BRT in Dar es Salaam

BRT in Dar es Salaam

Rendering of proposed BRT in Dar es Salaam, showing its close resemblance to Bogotá's Transmilenio (by Luc Nadal for ITDP, licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 2.5 License)


Bus rapid transit (BRT) has been promoted by various international consultants and experts as a viable transportation solution for Dar es Salaam’s worsening traffic congestion.  As a transport model, it offers the potential to improve mobility without excessive investment or operating subsidies that would be inappropriate given Tanzania’s poverty.  Yet despite readily available international funding and technical expertise for BRT, the existing fragmentation of bureaucratic and transit operating structures has delayed the implementation of BRT in Dar es Salaam.  A brief overview of DART’s Phase I is below, followed by a more extensive discussion of the project’s background and potential pitfalls.

Phase I Overview

  • 2002 – The Institute for Transportation and Development Policy (ITDP) drafts a bus rapid transit proposal with the Dar es Salaam City Council, predicting it will be Africa’s first bus rapid transit system
  • 2003 – The City Council endorses the BRT proposal
  • 2005 – BRT conceptual design completed
  • 2006 – Dar es Salaam Rapid Transit Agency (DART) formed under the Prime Minister’s Office
  • 2007 – Environmental impact planning completed
  • 2008 – The World Bank approves a $190 million International Development Association credit, partly towards the BRT project; The Japanese International Cooperation Agency (JICA) completes the Urban Transport Policy and System Development Master Plan for the City of Dar es Salaam
  • 2010 – President Jakaya Kikwete officially inaugurates construction on Phase I
    • 21 kilometeters of dedicated lanes, closed system
    • 29 stations, 5 terminals
    • Cost of $125 million (including $20 million for expropriations)
    • Cash and contactless card fare collection, baseline fare of 400 Tsh and transfer fare of 100 Tsh
  • 2012 – Projected opening
Phase 1 Trunk Routes

Phase 1 Trunk Routes (from DART Investors Document), compare this to the red Morogoro Rd. lines on the diagram I made


Current Need for Transit Improvements – Congestion “Without Mercy”

From the Center for Economic Prosperity

Traffic congestion in Dar es Salaam is horrendous, the worst I’ve observed in my travels.  I spent stretches of fifteen minutes at a time sitting in stationary daladalas on Morogoro Road (see my transit map of Dar es Salaam here).  Mwinyi/Bagamoyo Road, another artery, can be even worse, since it is only one lane on either side of a poorly enforced “reversible lane” (which leads to regular games of low-velocity chicken).  Average rush hour speeds on Mwinyi Road are about 7 mph, leading trips to take three times longer during rush hour than during normal flow.  As the Center for Economic Prosperity reports, “It is undisputed that congestion is slowing down economic activities in the city without mercy.”

The costs of the city’s worsening congestion extend beyond commuters’ wasted time.  Noise and air pollution from excessive congestion adversely affect public health.  Particulate matter from old diesel engines is of particular concern, especially given a recent study’s findings that “asthma is an important clinical condition in sub-Saharan Africa [and there are] major gaps in clinical care, particularly in urban areas.”


Concentrations of particulate matter are much higher near the Tazara Intersection than at other sites nearby. (Source: 2008 Presentation on Dar es Salaam Transport Policy and System DevelopmentMaster Plan Preliminary Feasibility Study, funded by the Japanese International Cooperation Agency)

There are also more immediately tangible harms inflicted by crashes.  Pedestrians and bicyclists face hazardous conditions; the city has few sidewalks, and those that do exist are often taken over by impatient drivers trying to bypass traffic congestion.  According to the African Center of Excellence for Public and Nonmotorized Transport (ACET), pedestrians account for 69% of the crash fatalities in Dar es Salaam.

Frustrated drivers often speed past traffic on sidewalks, creating dangers for pedestrians.

Frustrated drivers often speed past traffic on sidewalks, creating dangers for pedestrians.


