The Journal of Indian Botany/Volume 1/May 1920/A Contribution to the Ecology of the Upper Gangetic Plain

Layout 4

A CONTRIBUTION TO THE ECOLOGY OF THE

UPPER GANGETIC PLAIN

By Winfield Dudgeon,

Ewing Christian College, Allahabad.

---

}

Introduction.

There is little published information about the ecological relations of the vegetation of the Upper Gangetic Plain. In their monumental survey and summary of the vegetation of India, HOOKER AND THOMSON (4) have little to say of this area. They state that there are no forests except at the base of the Himalayas, and that uncultivated tracts are usually covered with loose "bush-jungle". As every writer must do, they call attention to the markedly periodic climate. SCHIMPER (10) concludes that a rainfall below 90 cm. per annum produce "xerophilous scrub", while above that amount produces "xerophilous woodland". Allahabad has a mean annual rainfall of very nearly 90 cm., but the connection of the vegetation with some type of a Schimper climatic climax is not so obvious.

The Upper Gangetic Plain, comprising roughly the United Provinces, is a very distinct ecological area. Eastward it passes into the much more humid climate of Bengal; southward it merges into the hilly regions of the Vindhias, where the climate is similar but the vegetation is distinct; westward it passes into the drier Panjab; and northward it is abruptly limited by the Himalayas. The rainfall diminishes westward, and increases northward and rapidly eastward. The climate is continental, with great range between winter and summer and between day and night.

This study is restricted to a small area about ten miles in radius with Allahabad at the center. Such a restriction is chosen because I am more familiar with this limited area, and because it is quite representative of the much larger area of the Upper Gangetic Plain.

Physical features of the area.

Allahabad is situated at 25° 26' N. Latitude and 81° 52' E. Longitude, at the southern edge of the great plain of the Ganges and Jamna Rivers, at their junction. The Meteorological Observatory is 319 feet above mean sea level, and this may be taken as approximately the level of the surrounding plains. The soil (9) is all alluvial, deposited within recent geological times. It ranges from sand, through a mixture of sand and clay, to fine clay. The older alluvium often contains deposits of calcium carbonate in irregular nodules, locally called "kankar". Both rivers have been depressed in the recent past, so that even the highest floods no longer inundate the surrounding plains. In the immediate flood-plains of the rivers there are limited patches of quite modern alluvial deposits, which are subject to redistribution when the streams are in flood. Near the rivers, and especially the Jamna, there are often deep ravines with precipitous sides, formed as the result of rapid erosion during the short monsoon season. During the rest of the year they, are stable.

Except where dissected by these ravines, the surrounding plains are monotonously level. Here and there are slight natural depressions which become shallow lakes during the rainy season, but which are dried up later by evaporation, and by use of the water for irrigation purposes. There are very numerous artificial pools, made for collecting a supply of water to be used for irrigation during the winter and spring. Some of them are deep enough and extensive enough that they retain some water throughout the year. Lastly, there are a very few deeper and more permanent bodies of water created by damming up ravines. Some of the larger natural depressions catch the drainage sufficiently large area to accumulate considerable amounts of salts, chiefly sodium sulphate, mixed with more or less sodium chloride and sodium carbonate, and hence become "alkali", known locally as "usar". Aside from variations in the amount of water present, and from the deposits of usar, the soil of the plains presents a very uniform substratum for the growth of vegetation.

The factors influencing and determining the vegetation.

Climatic factors usually are considered to be the most important factors in the ecology of an area, but in the Upper Gangetic Plain biotic factors are at least of equal importance. The climatic differ only in degree from those met with in other parts of the world. The climate is characterized by striking periodicity. No interpretation of the vegetation can be complete without due consideration of the biotic factors, of which the most important are the very dense agricultural population, and other factors associated with a crowded population.

1. The climatic factors.

The climatic factors that appear to be of greatest importance are rainfall, insolation, temperature, humidity, and air movements. These factors are so distributed as to result in a strongly periodic climate. The data I have made use of are mostly taken from the records of hourly observations at the Allahabad Meteorological Observatory (5), extending over a considerable number of years.

Rainfall. The mean annual rainfall for Allahabad, calculated in 1913 (7), is 37*51 inches; other calculations put it as high as 10 inches. The distribution of this fall is indicated in Table I and Fig.

Mean monthly rainfall in inches, and percentage for each month, for Allahabad.

From this table it will be seen that 91 per cent, of the rainfall occurs during the months of June to October, while only about 1 per cent, falls during March, April, and May. The rainfall is remarkably uniform from year to year (8). In 11 years a deviation of 33 per cent, or more above normal has been observed but seven times, and a corresponding diviat'on be'ow normal but six times. The greatest recorded rainfall was 76"25 inches in 1891, and the lowest 16'82 inches in 1861.

During the monsoon the rains are at times torrential, at other times gentle and lasting for one or more days- The soil is baked hard by the intense heat and drought of spring, and wherever drainage has been established the more violent rains quickly run off. Under such conditions erosion is heavy, and the soil is wet only to a slight depth. After unusually heavy rains the level areas become vast shallow seas ; one may travel for miles along the railways and see no land except the railway embankment and the mounds on which villages are situated. Water is retained in the shallow depres- sions, from which it disappears at length by evaporation and by use for irrigation. In consequence, where there is good drainage there is little accumulation of water in the soil, but where drainage is poor, the soil, or at least the surface soil exploited by plants, is saturated.

Insolation. Little has been done with this climatic factor. The Meteorological Observatory has kept hourly observations of cloudi- ness, using the usual scale of — 10, where indicates a cloudless sky, and 10 indicates the sky completely overcast by dense clouds or low fogs. Such a record is necessarily a matter of judgment, and is of no great value in the study of climatic factors. We have nothing to indicate what proportion of the solar energy is in the form of heat, what in the form of visible light, and what ultra-violet. Table II shows the observed records of mean cloudiness, based on hourly observations extending over a number of years ; these same records are plotted in graphical form in Fig. 3.

Mean monthly cloudiness on the scale of 0—10, for Allahabad.

Clouds become common in June, perhaps two weeks before the rains actually begin. June to September are the cloudy months. In June there are densely cloudy days interspersed with days of burning sunshine. During July and August the clouds are denser and more persistent, but even then there are breaks when the sun shines from a nearly cloudless sky with the intensity of the dry season. From September on the number of sunny days gradually increases, till in November the sky is cloudless or overspread with only thin hazy clouds for weeks at a time. This is broken at some period during December, January or early February by the " winter rains", when for a short time monsoon rains are repeated on a miniature scale. During most of the time from the end of the rains to the beginning of the next monsoon the sun shines down with unbroken violence, accentuated during March, April, and May by low humidity.

