Contents - Previous - Next

This is the old United Nations University website. Visit the new site at http://unu.edu

Operational conflicts of food aid at the recipient level: those who know don't plan and those who plan don't know

María A. Tagle
United Nations University World Hunger Programme, Tokyo, Japan

INTRODUCTORY NOTE

For several years, the author worked as Team Leader of a project whose purpose was to study the acceptability of a non-traditional food, experimentally added to the food basket in selected food-aid projects in Africa (Egypt, Liberia, Mali, Niger, and Senegal) and Asia (Indonesia, Pakistan, Sri Lanka, and the Philippines) (1-3). This provided the unique opportunity to observe the problem of food aid at the micro-level in nine countries: in four it dealt with institutional feeding; in four others, with the delivery of dry foods, and in one with both institutional feeding and delivery of dry foods. It has allowed the author to describe, analyse, comment on, and discuss the validity of this type of intervention. Both a case study and an overall analysis follow.

FEEDING IN RURAL PRIMARY SCHOOLS IN SENEGAL: A CASE STUDY

The Government of Senegal attaches the greatest importance to the development of education. It is particularly concerned about the low attendance rate in primary schools in rural areas as well as about the poor nutrition of the pupils. Therefore, it aimed a World Food Programme assistance project primarily at: a. increasing the attendance rate in primary schools from 36 to 45 per cent, and b. improving the health of the children through a wholesome and balanced diet.

The programme was carried out by the Ministry of Education through the School Canteen Service, where there were two nutritionists and one agriculture extension worker dealing with school feeding. The food assistance was intended to cover 13,000 primary school children from 6 to 15 years of age (in practice ages ranged from 6 to 16) who lived more than one kilometre from school; they were to receive one cooked meal at mid-day for 150 of the 160 school days per year.*

The Food Supplied

The food commodities imported as donations included sorghum, maize flour, canned meat, dried eggs, dried skimmed milk, butter oil, and pulse. However, the official agreement contained a safeguard: "Serious difficulties may be experienced in obtaining canned meat and pulses, in which case these commodities may have to be withdrawn from the food basket." 1 The nutritive value of the cooked meal provided to the pupils was set at 900 calories, 39 9 protein, and 20 9 fat. The provision of money for meeting nonfood costs as well as the small amount of fresh foods and condiments was the responsibility of the Government. In the school year 1974/ 1975, this amounted to CFAF 10 person/day (US$1.00 = CFAF 225) (4).

Sample Selection

Six schools were selected on the basis of four screening criteria: existence of a canteen; lunch programme in operation; our positive judgement of the local staff's capacity to collaborate with the tests, including interest in the feeding programme and alertness/interest in new activities; and relative accessibility by car. These conditions implied that the sample- although satisfactory for the acceptability study- was not representative of the rural primary schools in Senegal, and that whatever problems are described later were worse in other, less privileged schools.

TABLE 1. Distribution by Sex of a Sample of Beneficiaries of the School Lunch Programme in Three Different Regions of Senegal

The schools selected were located in three different regions: two in the coastal area of Cap Vert, two in the savannah area of Diourbel, and two in the forest area of Tambacounda. In each of the six, all the children participating ; in the school-lunch programme were included.

Description of the Sample

The distribution by region, age, and sex of the 290 children in the sample is shown in table 1. The proportion of boys was higher by far in all regions. Attendance of girls averaged 20 per cent- higher in the coastal areas than in the savannah region, and only 8 per cent in the forest area, where the people are still very much attached to their traditions and therefore less interested in girls' education.

The main ethnic groups ;in the sample were Serere, Ouolof, Toucouleurs, and Lebon; Peul, Bambara, and other ethnic groups were also present. Although theoretically they were to be selected because they lived at a distance of more than one kilometre from the school, the criteria for selection used in practice were more drastic. Children belonging to the poorest families and/or living very far from the school - more than eight kilometres- were the ones actually selected.

The children were also asked about their fathers' occupations and the number of persons in the family. The answers to the first question were not tabulated because in many cases it was very difficult to obtain and the validity of the answers was dubious. However, it is important to point out that most of the answers indicated unskilled labour. All children came from large families; the mean family size per region is shown in table 2. The largest mean family size was found in Diourbel and might reflect a higher Serere component of the sub-sampie. Some children were not able to provide this information; therefore, the total number is only 233.