Causes of the Standstill

A number of factors contribute to these congestion-related ills.  Poor planning, a lack of hierarchy in road sizes and uses, and underinvestment have severely curtailed the efficiency of the city’s road network.  Though there are traffic signals at some intersections, they are not adjusted to actual traffic conditions, so police officers manually direct traffic during peak periods.  The lack of available capital for transit operators leads them to use inefficiently small vehicles, increasing congestion.  Though rules prohibiting the smallest (16 seat) daladalas from entering the central business district have eased congestion slightly, relying on 35 seat minibuses still leads to unnecessary congestion.  For example, a 2007 traffic survey found that daladalas entered the central business district via Bandari Road more than 2300 times during one day.  To the north, about 1300 daladalas were counted on Maktaba Street (on the approach to New Posta) between 7:00 and 9:00 AM, accounting for half of the vehicles along that stretch of road.

The discomfort associated with these small vehicle also contributes to peoples’ desire to purchase and commute in their own private vehicles.  Car buyers can now order used vehicles from Japan from websites like Autorec.  The reconditioned vehicle will arrive at the Dar es Salaam Harbor a few months after ordering online. Strict emissions and safety standards in Japan create disincentives to driving older vehicles, so there is an abundant supply of affordable used vehicles that can be shipped to Tanzania.

To gain a wider perspective on these issues, I met with a retired infrastructure/road sector leader for the World Bank’s resident mission in Dar es Salaam.  In his view, the congestion I observed in Dar es Salaam was illustrative of similar problems across Africa; indeed, places like Luanda, Angola, have even worse traffic than Dar es Salaam.  He told me, “The big challenge for the world is cities in Africa,” since they face the challenge of confronting rapid growth and simultaneous poverty.  These cities have the opportunity to avoid repeating the automobile-centric mistakes of the United States by alleviating standstills while avoiding road expansion projects that will be eventually be overwhelmed by induced demand effects.

Currently, Dar es Salaam is home to about 30 registered motor vehicles per 1000 people.  The number of vehicles in the city is expected to grow by a factor of five in the next decade.  While the transport infrastructure is already totally insufficient for the current number of vehicles, expanding roadway capacity alone will only encourage increased use of private cars.

Working Towards Effective Institutional Structures

Institutional capacity is also insufficient to handle Dar es Salaam’s pressing mobility concerns.  Currently, fifteen separate agencies are in charge of various aspects of transportation policy within the city.  And, as a report commissioned by the Ministry of Infrastructure and Development states, “There is no coordinating mechanism that brings together all these institutions together to form a common set of goals and strategies that enable transport planning and regulation to be undertaken in a cohesive and integrated manner.”

Plans have been developed for a Dar es Salaam Urban Transit Authority (DUTA) that would integrate the various existing institutions, and international consultants have stressed the importance of such integration.  For example, the Japanese International Cooperation Agency (JICA), at the request of Tanzania’s national government and Dar es Salaam’s city council, provided technical assistance in the preparation of a transport master plan study between April 2007 and June 2008.  The study strongly backed an urban transit authority that would address the current fragmented regulatory and planning structure.  It was optimistic about previously completed BRT planning and the Dar es Salaam Rapid Transit Agency (DART), then recently established by the national government.

Particularly relevant in understanding the role of DART is that the Establishment Order outlines a semi-autonomous and commercial role for DART to manage the transport system role on a day-to-day basis; through a commercially viable operation, effective management to deliver quality service, operational efficiency and financial performance (surplus). These objectives are contained in the performance criteria set out in the charter to be monitored by the Ministerial Advisory Board. Its semiautonomous status implies that it is not subjected to direct political pressure or influence other than the strategy policy formally set by the Permanent Secretary and the Ministerial Advisory Board.

(Dar es Salaam Transport Policy and System Development Master Plan, Technical Report 3: Institutional Reform of Urban Transport Management)

According to DART’s website, its “vision is to provide a modern public transport system at a reasonable cost to users and yet profitable to the operators using quality, environmentally friendly, high capacity buses that meet international service standards and operate on exclusive lanes, reducing travel time.”  The agency’s ability to realize its vision may be compromised unless a larger coordinating authority like DUTA is actually created.  Technical and operational designs for BRT have been readily available in Dar es Salaam for years;  the limiting factor seems to be inefficiency in the existing bureaucratic system.

DART itself is intended to be an umbrella agency overseeing separate companies tasked with bus operations, fare collection, and fund management (see figure below).  This setup closely mirrors those of BRT systems, such as those in Bogotá (Transmilenio) and Santiago (Transantiago).

