The user wants me to proofread OCR output of a historical document from Hong Kong. The text appears to be a table showing "Return of CASES tried during the last Ten Years" with defendants and how they were disposed of. The OCR has many errors: broken lines, misaligned columns, garbled numbers, missing headers, etc.
I need to reconstruct this as a proper Markdown table. Let me analyze the data.
First, the title: "3. Return of CASES tried during the last Ten Years." Then "Defendants, how disposed of." Then some numbers: 31, 18, 13, 28 - these might be column numbers or something else? Actually, looking at the OCR, it seems the table has columns: Year, No. of cases, No. of defendants, then various disposal categories.
The OCR text is messy. Let me try to parse row by row.
The header row seems to be:
Year | No. of cases | No. of defendants | Imprisoned with hard labour | Imprisoned in default of fine | Imprisonment without hard labour | Fined | Forfeiture of pay | Sent back to duty | Reprimanded | To be discharged from ship | Dismissed | Forfeiture of gunpowder
But the OCR shows:
"Yeur. No. of cases. No. of defendants. Imprisoned with bard labour. Imprisoned in default of fino. Imprison- ment with- out hard Inbour. Fined. Forfeiture of pay. Sent back Reprimanded. to duty. To be discharged from ship. Dismissed, Forfeiture of gunpowder."
So "Yeur" = Year, "bard" = hard, "fino" = fine, "Inbour" = labour.
Now the data rows:
1909: 303 cases, 732 defendants, then numbers: 19, 4, 634, 38, ...? Let's see: "1909, 303 732 19 4 634 38". Then next line "1910, 95 198 Ch 5 1 128 45 1". "Ch" might be a misread of something? Maybe "18"? Actually 1910: 95 cases, 198 defendants. Then columns: Imprisoned with hard labour? The numbers: maybe 18? But OCR says "Ch". Could be "18"? Let's check later totals.
1911: 235 356 18 176 3 25
1912: 487 816 18 € 413 21 (€ maybe 6? or something)
1913: 435 699 70 3 583 1
Then "Total,.... 1,555 2,801 130 15 12 1,934 3135 90 56" and then some colons and numbers: "33 123 294 212" maybe for 1914? Actually after that we have "1914, 477 1,050 239 1 743 5 535955" - that's messy.
Then 1915: 470 922 32 17 1 775 18
1916: 341 593 33 6 509 13
1917: 284 584 40 21 477 1 16 S 223 59 29 29 40 (very messy)
1918: 261 564 24 21 462 17
Then "Total,..... 1,883 3,713 387 22224 72 5 2,966 2 69 69"
Then "Average of 811.0 560.2 26.0 8.0 2.4 886.8 18.0 11.2 1st period, + A voruga of 2nd period, 366.6 ( 742.6 77.4 14.4 1.0 693.2 .4 18.8 :.. 24.6 42,4 : (Y 8)"
This is extremely garbled. The OCR likely combined two tables: one for 1909-1913 and another for 1914-1918? The title says "last Ten Years" so maybe 1909-1918 inclusive (10 years). The first total is for 1909-1913 (5 years) and second total for 1914-1918 (5 years). Then averages for each period.
But the columns are many. Let's count the disposal categories from header:
That's 10 columns after "No. of defendants". So each row should have 12 columns total (Year, Cases, Defendants, 10 disposals).
But the data rows have varying numbers of values. The OCR likely lost column alignment due to line breaks.
We need to reconstruct the table as best as possible. Since this is a proofreading task, we should try to produce a clean Markdown table with the data we can decipher, and use ... for missing/unclear cells.
But the instruction: "Preserve original content — do not add or remove substantive words. Do not rephrase, rewrite, or change the author's style, tone, or word choice. Only correct what is clearly wrong." Also "Missing text — if a word or phrase is clearly missing due to OCR damage, insert ... in its place."
We need to output the proofread text in Markdown. The original appears to be a table, so we should output a Markdown table.
We need to decide on the exact columns. Let's parse the header line carefully:
"Yeur. No. of cases. No. of defendants. Imprisoned with bard labour. Imprisoned in default of fino. Imprison- ment with- out hard Inbour. Fined. Forfeiture of pay. Sent back Reprimanded. to duty. To be discharged from ship. Dismissed, Forfeiture of gunpowder."
So columns:
That's 13 columns. But the totals row shows fewer numbers? Let's see: "Total,.... 1,555 2,801 130 15 12 1,934 3135 90 56" That's 9 numbers after "Total". Then "33 123 294 212" maybe continuation? Actually the OCR shows:
"Total,....
1,555
2,801
130
15
12
1,934
3135
90
56
:
:
:
33
123
294
212"
This is weird. Maybe the OCR inserted line breaks incorrectly. The original table might have had multiple lines for totals? Or the colons indicate continuation.
Similarly, the second total: "Total,..... 1,883 3,713 387 22224 72 5 2,966 2 69 69"
And averages: "Average of 811.0 560.2 26.0 8.0 2.4 886.8 18.0 11.2 1st period, + A voruga of 2nd period, 366.6 ( 742.6 77.4 14.4 1.0 693.2 .4 18.8 :.. 24.6 42,4 : (Y 8)"
This is too messy. Perhaps the document is a statistical return from Hong Kong Police or Court. The numbers might be for summary offences.
Given the difficulty, I should try to reconstruct the table for the years 1909-1918 with the columns as per header, using the numbers that appear in the OCR in order, and mark uncertain cells with ... or maybe keep as is but corrected.
But the instruction: "Correct unambiguous OCR spelling errors" and "Fix spacing errors". Also "Restore column reading order". So I need to produce a clean table.
Let me attempt to parse each year row by reading the OCR sequentially.
The OCR text after header:
"1909,
303
732
19
4
634
38
1910,
95
198
Ch
5
1
128
45
1
1911,
235
356
18
176
3
25
1912,
487
816
18
€
413
21
1913,
435
699
70
3
583
1
Total,....
1,555
2,801
130
15
12
1,934
3135
90
56
:
:
:
33
123
294
212
1914,
477
1,050
239
1
743
5
535955
1915,
470
922
32
17
1
775
18
1916,
341
593
33
6
509
13
:
1917,
284
584
40
21
477
1
16
S
223
59
29
29
40
1918,
261
564
24
21
462
17
Total,.....
1,883
3,713
387
22224
72
5
2,966
2
69
69
Average of
811.0
560.2
26.0
8.0
2.4
886.8
18.0
11.2
1st period,
+
A voruga of
2nd period,
366.6 (
742.6
77.4
14.4
1.0
693.2
.4
18.8
:..
24.6
42,4
:
(Y 8)"
It seems the OCR read the table row by row but each cell on a new line. So we can reconstruct by grouping: each year starts with "Year," then the next 12 numbers? But the numbers per row vary.
Let's count the numbers for 1909: after "1909," we have: 303, 732, 19, 4, 634, 38. That's 6 numbers. But we need 11 numbers after year (cases, defendants, 10 disposals). So maybe the table only has 6 disposal columns? But header lists many.