Temperature. The climate of the Upper Gangetic Plain is distinctly continental. The nearest large body of water is the Bay of Bengal, more than 400 miles to the eastward. The temperature accord- ingly exhibits a large range between winter and summer, and between day and night, despite the fact that Allahabad is barely outside the tropics. The lowest temperature occurs in December, when the mean is 59.0° F. ; January is 0.3° warmer. From then on there is rapid rise to a maximum of 91.6° F. in May and June. The high point that should be reached in June and July is prevented by the increasing cloudiness of June, and the following monsoon. In July the mean temperature is 84.4°, falling to 83.0° in September. Here what would be the normal descending curve is met again, and the fall is rapid to the 59.0° of December. Table III shows the temperature in degrees Fahrenheit by months for the year (see Figs. 1 and 3).

The daily range of temperature is large throughout most of the year (Table III). During the monsoon it reaches a minimum of 8.8° R in July, then rises through the autumn and winter to a maximum of 29.2° in April ; from here it falls again rapidly to July. The period of maximum daily temperature range coincides roughly with the period of highest absolute temperature. Such temperatures as the above means alone would not be unfavorable for plants; it is the occasional extremes that make plant growth difficult. The mean annual extremes range from 39.6° to 114.9° F. On rare occasions the temperature of a winter night may drop down to or below the freezing point, but apparently these exceptional low temperatures are of little importance in determining either the character or the composition of the vegetation, even in depressions where frost is most liable to occur. The highest recorded temperature was 119. 8°F., on June 19, 1878 (8). It is the occasional unusually high temperature during the hot season, together with low humidity and a strong wind, that makes plant life difficult.

The temperatures given above were taken under open thatched sheds, at a height of 4 feet from the ground. The temperature at the soil surface at times rises very much higher than these figures (6) "The average temperature of the ground surface in India is, at the hottest time of the day in the cold weather from 10° to 20° (F.) above that of the air at 4 feet high. The difference increases until the months of April and May, when the excess is usually as high as 40° and sometimes 45° or 50° (F.)". Ifc falls rapidly during the rainy season, and is as small in August as in the cold season.

Humidity. Temperature and humidity seem to be the most important of the climatic factors influencing plant development. Perhaps humidity is the most important, for if the proper balance between water loss and water intake could be maintained, plants would be able to endure without much difficulty the maximum temperatures of the Gangetic Plain. In general, humidity depends on rainfall, but during the cold season the low temperature is an important factor in increasing the relative humidity.

From a mean of 70.4 per cent, in January, the relative humidity falls rapidly to a minimum of 34.7 per cent, in April, then rises rapidily to a maximum of 81.9 per cent, in July. There is a slow fall to 80.2 per cent, in September, then it becomes rapid with the cessation of the rains, down to 68.9 per cent, in October. Here it is overtaken by the falling temperature, and rises to 72.9 per cent, in November, and 70.5 per cent, in December and January. With the rapidly rising temperature and low rainfall of the hot season it falls very rapidly to the April minimum. See Table IV and Figs. 2 and 3. The periods of high relative humidity are optimum for plant growth. As in the case of high temperature, it is the periods of low humidity that cause trouble. The daily humidity range is great during most of the year. Maximum humidity usually occurs at about 6 o'clock and minimum at about 14 o'clock. July shows the least daily range and November the greatest, though the range is great throughout the cold season. The daily range during the hot season is intermediate.

Mean, mean maximum and mean minimum relative humidity, and mean daily range, in percent, by months, for Allahabad.

The mean relative humidity by months gives a very inadequate picture of the severe conditions to which plants are subjected in the hot season. During April, for example, the mean is 34.7 per cent, but the mean minimum attained about 14 o'clock, is 18.4 per cent, and there are many days when the extreme runs much below this ; even at night the mean maximum is only 53.2 per cent. During the rainy season humidity is generally high, and at all times very favorable for plants. During the cold season the mean maximum attains the highest point for the year, reaching 94.1 per cent, in November. Night after night there is a havy fall of dew. At the same time the daily range is greatest during this period, the 14 o'clock mean mini- mum dropping to 40 per cent. . The humid nights permit the vegetation to recover from the drought of daytime. During the hot season, however, there is little opportunity for recovery following the extremely dry day, and none but the most xerophytic of the herbaceous plants are able to survive this trying period.

Wind. The wind is an important climatic factor in two dis- tinct ways : by bringing in moisture during the rainy season ; and by accentuating the aridity of the hot season. In the latter case only the air currents near the earth are of importance. During most of the year the winds blow fitfully, with a large percentage of the time calm. During March, April and May, wind becomes a very im- portant factor influencing vegetation. Beginning about 11 and con- tinuing till 16 or 17 o'clock, there is a strong wind from the N.N.W., locally known as the " lu." Coming as it does at a time when the relative humidity is at the lowest point, it exerts a powerful dessicating effect on vegetation. The soil is dried out by wind and heat together, and little herbaceous vegetation that is not protected, or favorably situated with regard to water supply, is able to survive.

There is little or no air movement at night during the hot season, vegetational seasons. From the point of view of the vegetation it is

convenient to begin the year with the rainy season (Fig. 7). This season may be taken to begin about the 15th to the 20th of June, when the first scattering rains fall, and lasts to the end of Septem- ber. It is characterized by high rainfall, low insolation, high tem- perature, and high humidity (Table V, and Figs. 4 and 7). July may be taken as typical of this season, when the mean temperature range is from 80.6° to 89.2°F., while the relative humidity ranges from 89.3 to 71. 7 per cent. Such conditions are optimum for plant growth, and a luxurient herbaceous vegetation springs up.

The rainy season merges gradually into the cold season, which may be taken to extend from the first of October to the end of February. It is characterized by low rainfall, high insolation, low temperature, and relatively high humidity. December may be chosen as a typical month ; the mean temperature range is from 48.6 to 74.2°F., and the relative humidity runs from 88.4 to 40.5 per cent. (Table V, and Figs. 5 and 7). The lower temperature and the low

Fig. 5. Mean hourly temperature in degrees Fahrenheit and mean relative

humidity in per cent, for the month of December at Allahabad. Fig. 6. The same for the month of April.

humidity of midday result in a vegetation that is mesophytic, and which contains a larger proportion of temperate region plants. The hot season is ushered in with the rising temperature and decreasing humidity of spring, and extends from the first of March to the middle of June, It is characterized by low rainfall, high insolation, high temperature, low humidity, and strong winds. April and May are typical of this season. The April temperature ranges from 73.4° to 102.6° P., and the humidity from 53.2 to 18.4 per cent. (Table V, and Figs. 6 and 7). Both temperature and humidity are

Fig. 7. Mean temperature, mean sunshine (mean cloudiness curve inverted), mean rainfall, and mean relative humidity for the year at Allahabad, so calculated that all the maxima touch the top of the graph and the minima touch the bottom. This emphasizes the three climatic seasons, which determine three corresponding vegetational seasons.

slightly higher in May. The climatic conditions are distinctly xerophytic, with the result that the mesophytic vegetation of the cold season disappears, and only those plants that are adapted to withstand the severe drought can survive. There are limited local areas that for some reason or other are able to supply sufficient moisture to sustain mesophytic plants, and it is in such places only that the vegetation remains distinctly green.

