Last Friday, I searched the Incoming Passenger lists on Ancestry.co.uk and found Henry Hedges several times in the index. I did also find one of his return voyages by manually searching the voyages about the time he was due to return. This was the entry for the February 1926 trip, however I didn’t have the source citation, as I couldn’t find a way to find this from the image.

Common sense has prevailed. I searched for the person listed above him in the list, ‘H L Hall’, and once I had found him and confirmed it was the same image, I had my referance: Source Citation: Class: BT26; Piece: 811; Item: 76

Now all I need to do is find Henry returning in 1928 and 1930. This is when it starts to get frustrating!

Last week I did try finding him in 1928, but gave up when I hit two months that I expected him to be travelling where there were multiple incorrect images against the ship names. It made it impossible to search.

Searching in 1930, I was looking in March and April as these seemed to be the most likely months. I started checking the ships of the Elder Dempster Line, but came across the same problems. Here is an example of the problem:

Liverpool, March 1930, M.V. Apapa. There are 13 images in the set; they comprise of M.S. Apapa page 2, 4, 6, 5, 1, T.M.V. Aba (Apr 1930) page 9, 8, 10, 12, 11, M.S. Apapa page 14, 13 and 15. So not only are they out of order and pages missing, they are not even all the same ship or month. I have reported the errors to Ancestry like I did last week. The really annoying thing is, Henry would have been on page 3 if he was travelling on the Apapa in March 1930.

While the image quality of the scans is good, the quality control on the indexing is appalling for some months. I think it was March and April 1928 that were particularly bad last week, I gave up reporting in the end because almost every image was wrong. I guess this has been outsourced, but there should have been better supervision or quality inspection. I have also contacted Ancestry Support detailing the problems with examples and asking if reporting individual errors was the best way when there were so many.