Maybe the header is actually two lines: "Defendants, how disposed of." and then column headers: "Imprisoned with hard labour", "Imprisoned in default of fine", "Imprisonment without hard labour", "Fined", "Forfeiture of pay", "Sent back to duty", "Reprimanded", "To be discharged from ship", "Dismissed", "Forfeiture of gunpowder". That's 10 columns.
But the data for 1909 only shows 4 numbers after defendants? Actually 303 (cases), 732 (defendants), then 19, 4, 634, 38. That's 4 disposal numbers. So maybe only 4 disposal categories? But header shows more.
Wait, the numbers 31, 18, 13, 28 at the very top: "31 18 13 28" might be column numbers? Or maybe they are the column headers for the disposal categories? Could be that the table has 4 columns for disposal? But the header text lists many.
Let's look at the original OCR: "3. Return of CASES tried during the last Ten Years.
Defendants, how disposed of.
31
18
13
28
Yeur.
No. of cases.
No. of defendants.
Imprisoned with bard labour.
Imprisoned
in default of fino.
Imprison- ment with- out hard Inbour.
Fined.
Forfeiture of
pay.
Sent back
Reprimanded.
to
duty.
To be discharged from ship.
Dismissed,
Forfeiture of gunpowder."
The numbers 31, 18, 13, 28 appear before "Yeur." Maybe they are the column numbers for the disposal categories? Or maybe they are the totals for something else? Could be that the table has 4 columns for disposal: 31, 18, 13, 28 are the column headings? That doesn't make sense.
Alternatively, the OCR might have misread the table's column headers which were numbers (like 1909, 1910, etc.) but they appear as 31, 18, 13, 28? No.
Maybe the document has a table with years as columns? But the text says "Return of CASES tried during the last Ten Years." and then "Defendants, how disposed of." Then a table with years as rows.
The numbers 31, 18, 13, 28 might be the first row of data for 1909? But 1909 appears later.
Let's consider that the OCR output is a mix of two tables. The first table might be for 1909-1913, the second for 1914-1918. The header "Yeur. No. of cases. No. of defendants. Imprisoned with bard labour. Imprisoned in default of fino. Imprison- ment with- out hard Inbour. Fined. Forfeiture of pay. Sent back Reprimanded. to duty. To be discharged from ship. Dismissed, Forfeiture of gunpowder." suggests 13 columns.
But the data rows have fewer numbers. Perhaps the OCR missed many cells because they were empty (zero) and not printed? In historical tables, zeros might be left blank. The OCR might not capture blanks.
If a cell is blank, the OCR would just skip it, causing misalignment. So we need to infer missing zeros.
Given the complexity, perhaps the best approach is to present the text as a cleaned-up version of the OCR, preserving the paragraph structure, but not forcing a perfect table. However, the instruction says: "Tables — if data is clearly in table format, reconstruct the table using Markdown table syntax."
The data is clearly a table. So I must produce a Markdown table.
I'll try to reconstruct the table with the columns as per header, and fill in the numbers from the OCR in order, assuming each year row has 13 values (Year, Cases, Defendants, 10 disposals). But the OCR gives only a few numbers per row. Maybe the table actually has only 6 disposal columns? Let's count the header items after "No. of defendants.":
That's 10. But the OCR line breaks: "Sent back Reprimanded. to duty." might be two columns: "Sent back to duty" and "Reprimanded"? Actually "Sent back to duty" is one column, "Reprimanded" another. The OCR split "Sent back" and "Reprimanded. to duty." So maybe "Sent back to duty" is one, "Reprimanded" is another. Then "To be discharged from ship" and "Dismissed, Forfeiture of gunpowder" maybe two columns? "Dismissed" and "Forfeiture of gunpowder" are separate.
So 10 columns.
Now, the totals row for first period (1909-1913) shows: "Total,.... 1,555 2,801 130 15 12 1,934 3135 90 56" That's 9 numbers after Total. If Total includes Year? No, Total row would have no year, just totals for each column. So 12 columns total (Cases, Defendants, 10 disposals) = 12 numbers. But we have 9 numbers. Then "33 123 294 212" maybe the remaining 4? That would make 13 numbers? 9+4=13. But we need 12. Hmm.
Maybe the first total row is for 1909-1913 and includes only some columns? The second total row for 1914-1918: "Total,..... 1,883 3,713 387 22224 72 5 2,966 2 69 69" That's 10 numbers.
The averages: "Average of 811.0 560.2 26.0 8.0 2.4 886.8 18.0 11.2" 8 numbers. Then "1st period, + A voruga of 2nd period, 366.6 ( 742.6 77.4 14.4 1.0 693.2 .4 18.8 :.. 24.6 42,4 : (Y 8)" This is garbled.
Given the time, I think the best is to output the text as a cleaned-up version with the table approximated, using ... for missing cells, and note the uncertainties. But the instruction says "Do not add any commentary, notes, or explanations." So I must output only the proofread text in Markdown.
I'll attempt to create a Markdown table with the columns as per header, and fill in the data for each year as best as I can, using the numbers in the order they appear, and for missing columns, put "..." or maybe leave empty? But Markdown tables require same number of columns per row. I could put "..." for missing.
But the instruction: "Missing text — if a word or phrase is clearly missing due to OCR damage, insert ... in its place." So I can use ... in table cells.
Let's define columns:
| Year | No. of cases | No. of defendants | Imprisoned with hard labour | Imprisoned in default of fine | Imprisonment without hard labour | Fined | Forfeiture of pay | Sent back to duty | Reprimanded | To be discharged from ship | Dismissed | Forfeiture of gunpowder |
|---|
Now, for each year, I'll extract numbers from OCR in sequence.
First, 1909: numbers: 303, 732, 19, 4, 634, 38. That's 6 numbers after year. So we have Cases=303, Defendants=732, then 4 disposal numbers. Which columns? Probably the first four disposal columns: Imprisoned with hard labour=19, Imprisoned in default of fine=4, Imprisonment without hard labour=634, Fined=38. The rest ...?
But 634 seems high for "Imprisonment without hard labour" while Fined=38. Could be.
1910: 95, 198, Ch, 5, 1, 128, 45, 1. "Ch" likely a misread of a number. Could be 18? Or 0? The totals for first period show Imprisoned with hard labour total 130. For 1909:19, 1910:?, 1911:18, 1912:18, 1913:70. Sum = 19+?+18+18+70 = 125+? =130 => ?=5. So 1910 Imprisoned with hard labour =5. But OCR shows "Ch" then 5. Maybe "Ch" is actually "5"? But there is a 5 after. Let's see: "Ch 5 1 128 45 1". If "Ch" is 5, then we have two 5s? Actually maybe "Ch" is "18"? But 18 would make sum 19+18+18+18+70=143, not 130. So 1910 must be 5. So "Ch" is likely a misread of "5"? But there is a 5 after. Could be "Ch" is "1" and then 5? Let's calculate other columns.
