Chapter 234 - Chapter 224: The Night the Rails Broke

Chapter 224: The Night the Rails Broke

25 May 1976 — 26 May 1976

Between Kanpur Anwarganj and Lucknow Charbagh; the field near Unnao

23 May 1976 — Two Days BeforeDivisional Railway Manager's Office, Northern Railway, Lucknow Division

The file had been on the desk of the Divisional Railway Manager's Senior Assistant for eleven days.

It had come up from the Permanent Way Inspector at the Unnao section on May 12th. The Permanent Way Inspector's name was Ramesh Dayal, and he was forty-three years old and had been walking railway track in the Northern Railway's Lucknow Division for nineteen years, and he had a specific relationship with the track under his charge that was the relationship of someone who knew every bolt and every sleeper and every curve by the feel of his feet and the sound of his boots on the rail — the particular ring of well-seated track versus the dull, flat sound of something that had worked loose.

The section between Unnao and Kanpur Anwarganj had been ringing wrong since March.

Not catastrophically wrong. Not the obvious wrong of a missing bolt or a cracked rail that could be seen without looking for it. The specific, subtle wrong of a track that had been carrying more traffic than its maintenance schedule accounted for, in a monsoon season that had arrived early and wet and had softened the ballast under the sleepers in the way that early monsoons did in the Gangetic plain.

Ramesh Dayal had submitted three inspection reports through the chain since March. The first in March, noting the condition and requesting enhanced ballast tamping. The second in April, noting that the tamping had not occurred and requesting urgent attention. The third in May — the file now sitting on the Senior Assistant's desk — was specific. It named the kilometre mark: 94.3 from Lucknow, near the village of Daulatpur. It described the condition: differential settlement of the track formation producing a lateral alignment deviation of approximately 14 millimetres on the down-line curve, combined with a loose key on the 74th sleeper from the south end of the section. It requested: a temporary speed restriction to 50 km/h on the section until repairs were completed, and an emergency work gang deployed within 48 hours.

The Senior Assistant had read the file.

He had made a note on the file: For examination and decision.

He had sent it to the Assistant Divisional Engineer.

The Assistant Divisional Engineer had been on leave since May 15th. His office had received the file and logged it in the incoming register and placed it in the stack on the ADE's desk where it would wait for his return on May 26th.

The Divisional Engineer had not been informed because the ADE's office had assessed the file as routine — routine in the specific bureaucratic sense that it was a routine type of file, a Permanent Way inspection report, which was a category of file that passed through the ADE's office at a rate of forty to sixty per month and most of which described conditions that were monitored rather than emergencies that were acted upon. The forty to sixty files were not all read. Some were read. Some were filed. The system's implicit assumption was that the Permanent Way Inspector was the person who raised concerns, and the ADE's office was the system that allocated resources, and the allocation of resources was constrained by the maintenance budget, and the maintenance budget for the Lucknow Division in the financial year 1976-77 had been reduced by eighteen percent from the previous year's allocation.

The reduction had been decided in Delhi, at the Railway Board level, as part of a broader rationalization of the Northern Railway's budget following a revision of the five-year plan's railway investment priorities. The reduction was documented in the budget order. The budget order was filed in the Lucknow Division's accounts section.

The track at kilometre 94.3 was not mentioned in the budget order.

The track at kilometre 94.3 was not mentioned anywhere except Ramesh Dayal's three inspection reports.

The third inspection report, dated May 12th, was on the Senior Assistant's desk.

On May 23rd — two days before the accident — the Senior Assistant wrote a second note on the file: Please expedite. DRM attention may be required.

He sent it to the DRM's personal assistant.

The DRM's personal assistant was managing seventeen concurrent file movements on May 23rd. The file went into the DRM's incoming tray.

The DRM was in a meeting with the railway contractor whose company had resurfaced the Kanpur-Lucknow main line section in 1974. The meeting was about the contractor's proposal for the next phase of track maintenance. The contractor's name was Ghanshyam Wadhwa and his company, Wadhwa Rail and Infrastructure, had held Northern Railway maintenance contracts in the Lucknow Division for nine years. The company's work was, by the assessment of the Railway Board's annual audit, "satisfactory." The audit's definition of satisfactory was that the work had been completed and the contractor had submitted completion certificates and the divisional engineer had countersigned them.

The fish plates — the metal connectors that joined one rail to the next — installed by Wadhwa Rail and Infrastructure on the Kanpur-Lucknow main line section in 1974 were not the specification described in the contract. The contract specified IS Grade 880 high-tensile steel fish plates with a minimum tensile strength of 880 MPa. The fish plates installed were a commercially available lower-grade material with a tensile strength of approximately 650 MPa, which was less than the specification and which, under the cyclical load of a main line express train operating at speeds above 100 km/h, had a fatigue life that was significantly shorter than the IS Grade 880 specification.

The substitution had been possible because the inspection procedure for installed track components did not include material testing. The inspection recorded visual conformity — the plates were the right dimensions, the right colour, the right number of holes. The specification number stamped on each plate was not verified against the actual material composition.

Ghanshyam Wadhwa's company had billed the Northern Railway for IS Grade 880 fish plates.

The Northern Railway had paid for IS Grade 880 fish plates.

The fish plates installed were not IS Grade 880.

The DRM's meeting with Ghanshyam Wadhwa ended at four in the afternoon on May 23rd. The DRM's personal assistant placed the inspection file in front of the DRM at four-thirty.

The DRM was M.K. Srivastava, fifty-six years old, a 1945-batch officer of the Indian Railway Service of Engineers who had managed large railway divisions for eleven years and who had, in those eleven years, developed the specific administrative disposition of a man who had seen many inspection reports and who had learned that the most frequent category of inspection report was the report that described a condition that was being monitored and that would be addressed in the normal course of the maintenance programme.

He read the file.

He noted the kilometre mark. He noted the 14-millimetre alignment deviation. He noted Ramesh Dayal's request for a temporary speed restriction.

He thought about what a temporary speed restriction on the Kanpur-Lucknow main line at kilometre 94.3 would do to the 36-train-a-day schedule that the section carried. He thought about the Kashi Express, which was the busiest passenger train on the section, running two trips daily with an average of eighteen hundred passengers per trip. He thought about the Divisional Operations Superintendent's standing instruction about speed restrictions: Any sectional speed restriction below 80 km/h requires DRM approval and must be accompanied by a revised timing chart for all affected trains.

He thought about the maintenance gang. The Unnao section maintenance gang was currently engaged on the Lucknow Outer Crossing resurfacing, which was the priority work of the month and which could not be interrupted without setting back the crossing schedule by fourteen days.

He made a note on the file: Condition to be included in next scheduled maintenance programme. Normal speed to continue. Advise Permanent Way Inspector to maintain close watch.

He initialled the note.

He sent the file back down the chain.

On May 24th, Ramesh Dayal received the file back with the DRM's note.

He read the note.

He sat at his desk for a moment.

