Size of Units in Representation Election Cases Closed, Fiscal Year 1968

p. 249. 15 columns. The crop is cut off at the bottom: the table clearly continues past "200 to 299" (the listed "Number eligible to vote" figures sum to 316,165 against the printed total of 550,610, so size classes of 300 and over are on the part of the page not shown). The caption carries a raised footnote marker "1", but no footnote text is visible in the crop, so `footnotes` is empty. Decimal points: in this scan the period in the percent columns is not resolved, so the cells read like "23 9", "100 0", "0 6". These are decimal fractions (the per-column percents sum to 100.0 and the cumulative column runs 23.9, 45.0, 57.3, ...), so I have written them with the decimal point, e.g. "23.9". Likewise "8 524" in "Number eligible to vote" for row "190 to 199" shows no comma in the scan; written as "8,524". Blank/dash cells transcribed as "": the "Cumulative percent of total" cell of the "Total RC and RM elections" row (dash rule), the "Other local unions / Percent by size class" cell of row "120 to 129" (number 0, dash rule), and the "Other national unions / Percent by size class" cells of rows "170 to 179" and "190 to 199" (number 0, dash rule). Row "Under 10", column "Other national unions / Percent by size class": a small stray speck precedes the value (prints as "·20 1"); I read it as dirt, not a footnote marker or digit, and transcribed 20.1 (49/244 = 20.1 percent, consistent). Arithmetic cross-checks all pass: every size-class row's five outcome counts sum to its "Total elections" figure, and the five outcome totals sum to 7,618.

p. 250. CROP / STRUCTURE. This is a continuation page, so no column headings are printed anywhere on the image and header_rows is empty. The column count (15: stub + 14 data columns) was taken from the vertical column rules, which run the full height of both panels, and is confirmed by the "Total RD elections" row, which fills all 14 data columns. The first nine rows are the tail of the preceding section (size-of-unit classes, cut off above "300 to 399"); the centred rule-bounded line "B DECERTIFICATION ELECTIONS (RD)" is transcribed as a label-only row; everything below it is the RD panel. Data columns alternate number / percent in seven pairs, except that columns 3-5 are number / percent / cumulative percent. The running head "Appendix A" and page number 237 (printed sideways in the right margin) are omitted. DECIMAL POINTS — the main caveat. Almost every percent cell in this scan shows the decimal point as a blank gap rather than a dot (e.g. "1 5", "97 1", "100 0", "25 0"). I have transcribed these exactly as they appear, with a single space in place of the missing dot, rather than inserting a character I cannot see. Only four cells retain a visible dot: "2.5" (row "300 to 399" of the upper panel, column 11), "21.8" (row "10 to 19", column 4), "84.6" (row "90 to 99", column 5) and the "98.4"-like reading in row "400 to 499" of the RD panel, column 5, which I read as "98 4" because the mark there is a speck rather than a clear point. Every such cell is a one-decimal percentage, so the space can be converted to a decimal point mechanically in post-processing. The "98 0" / "98 4" distinction is not affected. DASH RULES. A cell holding only a dashed rule is "". These are frequent in the percent columns opposite a printed 0 count, and in the cumulative-percent column of the "Total RD elections" row. DOUBTFUL CELLS. (a) Row label "120 to 129": the final digit is broken and reads almost like a 0; I transcribe 9 because the surrounding sequence runs 110 to 119 / 130 to 139. (b) Row label "100 to 109": a stray speck sits above the first 0, giving an apostrophe-like appearance; the label is "100 to 109". (c) A small isolated ink speck sits in the gutter between columns 13 and 14 in the upper panel's "3,000 to 9,999" row; it is not a value and has been ignored. (d) In the upper panel's last two rows, columns 6-7 read "0" / dash rule for "3,000 to 9,999" and "1" / "0 0" for "10,000 and over"; the two are vertically close and worth a spot check. NO ARITHMETIC CHECKS were applied and nothing was corrected; the cumulative-percent column reaching exactly "100 0" at the last class in each panel is as printed.