The user wants me to proofread OCR output of a historical document about Hong Kong police pensions. The text appears to be a table/list of pensioners with various columns. I need to reconstruct this as proper Markdown with tables.
Looking at the OCR text, it seems to be a table with columns:
The OCR is quite messy with broken lines, misaligned columns, and garbled text. I need to reconstruct the table logically.
Let me parse the data row by row:
First, there's a header: "POLICE PENSIONS." then column headers.
Then "Brought forward, ..: 3,336 6 7 158,351.73" - this seems like a carried forward total.
Then individual pensioners:
Then there are dates and other columns mixed in.
The OCR seems to have read columns in wrong order. I need to reconstruct based on typical pension record format.
Looking at the data, there are two currency systems: £ s. d. and $ c. (likely Hong Kong dollars). Also some entries in Rupees (Rs.).
The columns appear to be:
But the OCR has mixed these up. Let me try to align.
From the text:
"Chu Ping, 81.00" - likely $81.00 pension amount
"Mahomed Shab. 64.82"
"Makhoul Shah, 65.65"
"Anokh Siugh, 52.70"
"Arthur Cornwall Lang- ley, 115 10 8" - this is £115 10s 8d
"Kesur Singh, 62.95"
"Kadar Bux.......................... 52.70"
"Lall Singh, 62.00"
"Munga Khan, 60.06"
"Bikh Mahomed, . 52.70"
"* Ram Kour (widow of L.S. 703 Blaggat Singh) Rs. 48,...... 3 4" - Rs.48 pension, £3 4s?
"Kam Kour (widow of P.C. 769 Jhanda Singh) Rs. 36, 3 4?" - not clear
Then there's "Ordinance No. 11 of 1900." which is likely the Authority for many.
Then dates: "1st Jan, 1912.", "180.00", "45", "Injured on duty." - this seems like a row: Date, Emolument, Age, Cause.
Let me try to reconstruct rows by matching data.
It appears the OCR read the table column by column incorrectly. I need to group by person.
Let me list all the data points I can see:
Names:
Amounts in $:
81.00, 64.82, 65.65, 52.70, 62.95, 52.70, 62.00, 60.06, 52.70, 3 4 (for Ram Kour), 3 4? (for Kam Kour), 60.00 (Lan Fuk), 53 6 8 (George Watt - likely £53 6s 8d), 0.44? (Suby Khan - maybe $0.44? unlikely), 110.19 (Chan Fan), 55.80 (Ahmed Deen), 47.12 (Atta Mohamed), 49.60 (Nathy Khan), 199 3 11 (Henry George Baker - £199 3s 11d), 52.70 (Jewand Singh), 55.80 (Hernam Singh), 55.80 (Mehr Khan), 42 0 0 (John Lenaghan - £42), 52.70 (Mahomed Shah)
Amounts in £:
115 10 8 (Langley), 53 6 8 (Watt), 199 3 11 (Baker), 42 0 0 (Lenaghan), 160.0.0 (maybe for someone), 145.0.0 (Lenaghan?), 3,336 6 7 (brought forward), 8,751,19% (carried forward - garbled)
Dates:
1st Jan, 1912
1st Mar, 1912
16th Mar., 1912
12th April, 1912
1st April, 1912
1st June, 1912
16th July, 1912
16th Aug., 1912
19th Aug., 1912
1st Sept., 1912
1st Nov., 1912
1st Jan., 1913
14th Jan., 1913
1st April, 1913
10th April, 1918 (likely 1913)
1st May, 1913
16th June, 1913
16th July, 1913
1st July, 1913
1st Oct., 1913
24th Oct., 1913
1st Jan. 1914
Emoluments ($):
180.00, 210.00, 210.00, 186.00, 1662.00, 186.00, 186.00, 186.00, 216.00, 186.00, 30? (for 19th Aug), 180.00, 246.00, 300.00, 186.00, 186.00, 2016.00, 186.00, 186.00, 60.00, 74.64, 180.00, 216.00, 186.00, 186.00, 166.00
Ages:
45, 51, 55, 17, 42, 56, 38, 56, 46, 41, 30, 31, 14, 46, 11, 45, 48, 39, 34, 35, 40, 45, 40, 41, 14, 17
Causes:
Injured on duty.
On Expiration of Term of Service.
Medical Certificate.
Killed on duty.
Re-married (remark)
Authority: Ordinance No. 11 of 1900 for many. "Granted Special Pensions by the Governor In Connell on the death of their husbands who were killed on duty." for the widows.
The text "GrantedSpecial Pensions by the Governor In Connell on the death of their ha bands who were killed on duty." appears for the widows.
Also "Re-married." as remark for one.
The carried forward: "Carried forward, ...£8,751,19%. 10%. $159,661.10|" - garbled.
I need to create a Markdown table. Since the OCR is messy, I'll reconstruct as best as possible, aligning rows by the order they appear.
The original likely had a table with multiple columns. I'll create a table with the columns as per the header.
Header from OCR:
"Name of Pensioner. Amount. Amount. Authority. Date from which the Pension £ 5. d. G. Brought forward, ..: 3,336 6 7 158,351.73"
This suggests columns:
But the OCR shows "£ 5. d." and "G." maybe for "£ s. d." and "$ c.".
Also "Brought forward" is a summary line.
I'll create a table with these columns. For each pensioner, I'll fill what I can.
Let me go through the text sequentially and assign rows.
The text starts with "Brought forward, ..: 3,336 6 7 158,351.73" - this is a summary line, not a pensioner. I'll include it as a row or note.
Then names with amounts: Chu Ping 81.00, Mahomed Shab 64.82, etc. These seem to be the $ amounts (column 3). The £ amounts might be blank for these (Indian/Chinese officers paid in dollars?).
Then Arthur Cornwall Langley has £115 10 8.
Then more names with $ amounts.
Then "Ordinance No. 11 of 1900." appears, likely authority for many.
Then dates and emoluments, ages, causes.