2. The biotic factors.

The term "biotic factors" includes all influences traceable to living organisms, but only animals and man are considered in this paper. Undoubtedly man and his domesticated animals are the most important of the biotic factors. For more than 20 centuries the Gangetic Plain has been populated with an agricultural people. It is difficult and perhaps impossible to form any adequate conception of the intensity of the human factor in times past, but since 1850 the population of the 2811 square miles of Allahabad District has fluctuated from 480 to the high level of 543 to the square mile at the 1891 census (8). At present it is about 530 per square mile. To the human population must be added, according to census of 1909, 331 cattle and buffaloes, 123 sheep and goats, and 8 horses and donkeys, or a total population of domestic grazing animals of about 463 per square mile. Therefore the area has to support a total population of about 1,000 per square mile of animals that gain most or all of their food directly from the vegetation.

Man influences the vegetation in a number of ways, mainly by cultivation, by grazing his animals, and by cutting for food and fuel.

Cultivation. The returns of 1907-08 (8) show that about 58 per cent, of the District was under cultivation. This figure probable fairly indicates the extent of cultivated land from year to year in the area immediately surrounding the city of Allahabad. Cultivation causes retrogression of the vegetation, and the more thorough it is, the greater the effect. Wild plants, both annuals and perennials, are rooted up and killed, and their place is taken by annual ruderals, usually native, but in many cases introduced. At the same time cultivation tends to make the area more xerophytic by removing the permanent plant covering, and substituting a cover of perhaps more mesophytic but short-lived annuals. "When the crop cover is harvested the soil is left practically bare, and dries out speedily.

Grazing animals. All of the uncultivated land in the area is closely grazed throughout the year. During the rainy season, the vegetation is able to keep pace with grazing, even though it is constantly kept eaten off close to the ground. As the cold season progresses and the growth of herbaceous vegetation is checked, the effect of overgrazing become more and more evident. Finally, during the dry season all grasses and associated plants are eaten down to the soil surface (Fig. 11). Annuals die under the combined hardships of the grazing and lack of protection. Only perennial grasses and a few other xerophytic herbs with strong perennating organs are able to survive. Where overgrazing progresses still further, even the perennial grasses are destroyed, and the ground left practically bare (Fig. 14).

In consequence of the severe grazing there is no cover of vegetation on the ground to aid in conservation of the monsoon rains. Where drainage has been established, runoff of the monsoon rainfall is very rapid and complete ; where the land is flat, the vegetation has no effect in preventing runoff.

The scattering thorny shrubs and small trees, with a few inedible exceptions, are kept eaten done to small dense rounded bushes wherever and as long as animals are able to reach them.

Cutting for food and fuel. In India a very large proportion of the plants comprising the flora is made use of by man for some specific purpose, especially for food and medicines. Perhaps the results of all such exploitation are not in the longrun very detrimental. Usually it is parts that are not essential to the life of the plant that are used for food. It is for forage and fuel that man works the greatest havoc with the vegetation. Throughout the year, and especially during the hot season, much of the available grass covering is cut off just below the surface of the soil for fodder for domestic animals. It is only the most persistent perennials that can survive such treatment. In the Upper Gangetic Plain no natural fuel remains except a few species of xerophytic shrubs and small tres3, and in most places these are periodically cut, sometimes almost to the extinction point. The fine groves about villages are all planted, and these alone escape the unremitting struggle of man to find fuel. There is no attempt locally at conservation of plant resources.

Wild fires are negligible as a factor influencing vegetation, for the reason that the land is so intensively cultivated and grazed that there is nothing left to burn.

Wild grazing animals are not numerous and are of little importance. Earthworms are abundant during the rains, and their burrows probably are of considerable importance in promoting aeration and water penetration. Numerous species of black ants probably render a similar service. In general, insect pests are relatively unimportant. It is uncommon to find the indigenous vegetation eaten to any great extent.

White ants, White ants are the only insects that exert any great influence on the vegetation. They are found everywhere and are almost unbelievably abundant. Doubtless their burrows function as do those of the black ants, and they reach a depth of 5 feet or more. It is only rarely that white ants attack living plants, though they closely follow dieing parts and keep them eaten away. When grasses and other plants, especially the annual vegetation, die, the white ants eat the remains. The great problem in this connection is whether and to what extent they prevent the accumulation of this dead vegetation as organic matter in the soil. It may be that the intense heat and dryness of the hot season would oxidize organic matter even if it were introduced into the soil, though the. experience from trenching indicates that this is not necessarily the case. It is probable that the organic matter eaten by white ants is oxidized within their bodies. This prevents the usual course of oxidation by the action of bacteria in the soil with the resulting increase of nitrates and solution of mineral constituents, and the soil is deprived of just so much fertility. In so far as they prevent the accumulation of organic matter in the soil, and thus impair its fertility and water-holding capacity, white ants are detrimental to the vegetation.

Almost literally every square inch of ground is either cultivated or grazed. Excessive grazing is by far the most important of the local biotic factors. It becomes increasingly evident that there can be no adequate interpretation of the vegetation of such an area as the Upper Gangetic Plain without due consideration of the tremendous pressure of the human factors. Neither climatic nor biotic factors alone, but all acting together, influence and determine the vegetation.

Ecological features of the vegetation.

Types of succession. It is well established that under natural conditions the vegetation of new local areas, such as ponds, stream margins, cliffs, bare rocks, cultivated tracts, and the like, passes through a succession of associations, finally culminating in a permanent climatic climax. Cowles (3) distinguishes two types of succession in any given region:— (l) Climatic, depending on slow changes in geologic climate, and which is at all times in the stage of climatic climax; and (2) Topographic, occurring as smaller cycles of development within the immensely large climatic succession, and dependent on local physiographic variations. The climatic succession progresses so slowly that it cannot be made the subject of exact study. Topographic succession progresses much more rapidly, but still is a slow process. It usually is assumed that the associations observed in passing from a new area to the climax fairly represents the succession as it occurs at any given spot through a long period of years. Each topographic succession area has as its goal the climatic climax vegetation, and is terminated when this stage is attained.

