840 and I874), was the author of a comprehensive "Geografia de la isla de Cuba." The work was planned to appear in more than 30 volumes or parts,6 but only four parts have been published (Habana, I 854-I 855). Parts 3 and 4 are extremely important for the historical geography of western Cuba. In the "Diccionario provincial casi-razonado de vozes Cubanas" (3rd edit., Habana, i862) Pichardo explains the meanings of place names with regard to the original vegetation. In i863 the first three volumes of the "Diccionario geografico, estadistico, hist6rico de la isla de Cuba" of Jacobo de la Pezuela made their appearance, and in i 866 the fourth and final volume was published. Despite the difference in treatment, it may be considered a welcome continuation of Pichardo's incomplete work. Here also much attention is paid to the distribution of sabanas and woods.

The American botanist Charles Wright studied the flora of Cuba during his explorations from i856 to i867, but, unfortunately, this famous plant collector paid little or no attention to the exact locations of plants. Of much greater value are the collections made in Cuba from I903 to I923 by various botanists of the New York Botanical Garden and described in the Garden's Journal.

In I926 the Cuban botanist Brother Leon7 gave an interesting resume of our knowledge of the flora and vegetation of the island; later8 he presented a classification of the vegetation formations; and now, in collaboration with Brother Marie-Victorin of the University of Montreal, he is at work on "Itineraires botaniques dans l'ile de Cuba," a series of great promise, the first volume of which appeared in I942.9

Finally, there is a work that has not been sufficiently appreciated by geographers, Bennett and Allison's "The Soils of Cuba,"'0 in which much attention is paid to vegetation, its relation to soil conditions, its geographical distribution, and the division of the island into natural regions. Turning

6 Salvador Massip: Esteban Pichardo, Rev. Soc. Geogr. de Ctiba, Vol. I4, No. 2, I94I, pp. 24-43;

reference on p. 28.

7 Cuba, in Naturalist's Guide to the Americas, prepared by [a committee of] the Ecological Society of America, edited by V. E. Shelford and others, Baltimore, I926, pp. 682-694.

8 Flora de Cuba, int Libro de la cultura, Barcelona, I936, Vol. 5, I, pp. 89-I02.

9 Contribs. l'Inst. Botani. de l'Univ. de Montre'al No. 41. IO H. H. Bennett and R. V. Allison: The Soils of Cuba, Tropical Plant Research Foundation,

Washington, I928; with separate "Soil Map of Cuba," I: 800,000. See also H. H. Bennett: Some Geographic Aspects of Cuban Soils, Geogr. Rev., Vol. I8, I928, pp. 62-82; with insert map, i: 3,400,000.


now to the classification of the vegetation formations of Cuba, I shall begin with the sabana, of prime interest for our purpose.

THE SABANAS

Hettner," writing on sabanas, says: "The Spanish word sabana, meaning tablecloth, originally connoted only the surface of a plain. As the latter, however, attains its characteristic appearance only if trees are scarce, the word has received its well known meaning." This derivation has been generally accepted. According to the Encyclopxdia Britannica (i ith edit.), for instance, "savanna or savannah (Span. savana, a sheet, . . . a linen cloth) [is] a term applied either to a plain covered with snow or ice, or, more generally, to a treeless plain." But, as I have already said, another definition came to my attention-in a quotation by Carl Troll"2 from the "Enciclo- pedia universal ilustrada" (I929). This Spanish encyclopedia derives "savanna" from a Carib word meaning "plain without trees, especially if it is extensive." This is the derivation given and explained by the Oxford English Dictionary.

If "savanna" is derived from an aboriginal word, sabanas must have been a feature of the pre-Columbian landscape and not a subsequent, man- made formation, as some botanists and geographers believe. After considerable research I arrived at the conclusion that the present-day "savanna" is of native West Indian origin. The word was formerly written iabana or zabana, with the stress on the second syllable. For example, Las Casas,'3 describing Higuey in Haiti, says: "This province has two parts, one with plains, which the Indians call gabanas, full of beautiful grass." He repeatedly explains that the stress is on the second syllable.

However, as early as the sixteenth century the letter s was substituted for the g and the letter v for the b. Oviedo,'4 for example, writes: "This name savana is applied by the Indians to all terrain without trees, but with much tall grass, or low." Later the spelling "savanna" or even "savannah" was used, but always with the stress on the second syllable. The Spanish word sabana, meaning "sheet," on the contrary, has the stress on the first syllable.

-Alfred Hettner: Vergleichende Landerkunde, 4 vols., Leipzig and Berlin, I933-I935, Vol. 4, p. 8o, footnote i.

I2 Termiten-Savannen, iu Landerkundliche Forschung: Festschrift zur Vollendung des sechzigsten Lebensjahres Norbert Krebs, Stuttgart, 1936, pp. 275-312, reference on P. 275, footnote 2.

3 Op. cit., Vol. 5, p. 258. I4 G. F. de Oviedo y Valdes: Historia general y natural de las Indias, edited by Jose Amador de

los Rios, 4 vols., Madrid, I85I-i855, Vol. I, pp. I44 and I83. Vol. I (i.e. Part I) was published for the first time in I535.

3 8o THE GEOGRAPHICAL REVIEW

It is understandable that non-Spaniards made the mistake of confounding the Indian sabana with the Spanish sabana and of transferring the actual sense of the former in a metaphorical manner to the latter. But even Span- iards made the same mistake, as is shown by the very common application of the term sabanilla (small sabana) to a small sabana.

Another question now arises: What is meant by a "treeless plain" as used by the first explorers? Does it mean the complete absence of tree or shrub? Or is sabana used only in contrast with forest land?

Oviedo distinguishes between three types of gold mines in Haiti. One type is found on sabanas, another in arcabuco (woodland on mountains and plains), and the third in rivers. Here we have the clear antithesis of woodland to sabanas. Thus we may assume that the primary meaning of the term sabana is indeed the reverse of "forest land."