The OCR seems to have listed all names first with their pension amounts, then the other columns separately. But that's unlikely; more likely the table rows are interleaved but OCR read column-wise.
Given the difficulty, I'll reconstruct by matching the order of names with the order of dates/emoluments/ages/causes.
List of names in order:
That's 27 pensioners.
Now dates in order:
That's 22 dates. Not matching 27. Some pensioners may share dates or some dates missing.
Emoluments in order:
26 emoluments.
Ages:
26 ages.
Causes:
The causes list has 22 entries.
Given the mismatch, I think the OCR has scrambled the table. As a proofreader, I should reconstruct the table as it logically should be, using the data in the order presented, but aligning rows by the sequence they appear in the text.
The text flows: first a list of names with pension amounts (in $ or £), then a block of dates, emoluments, ages, causes. But the names list includes 27 names, while the other columns have ~26 entries. Probably each name corresponds to one row, and the columns are just printed in a confusing way.
I'll assume the table rows are in the order of names as they appear, and the subsequent data (dates, emoluments, ages, causes) are in the same order. But the OCR has separated them.
Looking at the raw text: after "Bikh Mahomed, . 52.70" there is "* Ram Kour ... 3 4" then "Kam Kour ... 3 4" then "Ordinance No. 11 of 1900." then "GrantedSpecial Pensions by the Governor In Connell on the death of their ha bands who were killed on duty." then "has been paid." then "Amount of Emolument when last employed in Public Service." then "$ C." then "Present Age of Pensioner." then "Cause of Retirement." then "1st Jan, 1912. 180.00 45 Injured on duty." etc.
So the OCR read the header, then the first few rows partially, then the column headers again, then the data for the remaining columns.
This is a common OCR issue with tables: it reads the header row, then the first column for all rows, then the second column for all rows, etc.
But here it seems: it read Name and Amount columns for all rows, then Authority, then Date, then Emolument, Age, Cause.
But the Name list has 27 entries, while the Date list has 22. Maybe some rows are missing in the OCR.
Given the instruction to preserve original content and not add missing text, I should present the data as a table with the information available, leaving blanks where data is missing.
I'll create a Markdown table with columns: Name of Pensioner, Amount (£ s. d.), Amount ($ c.), Authority, Date from which Pension has been paid, Emolument ($ c.), Age, Cause of Retirement, Remarks.
I'll fill row by row using the sequential data.
Let me parse the text more carefully.
The OCR text:
"234
( L 20)
POLICE PENSIONS.
Name of Pensioner.
Amount. Amount. Authority.
Date from which the Pension
£ 5. d.
G.
Brought forward, ..: 3,336 6 7 158,351.73
Chu Ping,
Mahomed Shab.
Makhoul Shah,
Anokh Siugh,
81.00
64.82
65.65
52.70
Arthur Cornwall Lang-
ley,
115 10 8
Kesur Singh,
Kadar Bux..........................
Lall Singh,
Munga Khan,
Bikh Mahomed, .
L.S. 703 Blaggat
Singh) Rs. 48,......
Kam Kour (widow of
P.C. 769 Jhanda Singh) Rs. 36,
62.95
52.70
62.00
60.06
52.70
3 4
Ordinance No. 11 of 1900.
GrantedSpecial Pensions by the Governor In Connell on the death of their ha bands who
were
has been paid.
Amount of Emolument when last employed
in Public Service.
$
C.
Present Age of
Pensioner.
Cause of
Retirement.
1st Jan, 1912.
180.00
45
Injured on duty.
1st Mar, 1912.
210.00
51
On Expiration of
Term of Service.
Do.
210.00
55
17
16th Mar., 1912.
186.00
42
Merlical Cer. tificate.
12th April, 1912.
1,662.00
17
On Expiration of Terms of Service.
1st April, 1912.
186,00
56
*
Do.
186.00
38
Medical Cer- tificate.
1st June, 1912.
186.00
56
On Expiration of
Term of Service.
16th July, 1912.
216.00
46
Medical Cer-
tificate.
16th Aug., 1912.
186.00
41
=
19th Aug., 1912.
30
2 8 0
killed
Do.
on duty.
Lan Fuk,..
60.00
1st Sept., 1912.
George Watt,
53 6 8
Suby Khan,.............
*0.44
Chan Fan,
Ahmed Deen,
110.19
55.80
Atta Moliumed,
47.12
Nathy Khan,
49,60
Henry George Baker,. 199 3 11
Jewand Singh,
52.70
Hernam Singh,
Li Pui,
Abdullah,
Mehr Khan,
55.80
Ordinance No. 11 of 1900.
31
180.00
14
Melient Cer-
tifiente.
1st Nov., 1912. £ 160.0.0
1st Jan., 1913. $246.00
14th Jan., 1913. | 300,00
46
11
45
48
1st April, 1913.
186.00
Do.
186,00
39
Do.
186.00
34
"+
10th April, 1918. | 2,016.00
1st May, 1913.
16th June, 1913, 186,00
186,00
60.00
74.64
16th July, 1913.
1st July, 1913, 180.00 45
216.00
13
On Expiration of Term of Service.
Medical Cer- tificate.
57
45
On Expiration of Term of Service.
Medical Certificate.
40
•
40
35.80
1st Oct., 1913.
186.00
40
"
John Lenaghan,
42 0 0
24th Oct., 1913, | £145.0.0
41
14
Mahomed Shah,
52.70
1st Jan. 1914.
166.00
17
Carried forward, ...£8,751,19%. 10%. $159,661.10|
Re-married.
REMARKS."
This is extremely messy. It appears the OCR has read the table in a columnar fashion but interleaved.
Let me try to interpret the structure.
The document likely has a table with columns:
The "Brought forward" line is a summary from previous page.
Then the table rows.
The OCR seems to have captured:
But the names after the first 12 appear later in the text: Lan Fuk, George Watt, Suby Khan, Chan Fan, Ahmed Deen, Atta Mohamed, Nathy Khan, Henry George Baker, Jewand Singh, Hernam Singh, Li Pui, Abdullah, Mehr Khan, John Lenaghan, Mahomed Shah. That's 15 more names, total 27.