TABLE 2. Mean Family Size in the Sample, by Region

FIGURE 1

FIGURE 2

Figure 3

Figure 4

Figure 5

Figure 6

Nutritional Status

The children were weighed and their heights were measured at the beginning of the study. The ages were obtained from records of pupils' birth dates in the school, or- when records were not available- inquiries about birth date were made to the pupil, the mother, or neighbours, whoever seemed most reliable. Figures 1 and 2 show the boys' weights and heights by age compared with percentiles 50 and 3 from Harvard standards (5). The difference in the number of cases appearing in figures 1 and 2 and the numbers in table 2 are explained by absenteeism or because the precise ages were not available. Because the number of girls was so small, their heights and weights are not shown.

TABLE 3. Percentage of Boys with Weight or Height under 50 Percentile and 3 Percentile (Harvard Standards), by Region

Information on the nutritional status of the populations living in the Sahel is very scarce, and an effort is now being made to collect, review, and collate whatever surveys and/ or studies are available (6,7).

It may be seen (table 3) that both weight and height were impaired in an important part of the sample, but the percentage of boys with weight deficit is higher in all three places than the percentage of boys with a height deficit. Stunting of growth and emaciation were both seen, and the latter was common.

From this information, it is clear that the school-lunch programme was fully justified. In all schools, the number of qualified potential beneficiaries was higher than the capacity of the canteens that could serve only 40 to 50 lunches.

Operating Conditions of the Canteens

The School Canteen Service in Dakar had two drivers, but no vehicle was available for field work and we were told that there had been no regular visits to the school canteens from the main office during the previous two years. Supervision of six schools with a total of 290 children participating in the school-lunch programme implied at least seven working days. For monitoring the "ceiling" figure of 13,000 beneficiaries, an important part of the staff's time needed to be spent in the field. The problem of lack of transport affected both the supervision as well as the distribution of food itself.

In each region there was an Inspector of the Ministry of Education and also a person in charge of the canteens. The food was distributed from the capital to the regional storerooms and then to local areas where storage facilities either did not exist, or were in very poor condition. Where none existed, the food was stored in a room at the school or even in a bedroom of a private house. The director of the school was usually also the manager of the canteen. Some schools had only one teacher who took care of everything, including the canteen.

In most villages, there was a comité de gestion comprised of local people who took care of the principal problems of the village and also helped the director in managing the canteen. According to our experience, such a committee existed in five out of six schools sampled, and in two the committee was very active and interested.

There was no defined or closed space for cooking and eating in the school area. All but one of the kitchens visited consisted simply of a big pot over an open fire in a corner, or any other place in the schoolyard, either under a. roof or a tree.

The cook was an illiterate village woman. It was not easy to recruit someone for cooking because the salary was very low- 1,500 to 2,000 CFA/month (in that period, US$1.00 equalled CFA 225). This was one of the reasons why sometimes the wife of a teacher or the wife of the director did the cooking. It was not unusual to see the cook with a baby on her back and one or two pre schoolers in the vicinity of the boiling pot.

The diet was very monotonous, the same meal being served every day in some schools. In others there was some variation, though always limited because money to buy fresh ingredients and condiments was scarce. Basically, the lunch consisted of cereal boiled in water, either maize flour or rice, a bland dish having little flavour; other foods- whatever was available- went into the sauce that was served over the cereal. The sauce provided all of the appetizing ingredients and also helped the children to swallow the bland food. The quality and quantity of ingredients for the sauce were often grossly inadequate, as can be seen in the photograph just above.

Regional recipes with instructions on how to utilize the donated ration, tailored to regional food habits, were available, but few canteen managers and none of the cooks were familiar with the recipes. This is understandable because of poor communication between the people at the central level and those at the regional level, and again, between those two groups and the individual canteens. Written material sent by post is not useful for illiterate people.

Food distribution took place on the sandy ground and the children went to eat under any available shade. Generally, the meal was served in big bowls or basins and the children ate with their hands, four to 12 around a bowl. The children seemed very hungry, and if the smaller children left anything, it was eaten by the older boys.

Although the school year started in October, none of the canteens was yet functioning at the beginning of November. Theoretically, the number of meals served was to be five per week, but we found that three out of six canteens provided only four meals a week. On a surprise visit, we found that five out of six of the schools in our sample served no lunch at all, although there was no serious reason why the canteens were not functioning.