The use of the term sabana by Cuban scientists confirms this conception. According to Pichardo (Diccionario), the Indian word sabana means an "extensive plain without trees." Some, indeed, are so extensive that they reach the horizon. If less extensive, they are called sabanillas or sabanitas (Spanish diminutives); if smaller, sabanazos (Indian?); and if still smaller, sabaneton (Indian?).

Of the I69 place names of the map (Plate I and key), 5 i are sabana names; 26 are sabanilla names; 4, sabanita names; 2, sabanazo names; i, sabaneton name.

The vegetation'5 of these sabanas, according to Pichardo, consists of espartillo or wire grass (Sporobolus indicus); yabunas (Zamia sp.), a xeromorphic cycad with tuberous roots; some guanos, palms with flabellate leaves; yamagueyes (Belairia mucronata), a shrub, hardly I2 feet high, covered with thorns; and peralejos (Byrsonima crassifolia), a many-branched shrub with leaves coated with downy hairs. Other xeromorphic, small-leafed shrubs are maboa de sabana (Cameraria retusa), Brya ebenus, and Erythroxylum ro- tundifolium.

Typical sabana palms are palma cana (Sabal mexicana), jata (Copernicia hospita), palma jata (C. macroglossa), yarey (C. yarey), and, in Pinar del Rio Province and the Isle of Pines only, palma barrigona or bottle palm (Col- pothrinax [Pritchardia] wrightii). According to Bennett, the designation sabana in Cuba refers especially to land with palma cana.

15 Popular and scientific names used in this paper are based on the "Diccionario botanico de nombres vulgares cubanos" by J. T. Roig y Mesa, Estacion Exrperimental Agronomica Bol. No. 54, Santiago de las Vegas, Habana, I928.

VEGETATION OF CUBA 3 8 I

On the Isle of Pines and in Pinar del Rio Province, scattered pine trees are, or at least were formerly, found among the palm trees of the sabanas. This type of sabana may be designated "palm-pine tree sabana." Other sabanas consist only of coarse grass and low bushes; for instance, in the northern part of the peninsula of Cape Cruz.

It is clear, then, that the term sabana in Cuba refers not to one plant association of an open character but to various such associations. Marie- Victorin and Leon distinguished nine types of sabanas on the Isle of Pines, showing all stages of transition from a plain association of grasses and sedges to grassland with shrubs and low trees to an association of tall palms, pine trees, and low shrubs on a cover of grass.

For our purpose, however, it is sufficient to define the Cuban sabanas as open country, composed of grasses and sedges with scattered low shrubs, palms, and, in some areas, pine trees.

These sabanas do not show the luxuriant vegetation one would expect in a tropical country. The Camagiiey sabana, the most famous of all Cuban sabanas, is dry and waterless during most of the year, and palms, hard- leafed shrubs, and meager grass eke out a precarious existence. In the rainy season small herbaceous plants carpet the ground; in the dry season the sabana has a desertlike appearance. "Small wonder, that one may go for miles, without encountering a living soul, or a single dwelling to disturb the magnificent wildness of this savannah."'6

If we try to apply the modern phytogeographical terminology to these Cuban sabanas, we realize at once that they are not the true "savanna" of tall grass and isolated broad-leafed trees with evergreen hardwood forests along the permanent streams. They resemble, rather, xeromorphic scrub or thorn-shrub steppe. This conception is confirmed by travelers who com- pare the Cuban sabanas with the Mediterranean maquisI7 or the South African veld.'8 The pine sabanas of the Isle of Pines are even designated malpais or "badlands." To avoid confusion, I shall use only the old Indian name sabana for the natural grassland vegetation of Cuba.

According to Bennett, the sabanas have in the past been scrupulously avoided for agriculture, and even now the same attitude is generally preva-

x6 H. J. MacGillavry: Geology of the Province of Camaguey, Cuba,Geogr. eni Geol. Mededeelingetn, Physiogr.-Geol. Reeks No. 14, Rijksuniversiteit te Utrecht, I937, pp. I5-i6.

I7 C. J. Rutten Pekelharing: Botanische en plantengeografische indrukken in Cuba, Tijdschr. Koni. NJederl. Aardriksk. Genootschap, Ser. 2, Vol. 50, I933, pp. 930-945; reference on p. 942.

i8 A. G. Robinson: Cuba, Old and New, New York, London, etc., I9I5, p. I09.

3 82 THE GEOGRAPHICAL REVIEW

lent. Thus it seems that human influence has had relatively little effect on the original character of the Cuban sabanas.

PARKLAND

A sabana that is confined by gallery forests along the rivers is called ceja

(Spanish, eyebrow).I9 A sabana that is surrounded by forests and is uncon- nected or only narrowly connected with other sabanas is called ciego (Span- ish, blind). Examples are shown in Figure i, a to g. A small isolated patch of wood and thicket in the midst of a sabana is sao, an Indian term. Pichardo calls the saos "the oases of Cuba." Palms such as yarey and palma cana together with a hardwood tree, guayacan (Guaiacum officinale), are the main vegetation of the saos.

The Spanish cayo, "islet," is also often applied to small isolated forested areas; for example, in the Cienaga de Zapata. In the hardwood forestland, however, which extends south of the Cienaga, we find cayo applied to small isolated clearings (Fig. ia).

Of the I69 place names of the map, 3 I are ciego names; 7, sao names; 6, cayo names; 4, ceja names. These place names indicate two different types of vegetation: small spots of natural grassland or glades scattered through the hardwood forests; and areas where the spots of natural open country are so numerous that they touch and form a landscape in which patches of hardwood alternate with patches of grassland. The latter type has been designated by Bennett "semisavana" or "mixed savana and hardwood." As the aspect of the landscape is parklike, I prefer the term "parkland."