He wrote, in the bottom margin of his original inspection report, a single sentence: DRM decision noted. Risk assessment: significant. Speed restriction denied. Walk inspection scheduled for 26 May.

He filed his copy.

He did not know that the walk inspection on 26 May was going to be too late.

25 May 1976 — The Evening

The Kashi Express 3004, Varanasi to Lucknow, was due to depart Kanpur Anwarganj at 9:47 PM.

At 9:47 PM it was still in the platform, three minutes late, which was not unusual for the Kashi Express — the Kashi Express was late more often than it was on time, which was a fact that its regular passengers had incorporated into their planning the way that people incorporated all reliable unreliability.

The platform at Kanpur Anwarganj was a study in the specific density of Indian railway travel in the summer heat. The season was the pre-monsoon period — technically, the first rains had arrived, but they had not yet produced the sustained daily rainfall that was the true monsoon; what they had produced was the specific, oppressive quality of heat and humidity together, the air thick with moisture that had not yet fallen, the platforms smelling of hot concrete and chai and the particular combination of humanity and luggage that filled every available surface.

The train had twenty-two coaches. Its sanctioned capacity was 1,640 passengers across the sleeper and unreserved coaches. Its actual passenger load on the evening of May 25th was estimated, in the subsequent inquiry report, at approximately 2,100.

The overcrowding was the specific overcrowding of a train that had departed Varanasi with its unreserved coaches already full and had added passengers at every intermediate stop — Prayagraj, Kanpur Central, Kanpur Anwarganj — because the demand on the Varanasi-Lucknow route consistently exceeded the available rolling stock and because the ticket examiner's capacity to control boarding was finite in the face of crowds that needed to get home.

In sleeper coach S-4, in the lower berth of compartment 7, a man named Ramkhelawan Misra was sitting with his feet on the floor and his back against the compartment partition and his hands folded in his lap. He was sixty-two years old and had retired from the Kanpur district court's clerical service in January and was making, on May 25th, his first long journey since retirement — going to Lucknow to see his daughter, who had recently had her first child.

He had a small bag on the overhead rack that contained a tin of murabbha his wife had made and a package of clothes for the new baby. He had a ticket in his shirt pocket, folded in thirds, that he had bought three weeks ago because he had understood that advance booking was necessary and had done everything in advance because he was the kind of man who did things in advance.

He was looking out the window at the platform.

In the berth above him, a young woman named Priya Shankar was reading a paperback novel she had bought from the platform bookseller. She was twenty-three years old, from Allahabad, and was travelling to Lucknow for a job interview at the State Bank of India's regional office the following morning. She had her interview letter in her bag and a sari that she had spent forty minutes selecting and pressing that afternoon.

Across from Ramkhelawan Misra, in the lower berth of the facing bench, a family of four was managing the specific logistics of a family with two young children in a railway sleeper compartment: the father, Surendra Yadav, a forty-year-old farmer from a village outside Kanpur, with his wife Manorama and their children Sunil, age seven, and Geeta, age four. They were going to Lucknow for a family wedding — Manorama's elder sister's son was getting married on May 27th — and the children had been awake since five that morning and were achieving the specific state of exhaustion that produced simultaneous extreme tiredness and extreme resistance to sleep.

Geeta was in her mother's arms.

Sunil was looking at Ramkhelawan Misra with the frank curiosity of a seven-year-old who had run out of other things to look at.

"Kahan ja rahe ho?" Sunil asked. Where are you going?

"Lucknow," Ramkhelawan said.

"Hum bhi," Sunil said. Us too.

"Good," Ramkhelawan said. "Then we are going in the same direction."

Sunil considered this and found it satisfactory.

In the unreserved coach U-3, which was the third coach from the rear of the train, a man named Mohammed Ashraf was standing in the aisle with his back against a berth post and his bag between his feet. He was thirty-eight and had been working in a textile mill in Kanpur for four years and was going home to his family in the outskirts of Lucknow for the Eid celebrations that would begin in three days. He had not been home in six months. His bag contained four months' accumulated savings in cash — ₹640, which was the largest amount of money he had ever carried at once — and gifts for his three children.

The coach was so crowded that he could not have sat down if he had wanted to. He was standing in the aisle with forty-three other passengers. The temperature in the coach, with the windows open to the pre-monsoon air and the body heat of the crowd, was approximately thirty-eight degrees.

He was not thinking about the heat. He was thinking about his youngest daughter, who had been born two months after he had left for Kanpur and whom he had not yet seen. He was thinking about what she would look like. Whether she looked like his wife or like him.

The Kashi Express 3004 departed Kanpur Anwarganj at 9:51 PM — four minutes late.

The goods train that would collide with it had a different origin story.

The DFB-3 freight — a departure from Fazilnagar Bulk Yard, thirty cars of coal bound for the Panki thermal plant outside Kanpur — had been delayed at the Unnao loop since 7:30 PM. The delay was because the loop signal had not cleared when it was supposed to clear, which was because the point motor at the south end of the Unnao loop had a fault that had been reported to the Traction Distribution Section in March and not yet repaired.

The section controller at Lucknow Division's Control Room had been managing the DFB-3's delay since seven-thirty. The section controller was a man named Rajendra Prasad Sharma, forty-nine years old, who had been in railway operations for twenty-two years and who was managing, on the night of May 25th, a section that had two passenger trains running late, a goods train stuck in a loop, a point motor fault at Unnao, and a communications failure at the Daulatpur crossing gate that had been reported at 8:15 PM and not yet resolved.

At 9:55 PM, the Unnao stationmaster called Sharma on the control phone.

"DFB-3 is cleared," the stationmaster said. "The point is showing green. I'm ready to dispatch."

"What's the schedule for the Kashi?" Sharma said.

"Kashi passed Kanpur at nine fifty-one," the stationmaster said. "On the existing schedule, I have a clear window. Kashi won't reach here for another forty-five minutes."

The stationmaster was using the original schedule, not the current tracking. The original schedule had the Kashi arriving at Unnao at 10:36. The Kashi was four minutes late at Kanpur Anwarganj, but between Kanpur and Unnao, the driver had the Kashi running at 105 km/h — the scheduled maximum speed for the section — and was making up time on the straightaways.

The Kashi's current actual position was approximately 12 kilometres south of Unnao. It would reach kilometre 94.3, the Daulatpur curve, in approximately seven minutes.

Sharma said: "Dispatch DFB-3. Normal schedule. Kashi has clear."

He moved to the next issue on his control board.

The point motor at the south end of the Unnao loop had shown green when the stationmaster checked it. But the point had not fully completed its throw — the railway technical term for the physical movement of the track switch that directs a train onto one line or another. The point motor had moved the switch to approximately ninety-four percent of its full travel and had then stopped, producing a signal indication of 'Normal' because the detection circuit registered the switch as positioned, but leaving the actual rail gap in a position that was not quite the main line and not quite the loop.

In the darkness and the humidity and the operational pressure of a man managing multiple simultaneous issues, the stationmaster had not physically verified the point. He had read the signal.

