When Waratah was first sighted by crew of the Clan MacIntyre at 4 am, 27 July, Waratah's position was very close to shore off Isikombi Point 9 miles southwest of Port Edward. Waratah departed Durban at 8.15 pm, 26 July. What with exiting the harbour, clearing the Bluff and establishing cruising speed in a southwesterly direction, Waratah took about 7.5 hours to cover the distance of 88 miles between Durban and Isikombi Point. This translates into an average cruising speed of 11.73 knots which equates with Officer Phillips' initial assessment of Waratah making 12 knots 'over the ground'.
Up until this stage, Captain Ilbery had been making excellent passage times averaging 13.5 knots. Waratah's listed cruising speed was 13 knots. In fact on the voyage over from Adelaide Waratah consumed 15 tons of coal more per day, arriving a full day ahead of schedule. But after departing Durban Waratah's cruising speed was a full 1.77 knots slower. One could argue that something was said about the additional coal consumption and Captain Ilbery decided to take it slower. But this does not hold in view of the fact that Waratah must have increased her cruising speed between Isikombi Point and the Xora River mouth to have gone out of sight here at 9.30 am, 27 July.
This simple fact adds to the mystery of why Waratah was so close to shore at 4 am and her deck lights were going on and off......
It points to something amiss which counters the argument that a perfectly functional Waratah succumbed to the storm of 28 July. In fact it lends itself to the argument that problems had already manifested in the early hours of 27 July eventually warranting turning back for Durban during the course of that day. Whatever was amiss appears to have been temporarily dealt with when Waratah pulled ahead of the Clan MacIntyre just before and beyond Cape Hermes.
But as we know from anecdotal accounts, fires on board steamers could grumble and flare, persisting for many hours before reaching catastrophic proportions. This could account for the unexplained variables noted during the early hours of 27 July.