It is not always easy to make a distinction between natural and man- made parkland. On the Military Map a hardwood forest with many man- made clearings looks exactly like natural parkland. For example, the northern parts of sheets 3 3 and 58 show the same alternation of hardwood forest and open country that is found in the true parkland areas of sheets 48 and 49, but the clearings are man-made. In such cases the place names may afford a criterion. Of the 3 I ciego names shown on the map, 28 occur in areas whose vegetation must be classified as parkland according to soil conditions and the aspect of the landscape. Figure ib is illustrative. Sheet 35, on the other hand, shows no place names indicative of the nature of the clearings; but this region, because of its vegetation and soil, has been classified by Bennett and Allison as a "mixed savana and hardwood area."

i9 Antonio Perpinia': El Camagiuey, Barcelona, I889, p. 36I.

VEGETATION OF CUBA 3 8 3

=-X 26 3 ;sQ r( t==Ca~~~~~~~yo de los7-- i602 , yG<

2

g g== N e gfr . ra n Sabbana

-_

- = / - -----E Saban e od S trio !% o/- S.5Ya4

- S II -ai ''S - /X

1 2 ; Sao de CoroJod4 ,444X 3 Xayo'N uevo X . ,9

' ' 1'r-a~~~

SX~~~~~~4

C,3

FIGy I,atogE

amplsofloclte)nhsb rrltdpae ae.SePaeIadky nFgr d, Sa de Crojo,1Z4, or shet 60read 9; Saanila, Saroaset 0 ed I

It is logca toasmIht sstlmn sras odad hiki natra prkandara ndgrsslnd gin Insoe res ow raslndth plo ha unoveedoldstups nd ured ot ftes2 ua nlec

20 1. A. Wright: Cuba, New York, 1910, p. 428.


is clearly revealed when one compares the parklands of Santa Clara2' and Camagiiey Provinces on the Military Map. At the time of the reconnaissance not only hardwood forests but also parklands were relatively little disturbed by man over large parts of Camagiiey. The parkland belt that extends across the province from coast to coast in an east-west direction shows large tracts of woodland intermingled with even larger tracts of open country, some of which are entirely isolated (see sheets 48 and 49). The ratio of woodland to grassland is about I to 2. Santa Clara Province then had a much larger population than Camagiuey, and density of population decreased from west to east. Consequently, in the eastern part of the province the former parkland character is still recognizable on the southern part of sheet 33 and the northern part of sheet 34. In the western part of the province, however, very little woodland had been preserved, and only ciego names and soil conditions reveal the former parkland character on sheets 28 and 3I.

SMALL, SCATTERED CLEARINGS IN HARDWOOD FOREST

Large sabanas were so distinctive a feature of the original landscape of

Cuba that much information on them is found in the literature. It is quite

otherwise with the many small sabanas scattered through the hardwood

forestlands. Many of them, it is true, are mentioned in Pezuela's "Dic-

cionario geografico," but descriptions are vague as regards size, shape, and

location. Here again the Military Map proves its value. Because of its pur-

pose, particular attention was paid to these small spots of open country.

Not only were they carefully surveyed, but various pertinent remarks, on

categories of camp sites, water conditions, grazing opportunities, and the

like, were noted on the map. Some of the clearings are nameless; others

show one, two, or even three names. Some of the place names are indicative of the size of the clearing. Sabane-

ton de Satirio (42 on sheet 26) is a circular clearing of about 0.5 square

kilometer (Fig. id); Sao de Corojo (1 24, sheet 49) has an area of about I.2

square kilometers; Sabanilla (153), about 3.5 square kilometers; Sabana

Jibaro (142), about i6 square kilometers. In other names, however, there is

no such connection between area and name. Here again distinction must be made between natural and man-made

clearings. The most obvious assumption, that all clearings which have

sabana, sabanilla, ciego, etc. names are natural openings and others are man-

21 The name of the province was changed from Santa Clara to Las Villas in I940.


made, does not take two circumstances into consideration. The names applied by the Spaniards do not refer to the original landscape but only to the landscape of the early sixteenth century. It is possible that a study of the Indian place names of many of the clearings, especially on sheets 54, 57, and 58, would give some information on an older stage, perhaps even on the original landscape; this calls for linguistic study, however.

Furthermore, the place name Sabana was applied by the Spaniards from the very beginning not only to clearings but also to the cattle ranges established on them.22 This is the reason why a family name is so often coupled with a sabana name; for example, Sabana de Mateo Sanchez, Sabana Diego. These sabana names had to be eliminated for my purpose. Only seven appear on the map, and these for special reasons. Even the remaining 44 sabana names that I have used are, to some degree, suspect.

The situation is quite different with regard to such place names as Saba- nilla, Sabanita, Sao, and Ciego. These names are very rarely associated with a personal name, and, as a rule, they indicate a natural clearing, or at least an Indian clearing, in existence in pre-Columbian times.

Another feature that distinguishes a natural or old Indian clearing from a Spanish clearing is the location of the settlement. Many settlements with sabanilla or ciego names are located on the edge of a clearing; for example, Sabanilla (1 17) and Ciego-Yamagual (1 18), shown on Figure ib, and Sabanilla (153) and Sabana Jibaro (142), shown on Figure id. Where a place name does not have this significance, the settlement is generally not on the edge but in the midst of the clearing. Figure ie gives several examples. Fig- ure if also shows settlements with Spanish names located in the midst of clearings, and in this instance the names betray the origin of the clearings: Sabana Nueva, "new sabana," and El Quemado, a place where the woods have been burned.

A third criterion of the nature of a small clearing is the soil. Soils derived from serpentine or sandstone (Habana soil of Bennett), in particular, support a grassy vegetation. In easternmost Cuba, south of Baracoa, the place name Sabanilla (1 67) is found in a hilly region of serpentine rocks. Sabanilla (153) west of the Nipe Mountains is a clearing on a large body of intrusive serpentine; east of these mountains another clearing carries the names Loma Sabana Grande (157) and Sabanilla (158).