The DFB-3 freight — thirty coal wagons, approximately 1,800 tonnes total — began moving out of the Unnao loop at 9:57 PM.

The precise time of the first impact was subsequently established by the inquiry commission at 22:04:17, based on the time-stamp of the section controller's log entry at Lucknow Control Room recording the loss of the train detection circuit on the section between Unnao and Daulatpur.

In the reality of that moment, it was simply darkness and rain.

The monsoon had arrived early that night — the first sustained rain of the season, arriving in the specific, heavy, sheet-like way of the Gangetic plain's first real monsoon rainfall after the pre-monsoon humidity had built past what the air could hold. The rain had started at approximately 9:45 PM. By 10:00 PM it was falling hard enough that visibility from the locomotive cab of the Kashi Express was limited to approximately forty metres in the headlamp's illumination.

The locomotive driver of the Kashi Express was a man named Dharmendra Singh, forty-seven years old, with twenty-three years of service. He was running the Kashi at 97 km/h as he approached the curve at kilometre 94.3, which was within the section's authorized speed limit.

The Kashi Express's locomotive headlamp illuminated the rain and the track ahead. The track ahead was the Daulatpur curve, where kilometre 94.3 marked the beginning of the left-hand bend that the Permanent Way Inspector had noted in his third inspection report as having a 14-millimetre lateral alignment deviation.

At 22:04:17, what the headlamp also illuminated was the front of the DFB-3 freight locomotive, which had come out of the Unnao loop through the partially thrown point and was now on the main line, facing the Kashi Express, at a closing speed of approximately 210 km/h combined.

Dharmendra Singh applied the emergency brake.

The emergency brake on a passenger train running at 97 km/h on wet rails required approximately 800 metres to bring the train to a standstill. The distance between the headlamps of the two trains when Dharmendra Singh saw the freight was approximately 90 metres.

The collision occurred 0.9 seconds after he applied the brake.

The Kashi Express's locomotive and the DFB-3 freight locomotive met at the beginning of the Daulatpur curve, at a point approximately twelve metres into the left-hand bend where the track's 14-millimetre alignment deviation was located.

The impact was at a combined speed of approximately 200 km/h, reduced by the 0.9 seconds of emergency braking on both trains — the freight driver had also seen the Kashi and also applied his brake.

The impact was not a single event but a sequence of events that occurred in the first four seconds.

The two locomotives struck head-on and telescoped each other to a combined length approximately sixty percent of their original combined length. Both locomotive cabs were destroyed. Both drivers died instantly.

The force of the impact transmitted through the Kashi Express's locomotive into the first coach — the General Seating Coach GS-1 — which was empty of passengers but was structurally the weakest element of the train. The GS-1 was a 1962-era coach. It absorbed approximately thirty percent of the collision energy before its structural framework collapsed. The second coach, S-1, a 1968-era sleeper coach, received the balance of the energy and its forward end telescoped into the GS-1's wreckage.

The third, fourth, and fifth coaches — S-2, S-3, and S-4 — derailed on the curved track at the point of impact. The 14-millimetre alignment deviation, under the lateral force of eighteen hundred tonnes of freight train and the combined derailment dynamics, produced a track failure at the deviation point that threw S-3 and S-4 onto their sides.

S-4 was the coach where Ramkhelawan Misra was travelling to see his daughter.

In S-4, in the 1.3 seconds between the sound of the initial impact and the coach beginning to roll, the passengers had no time to do anything. Ramkhelawan Misra was thrown against the berth post. Priya Shankar fell from the upper berth. The family of four — Surendra, Manorama, Sunil, Geeta — were thrown across the compartment.

The coach rolled ninety degrees and came to rest on its left side, its windows now forming the floor and roof.

In coach S-4, in the dark and the rain and the sudden, total absence of the normal sounds of train travel, people began to understand what had happened to them.

The electrical system had failed. The emergency lighting — battery-powered, required by regulation — produced a dim orange illumination from three surviving units in the coach's now-horizontal ceiling, which was now the side wall.

Somewhere in the coach, a child was crying.

In the unreserved coaches at the rear of the train, which had not derailed and had not been directly affected by the collision, passengers sat in stunned silence for approximately forty seconds before the wave of understanding moved through the coaches and the sounds began — the crying, the calling of names, the shouting through the windows into the dark and rain outside, the first passengers jumping off and running forward along the track toward the smoke and the darkness where the front of the train had been.

Mohammed Ashraf, in U-3, was one of the first off the train. He was unhurt — the rear coaches had stopped abruptly but without the violence of the front. He jumped from the coach door onto the wet embankment and ran forward along the right-of-way, in the rain, toward the noise.

He ran for approximately ninety seconds before he saw what had happened.

He stopped.

He stood in the rain and looked at what was ahead of him and could not, for a moment, make his mind do anything useful with what he was seeing.

Then he started forward again because there was nothing else to do.

22:04 — 23:05The Crash Site; Lucknow Control Room; Chief Minister's Residence

The section controller at Lucknow Control Room, Rajendra Prasad Sharma, lost the train detection circuit at 22:04 and spent approximately ninety seconds assuming it was a detection failure — a common enough technical fault — before the calls started arriving.

The Daulatpur crossing gate keeper called first. He had heard the impact and could see the smoke and the lights of the derailed coaches from his position.

The Unnao stationmaster called second. He had lost contact with the DFB-3 freight at 22:05.

A passenger with access to the guard's emergency phone called third.

By 22:08, Rajendra Prasad Sharma knew there had been a collision.

He began the emergency response. He called the Divisional Control Officer. He called the Lucknow Central Station's Station Director. He called the Unnao station to block the section. He called the accident relief train depot at Lucknow.

The accident relief train at Lucknow was equipped with: one breakdown crane of fifty-tonne capacity, two rescue coaches with basic medical equipment, and a Supervisory staff of twelve railway employees.

It was not equipped with: hydraulic rescue cutters, large-capacity water tender for fire suppression, additional medical supplies beyond what the two rescue coaches carried, or the specific equipment for large-scale passenger rescue from overturned coaches.

The relief train was dispatched at 22:34 — thirty minutes after the collision.

Its estimated arrival at the crash site was 23:20, given the need to clear the blocked section and approach from the Lucknow direction.

The first call to the Chief Minister's residence came at 22:19.

Not from the railway. From the UP Police's district control room in Unnao, which had received calls from two constables at the Daulatpur crossing gate who had walked to the site through the rain and had called their control room with the information they had, which was: major train accident, many casualties, the first estimate of the number injured at the site was "more than a hundred."

The constables' estimate was conservative. The actual figure at the site was significantly higher.

Malhotra received the call at the security desk. He woke Karan at 22:21.

Karan was awake and sitting up in two seconds — not the groggy surface-waking of most people but the complete, immediate alertness that was one of the things about him that Sakshi had noted in the first year of their marriage and that had never changed.