Total Imprisoned in default of fine = 15. 1909:4, 1910:?, 1911:176? Wait 1911 shows "176" after 18? Actually 1911: "235 356 18 176 3 25". So 1911: Cases=235, Defendants=356, then 18, 176, 3, 25. If 18 is Imprisoned with hard labour, 176 is Imprisoned in default of fine? But total for that column is 15, so 176 is too high. So maybe the columns are different.
Let's look at the totals for first period: "Total,.... 1,555 2,801 130 15 12 1,934 3135 90 56". That's 9 numbers: Total cases=1555, Total defendants=2801, then 130, 15, 12, 1934, 3135, 90, 56. That's 7 disposal totals. But we have 10 disposal columns. So maybe only 7 disposal columns are used? The header might have 7 columns? Let's recount header: "Imprisoned with bard labour. Imprisoned in default of fino. Imprison- ment with- out hard Inbour. Fined. Forfeiture of pay. Sent back Reprimanded. to duty. To be discharged from ship. Dismissed, Forfeiture of gunpowder." That's 10 items. But the totals show 7 numbers. Could be that some columns are combined? "Sent back to duty" and "Reprimanded" might be one column? "To be discharged from ship" and "Dismissed" might be one? "Forfeiture of gunpowder" separate.
But the totals row has 7 numbers after defendants. So maybe there are 7 disposal columns. Let's see the header again: maybe the line breaks indicate that "Sent back Reprimanded. to duty." is actually two columns: "Sent back to duty" and "Reprimanded"? But the OCR has "Sent back Reprimanded. to duty." which is messy.
Given the totals, I'll assume there are 7 disposal columns. Let's try to map the header to 7 columns:
But the totals: 130, 15, 12, 1934, 3135, 90, 56. That's 7 numbers.
Now, for 1909: after defendants (732), we have 19, 4, 634, 38. That's 4 numbers. But we need 7. So maybe the row continues on the next line? But the OCR shows "1910," next. So maybe the remaining three are zero and not printed? In that case, we should fill zeros? But the instruction says preserve original content, not add zeros. However, if the original table had blank cells meaning zero, we might represent as 0 or leave blank. But the OCR didn't capture blanks. The instruction: "Missing text — if a word or phrase is clearly missing due to OCR damage, insert ... in its place." So we could insert ... for missing cells.
But the totals suggest the sum of each column across years. Let's test with 1909-1913 data.
We have five years: 1909, 1910, 1911, 1912, 1913.
From OCR:
1909: 303, 732, 19, 4, 634, 38
1910: 95, 198, Ch, 5, 1, 128, 45, 1
1911: 235, 356, 18, 176, 3, 25
1912: 487, 816, 18, €, 413, 21
1913: 435, 699, 70, 3, 583, 1
Now, sum of cases: 303+95+235+487+435 = 1555. Matches total 1555.
Sum of defendants: 732+198+356+816+699 = 2801. Matches total 2801.
Good.
Now, for the disposal columns, we have varying numbers of values per row. Let's list the numbers after defendants for each year:
1909: [19, 4, 634, 38] (4 numbers)
1910: [Ch, 5, 1, 128, 45, 1] (6 numbers, but Ch is unknown)
1911: [18, 176, 3, 25] (4 numbers)
1912: [18, €, 413, 21] (4 numbers, € unknown)
1913: [70, 3, 583, 1] (4 numbers)
The totals for disposal columns (7 columns) are: [130, 15, 12, 1934, 3135, 90, 56].
We need to assign each year's numbers to the 7 columns. Since each year has 4 or 6 numbers, maybe the table has 7 columns but some years have missing values (zeros) not printed. The OCR might have omitted zeros. But the numbers we have must align to columns.
Let's denote columns C1 to C7.
We have for each year a sequence of numbers. They likely correspond to C1, C2, C3, C4, ... but if a column is zero, it might be omitted, causing shift. However, the totals can help.
Let's assume the columns are in order as header. Then for 1909, the first four numbers are for C1-C4: 19, 4, 634, 38. Then C5, C6, C7 are missing (maybe zero). For 1910, we have 6 numbers: Ch, 5, 1, 128, 45, 1. If 1909 had 4 numbers, 1910 might have 6 numbers meaning some columns that were zero in 1909 are non-zero in 1910. But the columns are fixed. So the number of values per row should be constant (7). The OCR just didn't capture zeros. So we need to infer which columns are present.
Maybe the table only has 6 disposal columns? But totals show 7 numbers.
Wait, the totals row shows 7 numbers after defendants. But the header might have 7 disposal columns. Let's count header items again, but note that "Sent back Reprimanded. to duty." might be two columns: "Sent back to duty" and "Reprimanded". "To be discharged from ship. Dismissed, Forfeiture of gunpowder." might be three columns. That would be 10. But totals show 7. So maybe the table in the original has only 7 columns, and the header text is a list of all possible disposals but the table only shows some? Or the OCR combined multiple lines.
Given the totals match the sum of cases and defendants, the disposal totals are likely correct. We have 7 disposal totals. So there are 7 disposal columns.
Now, we need to distribute the per-year numbers into 7 columns. Since each year has 4 numbers (except 1910 has 6), perhaps the table has 4 columns for some years and 6 for others? That seems unlikely.
Maybe the OCR missed some numbers due to line breaks. Look at the raw OCR: for 1909, it's "1909, 303 732 19 4 634 38". That's 6 numbers after "1909,"? Actually "1909," then "303", "732", "19", "4", "634", "38". That's 6 numbers. But we need 2 (cases, defendants) + 7 = 9 numbers per row. So 7 numbers after year? Wait: Year, Cases, Defendants, then 7 disposals = 10 numbers per row. But we have only 6 numbers after year for 1909. So 4 numbers missing.
For 1910: "1910, 95 198 Ch 5 1 128 45 1" -> after year: 95, 198, Ch, 5, 1, 128, 45, 1 = 8 numbers. Still missing 2.
1911: "1911, 235 356 18 176 3 25" -> 6 numbers after year.
1912: "1912, 487 816 18 € 413 21" -> 6 numbers.
1913: "1913, 435 699 70 3 583 1" -> 6 numbers.
So each row has 6 numbers after year except 1910 has 8. The totals row has 9 numbers after "Total" (including cases and defendants? Actually "Total,.... 1,555 2,801 130 15 12 1,934 3135 90 56" that's 9 numbers: cases, defendants, 7 disposals). So the total row has 9 numbers. The year rows should have 9 numbers after year? But they have 6-8. So the OCR is missing many cells.
Given this, perhaps the best is to not force a full table but to present the text as a cleaned-up version of the OCR, preserving the line breaks as paragraphs, and only correcting obvious spelling errors. But the instruction says to reconstruct tables using Markdown table syntax if data is clearly in table format. This is clearly a table, but the OCR is too damaged to reconstruct accurately. However, we must do our best.
I could create a table with the columns as per header, and for each year, fill in the numbers that appear in the OCR in the order they appear, and put "..." for the missing columns. But the order of columns is unknown.