In the midst of the former hardwood forest of the Bayamo region are Sabanilla (143) and Sabana Grande de Caureje (144). According to Perpifia,

22 Pichardo, Geografia de la isla de Cuba, Vol. 3, p. I 5.


the second sabana is io kilometers by 4 kilometers. Bennett and Allison's soil map gives no indication of locally unfavorable soil, but in the text they speak (p. 214) of an extraordinarily tough and impervious soil with a sabana vegetation that occurs in this region. Hence Sabana Grande de Caureje is designated as naturally open country.

A circular outcrop of serpentine around the city of Ciego de Avila (84) is indicative of the natural character of this ciego. Sabanilla de Palma (40) is located on a small outcrop of serpentine near Marti, southeast of Bahia de Cardenas.

On the southern rim of the Sierra de los Organos in Pinar del Rio Province, Sabanilla occurs (8) in an area where the geological map shows a small outcrop of serpentine and the Military Map hardwood forest.

Soil conditions indicate that formerly a number of rather large natural openings existed in the hardwood forest of the northern part of Matanzas and Habana Provinces. On the east end of a large intrusion of serpentine in Matanzas Province are the place names Ciego (38), Corral Sabanilla Vivos y Muertos (37), and Sabanazo (36); on the west end is Pueblo Sabanilla (32).

Apparently the entire serpentine area included between these places must be considered naturally open country.

A long and narrow serpentine intrusion from Minas and Guanabo westward to the harbor of Habana carries a typical sabana palm, jata, but no sabana name is found on the map. There is, however, the name Habana or Havana, city and state. What does it mean?

THE NAME HABANA

Habana is the name of an Indian province more or less conterminous with the present province of Habana.23 According to Sven Loven,24 the name was applied to a natural rather than a political division. The Spaniards established their first settlement in the province on the south coast, either near Bataban6 or at the mouth of the Giiines River (Rio Mayabeque).25 Here, onJuly 25, I5I5, "la villa de San Crist6bal" was founded by Diego Velazquez. Because of the plague of mosquitoes on this site and the discovery of Mexico, the settlement was transferred in 15I9 to its present location on the north coast as "San Crist6bal de la Habana." The original

23 R. V. Rousset: Datos hist6ricos y geotopograficos de la isla de Cuba, Habana, I9I4 [on cover:

I9I5]-

24 Origins of the Tainan Culture, West Indies, Goteborg, 193 5, p. 83, footnote 3.

25J. M. de la Torre: Lo que fuimos y lo que somos, o La Habana antigua y moderna, Habana,

I857, p. 5.


name of this place, given by Sebastian de Ocampo in I508, was Carenas, ''careening place," indicating a good harbor and wood for ship repairs. Later the "San Crist6bal" was dropped and the name became simply La Habana.

However, we speak not only of La Habana but of Havana, Havanna, and even Havannah. In other words, the same confusion exists in regard to Habana that is found in regard to "sabana." This suggests that the names Sabana and Habana were originally identical. Elisee Reclus, indeed, says that this is so.26 Sven Loven has expressed his opinion that the name Habana is nothing more than the Haitian sabana, since h and s are interchangeable in the Tainan language. The probability is that Habana was a province in which sabanas were found. The northern part of the present province of Matanzas is identical with the ancient Indian caciqueship "Sabaneque," and the archipelago of small islands that stretches from Cardenas to Guajabo is even today called "the archipelago of Sabana and Camaguiey."

The fact that two provinces were named Sabana and Sabaneque by the Indians does not necessarily mean that natural sabanas were prevalent or even widespread. On the contrary! It is ancient wisdom that not the rule but the exception impresses the human mind. Place names in the old provinces of Cuba in which the largest sabana areas occurred, Cubanacan (Santa Clara) and Camagiiey, show no connection with natural sabana vegetation. In Sabana and Sabaneque Provinces, on the other hand, forest was the main vegetation cover; and where sabanas were infrequent enough to attract the attention, they appear in the nomenclature of the region.

It is significant that the capital of Habana Province received the name Sabana or Habana only after it had been transferred to the north coast. The conclusion that sabanas were found mainly in the northern part of the provinces of Habana and Matanzas is strongly confirmed by soil conditions and place names.

THE PINE FORESTS

Unfortunately, the Military Map does not distinguish between hardwood and pine forests. However, place names such as Pinal, Pinar, Pino, and Los Pinos testify to the occurrence and former distribution of pines throughout the island. Of the I69 place names on the map, 2I refer to pine trees.

Pine forests are now found in two widely separated parts of Cuba: on the Isle of Pines and in the piedmont belt of the Sierra de los Organos in

26 Nouvelle ge6ographie universelle, Vol. I7, Paris, I89I, p. 682.

3 88 THE GEOGRAPHICAL REVIEW

Pinar del Rio Province and on high mountains of eastern Cuba, especially the Nipe Mountains (sheets 6i and 62). The pines27 are scattered over a carpet of grass. As a rule the trees are small and hard, stunted by a combination of adversities, including poor soil and frequent fires. Low shrubs of many of the species characteristic of the sabanas-palmetto, for example- grow between the pines. Oaks (Quercus virginiana?) grow on the piedmont belt of the Sierra de los Organos.

Scattered pines were formerly found also on the sabanas of the coastal plain of the Isle of Pines and Pinar del Rio Province. The place names bear witness: Sabana Pinal Alto (6), Pinar del Rio (7), Santa Cruz de los Pinos (1o), Sabana de los Pinos (11), and Pinal de la Catalina (13). Pichardo2" refers to "ten sabanas with pine trees and one without pines" in the district of San Juan y Martinez. Since that time (I854), most of the pines along the roads and railroad have disappeared, and only in the remoter parts do any considerable stands remain. Brother Leon was certainly right when he wrote: "The chief modification of the savanna region [of Pinar del Rio Province] is the extinction of the pines, which, possibly, were growing formerly up to the limit of Havana province. '29

Nothing is said in Cuban literature, so far as I know, of the occurrence of pines in Habana Province. But, again, place names tell a different story. The Military Map shows only one place name indicating the former occurrence of pine in Habana Province-Pedro Pino (21). Many more names, however, are found on a map of Habana Province on the scale of I: 109,300

published by the Adjutant General's Office in I898: El Pino (1 7), northeast of Guanabacoa; Los Pinos (18), south of Guanabacoa; Pinar (19), near

Arroyo Naranjo; Pedro Pino (21), near Sabanilla (20); Pinos (22), north of Managua; Pinales (23), between Managua and Santiago de las Vegas (this name was changed on later maps to Piniales, "pineapple groves": apparently the compiler of the map could not believe that pines were found in this area); Los Pinos (24), north of La Salud; Pinar de Cazaz6n (25), east of La Salud; Los Pinos (26), south of San Antonio de las Vegas; Los Pinos (28),

south-east of Melena del Sur; Arroyo de los Pinos (31), within the Sa- bana de Robles (30).

