The holes and the divergences, printed whole

What this addendum does not settle

The holes. No Mishawaka row exists for 1904 — the city stood below the comparative table's class, searched and found absent above. The biennial censuses of 1921, 1923, 1925 and 1927, which published statistics for cities between the decennial counts, were not read; the 1923 volume — Census of Manufactures: 1923, statistics for industries, states and cities — is located in full view at HathiTrust, and the others stand unlocated. The 1958 and 1972 volumes' own full city rows were likewise not read from their own pages; both years print above only at the strength of the next census's comparative columns, and the full rows remain for the next pull, named with their volumes in this addendum's companion note. The 1909–1919 figures of the Abstract are stated in thousands of dollars as that volume prints them; the exact-dollar 1909 row stands from the Thirteenth Census itself. And no figure in this addendum comes from a page that was reached but not read: the failures are named, not papered.

The divergences. Four stand open. The 1899 capital components exceed their own printed total by $30,000, in the volume itself, unreconciled. The 1939 establishment count is 29 in the Sixteenth Census volume and 28 in the 1947 volume's comparative columns; both stand. The 1909 row is printed twice in the federal record — exactly, in the Thirteenth Census, and in thousands in the Abstract — and the two agree, which is stated because the $23,662,684 five-city total has already been mistaken once, in summaries of this build, for the city's own product value; Mishawaka's is $10,882,846. And the count article in the leg this addendum joins — "The count of the transformation" — says that no continuous series exists in the record reached and that no figure from the census's city tables is printed there. That was true of the record reached when it was written; this addendum is the pull that changes it, and at assembly the sentence is to be conformed to the series now standing, with this fragment folded in beside the ladder it frames. The ladder itself — the works-by-works employment record — stands untouched: the census counts the city whole, not works, and where the two instruments overlap they measure different things and are not to be harmonized.

Sources:

1. HathiTrust catalog record 101667888: Census of manufactures: 1923. Statistics for industries, states, and cities (item osu.32435000483883, full view) — located, not read.