"Train accident," Malhotra said from the doorway. "Kanpur-Lucknow section. Near Unnao. Multiple casualties. Police estimate more than a hundred."

"Which train?" Karan said.

"The Kashi Express."

Karan was already getting out of bed. "Get me the District Collector of Unnao. Get me the DIG Lucknow Range. Get me the ADMO of Unnao district." He was moving toward the bathroom as he listed the names. "Get me the Shergill logistics office — I need every ambulance vehicle in the Kanpur-Lucknow corridor told to move toward Daulatpur village, Unnao district, now. Get me the Director of Medical Education — I want the Lucknow Medical College and the KGMC on emergency intake status."

He was dressed in six minutes.

"The railway control room," he said, coming out. "Have they been reached?"

"I have their number," Malhotra said. "The Divisional Operations Control at Lucknow."

"Call them. Tell them the UP government's emergency response is being coordinated and ask what they have on the ground."

He went to his office in the residence.

The call to the Lucknow Division's Divisional Control Officer came through at 22:31.

The Divisional Control Officer was a man named S.C. Gupta who had been in the operations room since the accident information reached him at 22:12. He took Karan's call with the specific quality of a Central Government official receiving a call from a State government Chief Minister about a Central Government jurisdiction matter — not hostile, but careful.

"We have dispatched the accident relief train," Gupta said. "ETA at site is 23:20. The DRM has been informed. The General Manager's office has been informed."

"Medical support from your side," Karan said.

"The two rescue coaches carry basic first aid," Gupta said. "We're coordinating with the local hospital at Unnao."

"The Unnao district hospital has a bed capacity of ninety-two," Karan said. He knew this because the UP government's health infrastructure database was something he had personally reviewed during the sanitation programme planning. "It does not have the trauma surgery capacity for a major accident. I am putting the Lucknow Medical College and KGMC Lucknow on emergency intake."

"That's state government coordination," Gupta said. "The railway will coordinate with its own medical network."

"I am not asking for your coordination," Karan said. "I am telling you what the UP government is doing with state medical resources. If you have railway medical personnel to contribute, they should go to the site immediately."

A brief pause.

"Yes," Gupta said. "Understood."

Karan ended the call.

He called Trivedi.

The Chief Secretary of UP picked up on the second ring. He was already awake.

"I know," Trivedi said.

"The railway jurisdiction is Central," Karan said. "Everything on the ground tonight is ours. Police, medical, ambulances, shelter. I don't want a single person at that site waiting for the railway's approval to receive help."

"My disaster management team is already being alerted," Trivedi said.

"The buses," Karan said. "The Shergill Express bus network has twelve vehicles currently operational in the Kanpur-Lucknow corridor overnight service. I want them converted to passenger transport to Lucknow — take the uninjured from the rear coaches directly to Lucknow. The UP Police will escort."

"I'll authorize it," Trivedi said.

"The hospitals. All district hospitals within sixty kilometres of the crash site — Unnao, Lucknow KGMC, Lucknow Medical College, Kanpur Medical College — emergency intake status, all available surgical teams in. I want blood donation drives started by midnight."

"Done," Trivedi said.

"I'm going to the site," Karan said.

"Chief Minister—" Trivedi started.

"I'm going," Karan said. "Not for a press conference. Because I need to see what is there."

He ended the call.

He was in the car at 22:41.

The drive from Lucknow to Daulatpur, under normal conditions, was approximately forty-five minutes. With a police escort on the empty night highway, in the rain, it took thirty-two minutes.

He arrived at the crash site at 23:13.

23:13The Crash Site, Daulatpur, Unnao District

The scene that the headlights of the police vehicles illuminated as they came down the dirt track from the highway to the railway embankment was not a scene that permitted immediate comprehension.

The mind, confronted with a large-scale disaster, did not process it as a single image. It processed it in fragments. The fragments were specific and could not be generalized.

The Kashi Express's locomotive was not visible as a locomotive. What was visible was a mass of metal approximately four metres high that bore no obvious relationship to the shape of a railway locomotive. Adjacent to it was the DFB-3 freight locomotive in a similar condition. Between them and overlapping into both was the collapsed structure of the first two passenger coaches, GS-1 and S-1, their steel bodies folded and compressed in the specific geometry of catastrophic energy transfer.

S-3 was on its side beside the track. S-4 was on its side at approximately thirty metres from S-3, having traveled that distance during the derailment sequence. Both coaches had their windows — now the floor and the ceiling of a horizontal space — either broken by the impact or cut open from outside by the villagers who had arrived first with whatever tools they had brought.

The villagers had been there since approximately 22:15. They had come from Daulatpur and from two other villages within two kilometres of the crash site, in the rain, with torches and with whatever they had in their houses — hammers, axes, the iron rods used in construction work, lengths of rope, lanterns. They had broken through the coach walls and had been pulling people out for an hour before any official rescue team arrived.

A man named Bhushan Lal, a farmer from Daulatpur, had organized a team of eleven men to work on S-4. They had used a heavy iron crowbar to pry open a section of the coach's floor plate — now the side of the horizontal coach — large enough to allow a person to be lowered in on a rope. They had used the rope to lower the smallest man in the group, his nephew, who was twenty years old and slight enough to move through the interior of the overturned coach.

Bhushan Lal's nephew had been inside S-4 for forty minutes when Karan arrived.

He was still inside it.

Karan stepped off the embankment and walked toward S-4.

The rain was heavy enough that the ground between the track and the coach had become soft mud. He walked through it without adjusting his pace. A police constable attempted to guide him around the mud. He ignored the guidance.

At the opening in the coach's side, Bhushan Lal was pulling on the rope, assisted by three other men. Bhushan Lal looked at Karan — the white kurta, the official vehicles, the police escort — and registered the information and then turned back to the rope.

Karan crouched beside him.

"How many inside?" he said.

Bhushan Lal kept his hands on the rope. "Three we know. Maybe more. My nephew says he can hear sound from further in but he can't get through."

"Your nephew has been in there forty minutes?"

"Yes."

From inside the coach, muffled but audible through the opening: "Haath pakdo — grip my hand—" The nephew's voice, and beneath it a different sound, higher, the sound of a child.

The rope went taut.

Bhushan Lal's team pulled.

A child came through the opening — a girl, four years old, with a cut on her forehead and the specific blank expression of a child who has experienced something her mind has not yet categorized. She was wrapped in a cloth that someone inside had used to pad the pull.

She was not crying.

One of Bhushan Lal's men took her and carried her toward where the people from the rear coaches had gathered.

Bhushan Lal lowered the rope again.

Karan said: "What do you need?"

Bhushan Lal looked at him. "Something that can cut the berth post. My nephew says there is a man trapped under it. We've been trying to pull the post but it is welded into the frame."

Karan turned to the police officer behind him. "Find the SHO," he said. "Tell him I need cutting equipment. Check the accident relief train — if it doesn't have cutters, call the Gorakhpur complex maintenance depot. They have gas cutting torches. I want them here within the hour."