Locations 17 and 18 are near the long, narrow strip of serpentine that runs

27 Pinus ciibensis and P. occidenitalis are found in eastern Cuba, P. tropicalis and P. caribaea in western

Cuba. 28 Geografia de la isla de Cuba, Vol. 3, p. 95. 29 Cuba, p. 685.

VEGETATION OF CUBA 3 89

from Minas westward to the harbor of Habana. Similarly, names 19, 21,

22, 23, and 31 pertain to localities on such relatively infertile soils as the Habana series and the Mc type of the Matanzas series. However, locations 24,

25, 26, and 28 are found on Matanzas red clay, the most fertile soil of Cuba. These places may be near arroyos with sandy soils or on locally stony areas. Field investigation is needed here.

The maps do not show any place names connected with pines in Matan- zas. Such names may, however, occur in the northern part of the province; and here again field research might be fruitful.

In Santa Clara Province prehistoric existence of pines is proved by the finding of pine cones, together with fossils of Megalocnus rodens.30 That pines grew there up to the time when the Spaniards arrived is evidenced by the two place names Pino (53), in the southeastern part of Santo Domingo Sabana, and Arroyo de los Pinos (55), southeast of Cruces.

In Camagiiey Province the place-name evidence again is negative, but in my opinion this only means that pines had already disappeared when the Spaniards arrived. As Camagiiey was much more densely settled by Indians than Santa Clara or Matanzas and Habana Provinces, it is possible that the Indians early destroyed the scattered pines of the serpentine sabanas, either by fire or by cutting down the trees for their soft timber.

It should not be overlooked that pine sabanas are not a peculiar feature of Cuba but occur also in other parts of the Caribbean. In British Honduras pine sabanas are called "pine ridges" (the term "ridge" implies a vegetation belt and has no connection with topography). Pines (Pinus caribaea) and live oaks are the characteristic trees of the pine ridges. The ground between the scattered trees is carpeted with grass; and where grass is dominant, the land is called "dry savana."3' The coarse grass is subject to annual fires. Pine ridges occur in two areas with sandy soils: on the elevated coastal plain where a ferruginous hardpan forms the subsoil, and in the Maya Mountains where the granites and siliceous slates are capped with sand.

In Spanish Honduras similar pine ridges occur southeast of the Plantain River. "These savanas .. . are undulating plains of yellowish-red clay with pebbles of pure-white quartz (in places whole beds of them); in places a peaty soil supports tough grass, squat oaks, . . . and, especially, the pines (Pinus tenuifolia). . . . Next to the pine the most conspicuous tree on the

30 Leon, Cuba, p. 690.

3I D. Stevenson: Types of Forest Growth in British Honduras, Tropical Woods, No. I4, Yale University School of Forestry, I928, p. 21.


savanas is the squat palmetto (Sabal sp.),... which is confined to the humid depressions."32

A similar association of pine trees (Pinus caribaea), palmetto (Serenoa serrulata), and coarse grass is found in the "pine barrens" of the peninsula of Florida.

There seems to be no doubt that pines formerly grew throughout Cuba from east to west on the poorer soils, especially on serpentine sabanas.

HARDWOOD FORESTS

Although the floristic character of the hardwood forests of Cuba is well known, relatively little has been written on what I might call their "spatial and physical character"; namely height, density, mass, and seasonal aspect of the vegetation. The Military Map, it is true, makes the distinction of thick woods, thin woods, and brush. Thin woods and brush, however, seldom occur on the map, and, furthermore, the symbols used are not readily distinguishable.

Of some value as clues are the term monte bajo, "low forest," and certain corresponding (Indian?) terms. In western Cuba a low forest is called manigua if it occurs on limestone and cuabal if on serpentine. The corresponding term for cuabal in eastern Cuba is charrascal. Unfortunately these terms are not found as place names on the Military Map.

Many names, however, refer to characteristic hardwood trees, such as Ceiba or Seiba (Ceiba pentandra), Los Cedros (Cedrela odorata), Guayacan (Guaiacum officinale), and Guasimal (Guazuma tomentosa). But these names, interesting as they are, cannot be taken as a clear index to the physical character of the forests. Geographically the most important phenomenon of a hardwood forest is its foliage, whether deciduous or evergreen. With it is correlated not only the density of the canopy but also the growth and kind of underwood and ground cover. According to Leon,33 the Swedish botanist Erik Ekman, who studied Cuban plants for many years in the field, was of the opinion that "true tropical rain forest does not exist in Cuba, except, may be, in the Baracoa region [eastern Cuba], where one tree of the Brazil rain forest is growing, Carapa guianensis." Does this mean that Cuban hardwood forests are as a rule not evergreen but deciduous?

True deciduous forests, rather open and with a grass carpet, have been

32 V. W. von Hagen: The Mosquito Coast of Honduras and Its Inhabitants, Geogr. Rev., Vol. 30,

I940, pp. 238-259; reference on P. 246.

33 Cuba, P. 684.

VEGETATION OF CUBA 39I

described for very few regions of Cuba: leeward slopes of the high mountains of the east, and the Sierra de las Cubitas in Camagiiey Province.