The dates list has 22 entries. Emoluments 26. Ages 26. Causes 22.
Probably the table has 27 rows. The OCR captured the first 12 names with their pension amounts, then the next 15 names appear later with their pension amounts (in the second part of the text). And the date/emolument/age/cause columns are for all 27 but OCR only captured 22-26 entries.
Given the complexity, as a proofreader I should reconstruct the table as best as possible, aligning rows by the order of names as they appear in the text (first block of 12, then second block of 15), and for the other columns, use the sequential data in order, assuming they correspond to the same row order.
But the second block of names appears after some date data. Let's see the sequence:
After "Kam Kour ... 3 4" then "Ordinance No. 11 of 1900." then "Granted Special Pensions..." then "has been paid." then "Amount of Emolument..." then column headers for $ C., Age, Cause, then date/emolument/age/cause for several rows, then "Lan Fuk,.. 60.00" then "1st Sept., 1912." then "George Watt, 53 6 8" then "Suby Khan,............. 0.44" then "Chan Fan, Ahmed Deen, 110.19 55.80" then "Atta Moliumed, 47.12" then "Nathy Khan, 49,60" then "Henry George Baker,. 199 3 11" then "Jewand Singh, 52.70" then "Hernam Singh, Li Pui, Abdullah, Mehr Khan, 55.80" then "Ordinance No. 11 of 1900." then "31 180.00 14 Melient Cer- tifiente." then "1st Nov., 1912. £ 160.0.0" then "1st Jan., 1913. $246.00" then "14th Jan., 1913. | 300,00 46 11 45 48" then "1st April, 1913. 186.00 Do. 186,00 39 Do. 186.00 34" then "+ 10th April, 1918. | 2,016.00" then "1st May, 1913. 16th June, 1913, 186,00 186,00 60.00 74.64 16th July, 1913. 1st July, 1913, 180.00 45 216.00 * 5 9 9 9 13" then "On Expiration of Term of Service. Medical Cer- tificate. 57 45 On Expiration of Term of Service. Medical Certificate. 40 • 40 35.80 1st Oct., 1913. 186.00 40 " John Lenaghan, 42 0 0 24th Oct., 1913, | £145.0.0 41 14 Mahomed Shah, 52.70 1st Jan. 1914. 166.00 17 Carried forward, ...£8,751,19%. 10%. $159,661.10| Re-married. REMARKS."
This is a jumble. It seems the OCR read the table in a serpentine or column-major order.
Given the instructions, I need to output Markdown with the proofread text. I should reconstruct the table logically.
Perhaps the original document is a list of police pensioners with details. I'll create a table with the columns as per the header, and fill each row with the data that seems to belong together.
Let me try to match each name with a date, emolument, age, cause.
From the first block, we have 12 names. The first date is 1st Jan 1912. Likely the first pensioner (Chu Ping) has that date. But Chu Ping's pension amount is 81.00 (dollars). The emolument for first date is 180.00, age 45, cause Injured on duty.
Second: Mahomed Shab, 64.82, date 1st Mar 1912, emolument 210, age 51, cause On Expiration.
Third: Makhoul Shah, 65.65, date 16th Mar 1912, emolument 210, age 55, cause Do.
Fourth: Anokh Siugh, 52.70, date 12th April 1912, emolument 186, age 17, cause Medical Certificate.
Fifth: Arthur Cornwall Langley, £115 10 8, date 1st April 1912, emolument 1662, age 42, cause On Expiration of Terms of Service.
Sixth: Kesur Singh, 62.95, date 1st June 1912, emolument 186, age 56, cause Do.
Seventh: Kadar Bux, 52.70, date 16th July 1912, emolument 186, age 38, cause Medical Certificate.
Eighth: Lall Singh, 62.00, date 1st June 1912? Wait duplicate. Let's see: after 16th July 1912, next date is 16th Aug 1912. But we have 1st June 1912 already used. The dates list: 1st Jan, 1st Mar, 16th Mar, 12th Apr, 1st Apr, 1st Jun, 16th Jul, 16th Aug, 19th Aug, 1st Sep, 1st Nov, 1st Jan 1913, 14th Jan 1913, 1st Apr 1913, 10th Apr 1918, 1st May 1913, 16th Jun 1913, 16th Jul 1913, 1st Jul 1913, 1st Oct 1913, 24th Oct 1913, 1st Jan 1914.
That's 22 dates. For 27 pensioners, some may not have dates? Or some dates are for the same pensioner? Unlikely.
Maybe the first 12 names correspond to the first 12 dates? But there are 22 dates.
Let's count the pensioners in the first block: 12 names (including two widows). Then second block: Lan Fuk, George Watt, Suby Khan, Chan Fan, Ahmed Deen, Atta Mohamed, Nathy Khan, Henry George Baker, Jewand Singh, Hernam Singh, Li Pui, Abdullah, Mehr Khan, John Lenaghan, Mahomed Shah = 15. Total 27.
The dates: 22. Emoluments: 26. Ages: 26. Causes: 22.
Perhaps the table continues on next page, but we only have this snippet.
Given the instruction to preserve original content and not add missing text, I will create a table with the data as presented, but in a structured way. I'll include all names and the associated data that appears near them in the text.
But the text is not row-aligned. As a proofreader, I should reconstruct the rows by reading the text in order and grouping related information.
The OCR output is essentially a linear stream of the table cells in reading order (left to right, top to bottom). For a table, reading order is row by row. So the stream should be: Row1Col1, Row1Col2, Row1Col3, ..., Row2Col1, Row2Col2, ...
But the OCR has messed up because of column detection.
However, the user says: "Restore column reading order — if text was originally in multiple columns but OCR read it in the wrong order, reorder it to make logical sense (left-to-right, top-to-bottom)."
So I need to infer the correct table structure.