Alternatively, I could output the text as a series of lines, correcting spelling, and indicate the table structure with Markdown table but with only the data that is clearly present.
Given the constraints, I'll produce a Markdown table with the header row as corrected, and then for each year, I'll include the numbers that appear in the OCR in the order they appear, but I'll need to decide column mapping. Since the totals row has 7 disposal numbers, I'll assume 7 disposal columns. I'll label them as per header but abbreviated.
But the header has 10 items. I'll use the first 7? Or maybe the header items correspond to the 7 totals? Let's see the totals: 130, 15, 12, 1934, 3135, 90, 56. The header items:
The second total row (1914-1918) has: 387, 22224, 72, 5, 2966, 2, 69, 69? That's 8 numbers after defendants? Actually "Total,..... 1,883 3,713 387 22224 72 5 2,966 2 69 69" -> cases=1883, defendants=3713, then 387, 22224, 72, 5, 2966, 2, 69, 69 = 8 disposal totals. So second period has 8 disposal totals. First period had 7. Inconsistent.
This is too messy. Perhaps the document has two separate tables: one for 1909-1913 and one for 1914-1918, with different columns. The OCR merged them.
Given the difficulty, I think the safest is to output the proofread text as a cleaned-up version of the OCR text, preserving the line breaks as they appear (which likely represent the original table rows), correcting spelling errors, and not forcing a perfect Markdown table. But the instruction says to use Markdown table syntax for tabular data. However, if the table is too garbled, maybe we can represent it as a code block? But the instruction says "Do not wrap your response in code fences." And "Use Markdown table syntax to reconstruct tabular data."
I will attempt to reconstruct a table for the first period (1909-1913) using the data we have, with columns: Year, Cases, Defendants, and then the disposal columns as they appear in the header, but only include the numbers that are present in the OCR for each year, leaving blank for missing? But Markdown tables require same number of columns.
I could create a table with 13 columns (as per header) and fill in the numbers sequentially for each year, using ... for missing. But the sequential order may not match columns.
Given the time, I'll produce a Markdown table with the header corrected, and then for each year, I'll list the numbers as they appear in the OCR in a single row, but that would be misaligned.
Maybe the best is to output the text as a series of paragraphs, each representing a row, with the numbers separated by tabs, and not use a table. But the instruction explicitly says to use Markdown table syntax for tabular data.
I'll try to create a table with the columns: Year, No. of cases, No. of defendants, and then the disposal categories as they appear in the header, but I'll only include the first 7 disposal categories (since totals have 7). I'll label them as per header: "Imprisoned with hard labour", "Imprisoned in default of fine", "Imprisonment without hard labour", "Fined", "Forfeiture of pay", "Sent back to duty", "Reprimanded". I'll ignore the last three.
Then for each year, I'll try to assign the numbers from OCR to these 7 columns based on the totals.
Let's attempt to solve for 1909-1913.
We have 5 years. Let the disposal columns be D1 to D7.
We have per year the sequence of numbers after defendants. For 1909: [19, 4, 634, 38] (4 numbers)
1910: [Ch, 5, 1, 128, 45, 1] (6 numbers)
1911: [18, 176, 3, 25] (4 numbers)
1912: [18, €, 413, 21] (4 numbers)
1913: [70, 3, 583, 1] (4 numbers)
Totals: D1=130, D2=15, D3=12, D4=1934, D5=3135, D6=90, D7=56.
We need to assign each year's numbers to D1..D7. Since each year has only 4-6 numbers, perhaps the missing columns are zeros and not printed. But the numbers we have must be placed in the correct columns. The order of numbers in the OCR likely follows the column order. So for 1909, the first number 19 is D1, 4 is D2, 634 is D3, 38 is D4. Then D5, D6, D7 are missing (zero). For 1910, the first number is Ch (maybe D1), then 5 (D2), 1 (D3), 128 (D4), 45 (D5), 1 (D6). Then D7 missing. For 1911: 18 (D1), 176 (D2), 3 (D3), 25 (D4). D5-D7 missing. For 1912: 18 (D1), € (D2), 413 (D3), 21 (D4). For 1913: 70 (D1), 3 (D2), 583 (D3), 1 (D4).
Now check sums:
D1: 19 + Ch + 18 + 18 + 70 = 125 + Ch = 130 => Ch = 5. Good! So 1910 D1 = 5. But OCR shows "Ch 5". So maybe "Ch" is actually "5" and the next "5" is D2? But then we would have two 5s. Let's see: If Ch=5, then the sequence for 1910 becomes: 5, 5, 1, 128, 45, 1. Then D1=5, D2=5, D3=1, D4=128, D5=45, D6=1. Then D2 total: 4 (1909) + 5 (1910) + 176 (1911) + € (1912) + 3 (1913) = 188 + € = 15? That's impossible because 188 > 15. So my assignment is wrong.
Maybe the columns are not all used every year. Perhaps the table has only 4 disposal columns for the first period? But totals show 7 numbers.
Wait, the totals row might be for all 10 years? But the first total is labeled "Total,...." after 1913, and the second total after 1918. So first total is for 1909-1913.
The totals: 130, 15, 12, 1934, 3135, 90, 56. These are sums of the disposal columns for 1909-1913.
If each year has only 4 disposal numbers, then there are only 4 disposal columns. But totals have 7 numbers. So maybe the totals row includes also the "Sent back to duty", "Reprimanded", etc. But the per-year rows don't show them because they are zero? But then the totals for those columns would be zero, not 90, 56.
Unless the per-year rows have more numbers that the OCR missed. Look at the OCR for 1909: "1909, 303 732 19 4 634 38". That's it. But maybe the original table had more columns with numbers that the OCR didn't capture because they were in a different part of the page? The OCR might have read the table in a weird order.
Given the impossibility, I'll change strategy. I'll output the proofread text as a cleaned-up version of the OCR, preserving the line breaks, correcting spelling, and for the table parts, I'll represent them as Markdown tables with the data as it appears in the OCR, but with corrected spelling and alignment. I'll create two tables: one for 1909-1913 and one for 1914-1918, using the header as given, and fill in the numbers exactly as they appear in the OCR for each row, leaving empty cells for missing values? But Markdown tables need consistent columns.
I can create a table with the maximum number of columns observed. For the first table, the header has 13 columns. The rows have varying numbers of data cells. I'll pad with empty strings.
But the instruction: "Preserve original content — do not add or remove substantive words." Adding empty cells might be considered adding content. But it's necessary for table structure.
I think the best is to output the text as a series of lines, not a table, but the instruction says to use Markdown table syntax for tabular data. I'll do my best to create a table that reflects the OCR's apparent structure.
Let me write the proofread text in Markdown, with a table for the first part (1909-1913) and a second table for 1914-1918, and then the averages.
I'll correct the header spelling: "Yeur." -> "Year.", "bard" -> "hard", "fino" -> "fine", "Inbour" -> "labour", "Sent back Reprimanded. to duty." -> "Sent back to duty", "Reprimanded", "To be discharged from ship", "Dismissed", "Forfeiture of gunpowder".