Seasonal succession. A third typeof succession is a prominent feature in a strongly periodic climate. It is illustrated by the striking changes in the aspect and content of the vegetation from season to season, and may be called seasonal succession. As a result of the three well defined climatic seasons, rainy, cold, and hot, there are three equally well marked vegetational seasons. The traveller across the Upper Gangetic Plain is impressed with the numerous groves of stately trees. These groves are without exception planted, and are composed of species that either are not native to the immediate area or that cannot survive in the young stages without protection. Some of these trees would quite readily perpetuate -themselves were it not for the pressure of the human factors.

Aside from cultivated fields and the groves of planted trees, the aspect of most of the area is determined by herbaceous plants. Here and there are tracts dominated by shrubs and small trees, but across these the herbaceous vegetation extends unchanged. Obviously there is no change in the content of the woody vegetation in response to change in season, but there is response in the matter of leaf fall and time of blooming. Even these changes are relatively inconspicuous in comparison with the changing aspect of the herbaceous and under-shrub vegetation. It is periodicity in the small and very abundant annual and perennial plants that renders the fact of seasonal succession so conspicuous.

During the rainy season, when all conditions are favorable for maximum plant growth, the whole country is covered with a luxurient mantle of herbaceous vegetation. Much of it is annual. The perennial herbs and undershrubs attain their maximum development at this time. The vegetation approaches hygrophytic, with large soft thin leaves and tender stems. It is during the rainy season that the tropical aspect of the flora as well as its composition appears most clearly.

With decrease of rain, and the lower temperature of the cold season, much of the rank growth of the rainy season disappears. What survives is gradually eaten down by grazing animals. The hygrophytic annuals are replaced by mesophytic annuals, especially Compositae, of more temperate connections. Perennials lose their luxurient tropical appearance, and in every way the vegetation is less in amount, number of species, and conspicuousness.

With the advent of the hot season, the high temperature and extreme aridity complete the destruction of the annual vegetation that is not situated in favored edaphic spots, and little more than the persistent perennials remain. There are a few annuals, but the number is small. The perennials now present a very different appearance from that of the rainy season, and even the cold season. All but the youngest leaves fall, and the more delicate branches die. The hard resistent parts remain alive, and continue to put out a few new leaves and bloom profusely throughout this season. The herbaceous vegetation is characterized by well developed perennating organs. Prostrate and rosette forms abound. Much of the soil surface is exposed, and a pall of impalpable grey dust settles over everything. Both in aspect and structure the vegetation is distinctly xerophytic.

The vegetation of the cold season is more representative than at any other time of the year. Besides the many plants peculiar to this season, some of the more hardy of the rainy season annuals persist, and the vegetation exhibits a freshness and vigor that is lacking in the hot season. The outstanding features of succession are also more obvious at this time.

The influence of man on the vegetation. It is impossible to determine what was the condition of the vegetation in the Upper Gangetic Plain before the interference of man. Doubtless many plants have been directly exterminated, and others have been killed by removal of some sort of protecting forest cover; others have been eliminated in competition with plants introduced along with man, and by the changed conditions resulting from his various activities. But man has been present long enough, and has maintained his conditions of living sufficiently unchanged that the vegetation has become balanced against him. With the awakening in agriculture, the human factors are beginning to change, and in consequence we may expect more or less profound changes to take place in the vegetation again. The effect of the human factors is to interfere with the natural development of the vegetation, and to throw it back to a more primitive stage. If the human factors about Allahabad where to become much more intense, there would result the extinction of many valuable native plants, and a further regression in the vegetation; if the human factors where to relax in intensity, the vegetation would at once pass on to a more advanced stage; if man were removed entirely the vegetation would ultimately reach a climatic climax, though different in detail from the one present when man first arrived on the scene in significant numbers. This balance between the progressive tendencies of the vegetation and the retrogressive influence of man is one of the most striking features of the local vegetation.

Not only has man caused the vegetation to retrogress from the original climax, and is now in a state of balance with it, but he interferes with normal succession in edaphic areas so profoundly that it may actually be prevented. Pools, for example, that should show early stages in succession, serve as bathing places for man and beast. as the village water supply and laundry, as the village sanitary system, and finally as a source of water for irrigation. Ponds that appear promising places for study one year are devastated the next (Fig. 9). The result is that for the most part succession must begin as the beginning each year. Instead of the topographic succession, pools present year after year mUch the same appearance for a given season, and seasonal rather than topographic succession is conspicuous. This renders the study of topographic succession very difficult, and everywhere emphasizes the struggle that is going on between man and nature, and the balance that has become established between them. By far the largest part of the area is as far advanced in succession as is possible under existing conditions (Fig. 11).

Details of topographic succession.

In more favored parts of the earth where vegetation has a chance to develop naturally, it is possible to study with considerable precision the gradual development of the flora from the most primitive conditions to the most advanced association of plants that the area can support; For the most part studies in plant succession have been made in temperate regions, where there is but one vegetational season, the summer or growing season. This is followed by a winter season during which the vegetation is at more of less of a standstill. In the Upper Gangetic Plain there is no season in which growth is impossible. Growth is checked and the vegetation modified by the aridity of the hot season, and at all seasons and at every stage it is interferred with even to the extinction point by the human factors. Therefore it is not an easy matter to trace the steps of topographic succession. The problem is further complicated by the seasonal succession. Yet in spite of the difficulties, it has been possible to find a few favorable spots which give a clue to the early stages of succession, and others which indicate what the higher stages would be if the human factors were lessened or removed.

Stages is the topographic succession. It is convenient to divide the succession into the following stages:—

1. Aquatic stage.

2. Wet meadow stage.

3. Dry meadow stage.

4. Thorn scrub stage.

5. Pioneer monsoon deciduous forest stage.

6. Climatic climax monsoon deciduous forest stage.

Of these stages the first is quite commonly observed in pools and streams; the second is not well represented; the third is the present modified climatic climax; stage four is not found well developed anywhere, but fragments of it are to be seen in protected places, and in the thorny shrubs and small trees scattered over certain areas; stages five and six are largely hypothetical, and six would be the true climatic climax for this portion of the Gangetic Plain.

Aquatic stage. As has been said, the aquatic stage is to be found in the numerous pools that are scattered everywhere over the area. Most of these pools are very small and shallow (Fig. 9); a few may cover an area of several acres; some dry up after the close of the rains, while others retain their water up into the cold season, and a few are permanent. The smaller and shallower and less permanent pools show only the earlier phases of the aquatic stage, for the natural vegetation is always interfered with by man. The constant use of a pool for various domestic purposes destroys most or all of the plants, and irrigation finally removes the water and leaves a mud flat that quickly dries out without being able to support much in the way of plant growth. The green plants in proximity to water are eaten off by grazing animals. Shallow pools are planted with rice during the rains (Fig. 9), and this effectively prevents development of the characteristic wild aquatic vegetation. The small number of native plants that can grow under such conditions are more of the nature of ruderals. Even if a pool is relatively free from human disturbance, the water level is rapidly lowered by evaporation, so that plants that are strictly aquatic during the rains are finally stranded on mud flats to dry out and die, or to survive by means of perennating organs till inundated during the next rainy season (Fig. 8.).