DART Institutional Structure

Proposed DART Institutional Structure (from DART Investors Document)


Global Expertise – “A South-South Collaboration”

The close resemblance, in both physical and institutional infrastructure, between the proposed DART system and Transmilenio is unsurprising given the extensive involvement of Transmilenio officials in this “south-south collaboration,” as the retired World Bank expert described it.  Both Enrique Peñalosa, the mayor of Bogotá from 1998-2001 who implemented Transmilenio, and Edgar Sandoval, Transmilenio’s first manager, visited Dar es Salaam to assist with conceptual planning (see the picture of Peñalosa in Dar es Salaam on page 96 of ITDP’s BRT planning guide).  Brazilian consulting firm Logit, which worked on both Transmilenio and Transantiago, completed the operational planning for DART.  As Onésimo Flores writes in The Transmilenio Recipe (in Spanish), “Its promoters have been so successful that the ‘Transmilenio model’ is now a true export product, as is demonstrated by the growing list of cities in the process of implementing some closed corridor inspired by this Colombian experience.”

International financial entities have been enthusiastic about supporting this new South American “export product.”  At a meeting with DART in March, HSBC expressed interest in financing buses and fare collection, making use of its “expertise in Bus Rapid Transit (BRT) projects and accessing financing through involving Export Credit Agency (ECA) at affordable cost.”  The World Bank has provided $190 million in International Development Association credit to Tanzania, part of which will cover the construction costs for DART’s Phase I.  As reported in the Daily News, the World Bank’s top official in Tanzania, John McIntire, “said that Tanzania made the right choice when adopting DART system [and] that the country has emulated other mega-cities in the world that had introduced the similar systems.”

At a certain point, however, emulating other mega-cities and following the “recipe” can distract from important local concerns.  At the August 2010 inauguration of DART construction, Tanzania’s president referenced local concerns about the international nature of project financing somewhat dismissively:

“This is an opportunity for local investors, particularly the private sector, to team up and import quality buses that will operate once project construction is over. Don’t just remain idle complaining about the arrival of foreign investors,” he said.

The president thanked the World Bank for donating funds for the project and asked for more support as the country implements other city-based road projects.

BRT is trending globally, and the term has propagated rapidly without being clearly defined.  By indiscriminately aggregating a wide range of transportation projects in the nebulously defined category of BRT, planners risk overlooking fundamental differences in operating characteristics and contexts.   The 2009 Miami junket of three DART officials is illustrative:

The three DART officials who benefited from the program were Cosmas Takule, DART Chief Executive Officer, Enoch Kitandu, Director of Systems and Operation, and Peter Munuo, DART Operations Manager…

According to a statement from the U.S. Embassy in Dar es Salaam, Miami was chosen for this two-day professional program because of its well-planned and extensive “BusWay,” and its transportation system through various municipalities and that connects with adjoining counties. The 20-mile stretch of the Miami BusWay is the United States’ longest bus rapid transit line. The Tanzanian officials had a tour of several of Miami’s ethnically diverse neighbourhoods and rode the Metro mover which operates in a loop around downtown Miami. They also had an opportunity to meet with Miami’s chief transportation engineer to discuss environmental considerations relevant to the Dar transportation project…

Takule, who lead the Tanzanian delegation, reported, “The tour of Miami’s beach promenade gave us a chance to note that the paths along the beaches are potentially lucrative recreational areas if properly developed and cared for. As such promenades attract local and foreign tourists, they are a catalyst for other services such as restaurants, hotels, shops, marine transport and sports. The Dar es Salaam coastline has all the same potential develop its beach in the likeness of the Miami Beach promenade.”

“All said, the trip was an eye opener, educational, challenging as well as highly interactive with professionals in the Bus Rapid Transit industry worldwide. Every occasion was an opportunity to learn through observation, listening, dialogue and Q&A in an effort to get as much out of the trip as possible. The knowledge we gained will surely be reflected in better planning, designing and operations of the DART system,” he concluded.

The South Miami-Dade Busway operates in a former rail right-of-way, while DART’s Phase I will operate in the median of existing roads.  Miami-Dade County has a monopoly on public transit, while DART will have to negotiate with thousands of smaller operators.  Miami’s Busway feeds into a heavy rail system (at the Dadeland South Station), while Dar es Salaam lacks rail rapid transport.  Socioeconomic differences between the two cities are profound, yet these can be glossed over and politicians can justify such an international trip using the BRT label.  “Miami’s beach promenade” (which is miles away from the busway) can offer little insight into addressing the pressing institutional and bureaucratic issues that are delaying DART’s implementation.  But perhaps seeing how some portion of this Brazilian/Colombian recipe can be implemented both in developed Miami and developing Dar es Salaam can encourage policymakers to move forward in addressing the unique institutional and implementation challenges they face.