"Yes, sir."

He turned back to S-4.

"Keep working," he said to Bhushan Lal. It was not a thanks and it was not an instruction. It was acknowledgment of the fact that Bhushan Lal was already doing the only thing that mattered.

Bhushan Lal nodded and put his hands back on the rope.

Karan walked down the line.

He walked through what the crash had made of the Kanpur-Lucknow main line between kilometre 93.8 and kilometre 95.1. He walked past the medical teams that had arrived from the Unnao district hospital — two doctors, six nurses, limited equipment — who were working on the injured who had already been extracted. He walked past the police who were doing crowd control, keeping the accumulating crowd of villagers and passers-by from the embankment. He walked past the first ambulance, which had arrived six minutes before he did and was already departing with the first batch of critically injured.

He looked at each coach.

The rear coaches — U-3 through U-7, the unreserved coaches that had not derailed — were intact. Their passengers were frightened and in some cases bruised from the emergency stop but were not seriously injured. These were the people who were being organized by a young sub-inspector named Abhay Singh into groups for the Shergill buses that were on their way from the highway.

The middle coaches — S-2 through S-7 — had derailed but not overturned, and their passengers were in various states: some who had managed to get out through the doors and windows and were gathered beside the track, some who were still in the coaches and were being assisted out, some who were hurt badly enough that extraction required care.

The front coaches — GS-1, S-1, S-2, S-3, S-4 — were the disaster.

He stood at the boundary between the middle section and the front section for a moment.

He looked at the specific, identifiable object amid the wreckage of S-1: a school bag. Child-sized, the red and blue fabric of a mass-produced school bag, hanging by one strap from a protruding piece of metal at the compressed section of the coach where S-1 had telescoped.

The bag was intact. The metal that held it had been part of the overhead luggage rack. The bag had things in it — it had weight, it hung with the weight of books or a tiffin box or the accumulated contents of a school-going child's daily life.

The bag's owner was not visible.

Karan looked at the bag for a moment.

Then he turned and found the Unnao District Collector, who had arrived at 23:18 and was looking for the Chief Minister.

"How many?" Karan said.

The Collector was a man named P.K. Agarwal, thirty-nine years old, who was managing the worst night of his professional life with the specific, focused competence of someone who has discovered that the only way through a disaster of this scale is to make each individual decision correctly and not think about the total.

"Current confirmed," Agarwal said. "Thirty-one dead. The rescue teams think there are more in S-1 and the locomotive area. Injured — we're counting as they come out. Approximately two hundred, of varying severity."

Karan absorbed this.

"The hospital at Unnao," he said.

"Already at capacity," Agarwal said. "We've diverted the last four ambulances directly to Lucknow."

"KGMC and LMC?"

"Ready," Agarwal said. "Dr. Krishnaswami's department has been alerted." He paused. "Thirty-one is the number I have. I think it will be higher."

"I know it will be higher," Karan said. He said it without emphasis, as a fact about reality that did not require dramatising. "The families. They'll start calling the railway and calling the hospitals. I want a passenger information desk operational in Lucknow and in Kanpur by midnight."

"I'll set it up in both cities," Agarwal said.

"The press," Karan said. "They will arrive before morning. I will do one briefing at the site — tonight, before I leave — factual, no speculation. After that, all media briefings come through your office and my communication secretary."

Agarwal wrote.

"The DRM," Karan said. "Has he come to the site?"

"He's on his way," Agarwal said. "From Lucknow."

Karan said nothing for a moment.

"When he arrives," Karan said, "he will find the UP government has been managing this site for an hour with no railway support from the division headquarters. I want that documented. I want the time of each UP government resource's arrival at site documented against the time of each railway resource's arrival."

Agarwal understood the instruction precisely.

"The documentation is being maintained," Agarwal said.

"Good," Karan said. He turned back toward S-4. "I'm going back to the coaches. Find me when the DRM arrives."

The accident relief train arrived at 23:41 — forty minutes behind its estimated arrival.

The delay was because the section had to be cleared of additional loose debris, and because the relief train's crew had had difficulty locating one of the two relief coaches in the yard at Lucknow.

The relief train's fifty-tonne breakdown crane was the most valuable equipment it carried. With the crane, the extraction work on the compressed coaches could begin properly.

Before the crane could begin work, the electrical lines had to be confirmed dead — the overhead electric traction lines ran above the main line track and the section controller had been asked to confirm power-off at 22:45 but the confirmation had not been received due to a miscommunication between the section controller and the traction distribution section.

The confirmation of power-off at the site came at 00:17.

During the interval between the accident relief train's arrival and the confirmation of power-off, the crane could not be used.

During this interval, the villagers from Daulatpur and the surrounding area continued the manual extraction work they had been doing since 22:15. By midnight, they had extracted thirty-seven people from S-3 and S-4 using ropes, improvised tools, and the specific, exhausting physical effort of unarmed human beings engaging with the metal of a wrecked railway coach.

Bhushan Lal's nephew had found Ramkhelawan Misra pinned under the berth post.

The gas cutting torch from the Gorakhpur complex maintenance depot arrived at 00:03, brought in a Shergill logistics vehicle that had made the eighty-kilometre journey in fifty-one minutes. The maintenance team's lead technician, a man named Suresh Kumar, assessed the berth post situation and told Bhushan Lal's nephew, through the opening in the coach: fifteen minutes.

It took seventeen minutes.

When the post was cut and Ramkhelawan Misra was brought out through the opening at 00:20, he was conscious and had a broken collarbone and four broken ribs and a gash on his head that had bled significantly.

He was placed on a stretcher that the Unnao district hospital team had brought.

He was transferred to an ambulance.

Before he was put in the ambulance, he said — to the stretcher bearers, to no one specific — "My bag. The tin of murabbha. Is the bag—"

The stretcher bearer said: "Sir, the bag—"

Ramkhelawan said: "It was for my daughter's house. She just had a baby."

Nobody said anything.

The ambulance departed.

At 01:40, the crane began work on S-1.

What was in S-1 was the thing that changed the count from the initial estimate.

S-1 had been carrying passengers in its unreserved sections — the areas of sleeper coaches that were not in reserved compartments and where passengers without reservations sometimes occupied available space. In the collapse of the coach's forward section, this area had been compressed almost completely. The recovery from S-1 was slow and was conducted with the specific, terrible care that was required when the distinction between the structural elements of the coach and the remains of the people inside it was no longer simple.

The doctor from the Unnao district hospital, a man named Dr. Arvind Singh who had been at the site since midnight and who had been doing what doctors did when what was in front of them was too large for their training to have prepared them for — was checking each person extracted from S-1 for signs of life and directing them accordingly.

At 02:15, he approached Karan.

Karan was standing at the edge of the crane operation, watching.

"Sir," Dr. Singh said.

Karan turned.

Dr. Singh was forty-four years old and had done his residency at KGMC and had worked in Unnao's district hospital for twelve years. He had managed road accidents, farm equipment injuries, electrocutions, seasonal disease outbreaks, poisonings, drownings. He had never been at a scene like this.

