When Commander Cyril Mercer Duthie's airliner lifted off its Auckland runway, he would have been comforted by the roaring of the engine on his left wing.
Equally, he would have been horrified by the near silence on his right. He only had two engines. With one down, he was in deep trouble that would soon lead to his fiery death.
This was 1938 and the still-spinning engine on Duthie's Lockheed Electra plane only had the same maximum power as about three and a quarter Toyota Corolla engines.
But he couldn't run it at maximum power for long or it too would have failed.
A heavy rain squall had just eased off as Duthie and Second Officer John Peel revved up the Union Airways Electra to take off.
Soon after leaving the grass airstrip in the direction of the Manukau Harbour at Mangere Aerodrome - the forerunner of today's Auckland International Airport - the Electra was seen to twist to the right. The propeller on the right engine had stopped turning.
The plane, named Kotare, continued a low-speed curve to the right, only just getting over radio masts and the Auckland Aero Club hangar, whose occupants dashed outside. Still climbing and turning to the right, it appeared Duthie was trying to circle and land. The plane flew over Nixon Rd farmland about 1.6km from the aerodrome in what observers thought looked like a midair stall with its tail down.
The right wing tipped down and the plane dived. It hit a large pine tree, lost its left wing and crashed into a gully.
"We saw her hit the tree and cut the top off it, leaving a 40-foot [12m] stump," a farm worker who was among the first to get near the crash site told the Herald. "The machine dropped straight into the gully and there was an explosion when she struck."
Duthie, aged 30, and Peel, 23, were killed by either the impact or the intense fire which ensued.
The crash of the Kotare 80 years ago this Thursday, May 10 was New Zealand's first fatal accident on a scheduled airline service.
The second was on this day in 1942, when the Union Airways Electra Kereru crashed on Mt Richmond between Nelson and Blenheim. Three passengers and two crew members were killed.
When Kotare, which had 10 passenger seats, crashed in 1938 there were, unusually, no passengers on board with the two pilots - only bags of mail.
The Auckland-Wellington service had begun in June 1937, after the first Electras were shipped to New Zealand and assembled at the Hobsonville Air Force base. Initially the flights all included stops at New Plymouth and Palmerston North, but some later omitted New Plymouth.
On its fatal, 12.15pm flight, Kotare was returning to Palmerston North and Wellington, having made the northward journey in the morning. The flight even gave its five passengers a look over Auckland central.
"... the pilot who brought the machine from Palmerston North … took the Kotare over the city and she was seen by hundreds of people as she passed over the business area about 11.35 o'clock," the Herald wrote.
At the Government-appointed court of inquiry, it emerged that there was confusion over which one of Kotare's engines had been recently overhauled.
Union Airways chief Mangere engineer Andrew Patterson read in the pilots' report that the left engine had been recently overhauled, but he later learned it was in fact the right engine.
Upon turning the right propeller, its engine it was found to have a "very slight" stiff spot. The engine was started and ran "quite freely", Patterson said, who decided it was fit to fly.
"I told Commander Duthie that as it was a recently-overhauled motor, he had better keep his eye on it."
When asked if a test flight would have been prudent if it was known the right engine had not been recently overhauled, Patterson said: "After giving it a run up and external inspection, and finding no fault, the next procedure would be a short flight test as a final check."
Patterson had seen parts of the right engine after it had been dismantled following the crash and said there was no sign of its having seized. He had no theory on why it had stopped.
In a theory that was contested at the hearing, chief engineer Leonard Mangham, of Union's Palmerston North base, said an iced-up carburettor might have caused the engine to stop, but there was no definite evidence this had happened.
The inquiry blamed pilot error. It said: "... the effective cause of the accident was a turn to the right down wind into a dead engine when the speed of the aircraft was insufficient to avoid a stall and the height insufficient to enable successful recovery from the right-hand spin which resulted."
However, the book Electra Flying, by aviation historians David Phillips and Richard Waugh, describes the different findings of a Union Airways investigation, including an Electra test flight with one engine shut down the day after the crash. The book says the blame was shifted "more to the aircraft's single engine performance and the proximity to the runway of hazards such as buildings and aerials …"
The company report said Electra single-engine performance was not up to expectations and the test's conclusion was that the plane would climb on one engine only if carrying a light load.
An Electra's propeller pitch angle could not be changed at lower engine speeds. When Duthie's right engine stopped, its propeller was in the position that caused the most drag, slowing the plane. The flaps had also been put down, possibly by accident.
"By the time that he noticed that he was under the level of the trees his flying speed was not sufficient to lift the aircraft over them and with probably a burst of the good motor and a sharp turn to the right (towards the aerodrome) the right wing dropped, the live engine pulled the left wing over, and the aircraft stalled and nose-dived into the ground, hitting the trees as it went down."
After the Mangere and Mt Richmond tragedies, there was a third fatal Electra crash, on Mt Ruapehu in October 1948. All 13 people on board were killed.
These were among the nine fatal airliner crashes in the pioneering period of New Zealand airlines, from the 1930s to the 60s, in which 73 people were killed.
Waugh and colleagues oversaw a project from 1994 to 2011 to place plaques near the crash sites or at least in the regions where they happened. All the dead are named on the memorials.