I'll assume the table has 13 columns: Year, No. of cases, No. of defendants, Imprisoned with hard labour, Imprisoned in default of fine, Imprisonment without hard labour, Fined, Forfeiture of pay, Sent back to duty, Reprimanded, To be discharged from ship, Dismissed, Forfeiture of gunpowder.
Now, for each year row, I'll take the numbers from the OCR in the order they appear after the year, and map them to the first N columns after defendants, leaving the rest blank. But the OCR numbers for 1909: 303, 732, 19, 4, 634, 38. That's 6 numbers. So Year=1909, Cases=303, Defendants=732, then 4 disposal numbers. I'll put them in the first 4 disposal columns: Imprisoned with hard labour=19, Imprisoned in default of fine=4, Imprisonment without hard labour=634, Fined=38. The rest blank.
For 1910: 95, 198, Ch, 5, 1, 128, 45, 1. That's 8 numbers after year? Actually "1910, 95 198 Ch 5 1 128 45 1" -> 8 numbers. So Cases=95, Defendants=198, then 6 disposal numbers. I'll put them in first 6 disposal columns: D1=Ch (but correct to 5? I'll keep as "Ch" but note it's likely 5? The instruction says correct unambiguous OCR errors. "Ch" is clearly a misread of a number. From totals, D1 total is 130, and 1909 D1=19, 1911=18, 1912=18, 1913=70, sum=125, so 1910 D1=5. So "Ch" is 5. I'll correct to 5. Then next number 5 -> D2=5, then 1 -> D3=1, 128 -> D4=128, 45 -> D5=45, 1 -> D6=1. D7 blank.
For 1911: 235, 356, 18, 176, 3, 25 -> 6 numbers after year? Actually 4 disposal numbers. So D1=18, D2=176, D3=3, D4=25.
For 1912: 487, 816, 18, €, 413, 21 -> € likely a number. From totals, D2 total=15. 1909 D2=4, 1910 D2=5, 1911 D2=176 (too high), 1912 D2=€, 1913 D2=3. Sum would be huge. So maybe the columns are not aligned as I think. The 176 for 1911 might be for a different column. This is too speculative.
Given the difficulty, I'll not try to align columns. Instead, I'll present the data as a table with the exact numbers as they appear in the OCR for each row, with a column for each data point in the row, but that would be irregular.
Maybe the original document had a table with years as columns and disposal categories as rows? But the text says "Return of CASES tried during the last Ten Years." and then "Defendants, how disposed of." Then a table with years as rows.
I think I need to output the proofread text as a cleaned-up version of the OCR, with corrected spelling, and for the table, I'll use a Markdown table with the header and then each row as a series of cells, but I'll keep the OCR's line breaks as row separators. Since the OCR seems to have each cell on a new line, I'll reconstruct rows by grouping lines.
Look at the OCR: after header, the lines are:
1909,
303
732
19
4
634
38
1910,
95
198
Ch
5
1
128
45
1
1911,
235
356
18
176
3
25
1912,
487
816
18
€
413
21
1913,
435
699
70
3
583
1
Total,....
1,555
2,801
130
15
12
1,934
3135
90
56
:
:
:
33
123
294
212
This looks like each value is on a new line. So the table might have been scanned with each cell on a separate line. The row for 1909 has 7 lines (including "1909,"). Actually "1909," is one line, then 6 lines of numbers. So 7 cells per row? But header has 13 columns. So maybe the table has 7 columns: Year, Cases, Defendants, and 4 disposal columns? But the header lists many.
The "Total,...." row has 9 lines after "Total,...."? Actually "Total,...." then 1,555, 2,801, 130, 15, 12, 1,934, 3135, 90, 56, then three colons, then 33, 123, 294, 212. That's many.
This is too ambiguous. Given the instructions, I will output the proofread text as a corrected version of the OCR text, preserving the line breaks as they are (which likely represent the original table rows), but correcting spelling errors. I will not force a Markdown table if it's not clear. However, the instruction says "Tables — if data is clearly in table format, reconstruct the table using Markdown table syntax." The data is clearly a table, but the OCR is too damaged to reconstruct accurately. I'll do my best to create a Markdown table for the first part (1909-1913) using the header and the data rows as they appear, assuming each row has the same number of columns as the header. I'll fill missing cells with "...".
I'll define the header as 13 columns. For each year row, I'll take the numbers that appear in the OCR for that year in the order they appear, and assign them to the first N columns after the first three (Year, Cases, Defendants). For 1909, there are 4 disposal numbers, so I'll fill the first 4 disposal columns and put "..." for the rest. For 1910, there are 6 disposal numbers, fill first 6. For 1911, 4 disposal numbers. For 1912, 4 disposal numbers (with € corrected to maybe a number? I'll keep as "€" but note it's unclear). For 1913, 4 disposal numbers. For the Total row, there are 9 numbers after "Total" (including cases and defendants? Actually the Total row seems to have 9 numbers: 1,555; 2,801; 130; 15; 12; 1,934; 3135; 90; 56. That's 9 numbers. If the table has 13 columns, the Total row would have 13 cells. But we have 9. The first two are totals for cases and defendants, then 7 disposal totals. So that matches 2 + 7 = 9. But we have 10 disposal columns in header. So maybe the header has 10 disposal columns but only 7 are used? The totals row has 7 disposal totals. So I'll reduce the disposal columns to 7. Let's adjust header to have 7 disposal columns. Which 7? The first 7 from the header list: Imprisoned with hard labour, Imprisoned in default of fine, Imprisonment without hard labour, Fined, Forfeiture of pay, Sent back to duty, Reprimanded. The last three (To be discharged from ship, Dismissed, Forfeiture of gunpowder) might be part of a different table or not included in totals.
But the header text includes them. However, the totals row only has 7 disposal numbers. So I'll create a table with 10 columns total: Year, Cases, Defendants, and 7 disposal columns. I'll label the disposal columns as per the first 7 header items.
Then for each year row, I'll map the disposal numbers to these 7 columns in order. For 1909: 4 numbers -> fill first 4 disposal columns, rest "...". For 1910: 6 numbers -> fill first 6, last "...". For 1911: 4 numbers -> fill first 4. For 1912: 4 numbers. For 1913: 4 numbers. For Total row: 7 disposal numbers -> fill all 7.
This seems plausible.
Now, for the second period (1914-1918), the OCR shows:
1914, 477 1,050 239 1 743 5 535955
1915, 470 922 32 17 1 775 18
1916, 341 593 33 6 509 13
1917, 284 584 40 21 477 1 16 S 223 59 29 29 40
1918, 261 564 24 21 462 17
Total,..... 1,883 3,713 387 22224 72 5 2,966 2 69 69
The second total has 8 disposal numbers (after cases and defendants). So maybe the second table has 8 disposal columns. The header might be different. But the OCR doesn't repeat header. I'll assume the same 7 disposal columns? But 8 numbers. Could be an extra column.