"Sir," he said again.

"Tell me," Karan said.

"The count from S-1," Dr. Singh said. He stopped.

"Tell me," Karan said. His voice was quiet. Not impatient.

"Forty-seven confirmed in the coach section. Twelve survivors. The rest—" He stopped again. He was having trouble with the clinical vocabulary in the context of what he was standing in front of. "The rest did not survive."

Karan was quiet.

"The locomotive crew," Karan said.

"Both dead on impact," Dr. Singh said. "The freight crew as well."

Karan turned back to the crane operation.

The total death count at the crash site, by dawn, was seventy-three.

It would reach eighty-nine by the evening of May 26th as injuries proved fatal.

M.K. Srivastava, the Divisional Railway Manager, arrived at the crash site at 01:05.

He came in an official railway vehicle with three officers from the divisional headquarters and found, when he arrived, a site that had been managed for two hours and fifty-one minutes by the government of Uttar Pradesh.

The ambulances were UP government vehicles and Shergill logistics vehicles. The medical team at the site was Unnao district hospital staff and a team from Lucknow Medical College that had been dispatched by vehicle at 23:30. The police cordon was UP Police. The transport of the uninjured passengers to Lucknow was being managed by Shergill Express buses escorted by UP Police. The gas cutting equipment that had extracted Ramkhelawan Misra was from the Gorakhpur complex. The passenger information desk in Lucknow had been operational since 00:15.

The railway accident relief train had arrived at 23:41.

The DRM arrived at 01:05.

He found Karan standing near S-4.

The DRM was a large man, heavier than Karan by thirty kilograms, and had the bearing of someone who was accustomed to institutional authority. He arrived with the specific manner of a senior official who was coming to take charge of a situation.

He extended his hand.

"Chief Minister," he said. "M.K. Srivastava, Divisional Railway Manager. I apologise for the delay in my arrival. The road—"

"How long ago was the crash?" Karan said.

The DRM looked at him. "Approximately three hours."

"You arrived three hours after the crash," Karan said. "The UP government has had ambulances, medical teams, gas cutting equipment, and police at this site for two and a half hours." He paused. "The railway's accident relief train arrived two hours after the crash. You have arrived three hours after the crash." He paused again. "I would like to understand the sequence of decisions that produced that sequence of events."

The DRM was quiet for a moment.

"The standard emergency response protocol—" he began.

"I am not asking about the protocol," Karan said. "I am asking about the decisions."

"Chief Minister, the railway operates under central government jurisdiction. The emergency response has followed—"

"Eighty-nine people are dead," Karan said. He said it without raising his voice. He said it with the specific, cold precision of a man who was presenting a fact rather than making an accusation, and who was doing this because the fact was what it was and it did not require embellishment.

The DRM was quiet.

"I am not holding you responsible for the crash tonight," Karan said. "The investigation will determine responsibility. What I am holding you accountable for is being here three hours after it happened." He paused. "When you return to Lucknow tonight, I want a full briefing document on my desk by eight in the morning. The accident sequence, the emergency response timeline, and every file related to track maintenance on this section for the past twelve months."

"That is internal railway documentation—" the DRM began.

"I know what it is," Karan said. "I am not asking for it through informal channels. I am formally requesting it through the Disaster Management Act's provisions requiring state and central agencies to share information following a major industrial accident. My administration will file the formal request before six AM. You will have the documents ready."

The DRM said nothing.

"My District Collector will coordinate with you on the site operations," Karan said. "I am going to speak to the press."

He turned and walked toward where the press had gathered.

03:15The media holding area, Daulatpur embankment

There were twelve journalists at the site by 03:15 — print reporters, two radio correspondents, and the crew of the Shergill Cinema's news unit that had arrived from Lucknow, because the news unit was part of the Shergill Media division and was operational around the clock.

Karan stood in the rain in front of them.

He said:

"At approximately ten o'clock tonight, there was a collision between the Kashi Express 3004 and a goods freight train on the Kanpur-Lucknow main line near Daulatpur village in Unnao district. The full circumstances of the collision are under investigation. I will not speculate about causes while the investigation is ongoing.

"As of three in the morning, the confirmed death toll is seventy-three. Approximately two hundred and forty people have received medical treatment, of whom sixty are being treated for serious injuries at Lucknow Medical College and KGMC Lucknow. The Unnao district hospital is at capacity.

"The rescue operation is ongoing. Railway workers and UP government teams are working at the site. The villagers of Daulatpur and surrounding areas were at this site within fifteen minutes of the crash and have worked through the night. I want to be specific about that. The first people to pull survivors out of the wreckage tonight were farmers and labourers from Daulatpur village, using ropes and hand tools. I want them to be named and I want them to be recognized.

"The passenger information desk is operational in Lucknow and Kanpur. Families looking for information about passengers on the Kashi Express 3004 should call those numbers.

"The investigation into this accident will be conducted. I do not have the authority to conduct that investigation — the railway is a Central Government matter. What I do have authority over is what the UP government does in the aftermath, and what the UP government is doing is this: every state resource that can be deployed for rescue and medical support is deployed. I have asked for the relevant maintenance records and inspection files for this track section. When those files are made available, the UP government will study them and will take whatever position the facts require.

"I will not be making further statements tonight. My District Collector will be the point of contact for media queries."

He stepped back.

A journalist called out: "Chief Minister, is the railway responsible—"

"I said I will not speculate about causes," Karan said.

"Chief Minister, can you comment on—"

He walked back toward the crash site.

27 May — 2 June 1976Various offices; the Inquiry

The formal request for railway maintenance documentation was filed by the UP government's legal desk at 5:47 AM on May 26th.

The documents arrived at the CM's secretariat at 10:30 AM — not all of them, but the first batch: the maintenance records for the Kanpur-Lucknow main line section for the period January to May 1976.

The maintenance records included the Permanent Way Inspector's inspection reports.

Karan read all three of Ramesh Dayal's inspection reports in sequence. He read them in his office in the Lucknow Secretariat, with Trivedi and the UP government's infrastructure legal advisor sitting across from him.

When he finished the third report — the one dated May 12th, with the DRM's note in the margin: Condition to be included in next scheduled maintenance programme. Normal speed to continue — he set the file down.

He sat with it for a moment.

"The DRM's note on the file," Trivedi said carefully. "The date."

"May 23rd," Karan said. "Two days before the accident."

"The speed restriction was explicitly requested by the inspector," Trivedi said. "The DRM explicitly denied it."

"Yes," Karan said.

"The inquiry commission," Trivedi said. "The Centre will appoint a Commissioner for Railway Safety inquiry."

"I know," Karan said. "The UP government will submit this file to the inquiry commission as a party submission. We will not make it public before the inquiry has received it."

"There is also the matter of the fish plates," the legal advisor said. He had been reviewing a second set of documents — the track maintenance contracts for the 1974 resurfacing work.