Project Delays

Phase I of Johannesburg’s Rea Vaya BRT (more soon) was opened in 2009 after less than three years of planning and construction, despite the minibus industry’s violent opposition.  If Dar es Salaam’s system indeed opens in 2012, as is now planned, the project will have taken ten years to plan and construct.  Given the extensive technical and financial support described above, the reason for such a protracted implementation schedule is clearly not related to engineering or funding.  Project delays center on governance issues.  Expropriating property for bus terminals and depots is currently a hurdle.

In February, the Minister of State in the Prime Minister’s Office, Regional Administration and Local Government toured DART construction sites and voiced his concerns about expropriations:

In order for the DART project to be implemented faster than it is now, the Minister requested the people affected by project to collect their compensations rather than keeping on going to court to demand amounts of compensations whose evaluation standards do not adhere to the laws of Tanzania.

He warned those who are still nursing the idea of resisting to vacate the sites earmarked for the project on baseless grounds including ballooning of compensations and injecting ghost names in order to get illegal money. He said the adamant in this case will only force the government to apply force to evict them instead of using negotiations.

The Minister of State blames residents whose properties are being taken by eminent domain for making unreasonable demands, but government officials themselves have reportedly been equivocating about compensating these residents:

[City Council Director] Mr Kingobi said it was not the obligation of the city authority to compensate city residents, whose properties will be demolished to pave way for the project, saying: “It’s the government responsibility to pay compensations – not the city authority.” He added: “In fact we are also longing for the project to take place, believing that it will easy transport woes in the city.”

Last week, [National] Minister of Works Mr John Magufuli expressed concern over the delays by the city authorities in effecting compensations to city residents where the project will pass.  According to Mr Magufuli, the World Bank (WB) has already disbursed the project fund and contractors are in place ready to start the project, but dilly dallying in paying compensations is halting the project initiation.  He warned that if the city authority will not take immediate measures to pay compensations, his ministry will start to demolish places where the project is expected to pass.  The Dar Rapid Transit Agency Chief Executive Director was not immediately reached to explain the quandary on who should pay the compensations to allow the project implementation.

DART had authorization for its Resettlement Action Plan in 2008, but delays in compensating property owners are still holding up progress.  Clear lines of responsibility are lacking, and this deficient institutional structure will likely cause further delays.

Integrating existing operators with the new system may also lead to delays.  BRT has the potential to serve as “a means toward transit system regulatory reform,” but policymakers in Dar es Salaam are unclear on how BRT will or should affect the livelihoods of the city’s 3100 daladala owners.  At my meeting in the World Bank office, the transport expert asserted that the challenge was to “somehow make them involved in the management of the new system.” Dr. David Mfinanga, a professor associated with ACET who researches daladalas and is also a member of DART’s Board of Directors, told me that the government needed to address the sector’s “fragmented ownership,” since daladala operators will be a part of the nation’s transportation for a long time to come.  He described different scenarios of integrating daladalas with BRT, both as DART employees and as feeder services, and he cautioned that expecting them to form companies and compete in an international bidding contest is “insulting to them.”

Even if the city hires management consultants chosen by the minibus sector, as Johannesburg did, negotiations with daladala operators will continue decades into the future.  As discussed in The Transmilenio Recipe, ten years after the implementation of Transmilenio, three-quarters of passenger trips in Bogotá are still made with the old colectivo service:

There are many Mexican mayors and governors who promised or are now constructing BRT corridors.  Many of them made the decision after traveling to Bogotá.  They should understand that the inauguration of a BRT line represents just the beginning of the road and not the goal, an ingredient and not the complete recipe.

Political work is far more challenging and fundamental than new buses and fancy stations.  In short, the thousands of Tanzanians who are “[a]waiting Dar es Salaam rapid transit elatedly” will have to be content waiting a while longer.