It is my opinion that the extensive hardwood forests of the Cuban lowlands should be classified as "semideciduous," meaning that some, but not all, of the trees lose their leaves in the dry season. These forests are less dense and less imposing than the evergreen hardwood forests but denser and taller than the deciduous forests. The most important difference from the latter, however, is the absence of a grass ground cover.

Finally, there are the so-called "dry forests," which are also deciduous; but the trees are low, gnarled, and low-branching, many of them have thorns or thornlike branches, and the grass cover is more or less continuous.

At the present stage of our knowledge it is impossible to give a clear picture of the distribution of these various types of hardwood forest. Nor do we know enough to map the areas of second-growth forests in Cuba.

CACTI-THORNSHRUB FORMATION

A vegetation of thorny shrubs and tall cacti occurs along the south coast and on the lower foothills of the highest mountains of Cuba, the Sagua- Baracoa mountains, the Sierra Maestra, and the Sierra de Trinidad. In places cacti, 20 to 40 feet high, are dense enough to form veritable forests. In areal extent, however, this desertlike formation is very small.34

COAST VEGETATION

A rather xerophilous vegetation of shrubs, palmetto, and columnar cactus is also found on the rocky coasts and the higher parts of the many small fringing islands or cays. The cays also show occasional small areas of woodland intermingled with palmetto sabanas.35 This almost undisturbed strand and littoral vegetation is much the same throughout the West Indies. However, the coastal swamps extending along both coasts of the main island are an important feature of the Cuban landscape.

ECOLOGICAL CONDITIONS

Lack of space forbids the discussion of ecological conditions in any detail, but it may be said that the distribution of the vegetation formations of Cuba bears surprisingly little relation to climate, especially rainfall, except in the mountains. It is true that the network of meteorological stations

34 An excellent description of this type of vegetation is given by Marie-Victorin and L'on. 35J. C. T. Uphof: Vegetationsbilder von Cuba, Vegetationsbilder, Ser. I 8, No. 5, Jena, 1928.


is insufficient to permit any final conclusions from the records, but Table I is illustrative. In contrast with what one would expect, the highest rainfall is found in the sabanas, the lowest in the parklands and hardwood forests.

TABLE I-RAINFALL AND NATURAL VEGETATION

STATION VEGETATION ANNUAL RAINFALL (Ins.) DRY MONTHS t

A* Bt Pinar del Rio city Pine sabana 6i.78 59.75 6 Habana city Hardwood forest 43.00 48.28 5 Ciego de Avila Parkland 47-99 6 Camagiiey city Sabana 56.07 56.07 4

* 0. L. Fassig: Rainfall and Temperature of Cuba, Tropical Plant Research Fotndation Bull. NVo. 1,

1925, p. 17.

t T. W. Chamberlin: Rainfall Maps of Cuba, Monthly Weather Rev., Vol. 68, I940, pp. 4-I0,

reference on p. 6. t Number of months with less than 3 inches of rain.

Soil is the determining factor in the formation and distribution of the vegetation of the lowlands of Cuba. This has been proved unmistakably by Bennett and Allison. Furthermore, a comparison of their soil map with the geological map now available36 proves that the soils are primarily influenced by the rock formations and structure.

In sum, we may make the following statements: The fertile and friable red clays on Tertiary limestones support (semi-

deciduous?) hardwood forests. The rather shallow, locally gravelly and stony Habana clays on Eocene-

Oligocene and Upper Cretaceous sediments support a rather open hardwood forest intermingled with areas of natural open country.

The plastic clays on Quaternary sediments, if they are not too shallow and stiff, support semideciduous or deciduous hardwood forests; if they are very shallow and stiff, they support sabanas and parklands.

The Eocene and Oligocene sandstones of Pinar del Rio Province and the Isle of Pines, with their sandy and gravelly soils, support pine sabanas in the coastal plains and pine forests in the piedmont belt that surrounds the higher mountains.

Serpentine soils support sabanas on flat lowlands, parklands on rather hilly and rolling country, and pine forests on the red iron-ore lands of the

36J. W. Lewis: Geology of Cuba, Bull. Amer. Assn. of Petroleum Geologists, Vol. i6, I932, pp. 533-

555. The map, I : I,ooo,ooo, is dated 193 I. A simplified version on the scale of I : 2,200,000

appears in Charles Schuchert: Historical Geology of the Antillean-Caribbean Region, New York

and London, I935, facing p. 494.


high mountains. The widespread occurrence of intrusive serpentines is of the greatest importance: it distinguishes Cuba from the other Greater An- tilles.

As a rule, the sabana lands have at relatively shallow depths subsoils composed of material that affects the soil moisture unfavorably as regards the growth of many plants, including most of the cultivated ones. Some of these sabana soils are shallow because the parent rocks have not decayed deeply (serpentine sabanas). In other places the prevailingly flattish, and often almost level, surface makes underdrainage inadequate and is the main cause of the development of chalky and ferruginous concretions (coastal sabanas on Quaternary).

COMPILATION OF THE VEGETATION MAP

The map, Plate I, is a combination of a vegetation map and a map showing certain stages in the development of the cultural landscape of Cuba.

The most important feature that I took over from the Military Map is the distribution of hardwood forests. The map also indicates the former distribution of forests on the Habana-Matanzas red plains. We have clear proof that these fertile plains formerly were completely forested. The area of the Matanzas red clay on Bennett and Allison's soil map is thus indicated.

In eastern Cuba the areas left blank on the map evidently represent man-made clearings in former hardwood forests. In central and western Cuba they represent areas the vegetation of which could not be classified exactly according to the Military Map or place names. Soil conditions, however, indicate that here, too, the blank areas were originally covered with hardwood forest.

For the mapping of the pine forests I had to rely on travelers' descriptions, place names, and soil conditions.

Alternation of hardwood lands and sabana lands on the Military Map associated with place names such as Ciego and Cieguito and soil conditions enabled me to delimit the parkland areas. Distinction is made between areas that still had a parkland character in I906-I908 and areas where the original parkland character had been destroyed by human influence.