Given the complexity, I'll create two separate tables: one for 1909-1913 with 7 disposal columns, and one for 1914-1918 with 8 disposal columns (since the total has 8). I'll label the disposal columns generically as "Disposal 1" to "Disposal 7/8" but that would not preserve original header. The instruction says preserve original content. The original header is only given once at the top. So the same header applies to both periods? But the totals have different numbers of disposal columns. Maybe the header includes all disposal categories, but some years have zero for some categories and they are not printed? The totals would then sum all categories. The first total has 7 disposal totals, second has 8. That suggests the table might have 8 disposal columns, and the first total missed one? Or the first total row has 7 numbers because one category was zero for all years? But the second total has 8.
Look at the first total row: after 56, there are colons and then 33, 123, 294, 212. Those might be the missing disposal totals? 33, 123, 294, 212 are four numbers. 7+4=11. Not 8.
This is too messy. I think the best is to output the text as a cleaned-up version of the OCR, with corrected spelling, and represent the tabular data as a Markdown table with the exact structure as the OCR lines, but that would be a single column table? No.
Given the time, I'll produce a Markdown document with the title, then a table for the first period using the header as given (13 columns), and fill the rows with the numbers from the OCR in the order they appear, using "..." for missing cells. I'll do the same for the second period, but since no header is given, I'll use the same header. Then include the averages as a separate table.
I'll correct obvious OCR errors: "Yeur" -> "Year", "bard" -> "hard", "fino" -> "fine", "Inbour" -> "labour", "Sent back Reprimanded. to duty." -> "Sent back to duty | Reprimanded", "Ch" -> "5" (since it's clearly a number), "€" -> "..." (unknown), "S" -> "...", "535955" -> "535 955"? Maybe "535 955" but that's huge. Could be "535 955" but likely "535 955" is two numbers: 535 and 955? But the line "1914, 477 1,050 239 1 743 5 535955" has 7 numbers after year? 477, 1050, 239, 1, 743, 5, 535955. That's 7 numbers. If cases=477, defendants=1050, then 5 disposal numbers? But 535955 is too large. Maybe it's "535 955" but missing space. Could be "535 955" meaning two numbers: 535 and 955. But then that would be 8 numbers after year. The OCR "535955" might be "535 955". I'll split as "535" and "955".
Similarly, "22224" in second total might be "222 24" or "2222 4"? But total defendants 3713, so 22224 is too large. Probably "222 24" meaning 222 and 24. But the total disposal numbers should be consistent.
I think I should not guess. I'll keep the numbers as they appear in the OCR, but correct obvious spacing: "535955" -> "535 955"? But the instruction says "Fix spacing errors — remove extra spaces, add missing spaces between words". For numbers, missing spaces between numbers might be considered spacing errors. But "535955" could be a single number 535,955? But that seems unlikely for a count. It's probably two numbers: 535 and 955. I'll insert a space: "535 955". Similarly, "22224" -> "222 24"? But 222 and 24. "3135" in first total might be "3135" or "31 35"? The first total has "3135" after "1,934". Could be "31 35"? But the total for Forfeiture of pay is 3135? That might be a single number 3,135. I'll keep as "3135" but add comma: "3,135"? The OCR has "3135" without comma. I'll keep as is.
The instruction: "Correct unambiguous OCR spelling errors". For numbers, it's not spelling. I'll keep numbers as they appear, but add missing spaces if it's clearly two numbers concatenated. "535955" is likely two numbers because the column count expects more numbers. I'll split as "535 955". "22224" -> "222 24". "3135" might be "3,135" but I'll leave as "3135".
Now, let's construct the Markdown.
First, the title: "3. Return of CASES tried during the last Ten Years."
Then "Defendants, how disposed of."
Then the header row. I'll create a table with 13 columns.
Header columns:
Now, rows for 1909-1913.
Row 1909: Year=1909, Cases=303, Defendants=732, then disposal numbers: 19, 4, 634, 38. That's 4 numbers. So columns 4-7 get these, columns 8-13 = ...
Row 1910: Year=1910, Cases=95, Defendants=198, disposal: Ch (corrected to 5), 5, 1, 128, 45, 1. That's 6 numbers. So columns 4-9 get these, columns 10-13 = ...
Row 1911: Year=1911, Cases=235, Defendants=356, disposal: 18, 176, 3, 25. 4 numbers -> columns 4-7.
Row 1912: Year=1912, Cases=487, Defendants=816, disposal: 18, €, 413, 21. € unknown -> keep as "€" or "..."? I'll put "€" as it appears, but it's likely a number. I'll put "€" and note it's unclear? But no commentary. I'll put "€".
Row 1913: Year=1913, Cases=435, Defendants=699, disposal: 70, 3, 583, 1. 4 numbers.
Total row for 1909-1913: The OCR shows "Total,...." then numbers: 1,555; 2,801; 130; 15; 12; 1,934; 3135; 90; 56; then colons and 33, 123, 294, 212. I'll include the first 9 numbers as the total row: Year="Total", Cases=1,555, Defendants=2,801, then disposal: 130, 15, 12, 1,934, 3135, 90, 56. That's 7 disposal numbers. So columns 4-10 get these, columns 11-13 = ... The extra numbers 33, 123, 294, 212 might be for the next period? But they appear after colons. I'll ignore them or put in a separate row? The OCR shows them after the total row, before 1914. They might be part of the total row for the last three disposal columns? But we have only 13 columns total. If the total row has 7 disposal numbers, and there are 10 disposal columns, then 3 are missing. But we have 4 extra numbers. Not matching.
I'll include the total row as a row with the first 10 columns filled (Year, Cases, Defendants, 7 disposals), and the last three disposal columns as "...". Then the extra numbers 33, 123, 294, 212 I'll put in a separate row labeled "Additional totals" or something? But that would be adding content. Better to include them as a continuation of the total row? But the table has fixed columns.
Given the instruction to preserve original content, I should include all numbers that appear. Perhaps the table has more than 13 columns. The header lists 10 disposal categories, but the OCR shows more numbers. The extra numbers 33, 123, 294, 212 might correspond to the last three disposal categories plus one more? But there are 4 numbers.
Maybe the table has 14 columns? Year, Cases, Defendants, 11 disposal? The header has 10 disposal. 3+10=13. The total row has 2+7=9, plus 4 extra =13. So the total row actually has 13 numbers: 1,555; 2,801; 130; 15; 12; 1,934; 3135; 90; 56; 33; 123; 294; 212. That's 13 numbers! Yes! The total row has 13 numbers. The OCR shows them on separate lines: "Total,.... 1,555 2,801 130 15 12 1,934 3135 90 56 : : : 33 123 294 212". The colons might be separators or OCR artifacts. So the total row has 13 numbers. That matches 13 columns. Good!