The aquatic stage may be further divided into:—

Free-floating aquatic stage;

Attached submersed aquatic stage;

Attached emersed aquatic stage.

These stages overlap each other more or less, but they indicate the actual order of normal succession.

Free-floating aquatic stage. This stage is represented mainly by algae, of which there is great variety and considerable abundance. The principal algae are several species of Cyanophyceae, Spirogyra spp., and Cladophora spp; others very commonly found, but abundant only in very limited areas are Mougeotia spp., Oedogonium spp., Hydrodictyon and Vaucheria. The Cyanophyceae are specially abundant in the shallow rainy season pools, though they are rather common throughout the year. Various species of Spirogyra are to be found throughout the year. Mougeotia, Hydrodictyon, and Vaucheria are most abundant in later winter and the hot season.

There are a few free-floating aquatic vascular plants, the most important being Ceratophyllum demersum L., Azolla pinnata, Lemna minor L., Wolffia arrhiza Wimm. and Trapa bispinosa Roxb. The first named belongs primarily to the rainy season, the others mainly to the cold season.

Submersed attached aquatic stage. It is in this stage that the vegetation first becomes really abundant, and an important feature in succession. The plants of this stage root in the mud floor of the pool, and grow upward, though always remaining under water. Some of them growing in the rivers are attached 'many feet below the surface, and produce an enormous amount of vegetation. Potamogeton pectinatus L., is able to occupy the deepest water, and is one of the most abundant of the aquatics. Hydrilla verticillata Casp., Naias graminea Delile and Vallisneria spiralis L., are equally abundant, and are found in progressively shallower water. Others very commonly met with are Potamogeton crispus L., Zannichellia palustris L., Chara spp., and Nitella spp. Most of these plants are to be found throughout the year, though they reach their maximum development during the cold season. Chara and Nitella are not so abundant during the winter, because they are plants of shallow water, and are among the first to be destroyed by drying up of the pools.

Attached emersed aquatic stage. Plants of this stage are confined to rather shallow water (Fig. 8), and thus are very liable to be left stranded by lowering of the water level. Some of them are amphibious, and succeed in making fair growth even when exposed for long periods. The vegetative parts of others are killed by the exposure, and only subterannean perennating organs remain alive. The most important and characterestic plants of this stage, approximately in the order in which they appear in the succession, are Nymphaea lotus L., Eleocharis plantaginea Br., Scirpus maritimus L., and Marsilia sp. There is a number of other species that occur quite commonly, but are not present in large numbers, and take little part in giving character to the formation. A few characteristictly amphibious species root in shallow water, and produce extensive floating branches that grow out over the surface of the water (Fig. 9). Chief among these are Ipomaea reptans Poir., Panicum paspaloides Pers., and P. punctatum Burm. They never, however, are permitted to develop the dense floating mats that are so common a feature of aquatic vegetation growing under undisturbed conditions.

Wet meadow stage. This stage comprises the vegetation on the margin of damp soil between the water of pools and the extensive dry meadows that characterize the plains. It is perhaps the most difficult stage of all to unravel, because its green vegetation forms such an attractive grazing ground, and because it shifts position with the change of seasons. During the rainy season it encroaches on the dry meadow, and during the hot season it recedes far toward the center of the pools, or disappears entirely. I have found just one place where the water relations are sufficiently constant that the wet meadow is permanent enough to study in detail (Fig. 10). By supplementing this small area with observations on other places, it is possible to give a fairly consistent account of the wet meadow stage.

The typical association is characterized by a large development of Cyperaceae and Cynodon dactylon Pers. About the pools that are subject to extreme fluctuation in water level there is also intensive cultivation and grazing (Fig. 9), and the Cyperaceae and Cynodon are about the only plants of consequence that can survive from year ot year. Many of these plants are able to adapt themselves to a wide range of conditions, from shallow water to relatively dry banks. Cynodon dactylon, for example, is specially typical of later stages of the wet meadow, yet it can grow fairly well when completely submerged, and persists in many places into all but the dryest of the dry meadows. The following is a list of the common plants of the wet meadow is arranged approximately in the order of succession : Scirpus maritimus, S. quinquefarius Ham., Cijperus difformis L., Fimbristylis diphylla Vahl., Amniannia baccifera L., Ranunculus sceleratus L., Cynodon dactylon, and Eragrostis tenella R. and S. Usually Scirpus quinquefarius, Fimbristylis and Cynodon determine the general aspect of this stage (Fig. 10).

Where the wet meadow zone shifts back and forth with the change of season, where erosion produces fresh soil surfaces, and where the human factors are very intensive, perennial plants are replaced by an association of short-lived annuals. These come up from seed where- ever conditions are favorable, and persist as long as there is sufficient moisture. This group of wet meadow annuals reaches its greatest development during the late cold season and early hot season. It is such plants as these that make the seasonal succession so conspicuous. They are for the most part small plants, and contribute little toward the development of the permanent flora : Jimcellus pygmacus Clarke, Scirpus michelianus L,. Potentilla supina L., Gnaphalium indicum L., Grangea madcraspatana Poir., Bumex dentatus L., Ranunculus sceleratus, Polygonum plebsjum Br., Alternanthera scssilis Br., Bergia ammannioides Robx., Biccia sanguinca, Hydrolea zeylanica Vahl., and Glossostigma spathulatum Arn. Dry meadow stage. The wet meadow merges gradually into what may be called dry meadow (Fig. 11), a very extensive and uniform association dominated by perennial grasses. Under natural conditions it probably would be but an inconspicuous incident in the topographic succession, but due to the retrogressive influence of the very intense human factors, it has become for the time being the climatic climax over most of the area, This stage is connected with the wet meadow by Cynodon dactylon and Eragrostis tenella. The latter is a short-lived annual that flourishes everywhere during the rainy season, and almost completely disappears by the middle of winter.

The typical dry meadow (Fig. 11) is dominated by two perennial grasses, Andropogon intermedins Br., and Elusine aegyptica: Desf. Both are able to produce a luxurient grass cover, but under severe grazing they assume a dense compact tufted habit; both propagate freely by runners. Under excessive grazing the Andropogon is the more persistent. In slight depressions and in the shade of trees, where growth conditions are a little less severe, three perennial Leguminosae, Desmodium triflorum DC, Indigo/era enneaphylla L., and Alysicarpus monilifer DC. are very charateristic components of the dry meadow. They too have the tufted prostrate habit of growth. All are excellent pasture plants, and owe much of their value to their persistence under grazing and drought.