Dar es Salaam Transit Diagram

Schematic map of daladala routes in Dar es Salaam

Schematic map of daladala routes in Dar es Salaam (click to enlarge)

There are approximately 250 different daladala routes crossing Dar es Salaam, and depicting all of them on a schematic map is a challenge.  Routes and vehicles are color coded according to the major arterial streets over which they travel.  There are only seven such streets, and I used the corresponding seven colors in the diagram.  Different combinations of these arterials, as well as different origins and destinations along them, account for the hundreds of different routes. A high quality pdf version is available here.

"The Neglected Epidemic"

At the beginning of April, the US Department of Transportation estimated that traffic fatalities in the United States declined to the lowest rate (per vehicle miles traveled) since 1949. While it’s somewhat disturbing that the loss of 32,788 lives can be reported as good news, cars are indeed getting safer in the developed world. A number of factors, including increases in the prevalence of air bags, drunk driving enforcement, and seat belt usage, help explain this downward trend.

Unfortunately, traffic fatalities are skyrocketing in lower income countries (see this figure). An article from the Harvard Center for Population and Development Studies lists four factors contributing to such high injury and fatality rates in lower income countries:

  • The growing number of cars – In Tanzania, car buyers can order used vehicles from Japan online from sites like Autorec; the purchased vehicle will arrive at the Dar es Salaam Harbor a few months later. Strict emissions and safety standards in Japan create disincentives to driving older vehicles, so there is an abundant supply of affordable used vehicles that can be shipped around the world.
  • The high number of passengers per vehicle, especially in buses and minibuses
  • Poor enforcement of safety regulations – Widespread corruption means that it is cheaper for drivers and owners to pay a bribe for traffic or maintenance violations than to obey the law.
  • Inadequate health care systems.

The authors of this article go on to explain how socioeconomic status influences the breakdown of traffic injuries:

The choice of mode of transport in developing countries is often influenced by socioeconomic factors, especially income. In Kenya, for example, 27% of commuters who have no formal education were found to travel on foot, 55% usually used buses or minibuses, and 9% used private cars. By contrast, 81% of people with secondary level education or above usually travelled in private cars; 19% travelled by bus, and none walked. People with little formal education earn low incomes. For them, the affordable means of transport are walking, travelling by bus or truck, or cycling—all of which expose them to high risks for road traffic injuries.
People in developing countries are frequently aware of these risks. A regular commuter on the buses in Lagos, Nigeria—which are referred to locally as danfos, “flying coffins,” or molue, “moving morgues”—said, “Many of us know most of the buses are death traps but since we can’t afford the expensive taxi fares, we have no choice but to use the buses.”…
The injury profile for road traffic crashes in developing countries differs in important ways from the profile seen in developed countries, and it can provide guidance for making policies to improve prevention and control. Protection is needed for three main vulnerable groups—pedestrians, who in urban areas constitute up to 70% of the fatalities; passengers com­muting on buses, trucks and minibuses, who constitute the next largest population group affected; and cyclists.

Traffic fatalities are a leading cause of death in South Africa, as Mobility Magazine reports:

In the 2007/2008 research year, between 14-18 000 people were killed on the roads in South Africa (18 487 were murdered in that same time period). There are approximately 800 000 crashes per year — we’re the third worst in the world per capita. Road deaths are the second highest cause of accidental death in the 0-18 age group, and 22 people become permanently disabled every day. Our driving behaviour costs the country about R47 billion a year.

The Mobility article goes on to cite the recommendations of Dr. Marianne Vanderschuren from the University of Cape Town’s Centre for Transport Studies. The solution to road crashes is a combination of “education, enforcement, engineering, environment, encouraging innovation, and evaluation.”

In Tanzania, there were 10,000 traffic fatalities in 2002, amounting to 3.0 deaths per 10,000 people. In that same year, there were only 140 registered vehicles per 10,000 people. For every 10,000 people in the United States, in comparison, there are about 8,000 registered vehicles but only 1.1 traffic fatalities. In other words, per capita, the United States has 60 times the number of vehicles that Tanzania does and only 1/3 of the traffic fatalities.

A number of groups are trying to address this neglected epidemic. Tanzania’s transport regulator (SUMATRA), Association of Passenger Advocates (CHAKUA), Public Cyclists of Dar es Salaam (UWABA), Amend, and other groups joined together to sponsor Tanzania’s participation in the World Day of Remembrance for Road Traffic Victims.


