Although this Combe Martin Census is undated Gibson and Medlycott's Local Census Listings 1522-1930 (2nd ed.), Birmingham, Fed. of Fam. Hist. Soc. (1994) quotes the date as 1831, something which is confirmed by other internal evidence, namely the known age of Richard Burgess.
The census information is written in a ledger book approximately 8"x13" in size. Pages 1 and 2 have the full handwritten headings of Questions 1 to 12 spread right across both pages of the open book, and the various enumerations take up the whole of the pair of pages.
Pages 3 and 4 contain just Question 13, and the numbers of servants. It seems as though the unidentified rows of notations regarding the number of servants may correspond to some of the named rows on the previous pages. The names in the bottom half of this list are written in a scrawl, rather untidily, and crammed in possibly as an afterthought. These names correspond with some of the names on pages 9 and 10 of the census, but some are written in abbreviated form, e.g. "Mr Lover" instead of "Mr Lovering" and "Char Sen" and "Char Jun" instead of "Jno. Charley" and Jno. Charley Jun'r".
A number of names and numbers have been struck-through, representing the fact that they have been crossed out in the original.
On page 10, it appears that the enumerator got halfway through totalling the columns, and realised he had forgotten "Wm. Peak", added him in, and then started totalling the columns again.
The detailed headings given on pages 1 and 2 are abbreviated on subsequent pages, though are given in full here. The same question columns are on all the pages, except that page 2 has just question 13, and pages 5 to 10 omit question 8 (whose column is for convenience neverthless included here), while page 11 (the totals page) is written vertically, and page 12 seems to be just some rough notes made by the enumerator, maybe to remind himself what the questions were!