These observations indicated that the functional conditions left a great deal to be desired, and there was a general impression of poverty, unsatisfactory hygienic conditions, and insufficient knowledge of how to handle foods. The food programme was one of the principal incentives for pupils to attend school, and after days when no lunch was served, an immediate decrease in school attendance was observed.

The principal complaints heard during the visits were: lack of water; delay in the arrival and distribution of the food; insufficient quantity of food; delay in the arrival of the supplementary money; insufficient amount of supplementary money; lack of equipment, and difficulties in recruiting and keeping cooks because of the low salary, etc.

All the complaints implied serious problems with the programme, both in terms of its time sequence and in quality. It must be stressed that complaints came spontaneously. They show- once more- that people at the local level are well aware of the difficulties and bottlenecks, and they are willing to comment on them to outsiders, with the hope of being helped to solve them. In other words, they enjoy any possible opportunity to express themselves about their daily life problems and are prepared to co-operate in looking for solutions.

The lack of water deserves special comment. Some years ago, the school canteens were conceived of in combination with school gardens; i.e., the gardens were to provide an opportunity for training the children to grow vegetables, which, in turn, would be used in the school lunch.. The idea was very appealing, but those who planned did not realize that lack of water in rural areas in Senegal is a very serious problem, and the school gardens were doomed to failure from the start. In our tour through schools, we never saw a school garden, although gardening tools were still available in some schools.

Size of Portions

The theoretical individual portion would supply 900 calories and 39 g of protein per lunch, but it was clear that the portions were small and many times too diluted.

In carrying out the acceptability test, in each canteen we estimated the mean daily individually-consumed portion and the overall average individually-consumed portion for the full test period. The estimations gave the following results: 350, 551, 403, 358, and 647 wet weight/day/ pupil in five of the six canteens. Assuming that cereal-based cooked meals contain at the most one-third of their total weight in solids and two thirds water, the above wet weights reflect that dry contents were too low, and this multiplied by the caloric density factor gave results far from ideal for energy intake.

On some days we observed preparation of the lunch, including the quantities of the different ingredients used and the cooking method. This allowed us to do a rough estimation of the energy and protein supplied per lunch per portion. In the "model" canteen, the values were between 750 to 790 calories, and 16.5 and 18.2 g protein, i.e., 86 per cent of the theoretical supply of energy and 45 per cent of the protein.

Portions smaller than the theoretical ones could be due to: a. Losses throughout the process, from the time the food arrived in the area to the time it reached the consumer. Losses resulting in a specific canteen were highly probable because of general poverty and lack of storage rooms. Selective losses explain the low levels of protein referred to above. In one instance, we observed opened cans of meat whose contents were not added to the pot for the children. b. Extra people eating from the pot of cooked food prepared for the children. This was highly probable: for example, the cook, having spent all morning working for the canteen and having no time to cook for her own family, might take out extra portions. Six to eight portions taken from the pot prepared for 40 to 50 would immediately imply a decrease of up to 20 per cent in the mean individual portion. c. Insufficient food available locally. Our impression was that the "stored" food available on the spot was always in relatively or absolutely small amounts; it seldom happened that people in charge of feeding programmes tried to stretch the food resources, especially when they were not sure about the arrival of a new batch.

ANALYSIS OF PROJECT PERFORMANCE

In order for the project to help achieve the objective of increasing the attendance rate in primary schools, the school lunch should have been (a) made available to all the pupils, and (b) made available regularly every school working day. Neither condition was fulfilied. In all the schools, there was an unsatisfied demand for the lunch. A good school-lunch programme could have at least improved the attendance rate of the beneficiaries.

The improvement of the children's health through a wholesome and balanced diet (objective two) could be attainable if there were a way to make sure of the quality, quantity, and regularity of the school-lunch programme. In many poor families the member of the family who obtained some food from other non-family sources was considered the privileged one. Many times, when food was distributed in the home, the pupil in the school-lunch programme undoubtedly was in an unfavourable position with respect to his/her brothers and sisters. This fact of apparently world-wide validity has been commented on elsewhere (8).

Moreover, the programme was not carried out regularly in any of the schools of the sample; i.e., many days the children had taken the long walk to and from school, yet were fed less at home even when no lunch had been provided in the school.