Elaine, Pat, and I climbed Kilimanjaro, reaching the summit on 1/11/11. We climbed the Rongai Route, so our ascent was from the north of the mountain and our descent was to the south. The diversity of landscapes and habitats on different parts of the mountain was impressive.

I created a map of the route and added some of our photos to it. If you have Google Earth installed, download this file. Otherwise, you can view the map in your browser here. Click on each of the icons to see pictures of our climb.

Sights of Northern Tanzania

Pat and Elaine decided to visit me in Northern Tanzania, where our main destinations were Ngorongoro Crater and Kilimanjaro. We were also able to visit the towns of Arusha, Moshi, and Mto-wa-Mbu (Mosquito Village). Mto-wa-Mbu was especially enjoyable. We had the opportunity to learn about the wood carving of Mozambican immigrants and try the traditional banana beer of the Chagga people (it tasted kind of like beef jerky to me).

Generated by Facebook Photo Fetcher

Lions of Ngorongoro Crater (Graphic Pictures)

The best part of our time in Ngorongoro Crater was getting up close and personal with two brother lions. They had killed a wildebeest right next to the road, providing a rare opportunity to view their feeding. Our excellent guide, Justin, who has helped coordinate National Geographic shoots in the Serengeti, was so impressed that even he took some pictures with his phone. Though we didn’t see the kill happen, it was fascinating to watch the resulting drama unfold throughout the day.

The power of the lions tearing apart the wildebeest carcass was awe-inspiring. They were lucky enough to get two meals in one, since the wildebeest was pregnant. As the lions fell into food-induced naps, the circling hyenas would slowly close in. Sometimes the lions would stir, leading the hyenas to initiate a charade of walking by with no interest in the carcass. On two instances, the hyenas were able to get up to the carcass and start feeding, only to be chased off again by one of the brother lions. We were hoping to stay long enough to watch the circling vultures land, but the lions were too reluctant to leave their prey.

Generated by Facebook Photo Fetcher

Ngorongoro Crater

Ngorongoro Crater

Ngorongoro Crater

We spent two nights at the Simba campsite in the Ngorongoro Conservation Area, a UNESCO World Heritage Site. The number, variety, and accessibility of animals inside the crater was astounding. Within twenty four hours, we saw baboons, ostriches, hyenas, warthogs, jackals, zebras, wildebeest, elephants, a black rhino, flamingos, hippos, a leopard, and two amazing lions. The looming crater walls were an amazing backdrop, especially with rainclouds and a rainbow. When we returned to the campsite, a couple of elephants joined us and started drinking from the water cistern.

Special thanks go to Krisha African Safaris for arranging the amazing trip to the Crater, our Kilimanjaro climb, and our tour of Mto-wa-Mbu.  It seems that many other tour operators do either safaris or climbs, and it would have been inconvenient for us to have to deal with separate companies.  Krisha Safaris took care of it all, using their professional contacts across the industry to design a customized trip for us.  We appreciated the personalized service (airport pickups at the different airports we arrived at, briefings, and debriefings, all by the CEO) and look forward to trying one of their Serengeti hot air balloon safaris on our next visit.

Generated by Facebook Photo Fetcher

Daladalas of Dar es Salaam

Dar es Salaam’s three million residents rely heavily on daladalas (16-36 passenger minibuses) for transportation. Forty-three percent of trips are made by public transit, and of these, ninety-eight percent are by minibus. The second-highest transport mode is walking, which accounts for forty-two percent of trips.

Since the government deregulated urban transportation in 1983, thousands of used vehicles from Japan (mostly Toyota Hiaces, Toyota Coasters, and Mitsubishi Rosas) have entered the city via ship. There are about 7,000 daladalas operating in the city, serving various color-coded routes that range in length between 3 and 30 km. As of 2009, nearly sixty percent of daladalas were more than twenty years old.

Fares range between 250 and 350 Tanzanian Shillings (17 to 24 cents), depending on distance. Though these low fares result in little revenue for operators (the average daladala brings in a daily revenue of about $24), transport costs still account for about 17% of household expenses.

In my month in Tanzania, I took 32 trips on daladalas. The majority of these rides were hot, crowded, and slow. At least the horrendous traffic meant there were plenty of opportunities to buy cold water from the numerous vendors whistling and shouting “Majimajimajimaji!”

Generated by Facebook Photo Fetcher