The sabana areas are taken from the Military Map. Furthermore, with the exception of the sabana of the southern Camagiiey coastal plain and the Guantanamo sabana, they check with soil conditions. The former occurrence of pine trees in the sabana vegetation of Pinar del Rio Province and the Isle of Pines also is indicated.

Cacti-thornshrub formations were mapped according to travelers' de-

KEY TO PLACES ON THE VEGETATION MAP OF CUBA

The sheet number precedes the place number

3 i La Sabana Blanca 58 Sabanilla 48 113 ElCiegodeSantoDomingo 4 2 Los Ciegos 30 59 Sabana de Guanabanabo 114 Sabana la Mar

3 Pinar Ciego 3I 6o Sabana de Matezas 115 Sabanilla 6 4 Ceja del Torro 61 Quemado Grande 116 Sabanilla

5 Ceja Ana de Luna 62 La Sabana 48 117 Sabanilla 7 6 Pinal Alto 63 Cayo Obregon 118 Ciego

7 Pinar del Rio 64 Ceja de los Algodones 119 Ciego 9 8 Sabanilla 65 La Ceja 120 Ciego and Cieguita

9 Sabana del Ciego 66 Quemado Grande 121 Ciego Guaimaro 1o Santa Cruz de los Pinos 67 Colonia Quemados 49 122 Sabana Burro 1 Sabana de los Pinos 32 68 Sabanilla 123 Ciego

IO 12 Sabana Grande 33 69 Sabana de Neiva 124 Sao de Corojo 13 Pinal de Catalina 70 La Sabanita de Guayos 125 Las Ciegas

13 14 Sabana de Guanacaje 71 La Lita Sabana Grande 126 Sabana de Buena Ventura 15 Sabanalamar 34 72 Sabana de las Minas 127 Sabana la Mar 16 Sabana del Rosario (Sancti Spiritus) 128 Sabana Jobabe

15 17 El Pino 73 Sabanilla 50 129 Sabana La Mar i8 Los Pinos 74 Sabanilla de Lara 130 Sabana Guayabal 19 Pinar 75 Quemadito de Ruiz 131 Sabana Yuraguana 20 Sabanilla 76 Ciego Caballo 132 Sabana Grande 21 Pedro Pino 77 Ciego abajo de Caballo 52 133 Sabana La Mar

i6 22 Pinos 78 Quemado Grande 134 La Cieguita 23 Pinales 79 Ciego Majagua 135 Sabanilla 24 Los Pinos 8o Sabana de Silencio y 53 136 Sabanita 25 Pinar de Cazazon Corral del Mango 137 Cayo Largo 26 Los Pinos 35 S81 Sabana Grande 138 Cieguito 27 La Sabana 82 Sabana del Medio 139 Sabanoza 28 Los Pinos 36 83 Quemado Arriba 140 Sabana Tranquera 29 Sabana de Guarayuasi 84 Ciego de Avila 141 Sabana Cauto

I9 30 Sabana de Robles 85 Sabana la Mar 142 Sabana Jibaro 31 Arroyo de los Pinos 37 86 Ciego Moreno 54 143 Sabanilla 32 Sabanilla (de Otero) 38 87 Sao del Palma 144 Sabana Grande de Caureje 33 Sabana del Ajicon 88 Cayo del Toro 145 Sabana Venezuela 34 Sabana de Bagaez 39 89 Sabanita 146 Ciegos 35 Sabana Grande go Pueblo Nuevo 6 Quemado 147 Sabana de Nagua

22 36 Sabanazo 91 Las Ciegas 57 148 Sao Limpio 37 Corral Sabanilla de 92 La Sabana Nueva 149 Sao Arriba

Vivos y Muertos 93 Los Cayos 150 Sabanilla 38 Ciego 94 Ciego de Magarabomba 151 Sabanilla

24 39 Sabanas Anegadizas 95 Ciego Corojo 58 152 Sabana Salada 25 40 Sabanilla de Palma 96 Ciego Grande 6i 153 Sabanilla

41 Sabana Grande 97 Ciego de Escobar 154 Los Pinales 26 42 Sabaneton de Satirio 98 Sabanilla 155 Arroyo del Pino

43 Sabana Bonita 40 99 El Ciego 156 Pinal Colorado 44 Cayo Palma loo Sabanilla 62 157 Loma Sabana Grande 45 Cayo Ingles 1ol Ciego de Guanausi 158 Sabanilla

27 46 Reyes Sabanilla 102 Sabanita 159 Sabanilla 47 Sabana de Cafias 4I 103 Cieguito 64 16o El Pinar 48 Hato Quemado de Giiines 104 Embargo de Sabanilla 65 161 Rlo Sabana 49 Ciego Viamones 42 105 Sabana la Gloria 162 Sabana Abajo 50 Sabana de Palma 10 6 Las Sabanas de Minas 163 Sabana del Vinculo 51 Ciego de San Marcos 43 107 Sabanilla 68 164 Sabana del Mar 52 Quemado Hilaro 1 o8 El Sao 1 65 Sao Pepe

28 53 Pino 1og Sabanilla, south of 166 Sabana de Duaba 54 Ciego Montero Hato Sibanicu 70 167 Sabanilla 55 Arroyo de los Pinos 0lo Ciego de Najasa 168 Pueblo Sabana 56 Ciego Alonso 44 1ii Sao las Bocas 169 Sabana Vieja (Pueblo Viejo) 57 Sabana San Miguel 47 112 Sabanas de Mayanabo

TkeAmerica Geograp2hica.L Society ofr7ewYork

8,5? 840 83? 12 15

1/7

-23 /'

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~21? EEAINM LEO WAI BEl

X x~~~7

X, 4~ 10 17

X' X X

cg~ ~~~~ -akwns -90-

.__ 8 40~~~~~~~~~ 11-

XoyiSt 90,b tenerka Gorahci oitg fXeYX

-2-1

VEGETATION MIY LEO WAIBEI

SCALE I 2,000, 0 6 S KILOMJ

I Hardwood forests, /906-07

LI!?F7 Pine forests

[1.;Park/and's, /906-07

EE:Sabanas

20 Cacti- thornshru6 f0ormation

1-70, Sheet /num6ers of Mili'tary Alapo, /906-08

-4-- M---4--- RailwayVs,

Cayos (small islands) are

814 Copvyri9ht, 1,94-3, b,y the America,m Geo5graph2icaL Societ,y of New York

12 15 18 21 24 80?

