According to one source (I believe it is "Japanese Warships of WW2", but the pages I have copied don't include the title), in the entry on
Akagi:
"
Following on the experiments with the Hosho it was decided to dispense with a bridge, and Akagi was completed without one. This, however, was found to be impracticable, and shortly after completion a small bridge was built outboard of the starboard edge of the top flight deck and towards the bow.
...
In October 1935 the Akagi went into Sasebo Naval Dockyard for modernisation.
...
Lack of a decent bridge had also been felt, and this was remedied in the modernisation, a proper "island" superstructure being constructed amidships on the port side. The port side was chosen as it was felt that this would cause less obstruction to a pilot's vision when landing on, and improve control of flying operations.
A similar experiment was tried out on the new carrier Hiryu under construction at the same time."
It is interesting to note that the funnels remained on the starboard side after the 1935-1938 modernisation, although they
were angled downward and were totally below the flight deck. In fact, all of the IJN carriers had starboard funnels, regardless of island position.
Kaga was fitted with a starboard island during her 1934-1935 modernisation.
"
Completed in December 1937 the Soryu was built with what was to become the standard design of bridge for a medium carrier, which was sited on the starboard side towards the bow."
Hiryu was laid down 18 months after Soryu, and had a number of changes in addition to:
"As the design of Hiyru had been so altered it was decided to reposition the bridge as well, and it was planned on the same lines as that of the Akagi. Sited on the port side amidships it was a much larger affair than on the Soryu having many more control, communications and command positions than that of Soryu. It was hoped that the siting of the bridge to port would give the commander (air) better control of the flying operations and enable the aircraft to have a greater length of flight deck available for take-off. In practice it proved a failure. Although it did give greater control of flying operations and better take-off facilities, the bridge seriously restricted the length of deck for landing operations and caused reduction in aircraft parking space on the flight deck. Far more serious than these defects, however, were the conflicting air currents and turbulence which appeared immediately aft of the bridge, causing much distress to the pilots. So dangerous were these defects that the Japanese Navy never again tried siting the bridge on the port side of a carrier."
From this, it seems the much-cited "formation sailing using opposing flight paths" theory was not a part of their thinking... but was invented by US theorists to explain what they had not understood.
I suppose that the air current issue probably did relate to the engine-torque issue... I believe the Japanese engines rotated the same direction as British ones, as most of their early aviation industry was developed with British, not US, aid.
I don't understand why a port island restricted aircraft parking and landing deck length more than a starboard one would, though, as the funnels were all below the flight deck.