Given the header: "Name of Pensioner. Amount. Amount. Authority. Date from which the Pension £ 5. d. G. Brought forward, ..: 3,336 6 7 158,351.73"
This suggests the first row after header is "Brought forward" with amounts.
Then the pensioners.
The columns: Name, Amount (£), Amount ($), Authority, Date, Emolument ($), Age, Cause, Remarks.
But the header shows "£ 5. d." and "G." maybe for "£ s. d." and "$ c.".
Also "Date from which the Pension" then "has been paid." on next line.
Then "Amount of Emolument when last employed in Public Service." then "$ C." then "Present Age of Pensioner." then "Cause of Retirement."
So the table has 9 columns.
Now, the OCR text after header: "Brought forward, ..: 3,336 6 7 158,351.73" - this is likely a row spanning columns? Or just a note.
Then "Chu Ping, Mahomed Shab. Makhoul Shah, Anokh Siugh, 81.00 64.82 65.65 52.70" - this looks like four names then four dollar amounts. But where are the £ amounts? Maybe for these, £ amount is blank or zero.
Then "Arthur Cornwall Lang- ley, 115 10 8" - name and £ amount.
Then "Kesur Singh, Kadar Bux.......................... Lall Singh, Munga Khan, Bikh Mahomed, . 62.95 52.70 62.00 60.06 52.70" - five names, five dollar amounts.
Then "* Ram Kour (widow of L.S. 703 Blaggat Singh) Rs. 48,...... Kam Kour (widow of P.C. 769 Jhanda Singh) Rs. 36, 3 4" - two names with Rupee amounts and £ amounts 3 4.
Then "Ordinance No. 11 of 1900." - authority.
Then "Granted Special Pensions by the Governor In Council on the death of their husbands who were killed on duty." - authority for widows.
Then "has been paid." - part of date column header.
Then "Amount of Emolument when last employed in Public Service. $ C. Present Age of Pensioner. Cause of Retirement." - headers for next columns.
Then data: "1st Jan, 1912. 180.00 45 Injured on duty. 1st Mar, 1912. 210.00 51 On Expiration of Term of Service. Do. 210.00 55 17 16th Mar., 1912. 186.00 42 Merlical Cer. tificate. 12th April, 1912. 1,662.00 17 On Expiration of Terms of Service. 1st April, 1912. 186,00 56 * Do. 186.00 38 Medical Cer- tificate. 1st June, 1912. 186.00 56 On Expiration of Term of Service. 16th July, 1912. 216.00 46 Medical Cer- tificate. 16th Aug., 1912. 186.00 41 = 19th Aug., 1912. 30 2 8 0 killed Do. on duty."
This is 11 rows of date/emolument/age/cause.
Then "Lan Fuk,.. 60.00 1st Sept., 1912." - name, dollar amount, date.
Then "George Watt, 53 6 8" - name, £ amount.
Then "Suby Khan,............. *0.44" - name, dollar amount? 0.44 seems low.
Then "Chan Fan, Ahmed Deen, 110.19 55.80" - two names, two dollar amounts.
Then "Atta Moliumed, 47.12" - name, dollar amount.
Then "Nathy Khan, 49,60" - name, dollar amount.
Then "Henry George Baker,. 199 3 11" - name, £ amount.
Then "Jewand Singh, 52.70" - name, dollar amount.
Then "Hernam Singh, Li Pui, Abdullah, Mehr Khan, 55.80" - four names, one dollar amount? Or each 55.80?
Then "Ordinance No. 11 of 1900." - authority.
Then "31 180.00 14 Melient Cer- tifiente." - age? emolument? age? cause?
Then "1st Nov., 1912. £ 160.0.0" - date, £ emolument?
Then "1st Jan., 1913. $246.00" - date, $ emolument.
Then "14th Jan., 1913. | 300,00 46 11 45 48" - date, emolument, ages?
Then "1st April, 1913. 186.00 Do. 186,00 39 Do. 186.00 34" - date, emolument, cause, emolument, age, cause, emolument, age.
Then "+ 10th April, 1918. | 2,016.00" - date, emolument.
Then "1st May, 1913. 16th June, 1913, 186,00 186,00 60.00 74.64 16th July, 1913. 1st July, 1913, 180.00 45 216.00 5 9 9 9 13" - messy.
Then "On Expiration of Term of Service. Medical Cer- tificate. 57 45 On Expiration of Term of Service. Medical Certificate. 40 • 40 35.80 1st Oct., 1913. 186.00 40 " - causes, ages, date, emolument, age.
Then "John Lenaghan, 42 0 0 24th Oct., 1913, | £145.0.0 41 14" - name, £ amount, date, £ emolument, age, ?.
Then "Mahomed Shah, 52.70 1st Jan. 1914. 166.00 17" - name, $ amount, date, emolument, age.
Then "Carried forward, ...£8,751,19%. 10%. $159,661.10| Re-married. REMARKS."
This is extremely disordered. It seems the OCR read the table in multiple passes: first the name and pension amount columns, then the date/emolument/age/cause columns, but also interleaved with more names.
Given the difficulty, I think the best approach is to present the data in a structured Markdown table with the columns as defined, and fill in the information for each pensioner as can be deduced, leaving unknown cells blank or with "...".
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."
And "Missing text — if a word or phrase is clearly missing due to OCR damage, insert ... in its place."
So I should not invent data. I should present the text as a corrected version of the OCR output, with paragraphs and sentences fixed. But this is a table, not prose.
The user says: "Return the proofread text in standard Markdown: Use Markdown table syntax to reconstruct tabular data."
So I must reconstruct the table.
Given the OCR is from a historical record, likely the table is regular. I'll assume the table has 27 rows (one per pensioner). I'll use the order of names as they appear in the text (first 12, then the 15 later). For each, I'll try to assign the corresponding data from the subsequent columns based on sequential order.
But the subsequent columns (date, emolument, age, cause) appear in a block after the first 12 names. That block has 11 rows. Then more names appear, then more data.
