Roy (& Sackett List), Many thanks for sending me the page from the O'Callaghan book which was your source for the Richard Sackett 1703 New York Census data. ("The Documentary History of the State of New York", by E.B. O'Callaghan (1849), p614, table headed "Census of the City of New York about the year 1703"). If, as seems likely, the original of this census was lost in the 1911 library fire, then it stands to reason that secondary sources could only have been compiled before that date. We should therefore rely on O'Callaghan's transcription (1849) rather than on P D Smith's transcription (1998) on the USGenWeb site. I note that P D Smith does not give a source for his/her work, but clearly it must itself have been taken from a secondary source. I have made a comparison between the two transcriptions for the Richard Sackett entry and for a number of entries before and after his entry. There are differences in the records of 18 householders forming a block from Widd Vontylborough who appears 13 lines before Richard Sackett to Isaac Ferbergin who appears 4 lines after Richard. I do not have the previous page of O'Callaghan to check entries before Widd Vontylborough but entries after Isaac Ferbergin look to be the same in both transcriptions. Of the 9 columns of figures, the first 3 are the same in both transcriptions but the next 6 are one line different, the figures in Smith appearing one line below those in O'Callaghan. Thus for Richard Sackett, both transcriptions have 1 male 16-60; 1 female; and 2 male children (these being the first 3 columns of figures). O'Callaghan then shows for Richard (in the next 6 columns): 2 female children; 3 male negroes; 1 female negro; 0 male negro children; 0 female negro children; 0 all above 60. These figures are found in Smith one line down, in the entry for Elener Eleworth. Smith shows for Richard (in the same 6 columns): 0 female children; 0 male negroes; 0 female negro; 1 male negro children; 0 female negro children; 0 all above 60. These figures are found in O'Callaghan one line up, in the entry for Levenus Deuind. What seems to have happened in Smith's version (assuming O'Callaghan to be correct) is that a block of numbers 18 rows by 6 columns (columns 4 to 9) has slipped down by one row. I will write to USGenWeb to see if they can confirm an error in their transcription. Meantime I would think we are safe to rely on O'Callaghan which, of course, agrees with the version quoted by Weygant - albeit Weygant has 1708, which I guess we must assume was a typo. This means that Richard and Margery either had four children between 1699 and 1703 (possible of course - twin daughters would help!), or perhaps one or both daughters was/were from a previous marriage(s). I note your comment that the Census was from "about" 1703. Thanks again for your help, Best wishes, Chris