Limited anthropometric measurements in one school confirmed that the indicators of nutritional impairment were as poor at the end of the study as at the start of the programme. We did not discover any type of organized control or any system to monitor the fulfillment of the objectives of the project. The objectives stated in the project document turned out to be no more than expressions of good intentions.

Common Problems of Food Aid at the National Level

Many of the problems faced at the micro-level were found to be common in different types of food aid projects in the eight other countries visited. They result from: a. a national infrastructure with only limited capacity to absorb food aid, and no guarantee that the food will reach the intended consumer; b. inadequate planning capacity that does not take the lack of infrastructure into account.

The following problems were found- in greater or lesser degree- in most of the countries we visited.

1. Lack of awareness of the peculiarities of the beneficiaries and their settings

There was generally ignorance of the country's social and physical conditions, underestimation of regional characteristics and differences, and of many other practical aspects, such as the state of the roads throughout the year, that affect the implementation. This deplorable ignorance results in unrealistic and unfeasible projects, created out of desk-work without proper background knowledge. This ignorance is found in both internationally-recruited personnel as well as high-level national officers. Very few of the national personnel have an in-depth knowledge of their own countries.

The relevant information concerning local foods, food habits, taboos, and other cultural restrictions, either are unavailable in the country- for example, for French speaking African countries, the best sources of information are the university libraries in France- or, when available, are not sufficiently taken into account in the planning process.

Misinformation may be even more dangerous than ignorance. In our experience in some countries, we received information from officers of the upper hierarchy that was found to be wrong as soon as we went to the field.

2. The passive role of the so-called beneficiaries and the people directly related to them

In the process of planning food aid, the beneficiaries are considered only as potential recipients, and therefore, whatever aid results does not necessarily reflect their aspirations or their needs. For instance, in the case of Senegal, I am sure that if some canteen managers and committees (comités de gestion) had been given the chance to express their needs, both at the initial stage of planning as well as during implementation, a much better project could have been developed.

Here, also, it should be mentioned that in many places we observed packed cooking equipment, donated by an international organization, that had never been taken out of the original containers. Was that equipment appropriate for the setting? Was it delivered with instructions? To whom? Was it really delivered with a purpose, or just delivered? In one case, we saw a container with a potato-peeling machine in an area where there were no potatoes!

3. Personnel

The number of professionals in charge of food aid at the national level may go from a small number- two to three - to a full bureaucratic structure, as in the case of Pakistan. However, neither the number nor quality is strictly related to the success or failure of a project. The national infrastructure coping with food aid is the bottleneck. Personnel need to be trained in logistics and supply needs according to the conditions of the country because they generally lack this capacity.

When food aid consists of delivery of dry foods, the final operation is more an activity of disposal and record-keeping, which is not very time consuming because it happens periodically. Therefore, according to our observations, it presents few serious problems of personnel. However, when the food aid is delivered in the form of cooked food to schools or hospitals, the personnel problem becomes important, particularly if there is illiteracy and inability to follow written instructions, to write menus, or to keep records, and if salaries are poor, implying lack of stimulus to go to work, absenteeism, and high turnover.

4. Transport

Transportation problems make it difficult to distribute the food and to control use/misuse of commodities in all the countries visited. The problem is aggravated by the long distances that have to be covered. Other aspects to be considered include: lack of roads, as is the case in areas of Agadez (Niger), where camels are used to transport the food commodities; blockage of roads because of rain, floods, etc., that we happened to experience in Senegal and Liberia, and other vicissitudes of climate and weather.

5. Storage

With one exception - the storeroom of Solo, Central Java, Indonesia, which could be considered a model- in all of our experience, the storage conditions were either mediocre or deplorable. It was easily seen that people in charge of storerooms lacked even minimal knowledge with regard to food-handling and the stacking and general care of foods. Fumigation was considered to be an unnecessary foreign practice, but even if it was practised, it was not done regularly. In general, different types of insects bred inside the storerooms and rat and mice feces were seen. At the village level, in some areas, we observed a fatalistic attitude regarding rodents and insects: they come as the storm or earthquake comes, and therefore cannot be avoided; also, there are good and bad years, i.e., with fewer or greater problems of this type of food waste.