Perhaps the table is split into two parts: first part for Indian/Chinese officers (paid in dollars), second part for European officers (paid in pounds). But the names include both.
Let's look at the names: Chu Ping, Mahomed Shab, Makhoul Shah, Anokh Siugh - likely Indian/Chinese. Arthur Cornwall Langley - European. Kesur Singh, Kadar Bux, Lall Singh, Munga Khan, Bikh Mahomed - Indian. Ram Kour, Kam Kour - widows. Lan Fuk - Chinese. George Watt - European. Suby Khan - Indian. Chan Fan - Chinese. Ahmed Deen - Indian. Atta Mohamed - Indian. Nathy Khan - Indian. Henry George Baker - European. Jewand Singh - Indian. Hernam Singh - Indian. Li Pui - Chinese. Abdullah - Indian. Mehr Khan - Indian. John Lenaghan - European. Mahomed Shah - Indian.
So mixed.
The pension amounts: for first 12, mostly dollars. Langley in pounds. Widows in rupees and pounds.
Second block: Lan Fuk $60, George Watt £53 6 8, Suby Khan $0.44? (maybe 60.44?), Chan Fan $110.19, Ahmed Deen $55.80, Atta Mohamed $47.12, Nathy Khan $49.60, Henry George Baker £199 3 11, Jewand Singh $52.70, Hernam Singh $55.80, Li Pui ?, Abdullah ?, Mehr Khan $55.80, John Lenaghan £42, Mahomed Shah $52.70.
The date/emolument/age/cause block after the first 12 names has 11 entries. That might correspond to the first 11 pensioners (excluding the widows? or including). 11 entries for 12 pensioners? The widows might have different authority.
Then after that, we have more data for the remaining pensioners.
The text "Lan Fuk,.. 60.00 1st Sept., 1912." suggests that for Lan Fuk, the date is 1st Sept 1912. Then "George Watt, 53 6 8" no date. Then "Suby Khan,............. 0.44" no date. Then "Chan Fan, Ahmed Deen, 110.19 55.80" no dates. Then "Atta Moliumed, 47.12" etc. Then "Ordinance No. 11 of 1900." then "31 180.00 14 Melient Cer- tifiente." then "1st Nov., 1912. £ 160.0.0" then "1st Jan., 1913. $246.00" then "14th Jan., 1913. | 300,00 46 11 45 48" then "1st April, 1913. 186.00 Do. 186,00 39 Do. 186.00 34" then "+ 10th April, 1918. | 2,016.00" then "1st May, 1913. 16th June, 1913, 186,00 186,00 60.00 74.64 16th July, 1913. 1st July, 1913, 180.00 45 216.00 * 5 9 9 9 13" then causes and ages.
This is too messy to perfectly reconstruct.
Given the constraints, I will produce a Markdown table that captures all the legible data in a structured way, with rows for each pensioner, and columns as per the header. I'll use the order of names as they appear in the text, and for each, fill in the data that appears associated with them in the text (like pension amount, date, emolument, age, cause) based on proximity.
But the instruction: "Rejoin broken sentences — within a paragraph, if a sentence is split across multiple lines, merge the pieces back into one flowing sentence." This is for prose. For tables, "reconstruct tabular data".
I think the best is to create a table with the columns, and for each row, put the information that belongs to that pensioner as per the original document's row. Since I cannot be sure, I will make a reasonable reconstruction.
Let me attempt to list the pensioners in the order they appear in the OCR text, with all data points that appear near their name.
The date rows:
That's 11 rows. We have 12 pensioners in first block (including two widows). The widows might be the 19th Aug entry? "19th Aug., 1912. 30 2 8 0 killed Do. on duty." Age 30, emolument £2 8 0? Cause killed on duty. That could be for Ram Kour (widow of Blaggat Singh killed on duty). And Kam Kour similar.
But the text says "Granted Special Pensions by the Governor In Council on the death of their husbands who were killed on duty." So the widows' pensions are special.
The entry "19th Aug., 1912. 30 2 8 0 killed Do. on duty." might be for Ram Kour. Then Kam Kour might be next? But next date is 1st Sept 1912 for Lan Fuk.
Let's continue.
After the 11 date rows, we have "Lan Fuk,.. 60.00 1st Sept., 1912." So Lan Fuk: pension $60, date 1st Sept 1912. Then emolument? age? cause? Not given directly. But the date row for 1st Sept 1912 is not in the previous block? The previous block had up to 1st Nov 1912. Actually the block had 1st Nov 1912 as 11th. But 1st Sept 1912 is missing from that block. The block had: 1st Jan, 1st Mar, 16th Mar, 12th Apr, 1st Apr, 1st Jun, 16th Jul, 16th Aug, 19th Aug, 1st Sep? Wait, the text: "16th Aug., 1912. 186.00 41 = 19th Aug., 1912. 30 2 8 0 killed Do. on duty. Lan Fuk,.. 60.00 1st Sept., 1912." So the block includes 16th Aug, 19th Aug, then Lan Fuk with 1st Sept. So the date rows are interspersed with names.
This suggests the OCR read the table row by row but with column misalignment.
Let's try to parse the text as a sequence of rows, each row having multiple fields.
The text after "Cause of Retirement." starts the data rows.
I'll split the text from "1st Jan, 1912." onward into tokens.
But perhaps it's better to treat the entire OCR output as a single text and proofread it as a document, preserving the table structure by converting to Markdown table.
Given the complexity, I will produce a Markdown table with the columns: Name, Pension (£), Pension ($), Authority, Date, Emolument ($), Age, Cause, Remarks.
I'll include all names found, and for each, fill in the data that appears in the text in the same relative order.
I'll make a judgment call to align the first 11 date rows with the first 11 pensioners (Chu Ping to Lall Singh? or to Bikh Mahomed?). The 12th pensioner (Bikh Mahomed) might not have a date row in that block. The widows have a separate entry.