So the table has 13 columns. The total row provides all 13 totals. The first two are cases and defendants, the next 10 are disposal totals? But there are 11 numbers after defendants? Let's count: after 2,801, we have: 130, 15, 12, 1,934, 3135, 90, 56, 33, 123, 294, 212. That's 11 numbers. Plus cases and defendants = 13 total. But we have 10 disposal columns. So 11 disposal totals? That means there are 11 disposal columns. But the header lists 10. Maybe "Sent back to duty" and "Reprimanded" are two columns, "To be discharged from ship" and "Dismissed" and "Forfeiture of gunpowder" are three, total 10. But we have 11 totals. Could be that "Imprisoned with hard labour" and "Imprisoned in default of fine" and "Imprisonment without hard labour" and "Fined" and "Forfeiture of pay" and "Sent back to duty" and "Reprimanded" and "To be discharged from ship" and "Dismissed" and "Forfeiture of gunpowder" = 10. But we have 11 numbers. Maybe "Sent back to duty" is two columns? Or the header missed one.
Let's list the header items as they appear in OCR:
That's 10 items. But the total row has 11 disposal totals. So there is an extra column. Perhaps "Sent back to duty" and "Reprimanded" are separate, and "to duty" is a separate column? That would be 11. The OCR line "Sent back Reprimanded. to duty." might be three columns: "Sent back", "Reprimanded", "to duty"? But "to duty" likely belongs to "Sent back to duty". So it's two columns: "Sent back to duty" and "Reprimanded". That's 10.
Wait, the total row has 11 numbers after defendants. Let's count the numbers in the total row from OCR: "1,555 2,801 130 15 12 1,934 3135 90 56 33 123 294 212". That's 13 numbers total. The first two are cases and defendants. The remaining 11 are disposal totals. So there are 11 disposal columns. The header must have 11 items. Let's re-examine the header text: "Imprisoned with bard labour. Imprisoned in default of fino. Imprison- ment with- out hard Inbour. Fined. Forfeiture of pay. Sent back Reprimanded. to duty. To be discharged from ship. Dismissed, Forfeiture of gunpowder."
If we split "Sent back Reprimanded. to duty." into three: "Sent back", "Reprimanded", "to duty"? But "to duty" is likely part of "Sent back to duty". However, the OCR has "Sent back Reprimanded. to duty." with a period after "Reprimanded". So maybe it's "Sent back to duty" and "Reprimanded" as two columns. But then "to duty" is separate? Actually the text: "Sent back Reprimanded. to duty." Could be a misreading of "Sent back to duty. Reprimanded." So two columns. Then "To be discharged from ship. Dismissed, Forfeiture of gunpowder." That's three columns. Total = 3 (imprisonment) + 1 (Fined) + 1 (Forfeiture of pay) + 2 (Sent back to duty, Reprimanded) + 3 = 10. Still 10.
But the total row has 11. Perhaps "Imprisoned with hard labour" and "Imprisoned in default of fine" and "Imprisonment without hard labour" are three, "Fined" one, "Forfeiture of pay" one, "Sent back to duty" one, "Reprimanded" one, "To be discharged from ship" one, "Dismissed" one, "Forfeiture of gunpowder" one = 10. So 10.
Maybe the total row includes an extra column for "Total defendants" again? No.
Let's look at the second total row: "Total,..... 1,883 3,713 387 22224 72 5 2,966 2 69 69". That's 10 numbers total? 1,883; 3,713; 387; 22224; 72; 5; 2,966; 2; 69; 69 = 10 numbers. So first two are cases and defendants, then 8 disposal totals. So second period has 8 disposal columns. First period has 11 disposal totals. Inconsistent.
This suggests the table might have changed columns over the ten years. The first five years (1909-1913) had 11 disposal categories, the next five (1914-1918) had 8. The header shown might be for the first period only.
Given this, I'll create two separate tables: Table 1 for 1909-1913 with 13 columns (Year, Cases, Defendants, 11 disposal). Table 2 for 1914-1918 with 10 columns (Year, Cases, Defendants, 8 disposal). But the header is only given once. I'll use the header for the first table, and for the second table, I'll use generic disposal column names or repeat the header but only first 8? But the instruction says preserve original content. The original content only has one header. So maybe the second period uses the same header but some columns are not used? But the totals differ.
I think the best is to present the data as it appears in the OCR, in a single table with 13 columns, and for the second period rows, they have fewer disposal numbers, so I'll fill the first N disposal columns and leave the rest blank. The second total row has 8 disposal numbers, so I'll fill first 8 disposal columns.
But the second period rows have varying numbers: 1914: 7 numbers after year? Let's parse 1914: "1914, 477 1,050 239 1 743 5 535955". After year: 477, 1050, 239, 1, 743, 5, 535955. That's 7 numbers. If cases=477, defendants=1050, then 5 disposal numbers? But 535955 is likely two numbers: 535 and 955. So that would be 6 disposal numbers. 1915: "470 922 32 17 1 775 18" -> after year: 470, 922, 32, 17, 1, 775, 18 = 7 numbers. Cases=470, defendants=922, then 5 disposal numbers? 32, 17, 1, 775, 18 = 5. 1916: "341 593 33 6 509 13" -> 6 numbers after year? 341, 593, 33, 6, 509, 13 = 6. Cases=341, defendants=593, then 4 disposal numbers. 1917: "284 584 40 21 477 1 16 S 223 59 29 29 40" -> many numbers. 1918: "261 564 24 21 462 17" -> 6 numbers after year.
This is too variable. I think the OCR has completely mangled the table structure. Given the instructions, I will output the proofread text as a cleaned-up version of the OCR text, with corrected spelling, and for the tabular data, I will present it as a series of lines (preserving the line breaks) but not as a Markdown table, because the table structure is not clear. However, the instruction says to use Markdown table syntax for tabular data. But if the data is not clearly tabular due to OCR damage, maybe I can use a code block? But code fences are prohibited. "Do not wrap your response in code fences."
I could use a Markdown table with a single column "Text" and put each line as a row? That would preserve the line breaks. But that's not reconstructing the table.
Given the dilemma, I'll assume the table is as the OCR presents: each value on a new line, but the original table had rows. I'll reconstruct rows by grouping lines that belong to each year. The OCR shows a pattern: a line with year (e.g., "1909,"), then several lines of numbers. The number of lines per year varies. But the total row has many lines. I'll group them as rows based on the year labels.
I'll create a Markdown table with two columns: "Year" and "Data". But that loses column structure.
Alternatively, I'll create a table with the header as given, and then for each year, I'll put all the numbers that appear for that year in a single cell, separated by commas. That would preserve the data but not the column alignment.
The instruction: "Reconstruct the table using Markdown table syntax." So I must attempt a proper table.
I'll make a decision: The table has 13 columns as per the header. I'll fill the rows for 1909-1913 using the numbers in the order they appear in the OCR for each year, assuming they correspond to the first N disposal columns. For the total row, I'll use all 11 disposal totals. For 1914-1918, I'll create a new table with the same header but only 8 disposal columns? But the header has 10 disposal columns. I'll keep 10 disposal columns for both tables, and for the second period, I'll fill the first 8 disposal columns with the totals from the second total row, and for the year rows, I'll fill as many as they have.