6. Losses

Apparently, losses occur frequently throughout the chain of operations before the food reaches the consumers. It is not exceptional to find food aid commodities being sold in the markets. When there are only a few units, it may be that individual beneficiaries are selling their dried food rations, but when a shop has a stock of such commodities, it is probable that the food aid was misdirected intentionally. The question that arises is: Is the recording and monitoring system adequate?

7. Inadequate packaging of food commodities

Packaging that is good according to developed-country standards does not necessarily prove to be so in tropical areas, nor is it resistant to the storage conditions already described.

At the beginning of the rainy season in Monrovia, Liberia, 90 per cent of the remaining fish protein concentrate, stored in the central storeroom for food aid, was lost because the product was fully infested by Dermestides and had to be burned. A three-layer laminate (9/1,000 mm aluminium foil, 60 9 paper, and 20 to 25 9 polyvinyl-idenchlorid) considered adequate in Europe, failed completely in Liberia. Also, the seals of the sacks produced with the above-mentioned material or with similar material, did not resist the transport conditions of the developing world (too much bumping on the roads?) and broken seals were not exceptional.

The basic question is: Do the donors care about protecting the products they give away? If the answer is yes, then they do not know the conditions in developing countries and my first argument- ignorance of conditions in the setting where the food aid is sent- prevails. If the answer is no, then the conclusion is that donors take advantage of the political and economic openings that food aid provides in the international scene today, without caring whether the "beneficiaries" actually receive the food.

8. Food commodities

In our travels, we observed that there were many commodities that were not at all suitable for the potential consumers, either in quality or quantity, or both. To illustrate this point, canned fish was seen piling up in storage in north and west Sudan (1977) in areas where fish is not eaten. Fish reminded some people of lizards, which were not regarded as food fit for human consumption (2).

In the storeroom at Agadez, Niger (19751, which distributes food aid to the nomadic primary-school children, a stock of fish was seen, although Sahelian nomads do not know fish and refuse to eat it. In the same place, another curious commodity was found: out-dated emergency rations intended for a northern European army, consisting of small blocks of highly caloric, pressed food. This was distributed as snacks and the children ate it; however, it was far from ideal under conditions of hot temperatures and scarce water.

Fruit purées (peach, apricot, etc.), intended as baby foods in the developed world, arrived in Niger as food aid, and it was presented as a dessert in a school for girls in Niamey (1975). The idea of dessert was strange to them in the first place, and worst of all, fruit purées were completely unknown to them.

In Central Java (1976) we found European fish canned in brine as part of a food-aid ration. Unfortunately, the can contained 460 9 comprising 290 9 of fish and 170 9 of brine, and the individual ration was calculated on the net contents of the can, not on the fish content. This example makes one wonder how much water is being canned by developed countries on behalf of the underdeveloped ones.

9. Distribution

From the central stores to the places of operation, distribution is seriously affected by the transport problems referred to earlier, implying uncertainty about the dates of arrival of commodities, frequent delays, and general lack of regularity.

10. Distribution to the beneficiaries

Rations received by the beneficiaries present different problems, depending on whether the project deals with cooked food or with delivery of products. In the first case, a very common situation is that the cooked meal should consist of the "theoretical" individual ration plus fresh foods locally supplied; but in practice, the second component is either small or nil, and the practical portion is smaller than the "theoretical" one.

Although in general the beneficiaries received less than expected, situations of excess also existed. In the years of drought, the Sahelian countries received so much food aid that in some places the local food distributors gave it away in indiscriminate ways. This was observed in Mali and Niger in 1975; in one school in Niamey the children were served portions of more than 1.1 kg of cooked rice each day at lunch time (1).

In the case of dry-food delivery, the items were not necessarily the same at each distribution, quantities varied, and intervals between two distributions were also inconsistent, i.e., once every three to four months, or once or twice in one month. In other words, "beneficiaries" receiving foods as part of their salaries could not count on what kind of food, how much of it, or when they would receive it.

A food for work project might have a target expressed in man days per year, or a given number of workers constantly employed. However, because of high unemployment and strong competition for jobs, the available man-days were spread out over a much larger number of workers, implying that the majority did not gain full employment or the full quota of food for the family.

11. Conditions of the kitchens

In general, kitchens ranged from mediocre to deplorable, with unsatisfactory hygienic conditions being the prevailing characteristic, followed by insufficient knowledge of food handling and scarcity of cooking equipment. Even within the limitations imposed by scarce economic resources, everywhere there was tremendous room for potential improvement.