Then for the remaining pensioners (Lan Fuk onward), there are more date rows later.
Let's list all date-like entries in order:
22 dates.
Now pensioners: 27. So 5 pensioners might not have dates in this extract (maybe continued on next page).
But the "Carried forward" suggests this is a page of a ledger.
I'll create a table with 27 rows, filling dates sequentially for the first 22, leaving last 5 dates blank.
Similarly for emolument, age, cause.
But the emolument and age lists are longer.
Given the time, I'll produce a cleaned-up version of the text as a Markdown table, using the data as it appears in the OCR, but organized into rows based on the sequence of names.
I'll write a script in my mind to parse, but manually.
Let me create the table row by row as per the order of names in the text.
Order of names as they appear in the OCR text (first occurrence):
Now, for each, I'll try to find associated data.
From the text, the pension amounts (in $ or £) appear near the names in the first part for 1-12, and in the second part for 13-27.
For 1-4: $81.00, $64.82, $65.65, $52.70 (listed after the four names).
For 5: £115 10 8 (after name).
For 6-10: $62.95, $52.70, $62.00, $60.06, $52.70 (listed after the five names).
For 11: Rs. 48, and £3 4? ( "3 4" after)
For 12: Rs. 36, and £3 4? ( "3 4" after? The text: "Kam Kour (widow of P.C. 769 Jhanda Singh) Rs. 36, 62.95 52.70 62.00 60.06 52.70 3 4" - actually the numbers after are the dollar amounts for 6-10, then "3 4" for Ram Kour? Then Kam Kour has no separate £ amount? The "3 4" might be for both? Or only for Ram Kour.
Then "Ordinance No. 11 of 1900." and "Granted Special Pensions..." for the widows.
Then the date/emolument/age/cause block.
Then for 13 Lan Fuk: "Lan Fuk,.. 60.00 1st Sept., 1912." So pension $60, date 1st Sept 1912.
14 George Watt: "George Watt, 53 6 8" - pension £53 6 8.
15 Suby Khan: "Suby Khan,............. 0.44" - pension $0.44? That seems too low. Maybe it's $60.44? Or the "0.44" is something else.
16 Chan Fan: "Chan Fan, Ahmed Deen, 110.19 55.80" - Chan Fan $110.19, Ahmed Deen $55.80.
17 Ahmed Deen: $55.80.
18 Atta Moliumed: "Atta Moliumed, 47.12" - $47.12.
19 Nathy Khan: "Nathy Khan, 49,60" - $49.60.
20 Henry George Baker: "Henry George Baker,. 199 3 11" - £199 3 11.
21 Jewand Singh: "Jewand Singh, 52.70" - $52.70.
22 Hernam Singh: "Hernam Singh, Li Pui, Abdullah, Mehr Khan, 55.80" - likely each $55.80? Or only Mehr Khan $55.80? The text lists four names then one amount. Probably each has $55.80.
23 Li Pui: $55.80
24 Abdullah: $55.80
25 Mehr Khan: $55.80
26 John Lenaghan: "John Lenaghan, 42 0 0" - £42.
27 Mahomed Shah: "Mahomed Shah, 52.70" - $52.70.
Now for dates, emoluments, ages, causes.
The first block of date rows (after the authority lines) seems to correspond to the first 11 pensioners (1-11). Let's assume:
Row1: 1st Jan 1912, emolument $180, age 45, cause Injured on duty.
Row2: 1st Mar 1912, $210, age 51, cause On Expiration of Term of Service.
Row3: 16th Mar 1912, $210, age 55, cause Do.
Row4: 12th Apr 1912, $186, age 17, cause Medical Certificate.
Row5: 1st Apr 1912, $1662, age 42, cause On Expiration of Terms of Service.
Row6: 1st Jun 1912, $186, age 56, cause Do.
Row7: 16th Jul 1912, $186, age 38, cause Medical Certificate.
Row8: 16th Aug 1912, $186, age 41, cause Medical Certificate (Do).
Row9: 19th Aug 1912, £2 8 0? age 30, cause Killed on duty. (This for Ram Kour)
Row10: 1st Sep 1912? But the text says "Lan Fuk,.. 60.00 1st Sept., 1912." So maybe row10 is for Lan Fuk? But Lan Fuk is pensioner 13. The date block has 11 rows, but the 10th row in block is 1st Sep? Actually the block includes "16th Aug., 1912. 186.00 41 = 19th Aug., 1912. 30 2 8 0 killed Do. on duty." Then "Lan Fuk,.. 60.00 1st Sept., 1912." So the date rows in the block are up to 19th Aug. That's 9 rows? Let's count:
That's 9 rows. Then Lan Fuk with 1st Sept. Then later "1st Nov., 1912." etc.
So the first block has 9 date rows. That could correspond to first 9 pensioners (Chu Ping to Lall Singh). Then Munga Khan and Bikh Mahomed (10,11) might not have dates in this block. The widows (11,12) have the 19th Aug entry? But 19th Aug is the 9th row. Ram Kour is 11th pensioner. So not aligned.
Maybe the date rows correspond to the pensioners in the order they appear in the ledger, which might be different from the name list order.
Given the difficulty, I will create a table with the data as presented in the OCR, but in a structured format, using the sequence of the text to define rows. Essentially, I will treat the OCR output as a series of lines and group them into rows based on the pattern: Name, then pension amount, then maybe date, etc.
But the OCR is not line-based.
Given the instructions, I think the expected output is a cleaned-up Markdown version of the table, with corrected OCR errors (like "Merlical Cer. tificate" -> "Medical Certificate", "Melient Cer- tifiente" -> "Medical Certificate", "Anokh Siugh" -> "Anokh Singh", "Makhoul Shah" -> "Makhoul Shah" (maybe "Makhoul Shah" is fine), "Kadar Bux" -> "Kadar Bux", "Blaggat" -> "Blaggat" (maybe "Blaggat" is "Blaggat"?), "Jhanda Singh" -> "Jhanda Singh", "Lan Fuk" -> "Lan Fuk", "Suby Khan" -> "Suby Khan", "Atta Moliumed" -> "Atta Mohamed", "Nathy Khan" -> "Nathy Khan", "Hernam Singh" -> "Hernam Singh", "Merlical" -> "Medical", "Cer. tificate" -> "Certificate", "Do." -> "Ditto" or "On Expiration of Term of Service", "Re-married" -> "Re-married".
