>Comments made on various other lists would suggest that I am not alone in waiting for 'ommitted' census pages.< Writers to other mailing lists are speculating on what Ancestry should do next. My suggestion would be to improve what is there already before they jump into something else. Other companies learn from their mistakes and improve the next batch of indexing, but not so with Ancestry. There are (proportionally) just as many female Williams in the 1841 census (6224) as there are in the other censuses - more in fact. Just as many Geroges (1901) and just as many wierd surnames and first names. Of course, the indexes could be improved considerably if they would only improve the correction process, admit that the indexing is bad and allow corrections rather than "alternative names". They should also allow corrections to all fields - not just the surname. Oh well, back to my "Why Can't I find Them?" lecture where I can include a whole new batch of examples from the 1841 census! Jeanne Bunting

At 18:35 24/04/2006, Jeanne Bunting wrote: >Of course, the indexes could be improved considerably if they would only >improve the correction process, admit that the indexing is bad and allow >corrections rather than "alternative names". They should also allow >corrections to all fields - not just the surname. Clearly Ancestry are erring in the direction of:- Quickly produced, somewhat inaccurate, indexes pointing at scans of the original documents. With the benefit of hindsight it might have been quite sensible for them to provide some sort of secondary index that can be freely "improved" by anybody directly on line. To avoid the effects of gross abuse, the main search facility would have the option to use or exclude the secondary index. The problem is that we all like to use the indexes, not just as a pointing aid to the original documents, but as data in its own right. eg it is probably rather dangerous to use the index to work out ratios of surname variants etc. In the case where an original document has obscure handwriting and a name or a place could be one of several possibilities, I would rather that the index "integrity" was compromised by indexing all possible options, which of course is done at the price of limiting the value of any statistics derived from the index alone. +Z_______________+Z_______________+Z__________________________+Z <www.adshead.com> Gordon Adshead Manchester Design Technology Beaumont House, 2 Goodrington Road, Handforth, Cheshire, SK9 3AT, England Tel:Fax:Msg:+44-1625-549770 Mob:+44-777-6145602 <gordon@adshead.com>