12. The beneficiaries

In general, beneficiaries are poor people. However, in some countries, the selection of the beneficiaries was not so clear. For example, in the years 1976 - 1977 in Pakistan we found that up to 17 per cent of the families could be classified as middle class; in the Philippines we found that 12 per cent of our sample belonged to the medium or well-to-do strata; and in Sri Lanka and Egypt these figures were 17 and 12 per cent, respectively. In the hospitals in Mali (1975) the "first-class" patients also ate donated food. At the grass roots level, the delivery of foods was sometimes used as a tool for reward or punishment; i.e., as a means the local distributor(s) may use to gain power over the recipients. On the other hand, people having "different" ideas may find themselves off the list for food distribution without any apparent reason. This fact was evident in Pantabangan in the Phillippines and in Uda Walawe in Sri Lanka (both in 1976).

In Monrovia, Liberia (1975), there were two schools where food prepared with food-aid commodities- plus other ingredients- was sold to students. No doubt, this was locally a well-known fact and should have been based on a special agreement. However, the nutritional intention of the programme was altered because the most needy could not purchase any food.

Suggestions for Improvement

Many specific recommendations are suggested by the preceding descriptions. They are self-evident, and most of them have been well-documented elsewhere (9). However, a few suggestions have not yet been developed sufficiently or at all.

1. Governments should make sure that at least all officers who have responsibilities for planning and programme designs in all ministries understand the country and its population characteristics adequately. This awareness is a national responsibility and the negotiation of an agreement is a national prerogative that, if fully used, may bring about important advantages.
2. Because recipient developing countries have limited capacity to absorb it, food aid without support for building and/or reinforcing infrastructure is doomed to be used inefficiently. International food aid should always be accompanied by significant economic resources for (i) training of national personnel at all levels, and (ii) material expressions of infrastructure such as construction of wells, improvement of kitchens, building of storerooms, etc.
3. The decision-makers of the recipient countries should take into account the potential beneficiaries or their representatives as participants for nutrition programme planning decisions, and later, the agencies in charge of implementation should take the beneficiaries into account for monitoring purposes.
4. The governments of the donor countries and/or their representatives should also consider the potential beneficiaries and respect their needs, preferences, and values.

REFERENCES

1. M.A. Tagle, "Acceptability Testing of FPC Type B," Rome, FAO, Norway Funds-in-Trust, Fl:TF/INT 120 (NOR) PHASE I (1976).

2. M.A. Tagle, "Acceptability Testing of Fish Protein Concentrate (FPC) Type B," Rome, FAO,, Norway Funds-in-Trust, Fl:TF/INT 120 (NOR) PHASE II (1978).

3. M.A. Tagle, I. Ofstad, A. Smits, and K. Vesterhus, "Factors Affecting the Acceptability of Fish Protein Concentrate (FPC) Type B in Food Aid," Food and Nutrition (FAO) vol.. 6 (1980, in press).

4. World Food Programme, Intergovernmental Committee, Project Summary: Senegal 597. Feeding in Rural Primary and Training Schools, Rome,WFP/IGC: 17/9, Add. 12. (1970).

5. W.E. Nelson, V.C. Vaughan, and R.J. McKay, Textbook of Pediatrics (9th ed., Saunders, Philadelphia, 1969).

6. J.C. Dilion and N. Lajoie, Rapport sur l'Evaluation de la Situation Nutritionelle des Populations Rurales du Sahel a la Lumiére des Enquêtes Effectuées entre 1960 et 1979, IDRC Manuscript
Reports (January 1980).

7. Proposal for a Joint IDRC/UNU/IUNS Workshop on Food and Nutrition in the Sahel, to be held in 1980.

8. Universidad de Chile, Alimentación del Escolar, Departmento de Extensión Universitaria y Acción Social, (Santiago, 1971).

9. J. Austin, M.A. Anderson, R. Goldman, J. Heimendinger, C. Overholt, D. Pyle, B. Rogers, J.D. Wray, and M. Zeitlin, Nutrition Intervention Assessment and Guidelines. Harvard Institute for International Development, Cambridge, Massachusetts (1978). Submitted to United Nations ACC Sub Committee on Nutrition.

Contents - Previous - Next