He set a document in front of Karan.

It was the contract between the Northern Railway and Wadhwa Rail and Infrastructure for the 1974 Kanpur-Lucknow main line resurfacing. Appended to it was the completion certificate signed by the divisional engineer. Appended to that was the inspection record.

The inspection record did not include material testing.

The contract specified IS Grade 880 fish plates.

The legal advisor had, the previous evening, acquired from the Northern Railway's procurement division the invoice from Wadhwa Rail and Infrastructure for the fish plates used in the 1974 contract. The invoice listed IS Grade 880 fish plates.

"If the fish plates installed were not the specification," Karan said.

"The substitution would be a criminal fraud against the government," the legal advisor said. "The Indian Penal Code, Section 420 and 120-B. Criminal breach of trust and conspiracy."

"Can the substitution be proven?" Karan said.

"The inquiry commission will need to test the fish plates at the accident site," the legal advisor said. "The fish plates were part of the track that failed. They are in the wreckage. If the inquiry commission tests them—"

"They will be tested," Karan said. "I will submit a specific request to the inquiry commission that the fish plates at the failure point be materially tested as part of the investigation."

He stood.

"Ramesh Dayal," he said.

"The Permanent Way Inspector," Trivedi said.

"He filed three reports," Karan said. "He was told to do nothing. He documented the denial and he documented the risk assessment. He was correct about everything." He paused. "When the inquiry commission takes evidence, Ramesh Dayal's testimony is the centre of the factual record. I want the UP government's legal advisor to contact him directly and offer him whatever support is appropriate to ensure that his evidence reaches the commission completely."

"He is a railway employee," Trivedi said. "He will face institutional pressure—"

"I know," Karan said. "Which is why I want the UP government's legal advisor on record as being available to him."

He moved toward the door.

"The families," he said. "Seventy-three dead as of this morning. The railway's accident compensation process will be slow." He looked at Trivedi. "I want the UP government to make an immediate ex-gratia payment to every family of a deceased passenger from UP. The amount is separate from whatever the railway eventually pays. It comes from the UP CM Relief Fund. I will determine the amount today."

"The railway may object that it creates a precedent—"

"The railway's objection is not my concern," Karan said. "The families' situation is."

He left.

3 June 1976Cabinet Room, UP Secretariat, Lucknow

Eight days after the accident.

The death toll had reached eighty-nine by the evening of May 26th. All recovery from the wreckage had been completed by May 28th. The inquiry commission had been appointed by the Railway Ministry on May 27th: the Commissioner of Railway Safety for the Northern Circle, with a specific mandate to examine the crash circumstances and the pre-crash maintenance record.

The UP government had submitted its documentary evidence to the inquiry commission on May 30th: Ramesh Dayal's three inspection reports, the DRM's note on the third report, the 1974 maintenance contract with Wadhwa Rail and Infrastructure, and a formal request for material testing of the fish plates.

The cabinet meeting on June 3rd was not scheduled for this subject. It was the normal biweekly cabinet meeting. But the railway accident was on the agenda because it was on everyone's mind.

Before the formal agenda, Karan sat at the head of the cabinet table and looked at his ministers.

He said:

"Steel did not fail on the night of May 25th. Men failed."

The cabinet was quiet.

"The track failed because a specific man decided that the speed restriction request from a specific inspector was an inconvenience he did not want to manage. The fish plates may have failed because a specific contractor chose to substitute lower-grade material and pocket the difference and sign a certificate saying the work met specification. A specific engineer may have countersigned the completion certificate without testing. These are decisions made by specific people. They are not the system failing. They are people choosing to do the wrong thing and being enabled to do the wrong thing by a system that did not make the right choice more rewarding than the wrong one."

He paused.

"The railway is Central jurisdiction. I cannot change what the railway does. What I can change is what happens in UP when something goes wrong on any infrastructure managed by any entity, central or state, within our geography."

He looked around the table.

"We are implementing three things. First: the UP State Rail Safety Research Institute. We do not manage the rails, but we can research the standards. We can produce the analysis that holds the inquiry commission accountable. We can become the organisation that the Central government cannot ignore because our data is better than theirs."

He looked at the Infrastructure Minister — Sreedharan, who was sitting at the table with the specific expression he wore when information was being received that had direct implications for work he was doing.

"Sreedharan Sahab," Karan said. "The expressway grade separations. The bridges. The utility corridors. Everything we build in UP is built to a standard that includes material testing. Every contractor who provides material for infrastructure in UP is subject to an independent laboratory test that is not the contractor's own test."

"We have this system in place," Sreedharan said.

"I want it strengthened," Karan said. "Every material specification in every infrastructure contract in UP, from today, includes a clause that the material supplier's certificate is insufficient and that independent testing is mandatory before payment is released." He paused. "What happened with those fish plates — the specific mechanism of a contractor billing for a higher-grade material and supplying lower-grade material and having a completion certificate signed without testing — that mechanism must be impossible in any contract that carries the Government of Uttar Pradesh's name."

Sreedharan wrote.

"Second: disaster medicine training. The doctors and paramedics who were at that site on the night of May 25th did everything humanly possible. They were not trained for mass casualty rescue. They did not have the equipment for mass casualty rescue. Dr. Nalini Krishnaswami's department will develop a mass casualty response training programme for every district hospital in UP. The equipment — hydraulic cutters, rescue equipment, emergency medical kits adequate for a major incident — will be in every district hospital and every emergency response depot by the end of this year."

He looked at the Health Minister.

"The training programme is your department's work," Karan said. "The equipment procurement — coordinate with Sreedharan Sahab's ministry on the specifications and with Aditya on the budget. I want a programme document on my desk by July 15th."

The Health Minister wrote.

"Third: the Shergill construction equipment. The gas cutting torches that extracted Ramkhelawan Misra from the wreck of coach S-4 came from the Gorakhpur complex maintenance depot because the accident relief train did not have them. I am not going to wait for the railway to upgrade its relief train equipment. The UP government will position gas cutting equipment, hydraulic rescue tools, and emergency power generators at five strategic locations along the major rail routes passing through UP. These are state assets. If there is an accident anywhere on any rail line in UP, the state will have the tools at the site faster than the railway can."

He looked at the cabinet.

"The families of the eighty-nine who died have received the UP government's ex-gratia payment. The railway's compensation process will take months. The railway's institutional process will determine what happened and assign responsibility. The inquiry commission will produce its report."

He paused.

"I want to say something about Ramesh Dayal. The Permanent Way Inspector who filed three reports and was told to do nothing. He will testify to the inquiry commission. His evidence is the most important evidence in this case. He did his job correctly, in the face of a system that did not want to hear what he was saying, and he documented everything. That is what every inspector, every engineer, every official in any government capacity is supposed to do."

He looked at the table.

"If there is any officer in the UP government who has a file on their desk right now that documents a risk that they have been told not to act on — bring it to the Chief Secretary's office before the end of this week. Not because I will necessarily act on every one of them today. But because I want to know that they exist. I do not want another man like Ramesh Dayal having to wait for a disaster to prove that he was right."