~/23' ~ / 16 "~73 22 730

"7 ~~~~1 V

tO ,~ , ~ I~ 7/ /0 MIE

,23sEs, //36-0 /7a Swam/

\7\\

__ * ~~~~5/\\\\ k*

10-

7

___

ScEe 23in frees

14 o \\\\6 \\2N5

LTION M4AP OF CUBA LEO WAIBEL

SCALE I 2,OOO,OOO 0 SO MILES

0 so KILoMETERS

wests, /906-07 Swamp,s

Formner hardwood f"orests

906-07 L3Z Scattered pine 6Srees

Former park/ands

,hru6 formati'on L Cultivated lands or on/qinal ve9etation unc/assi/'iecf

s of MoWilit6ary Map, /906-068 -/69 Place names (see Table)

- *

a --- Railways, /906-07

Cayos (small islands) are not shown

80'

790 78

Ha

30 .n rd

dePin S~-

\\\ w<<X 33 JAMa p

I_ . __ . ~ |

........ ~43.~/ ~ ~ 48 52

. . . . . . . .. . .~~~~~*-*.**~~~k~1A1~

a2-~ ~ ~ ~~~~~~~~~~~2

/30 ~ ~ ~ 5

- r44/ )X~~~~~~~~~?~~X~~~ r

79'

The GeoqraphicaZ Review, Vo Xl. fo.3,]39#*3,P1.I

Habana SCALE: 1: 8,000,000

I Habana 4J0 o 00 MILES

HBANA\* Mata,zaO 100 KILOMETERS

____ ~~~~?,o Batabanci

2Z- _.Pi deb R(o MATANZA SS Clara

GolC,O de LA ataband A

I.:dePinos---: i5am~~~~~amauey .< t\

M |a r C: a r / 6 e Guafcanayeo anzanillo \ .Guantnam;)

de Pi8i' 810 78 716 tcIago de Csba |

r- -- z 1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~220

47

_25

ZA~~~~Ve 553 ~~~~~~ * 57~~~~~~~7Md C 3~~ ~ ~~w>f~ ?~~?w :;49.E~~~7o7 ga 4

47~~~~~~~~~~~~~s~t

SW 62 /16 b ~ 87

-J~~~~~~~~~I c73(I

- /

'4 139~~~~~~~~~~~~~~~~~~~I

Xq ,4~~~~~~~~~~~~~


scriptions only. The distribution of swamps follows the Military Map. The coast and strand vegetation is not shown, since it has no relation to the problem of the orginal landscape.

This map, the first detailed vegetation map of Cuba so far as I know, is offered as a first essay only. A great deal of work in the field and the archives of Cuba lies before us, and it is a job that calls for cooperation by many workers in botany, geography, geology, history, Indian linguistics, arche- ology, and the like. All I have tried to do here is to call attention to the problems and thus make a contribution to their eventual solution.

SUMMARY AND CONCLUSION: THE ORIGINAL LANDSCAPE OF CUBA

The results of my studies on the original landscape of Cuba are sum- marized in the following areal figures:

AREA AREA

Sq. Km. % Sq. Km. %

Hardwood forests 68,500 60 Forest 73,000 64 Pine forests 4,500 4 Parklands I7,000 I5 Sabanas I3,000 II Open country 3I,000 27 Cacti-thornshrub formation I,000 I Swamps 7,000 6 l Cays 3,500 3 Swamps & cays 30,500 9

Total area, including the II4,500 Isle of Pines

Ram6n de la Sagra's question can now be answered more definitely. The original landscape of Cuba consisted of a variety of vegetation formations, including types of hardwood forest, open pine forests, parklands where hardwoods alternated with grasslands, sabanas with scattered palms and pines, desertlike cacti-thornshrub formations, and swamps. So unusual indeed was the variety in so relatively small an island that in this respect Cuba may be described as a "dwarf continent." The division of the island into three entirely different natural regions is truly continental.

Only in recently uplifted eastern Cuba is found that contrast between wet windward slopes with dense evergreen hardwood forests and dry leeward slopes with rather open deciduous forests and cacti-thornshrub formations along the coast which is characteristic of the other Greater Antilles.

Western Cuba, including the Isle of Pines, is composed of folded sediments heavily degraded and weathered to poor sandy soils. The original


vegetation consisted of pine forests and pine sabanas similar to those of the

southern United States. In central Cuba the original vegetation consisted of the alternation of

hardwood forests and grasslands with transitional parklands that is charac-

teristic of Central America. But whereas in Central America climate is the

main factor in determining the distribution of the original vegetation, in

Cuba soil plays the dominant role.

Forests originally occupied about two-thirds of the surface of the island;

if parklands are added, the proportion rises to nearly four-fifths. Now we

understand the saying of Las Casas that in I5II Cuba was so well wooded

that one could walk under trees for 300 leagues. He did not say that it was

one immense forest or that the forests were all thick and heavy; he simply

spoke of the abundance of trees, doubtless trees of the parklands and sabanas

as well as trees of the hardwood forests.

Certainly the island originally was richer in woods than it was in i5ii

and much richer than it was at the beginning of this century. But the prin-

cipal pattern of vegetation, the fundamental contrast between forestland

and natural open country, remains unchanged by man, since it is condi-

tioned by soil and underlying rocks. This most characteristic feature of the

island is highly provocative and raises a host of questions with regard to

animal life, course of settlement, and economic development.

place names as an aid in the reconstruction of the original

Documents