Also fix numbers: "186,00" -> "186.00", "49,60" -> "49.60", "300,00" -> "300.00", "2,016.00" -> "2016.00", "158,351.73" -> "158351.73", "8,751,19%" -> "8751.19"? "10%." -> "10."? "159,661.10" -> "159661.10".
Also "£ 5. d." -> "£ s. d.", "G." -> "$ c.".
And "Brought forward, ..: 3,336 6 7 158,351.73" -> "Brought forward: £3,336 6s 7d, $158,351.73"
"Carried forward, ...£8,751,19%. 10%. $159,661.10|" -> "Carried forward: £8,751 19s 10d, $159,661.10"
"GrantedSpecial Pensions by the Governor In Connell on the death of their ha bands who were killed on duty." -> "Granted Special Pensions by the Governor in Council on the death of their husbands who were killed on duty."
"Merlical Cer. tificate" -> "Medical Certificate"
"Melient Cer- tifiente" -> "Medical Certificate"
"On Expiration of Terms of Service" -> "On Expiration of Term of Service"
"Killed on duty" -> "Killed on duty"
"Re-married" -> "Re-married"
Also fix "L.S. 703 Blaggat Singh" -> "L.S. 703 Blaggat Singh" (maybe "L.S. 703 Blaggat Singh" is "L.S. 703 Blaggat Singh"? Could be "L.S. 703 Blaggat Singh" but "Blaggat" might be "Blaggat"?)
"P.C. 769 Jhanda Singh" -> "P.C. 769 Jhanda Singh"
"Ordinance No. 11 of 1900." appears multiple times.
Now, for the table, I'll create a Markdown table with the columns as per the header.
I'll include a row for "Brought forward" and "Carried forward" as summary rows.
But the instruction: "Reconstruct tabular data." So I'll make a table for the pensioners.
Given the ambiguity, I'll produce a table with the 27 pensioners, with columns: Name, Pension (£), Pension ($), Authority, Date, Emolument ($), Age, Cause, Remarks.
I'll fill in what I can, leaving blanks for missing.
But the user might expect the proofread text to be the OCR text corrected, not a reconstructed database. However, they said "Use Markdown table syntax to reconstruct tabular data."
So I will output a Markdown table.
Let me write the table.
First, the header row.
Then rows.
I'll use the data as aligned above for first 11 with the first 9 date rows? I'll align by the order of the date rows in the text.
The date rows in text order:
Also causes: "On Expiration of Term of Service.", "Medical Certificate.", "Killed on duty.", "Re-married."
And ages: 45,51,55,17,42,56,38,41,30,31,14,46,11,45,48,39,34,35,40,45,40,41,14,17,57,45,40,40,35,80? etc.
This is too messy.
Given the time, I will output a cleaned-up version of the OCR text as a Markdown document with a table that represents the data as best as possible, with corrected OCR errors, and with the table structure inferred.
I'll create a table with the columns: Name, Pension (£), Pension ($), Authority, Date, Emolument ($), Age, Cause, Remarks.
I'll populate with the 27 names, and for each, the pension amount from the text, and for the first 9, the date/emolument/age/cause from the first block. For the rest, I'll use the later data.
But to be honest, this is guesswork.
Perhaps the user expects me to simply correct the OCR text and present it as a formatted table with the same rows as the OCR but with corrected spelling and formatting.
Looking at the OCR, it might be that the original table has two sections: one for "Police Pensions" with columns Name, Amount (£), Amount ($), Authority, Date, Emolument, Age, Cause, Remarks. And the OCR has read it in a way that the first few columns are listed for all rows, then the next columns.
But the instruction: "Restore column reading order — if text was originally in multiple columns but OCR read it in the wrong order, reorder it to make logical sense (left-to-right, top-to-bottom)."
So I should reorder the text to form proper rows.
The OCR text is essentially a linear stream. If the table has 9 columns, the stream should be row1col1, row1col2, ..., row1col9, row2col1, ...
But the OCR has: row1col1, row2col1, row3col1,... (all names), then row1col2, row2col2,... (all £ amounts), then row1col3, row2col3,... (all $ amounts), then authority, then date, etc. That is column-major order.
So to restore, I need to transpose: take the list of names (27), list of £ amounts (some), list of $ amounts (some), list of authorities (mostly Ordinance No. 11 of 1900), list of dates (22), list of emoluments (26), list of ages (26), list of causes (22), list of remarks (few).
But the lists are not complete and not aligned.
Given the difficulty, I will assume the table has 27 rows, and the columns are as per header. I will fill each column with the data from the OCR in the order they appear for that column.
For Name column: the 27 names in order.
For £ Amount column: the £ amounts appear: for Langley 115 10 8, for Ram Kour 3 4, for Kam Kour 3 4? (maybe), for George Watt 53 6 8, for Henry George Baker 199 3 11, for John Lenaghan 42 0 0. Also "Brought forward 3,336 6 7" and "Carried forward 8,751 19 10". But those are summaries. For the pensioners, only those 6 have £ amounts. The rest are blank or zero.
For $ Amount column: the $ amounts appear in two batches: first batch for first 10 pensioners (Chu Ping to Bikh Mahomed) 10 amounts: 81.00, 64.82, 65.65, 52.70, (Langley?), 62.95, 52.70, 62.00, 60.06, 52.70. That's 10. But there are 12 pensioners in first block. The widows have Rupees. Then second batch: Lan Fuk 60.00, Suby Khan 0.44, Chan Fan 110.19, Ahmed Deen 55.80, Atta Mohamed 47.12, Nathy Khan 49.60, Jewand Singh 52.70, Hernam Singh 55.80, Li Pui 55.80, Abdullah 55.80, Mehr Khan 55.80, Mahomed Shah 52.70. That's 12. Total 22 $ amounts. But 27 pensioners. Missing for Langley, Ram Kour, Kam Kour, George Watt, Henry George Baker, John Lenaghan. Those have £ amounts.
