Disposition of unfair labor practice cases of C. I. O. unions on docket during the fiscal year ending June 30, 1939

p. 46. Seven columns: stub plus "Number of cases", two "Percentage of—" columns (Total cases closed; Total cases on docket), "Number of workers involved", and two more "Percentage of—" columns (Total number of workers involved in cases closed; ... in cases on docket). The stub column has no printed heading, so cell 0 of both heading levels is empty. "Number of cases" and "Number of workers involved" are tall single headings spanning both levels and are repeated in each. Apparent misprints, transcribed as printed: (1) the row label "Total cases closed July 1, 1938, to 30, 1939" is missing the month before "30, 1939" (the other rows read "June 30"); the second printed line begins "30, 1939" after a line break. (2) Row "By intermediate report finding no violation", last column (workers involved, percent of total on docket) reads ".3"; 263 of 948,456 is about 0.03 percent, so this value looks wrong on the page, but ".3" is what is printed. (3) Row "Cases pending June 30, 1939", last column reads "72.9" while 692,360/948,456 rounds to 73.0. Faint/doubtful glyphs resolved from arithmetic: "By settlement before hearing", Number of cases, prints like "1S" and is read as 18 (18/1,924 = 0.9 percent, matching the adjacent cell); "Cases received July 1, 1938, to June 30, 1939" prints the day like "3C" and is read as 30. In the row "Total cases closed before issuance of complaint" the last cell "22.4" is followed by a small ink mark at the right rule, which appears to be a scanning artifact rather than a footnote marker. Dash-rule cells (blank percentage cells for the pending/received/docket rows, and the two percentage cells on the "Total cases closed July 1, 1938, to 30, 1939" row) are given as empty strings. All internal sums and the column percentages otherwise check out against the printed figures.