During the rainy season there is a conspicuous development of Eragrostis tenella, and of two annual Leguminosae, Cassia obtiisifolia L. and Grotalaria medicagitiea Lamk. The last two are often so abundant as to give character to the dry meadow vegetation, but they die at the beginning of the cold season, and little trace of them remains in the hot season.

Overgrazing and the intense aridity of the hot season greatly reduce the Iuxurience of the grasses of the dry meadow (Fig. 11), and another important constituent then becomes prominent. This is a series of small xerophytic and very persistent perennials that are able to survive both the grazing and aridity because of the development of effective perennating organs. All have deep tap roots with a perennial crown of stem, and most of them have a well developed rosette habit of growth. As the cold season advances and gradually merges into the hot season, the older and larger leaves fall, the more delicate stems are grazed off or die, and the aerial parts become reduced to a small compact crown of very resistent vegetation. In this condition they are able to bloom and produce seed abundantly. Some of the more common of these plants are:—Convolvulus pluricaulis Chois., LepidagatJiis trinervis'.Nees, Justicia simplex D. Don, Euphorbla thymifolia Burm., Boerhaavia repens L., Vernonia cinerea Less., Corchorus antichorus Eoeusch., Launea asplenifolia Hook, and Heliotropium strigosum Willd. Equally persistent and xerophytic are two undershrubs, Calotropis procera Br. and Tephrosia purpurea Pers., and they show a similar reaction to the extreme aridity of the hot season.

It must be emphasized that the entire area is subjected to merciless grazing. Where overgrazing progresses too far, the dry meadow grasses are nearly or quite exterminated (Fig. 14), and a com- paratively worthless grass, Aristida adscenscionis L. becomes domi- nant. It grows up quickly at the beginning of the rainy season, and dies out with the increasing dryness of cold season. It is fairly good for grazing while green, but when mature the awned fruits render it worse than worthless. The appearance of this grass marks the last stage of the exploitation of the plant resources by man. When it dies the ground is left bare except for a few scattering half dead tufts of the typical dry meadow grasses, and some of the more persistent of the rosette weeds.

Thorn scrub stage. Under existing conditions by far the larger part of the uncultivated area about Allahabad is in the dry meadow stage, the modified climatic climax. There are a few scattered areas that give a clue to what the next stage in the vegetation would be if left free to develop naturally (Figs. 12 — 15). This stage is here called provisionally " thorn scrub", for that probably -"is whafc it most closely resembles in its present condition. Nowhere is it to to be found fully developed. To complete the picture we must patch together two quite separate types : (l) areas protected from grazing, such as military and dairy grass farms (Fig. 12), and (2) uncultivated areas that bear a scattering growth of thorny shrub and small trees (Figs. 13-15).

The grass farms owe their more advanced ecological state to protection from grazing. The vegetation is practically limited to grasses and a small number of associated annual and perenrjial herbs. The grass is permitted to grow undisturbed to maturity, when it is cut for hay. All the cutting is done by hand, and not only the grasses but all vegetation impartially is cut off close to the ground. Such a method of harvesting very effectually prevents the develop- ment of shrubs and trees. The characteristic grasses of these pro- tected meadows (Fig. 12j are Apluda varia Hack., Cenchrus biflorus Roxb., Andropogon annulatus Forsk., A. contortus L., Paspalum sanguinale Lamk., Anthistiria imberbis Retz., Iseilema laxum Hack., and I. wightii Anderss. It is difficult to determine which of these grasses is the most important. Apluda varia, a tall coarse grass, ig perhaps the first to assume a place of prominence, and is the most xerophytic. It is closely followed in point of time and in importance by Cenchrus and the two Andropogons. The remaining species are of secondary importance. Along with the grasses of such a protected area there come in a new set of herbaceous annual and perennial dicotyledons, such as Rhynchosia minima DC, Alysicarpus bupleuri- folius DO., and Grotalaria mysorcnsis Eoth. At times one finds a straggling Zizyphus jujuba or Z. rotundifolia, Lamk., an Acacia arabica, or some other woody plant, but they are not numerous.

That the grasses cf the protected meadows actually are the result of protection, and not a matter of special soil or moisture relations, is indicated by the fact that they are found all over the area, wherever grazing is lessened or difficult. They flourish even on the dryest of the cliffs, in positions where cattle and goats cannot readily reach them. They already potentially occupy the area, and await only protection from excessive grazing to develop into a more advanced plant formation.

The areas of thorny shrubs and small trees, which are used to give name to the stage following the dry meadow, are remnants of a pioneer type of forest that has persisted on tracts that have remained uncultivated for considerable time (Figs. 13 and 14). In some places the growth is scattering and savannah-like, in others it is more ' dense (Fig. 15) ; but everywhere the ground is occupied by a typical dry meadow association of grasses (Fig, 13). The principal species comprising this pioneer forest are Capparis sepiaria L., Acacia arabica Willd., A. leucophloea Willd., Balanites aegyptica Delile, Justicia adhatoda L., Flacourtia sepiaria Roxb., Jatropha gossypi- folia L., Zizyphus jujuba Lamk. and Alangium lamarckii Thw. Most of these plants are conspicuously thorny, and are thus protected from complete destruction by grazing animals. Justicia and Jatropha, are the only abundant ones with no obvious protection, yet animals will not eat them even when other vegetation is scarce ; the latter has been introduced from Brazil, and is now completely naturalized and competes successfully with the other plants of the formation. Alangium also is sparingly eaten. Balanites is the most successful as a pioneer in dry unpromising conditions, though the Acacias are not much behind it in this respect.

Some distance north and west of Allahabad Butea frondosa Roxb., becomes one of the most important of the thorn scrub trees ; there are only straggling outliers to be found in the local area. It is not thorny, but is the ecological equivalent of the thorn scrub plants.

Wherever such woody vegetation occurs it is more or less con- stantly cut for fuel, and the smaller plants are pruned by grazing

2563—41 animals to dense oval bushes (Fig. 13). It is only when the bushes finally spread out so wide that the animals cannot reach the center that shoots spring up into trees. The thorn scrub is destroyed only by cutting combined with cultivation ; cutting alone does not destroy it, for all the species coppice freely, and most of them propagate by shoots from roots. Where cutting is restricted, the Zizyphus jujuba and Acacia arabica develop into a very fair forest (Fig. 15).