The cabinet was quiet.

He picked up his pen.

"Next item," he said.

Aftermath

The Commissioner of Railway Safety's inquiry report was submitted to the Railway Ministry in August 1976.

The report found:

That the collision was caused by the incorrect dispatch of the DFB-3 freight from the Unnao loop due to a partially completed point throw and an incorrect signal indication.

That the section controller's decision to clear the DFB-3 without current train position information for the Kashi Express was a contributing factor.

That the DRM's denial of the Permanent Way Inspector's request for a temporary speed restriction at kilometre 94.3 had not caused the collision but had contributed to the severity of the derailment sequence.

That the fish plates at the failure point on the Daulatpur curve, when tested by the Central Railway Mechanical Engineering Institute, were found to have a mean tensile strength of 642 MPa — significantly below the IS Grade 880 specification of 880 MPa.

That the 1974 completion certificate signed by the divisional engineer for the Wadhwa Rail and Infrastructure contract was signed without material testing, in violation of the contract's own inspection requirements.

That Ghanshyam Wadhwa, proprietor of Wadhwa Rail and Infrastructure, had submitted invoices for IS Grade 880 fish plates while supplying lower-grade material, constituting a prima facie case of fraud.

The report recommended: criminal proceedings against Ghanshyam Wadhwa. Departmental proceedings against the divisional engineer who had signed the completion certificate. A review of the DRM's decision on the speed restriction request. Mandatory material testing for all future track maintenance contracts.

The CBI was asked to investigate Ghanshyam Wadhwa.

The CBI's investigation took seven months. In March 1977, Ghanshyam Wadhwa was arrested.

In August 1977, the divisional engineer was suspended pending departmental proceedings.

The DRM, M.K. Srivastava, was transferred to a non-operational posting.

Ramesh Dayal, the Permanent Way Inspector, testified before the inquiry commission on June 28th. He testified for four hours. The UP government's legal advisor was present in the inquiry room as a non-party observer, which meant that Ramesh Dayal was aware that there was someone present who had no institutional interest in limiting what he said.

He said everything.

The inquiry commission's report cited his three inspection reports as the primary documentary evidence of the pre-crash maintenance failure.

The report noted, specifically: The Permanent Way Inspector's third inspection report dated May 12th, 1976, identified the exact failure point at kilometre 94.3, requested the specific intervention that would have reduced the severity of the derailment, and documented the institutional denial of that request. This report represents a model of accurate, timely, professionally sound safety reporting that was failed by the administrative system above it.

Ramesh Dayal received the Railway Board's safety commendation in 1977.

He had asked, when the commendation was announced, whether the commendation changed the fact that eighty-nine people had died.

The Railway Board's officer who informed him of the commendation had not known how to answer this.

Ramesh Dayal had said: "I know it doesn't. I just wanted to say it."

He continued his walk inspections on the Unnao section. He filed his reports on time. He continued to request speed restrictions when the track required them, and the track, under the revised Northern Railway maintenance protocols that followed the inquiry commission's recommendations, was maintained to the specifications that the existing protocols had always required but had not previously been enforced.

Bhushan Lal, the farmer from Daulatpur who had organized the first rescue effort, received a commendation from the Chief Minister's office in June 1976. He was invited to Lucknow. He came with his nephew, who had been inside the coach.

They sat in Karan's office.

Bhushan Lal was fifty-three years old and wore the clean clothes he kept for occasions that required them. His nephew was twenty years old and quiet in the way of young people who have been through something that has used up all the words for a while.

Karan said: "You were there before anyone else."

Bhushan Lal said: "The sound was very loud. My house is near the railway. My wife said there is an accident. I went."

"You organized the rescue."

"There was no one else to organize it," Bhushan Lal said. "What was there to wait for? The people were inside."

Karan looked at him.

"The little girl," Karan said. "Who came out of S-4."

"Geeta," Bhushan Lal said. He knew the name because the family had come back to Daulatpur, a week after the accident, to find the people who had pulled them out. Surendra Yadav and Manorama and Sunil had come, and Geeta had come. They had sat in Bhushan Lal's courtyard. "She is four years old."

"I know," Karan said.

Bhushan Lal was quiet for a moment. Then: "The tin of murabbha," he said. "The old man. Was his bag found?"

"I don't know," Karan said.

"He was asking about it," Bhushan Lal said. "When we brought him out. He was asking about a tin of murabbha for his daughter."

Karan was quiet.

"He survived," Karan said. "He has a broken collarbone. He's recovering in Lucknow. He went to see his daughter."

Bhushan Lal absorbed this.

"Good," he said.

"Was the tin found?" Karan said.

"No," Bhushan Lal said. "We looked. The coach was—" He paused. "We looked."

Karan nodded.

He sat with these two men from Daulatpur — the farmer who had gone to the crash with an iron crowbar and a rope in the rain because there was no one else, and his nephew who had lowered himself into the dark of an overturned coach because someone had to — and understood something that he had understood before but that required periodic re-understanding, the way certain truths did.

The system had failed. The inspection file had sat unsigned. The fish plates had been wrong. The speed restriction had been denied. The relief train had taken forty minutes too long.

But before the system arrived — before the ambulances and the police and the railway crane and the Chief Minister's car and the gas cutting equipment from the Gorakhpur complex — before any of it, at 22:15 when the echo of the collision was still in the air and the rain was heavy and there were no instructions and no protocol and no authority, Bhushan Lal had walked from his house with a crowbar and had begun.

He said, as he was leaving: "The railway should fix the track better."

"Yes," Karan said. "It should."

Bhushan Lal nodded.

He left.

Karan sat in his office after they had gone.

He thought about the file that had sat on the Senior Assistant's desk for eleven days.

He thought about the DRM's note: Condition to be included in next scheduled maintenance programme. Normal speed to continue.

He thought about Ramesh Dayal writing in the bottom margin of his own report: Walk inspection scheduled for 26 May.

He thought about the eighty-nine.

He sat with the number the way the number deserved to be sat with — not quickly, not as a statistic to be processed and filed, but as what it was. Eighty-nine specific people who had been on a train on the night of May 25th because they had somewhere to go and someone to see. A man going to see his newborn grandchild. A woman going to a job interview. A farmer taking his family to a wedding. A factory worker going home for Eid with four months' savings and gifts for children he hadn't seen in six months.

The system had killed them as surely as any deliberate act, and the system had killed them through the accumulation of small, individually deniable decisions that were each, in isolation, defensible: a budget reduction that was nationally rationalized; a maintenance schedule that followed established practice; a file notation that was within the DRM's discretion; a completion certificate that followed the inspection protocol as it was understood.

Each decision defensible. The accumulation lethal.

He would not let the same lethal accumulation exist in the infrastructure he controlled.

He would not.

He picked up the files on his desk.

He got back to work.

There was still work to do.

There always was.

End of Chapter 224