But the second period year rows have many numbers. For 1914, after cases and defendants, there are 5 numbers? Actually 1914: 477, 1050, 239, 1, 743, 5, 535955. If 535955 is two numbers, that's 6 disposal numbers. I'll split 535955 into 535 and 955. So 1914 has 6 disposal numbers. 1915: 32, 17, 1, 775, 18 = 5 disposal numbers. 1916: 33, 6, 509, 13 = 4 disposal numbers. 1917: 40, 21, 477, 1, 16, S, 223, 59, 29, 29, 40 = many. 1918: 24, 21, 462, 17 = 4 disposal numbers.
The second total row has 8 disposal totals. So maybe the second period has 8 disposal columns. I'll create a table for 1914-1918 with 10 columns (Year, Cases, Defendants, 8 disposal). But the header from the original only has 10 disposal columns? Actually the header has 10 disposal columns. So 13 columns total. For the second period, I'll use the same 13 columns, but only the first 8 disposal columns have data; the last two are blank.
But the 1917 row has many numbers, more than 8. That suggests the OCR for 1917 is garbled, maybe combining multiple rows.
Given the time, I'll produce a proofread text that includes the title, the header, and then the data rows as a Markdown table for the first period (1909-1913) with 13 columns, using the total row to define the 11 disposal columns. I'll label the disposal columns as D1 to D11, but use the header names for the first 10, and for the 11th, I'll use "Extra" or something. But the header only has 10 names. The 11th might be "Forfeiture of gunpowder" is the 10th, but we have 11 totals. Maybe "Sent back to duty" and "Reprimanded" are two, and "to duty" is a third? I'll assume the header has 11 items if we split "Sent back Reprimanded. to duty." into three: "Sent back", "Reprimanded", "to duty". But "to duty" is likely part of "Sent back to duty". However, the OCR has "Sent back Reprimanded. to duty." with a period after Reprimanded. So it could be "Sent back to duty. Reprimanded." That's two. Then "To be discharged from ship. Dismissed, Forfeiture of gunpowder." That's three. Total 3+1+1+2+3=10. Still 10.
Maybe "Imprisoned with hard labour" and "Imprisoned in default of fine" and "Imprisonment without hard labour" are three, "Fined" one, "Forfeiture of pay" one, "Sent back to duty" one, "Reprimanded" one, "To be discharged from ship" one, "Dismissed" one, "Forfeiture of gunpowder" one = 10. The total row has 11. Could there be a column for "Total" or something? No.
Let's count the total row numbers again: 1,555; 2,801; 130; 15; 12; 1,934; 3135; 90; 56; 33; 123; 294; 212. That's 13 numbers. If the first two are cases and defendants, then 11 disposal totals. So there are 11 disposal columns. The header must have 11 items. Let's split the header text by periods. The OCR header: "Imprisoned with bard labour. Imprisoned in default of fino. Imprison- ment with- out hard Inbour. Fined. Forfeiture of pay. Sent back Reprimanded. to duty. To be discharged from ship. Dismissed, Forfeiture of gunpowder."
If we consider each period as a separator, we have:
That's 9 items. But "Dismissed, Forfeiture of gunpowder" might be two: "Dismissed" and "Forfeiture of gunpowder". So 10. "Sent back Reprimanded" might be two: "Sent back" and "Reprimanded". But "Sent back" likely is "Sent back to duty". The "to duty" is separate item 7. So maybe "Sent back" is one, "Reprimanded" is two, "to duty" is three. That would give 11 items: 1,2,3,4,5,6(Sent back),7(Reprimanded),8(to duty),9(To be discharged from ship),10(Dismissed),11(Forfeiture of gunpowder). That's 11! Yes! Because the OCR has "Sent back Reprimanded. to duty." If we split at the period: "Sent back Reprimanded" and "to duty". But "Sent back Reprimanded" could be two items: "Sent back" and "Reprimanded". However, there is no period between "Sent back" and "Reprimanded". The OCR has "Sent back Reprimanded." with a period after Reprimanded. So "Sent back Reprimanded" might be a single phrase? But "Sent back Reprimanded" doesn't make sense. It's likely "Sent back to duty. Reprimanded." The OCR merged "to duty" with the next line? Actually the OCR shows: "Sent back Reprimanded. to duty." So "Sent back Reprimanded." then "to duty." So "Sent back Reprimanded" is one line, "to duty" is next. But "Sent back Reprimanded" could be a misreading of "Sent back to duty. Reprimanded." So two columns: "Sent back to duty" and "Reprimanded". Then "to duty" is a fragment. But the total row has 11 disposal totals, so there must be 11 columns. Let's assume the columns are:
Given the total row has 11 disposal totals, I'll create 11 disposal columns. I'll name them based on the header fragments:
Col4: Imprisoned with hard labour
Col5: Imprisoned in default of fine
Col6: Imprisonment without hard labour
Col7: Fined
Col8: Forfeiture of pay
Col9: Sent back to duty
Col10: Reprimanded
Col11: To duty
Col12: To be discharged from ship
Col13: Dismissed
Col14: Forfeiture of gunpowder
That's 11 disposal columns (4-14). But the header only shows up to "Forfeiture of gunpowder". That would be 11 disposal columns? Let's count: 4 to 14 inclusive is 11 columns. Yes.
But the header text has "To be discharged from ship. Dismissed, Forfeiture of gunpowder." That's three columns. So total disposal columns = 3 (imprisonment) + 1 (Fined) + 1 (Forfeiture of pay) + 2 (Sent back to duty, Reprimanded) + 1 (To duty?) + 3 = 11. The "to duty" might be a separate column "To duty" meaning something else? Or it's part of "Sent back to duty". But the OCR has it as a separate line. I'll include it as a column "To duty".
Thus, the table has 14 columns total: Year, Cases
Defendants, how disposed of.
31
18
13
28
Yeur.
No. of cases.
No. of defendants.
Imprisoned with bard labour.
Imprisoned
in default of fino.
Imprison- ment with- out hard Inbour.
Fined.
Forfeiture of
pay.
Sent back
Reprimanded.
to
duty.
To be discharged from ship.
Dismissed,
Forfeiture of gunpowder.
1909,
303
732
19
4
634
38
1910,
95
198
Ch
5
1
128
45
1
1911,
235
356
18
176
3
25
1912,
487
816
18
€
413
21
1913,
435
699
70
3
583
1
Total,....
1,555
2,801
130
15
12
1,934
3135
90
56
:
:
:
33
123
294
212
1914,
477
1,050
239
1
743
5
535955
1915,
470
922
32
17
1
775
18
1916,
341
593
33
6
509
13
:
1917,
284
584
40
21
477
1
16
S
223
59
29
29
40
1918,
261
564
24
21
462
17
Total,.....
1,883
3,713
387
22224
72
5
2,966
2
69
69
Average of
811.0
560.2
26.0
8.0
2.4
886.8
18.0
11.2
1st period,
+
A voruga of
2nd period,
366.6 (
742.6
77.4
14.4
1.0
693.2
.4
18.8
:..
24.6
42,4
:
(Y 8)
No comments yet.
Private notes are available after approval.