As might be expected, there are no bulbous plants, and no epiphytes in the thorn scrub. There are a few lianas, the most com- mon being species of Asclepiadaceae, notably Bemklesmus indicus Br. Where there is more protection, as in planted groves and in hedges around orchards, Cocculus villosus DC, Tinospora cordi/olia Miers. Vitis trifolia L., and a number of Cucurbitaceae are common, but they are unable to survive unrestricted grazing. The parasite, Guscuta rejicxa Roxb., is very common, and whileoccurcing most frequently on Acacia arabica and Zizyphns jicjioba, it is able to grow impartially on almost any available host plant.

Seasonal succession is not such a conspicuous feature of the thorn scrub, though it is well shown in the dry meadow herbaceous vegetation that extends over all such areas. Periodicity is prominent only in leaf fall and in time of blooming. Many of the species are deciduous and have their flowering period during the hot season.

There is little doubt that the meadows of the grass farms and the thorn scrub woody vegetation belong to the same general stage in topographic succession. The former are shrubless and treeless because of the thoroughgoing annual cutting, the wooded areas lack the more advanced grasses because of overgrazing. If left undisturbed, both the protected meadow grasses and the thorny shrubs and trees together would very completely occupy most or all of the area now covered by dry meadows, and would become a dense thorn forest. This actually is happening in the Fisher Forests at Etawah, 200 miles to the west of Allahabad. This is an area along the Jamna Eiver recently placed under Government supervision for afforestation experiments. It is a series of deep ravines developed as the result of the removal of the soil-holding vegetation by overgrazing and unrestricted cutting for fuel. With only five years protection, both the grasses and the woody plants of the thorn scrub stage have developed luxuriently (Fig. 12).

Nowhere over the area about Allahabad has the vegetation been able to develop without protection beyond a poor display of the thorn scrub trees and shrubs, and most of the area is in the dry meadow stage. This stage, so conspicuous under existing conditions, prob- ably would be but a short stage following the wet meadow, or it might even not occur as a distinct stage at all, if the vegetation could develop naturally. Such a flora now represents the climatic climax as degraded and modified by man. It is the resultant of the struggle between the vegetation on the one hand developing toward some type of climatic climax forest, and the retrogressive influence of the intense human factors on the other hand, continually interfering with and destroying it. This balance between man and the natural vegetation is very delicate, and may perhaps help to explain many of the very serious human problems of the area. After a series of favourable years, the slightly increased vegetation allows a corres- ponding increase in the human and animal population, and results in increased demands on the plant resources. Then subsequent bad years leave the men and animals with insufficient food, and if many do not die as the direct result of famine, they are left so weakened as to become easy prey to- pestilence, and are again reduced in numbers. In other words, under present conditions, the human and animal population is about as large as the area can support. Improved methods of agriculture, and intelligent protection and conservation of the plant resources appear to the only solution of the problem.

Pioneer monsoon deciduous forest stage. If left to itself, it is very probable that the thorn scrub would in turn be replaced by some form of higher and more mesophytic forest. Schimper concludes (10) that if the annual rainfall is below 90 cm., " xerophi- lous scrub", especially " thorn forest" and "thorn bush," prevails ; with 90 — 150 cm. there is a struggle between "xerophilous woodland" and grassland, with the former prevailing when there are greater heat and longer rainless periods during the vegetative season. With 180 cm. or more rainfall, a high forest is produced. Brandis (1) says that " really thriving forests are only found where the fall exceeds 40 inches, and a rich luxurient vegetation is limited to those belts which have a much higher rainfall. "

The highest type of vegetation about Allahabad, the scrubby xerophytic shrubs and trees, would, I think, correspond roughly to Schimper's " thorn scrub. " Doubtless in most regions treated by Schimper, the vegetation described as climax actually is climatic climax. He was recording situations as they actually are ; some of them at least are not necessarily climatic climaxes but modi- fied claimaxes due to retrogressive influences. To class the thorn scrub of the Upper Gangetic Plain as the true climatic cli- max is a mistake. On Schimper's classification, the rainfall in the area about Allahabad should produce a forest somewhere between ' xerophilous scrub" and "xerophilous woodland ". It is difficult to determine exactly what these terms mean. Probably each writer must provisionally fix his own limits to them. Certainly the Allahabad thorn scrub is not truly xerophilous. According to Brandis the area ought to be able to support a "really successful forest". In adjacent hilly regions, especially to the south, where there is a much smaller amount of land suitable for cultivation, and the population consequently is much less, a successful forest occurs (Fig. 16). It appears as if forests and density of population may be definitely correlated with each other in India.

We must assume then that if the human factors were lessened or eliminated, the area would more or less rapidly pass into the complete thorn scrub stage, or even into a true thorn forest. In the protection of the grasses, shrubs, and trees of this forest,' a more mesophyfcic type of forest would develop. Some of the thorn scrub trees, as Zizpyhus jujuba, Acacia arabica, and Alangium lamarckii, would develop to much larger dimensions (Fig. 15). Butea frondosa would become abundant. Such a forest is pioneer to the climatic climax in the mountainous region to the south. Other trees, as Dalbergia sissoo Eoxb, and Holoptelea integrifolia Planch, at present exclusively culti- vated would quite readily become self-perpetuating components of such a forest. Finally, there would be the slower influx of truly climax trees from neighboring forested areas.

The development of such a forest might result in some increase of rainfall, making conditions still more favorable for the development of a high forest. Brandis records (2) an instance of slight increase of rainfall following thorough-going protection of a large forest tract in Central India. Certainly a dense intermediate forest, with its accompanying herbaceous and shrubby floor vegetation, would greatly conserve rainfall by preventing rapid run-off of the monsoon rains, and by increasing the relative humidity during the hot season.

Climatic climax monsoon deciduous forest stage. Probably the dominant constituents of the ultimate climatic climax forest would be immigrants from neighboring forested areas to the north and south. We should expect to find Terminalia tomentosa Bedd, and Tectona grandis L., dominant. Almost certainly species of Sterculia, Bombax malabaricum DC, Anogeissus latifolia Wall, Stephegyne parvifolia Korth., Buchanania latifolia Eoxb., Eugenia jambolana Lamk., and other fine trees, and Dendrocalamus strictus Nees would invade the area from the forests both to the north and to the south. Acacia catchu Willd., and even Shorea robusta Gaertn, might finally be represented. With such immigrants, and in the protection of their shade, a new herbaceous and shrubby vegetation, at present unknown in the area, would spring up. Perhaps several species now found only in favored places in protected planted groves would become compon- ents of this undergrowth. Such a climax forest as has been sketched above is purely hypothetical for the area, but all natural conditions seem to warrant the assumption that it would develop. It would be a typical deciduous monsoon forest, decidedly tropophytic, leaf- less during some portion of the hot season. It would contain few bulbous herbs, few epiphytes, abundants lianas, and would have a great abundance and variety of grasses in the more open places. The fine forests at the north side of the Upper Gangetic Plain adjacent to the Himalayas, and in the more inaccessible parts of the Vindhias to the south (Fig. 16) may be taken as a picture of this hypothetical forest.