So $ column: for those with £ pension, $ pension might be blank.
Authority column: mostly "Ordinance No. 11 of 1900." For widows: "Granted Special Pensions by the Governor in Council on the death of their husbands who were killed on duty."
Date column: 22 dates in order.
Emolument column: 26 emoluments in order.
Age column: 26 ages in order.
Cause column: 22 causes in order.
Remarks column: "Re-married" for one.
To align, we need to decide which pensioner gets which date. The dates are in chronological order. The pensioners might be listed in order of date. So the first pensioner (Chu Ping) has earliest date 1st Jan 1912. Second (Mahomed Shab) 1st Mar 1912, etc. That would align the first 22 pensioners with the 22 dates. The last 5 pensioners (Li Pui, Abdullah, Mehr Khan, John Lenaghan, Mahomed Shah) would have dates after 1st Jan 1914? But the last date is 1st Jan 1914 for Mahomed Shah. John Lenaghan has 24th Oct 1913. So maybe the order is not strictly by name list.
The name list order might be the order in the ledger, which could be by date. So the first name Chu Ping has first date, etc. But the name list includes 27 names, and there are 22 dates. So the last 5 names might not have dates in this page (carried forward).
But the text shows John Lenaghan and Mahomed Shah at the end with dates.
In the name list, John Lenaghan is 26th, Mahomed Shah 27th. In the date list, 21st is 24th Oct 1913 (John Lenaghan), 22nd is 1st Jan 1914 (Mahomed Shah). So if we align the date list to the name list in order, the 21st name should be John Lenaghan, 22nd Mahomed Shah. But in the name list, John Lenaghan is 26th. So the name list order is not the same as date order.
The OCR name list is from the first column read top to bottom. The date column is also read top to bottom. So they should correspond row by row. Therefore, the first name corresponds to first date, second name to second date, etc. But there are 27 names and 22 dates. That means the last 5 names have no date in this extract (maybe continued on next page). But the text shows dates for John Lenaghan and Mahomed Shah later in the text, but those might be from the date column read later? Actually the date column is read in one go? The OCR text shows dates interspersed.
Given the confusion, I will not attempt to perfectly align. Instead, I will present the data as a series of rows as they appear in the OCR text after correcting OCR errors, using a table for the main pensioner list.
Given the instruction to "reconstruct tabular data", I'll create a table with the columns and fill each row with the information that appears in the text for that pensioner,
234
( L 20)
POLICE PENSIONS.
Name of Pensioner.
Amount. Amount. Authority.
Date from which the Pension
£ 5. d.
G.
Brought forward, ..: 3,336 6 7 158,351.73
Chu Ping,
Mahomed Shab.
Makhoul Shah,
Anokh Siugh,
81.00
64.82
65.65
52.70
Arthur Cornwall Lang-
ley,
115 10 8
Kesur Singh,
Kadar Bux..........................
Lall Singh,
Munga Khan,
Bikh Mahomed, .
L.S. 703 Blaggat
Singh) Rs. 48,......
Kam Kour (widow of
P.C. 769 Jhanda Singh) Rs. 36,
62.95
52.70
62.00
60.06
52.70
3 4
Ordinance No. 11 of 1900.
GrantedSpecial Pensions by the Governor In Connell on the death of their ha bands who
were
has been paid.
Amount of Emolument when last employed
in Public Service.
$
C.
Present Age of
Pensioner.
Cause of
Retirement.
1st Jan, 1912.
180.00
45
Injured on duty.
1st Mar, 1912.
210.00
51
On Expiration of
Term of Service.
Do.
210.00
55
17
16th Mar., 1912.
186.00
42
Merlical Cer. tificate.
12th April, 1912.
1,662.00
17
On Expiration of Terms of Service.
1st April, 1912.
186,00
56
*
Do.
186.00
38
Medical Cer- tificate.
1st June, 1912.
186.00
56
On Expiration of
Term of Service.
16th July, 1912.
216.00
46
Medical Cer-
tificate.
16th Aug., 1912.
186.00
41
=
19th Aug., 1912.
30
2 8 0
killed
Do.
on duty.
Lan Fuk,..
60.00
1st Sept., 1912.
George Watt,
53 6 8
Suby Khan,.............
*0.44
Chan Fan,
Ahmed Deen,
110.19
55.80
Atta Moliumed,
47.12
Nathy Khan,
49,60
Henry George Baker,. 199 3 11
Jewand Singh,
52.70
Hernam Singh,
Li Pui,
Abdullah,
Mehr Khan,
55.80
Ordinance No. 11 of 1900.
31
180.00
14
Melient Cer-
tifiente.
1st Nov., 1912. £ 160.0.0
1st Jan., 1913. $246.00
14th Jan., 1913. | 300,00
46
11
45
48
1st April, 1913.
186.00
Do.
186,00
39
Do.
186.00
34
"+
10th April, 1918. | 2,016.00
1st May, 1913.
16th June, 1913, 186,00
186,00
60.00
74.64
16th July, 1913.
1st July, 1913, 180.00 45
216.00
13
On Expiration of Term of Service.
Medical Cer- tificate.
57
45
On Expiration of Term of Service.
Medical Certificate.
40
•
40
35.80
1st Oct., 1913.
186.00
40
"
John Lenaghan,
42 0 0
24th Oct., 1913, | £145.0.0
41
14
Mahomed Shah,
52.70
1st Jan. 1914.
166.00
17
Carried forward, ...£8,751,19%. 10%. $159,661.10|
Re-married.
REMARKS.
No comments yet.
Private notes are available after approval.