Summary

1. The area under discussion lies in a 10 mile radius about Allahabad, and is representative of a large part of the Upper Gangetic Plain.

2. The vegetation is influenced and determined as much by biotic as by climatic factors.

3. The climatic factors, rainfall, insolation, temperature, humi- dity, and air movements, are periodic in distribution, and produce three distinct seasons :

(a) Rainy season, from the middle of June to the end of September, with high rainfull, low insolation, high tempe- rature, and high humidity. .

{b) Cold season, from the first of October to the end of Febru- ary, with low rainlall, high insolation, low temperature, and high humidity.

(c) Hot season, from the first of March to the middle of June, with low rainfall, high insolation, high temperature, low humidity, and large air movement.

4. The biotic factors are an agricultural population of about 530 er square mile, and associated domestic grazing animals of about

4?0 per square mile. Cultivation, grazing, and cutting for food and fuel have profoundly modified the original vegetation ; constantly interfere with the normal development of the vegetation ; and cause retrogression from the original climatic climax. White ants also prob- ably exert marked influence on the vegetation. 5. The vegetation is characterized by :

{a) Distinct seasonal succession, as the result of the pronounced climatic seasons.

(b) A modified climatic climax degraded several stages from the

true climatic climax for the area. The vegetation is now balanced against the intense human factors, at about the dry meadow or early thorn scrub stage.

(c) Obscure topographic succession, due to interference by the

human factors, which at times and in places are so severe as to entirely prevent succession. G. Topographic succession presents the following stages : —

(a) Aquatic stage.

(b) Wet meadow stage.

(c) Dry meadow stage, which occurs over most of the area,

and is the modified climatic climax. (c7) Thorn scrub stage, occurring in poorly developed form in many places.

7. It seems clear from the evidence that if the retrogressive in- fluence of the biotic factors were removed, the vegetation would pass through progressively higher forest stages : — (e) Fully developed thorn scrub stage. (/) Pioneer monsoon deciduous forest stage. {g) Climatic climax monsoon deciduous forest, a forest of con- siderable density and luxurience.

Literature cited.

1. Brandis, D. On the distribution of forests in India. Ocean Highways

1872 : 8S-113. 1872. 2. Influence of forests on rainfall. Indian Forester 14 : 10-20.

18S3.

3. Cowles, H. C. The cause of vegetative cycles. Bot. Gaz 51: 161-183.

1911.

4. Hooker, J. D. and T. Thomson. Flora Indica, Vol. I., Introductory

essay, pp. 1-260. London. 1855.

5. Indian Meteorological Memoirs 5 : 187-226. 1894.

6. Imperial Gazetteer of India, Vol. I., Chapter Meteorology, pp. 104-

156. London, 1907.

7. Meteorological Department of the Government of India : Rainfall

of India. Calcutta, 1914.

8. Nevill, H. K. Allahabad District Gazetteer. Allahabad, 1911.

9. Oldham, R. D. A Manual of the Geology of India. 2nd Ed..

Calcutta, 1893. 10. Schimper, A. F. W. Plant Geography. English Ed Oxford, 1903.

Explanation of Plates.

Plate XIX

Fig. 8. Permanent lake near Allahabad, formed by damming up a ravine. Scirpus mar.itimus is dominant in the attached emersed aquatic and wet meadow stages. Note the bare wet meadow shore left by the receding water. In the back-ground is a village grove cf planted trees. Photographed January 30, 1920.

Fig. 9. Shallow temporary pcol near Allahabad, showing the result of cultivation and grazing on early stages of succession. In the water and stranded on the wet meadow shore are the remains of a crop of Oryza saliva, along with Panicum punctatum, Ipomaea reptans, and Cynodon dactylon. Spontaneous Acacia arabica in the background. This view shows the flat character of the plains. November 4, 1917.

Fig. 10. Well developed wet meadow stage in the seepage area from the permanent lake in Fig. 8. In the shallow open water are submersed aquatics ; in the edges of the pools is tall Scirpus maritimus ; over the damp soil of the wet meadow stage are Scirpus maritimus and Cypcrus difformis, fringed about by a dense mat of smaller Fimbris- tylis diphylla ; outside this is a later stage of the wet meadow, dominated by Cynodon dactylon ; dry meadow plants on the higher banks ; and spontaneous Acacia, arabica and planted village trees in the background. The whole area is closely grazed. January 30, 1920.

Plate XX

Fig. 11. Well drained area west of Allahabad, showing closely grazed dry meadow stage, with cultivated fields in level places. The meadow is dominated by Andropogon intermedins and a good represen- tation of small perennial prostrate weeds. Spontaneous Acacia arabica in the background. April 20, 1918.

Fig. 12. Rank growth of thorn scrub grasses, Aplnda varia, Ccnchrus biftorus, -Andropogon contort us, and A. intermedins, in the Fisher Forest, Etawah, as a result of five years protection from grazing. The trees are planted Acacia arabica. October 24, 1919.

Fig. 13. Closely grazed thorn scrub area north of Allahabad. The trees are Acacia arabica, the oval bushes are Capparis sepiaria pruned by grazing animals, and the straggling bushes aie Justicia adhatoda. Over the ground is a poor development of the dry meadow. April 20, 1918.

Plate XXI

Fig. 14. Overgrazed thorn scrub area north of Allahabad, show- ing untouched Justlcia adhatoda and pruned Capparis scpiaria. The soil is bare and strewn with calcium carbonate nodules (kankar). The principal rainy season herbaceous vegetation is Aristida adscensci- onis. April 13, 1919.

Fig. 15. Closely grazed thorn scrub stage near Allahabad. It is protected from cutting and the Acacia arabica and Zizyphus jujuba have developed into a fair forest. January 30, 1920. (This forest was cut off about March 1, 1920).

Fig. 16. Climatic climax monsoon deciduous forest at Ghansore, Central India. The trees are Sterculia icrens Kobx. ,* Dalbergia panic- ulata Kobx., Anogeissus la-tifolia, Tectona grandis, Semecarpus ana- cardium L , Bosiucllia serrata Kobx., and Hymenodictyon excelswn Wall. The entire area is closely grazed. Such a forest as this would be expected to develop over much of the Upper Gangetic Plain follow-

ing elimination of the intense human factors. January 18, 1920.

Fig. 8

Fig. 9

Fig. 10

Fig. 11

Fig. 12

Fig. 13

Fig.14

Fig.15

Fig.16