1928 — Page 322

The user wants me to proofread OCR output of a historical document about pensions payable in Hong Kong. The text appears to be a table with columns for Name of Pensioner, Amount of Pension in Sterling, Amount in Dollars, Authority, Date, Service, Emolument, Age, Cause of Retirement.

I need to reconstruct this as a proper Markdown table. The OCR has many errors: broken lines, misaligned columns, garbled text. I'll need to parse the data carefully.

Let me analyze the text line by line.

First lines appear to be headers:

"PENSIONS PAYABLE OUT OF THE REVENUE OF THE COLONY."

Then column headers:

"Amount of the" (maybe "Amount of the Pension in Sterling, per annum")

"Amount of the" (maybe "Amount of the Pension in Dollars, per annum")

"Name of Pensioner."

"Pension in Sterling, per annum."

"Pension in Dollars, per annum."

"Anthority under which the Pension was granted." (Authority)

"Date from which the Pension has been paid."

"Servico for which the Pension was granted." (Service)

"Amount of Emolument when last employed in Public Service."

"Present Age of Pensioner."

"Свине" (looks like Cyrillic? maybe "Cause of Retirement")

"of Retirement."

Then data rows. There's a "Brought forward" line with "$ c. 11,477.96" and "C.S.O. No." maybe.

Then names: "Cheung Wan-tɑai," "Wadowa Singh." "J. A. Lowson," "Fat Ngan," etc.

Numbers: 230.00, 110.83, 1,634.00, 2210 of 1901, 2501 of 1901, 2062 of 1900, 30,80, 3463 of 1901, dates: 1901. 6th August, 1st September, 26th December, 1902. 1st January.

Services: "Second Shroff, Treasury," "1st Class Assistant Warder, Victoria Gaol," "Assistant Surgeon," "Dispensarymat,..." (Dispensary Matron?), "£ 2 3 4" maybe? "80" "Ill-health." "63" "N.J" "Ill-health." "Age." "Mrs. Jane Ackers (now Mrs, Wahh)," "456.16" "John Lee," "1,158,10" "Fonja Singh," "F. F. Remedioa," "Fung Fu........" "Lo Sik-ling," "†C. Wagner," "Sir Wm. Meigh Goodman Kl." "Janlah Singh." "Tang Shi-kit," "Elizabeth Annie Bateman" "Karwo Dad," "87.26" "3203 of 1901," "57 of 1902." "2782 of 1902." "17th February," "Matron, Civil Hospital," "1,190.00" "71" "Ill-health." "1st April." "19th November." "Inspector under the Women & Girls Protection Ord.," "+" "1,878,00" "81" "2nd Class Assistant Warder," "308,00" "59" "Age." "Ill-health" "400.00" "336.40" "318,00" "47 11" "2349 of 1902." "3118 of 1902." "2911 of 1902," "6432 of 1901." "24th December. 1903." "3rd Clerk, Colonial Secretary's Office," "16th February," "Head Master, Sai-Ying-Pan School," "1,592.26" "696.00" "63" "11" "I" "1st July. 1904. 17th May." "Clerk, Education Departineut.." "720,00" "Age." "Ill-health." "1st Clerk, Police Department,." "1,440,00" "76" "Age." ":" "5,400,00" "57.50" "343 of 1905." "849) of 1905." "1905. 2716 April." "Chief Justice, *********" "13,500,00" "81" "17th November, † 2nd Class Assistant Wunder. Vietoria Gaol," "230,00" "H-health." "12.00" "J" "455.00" "1" "66.00" "901 of 190G." "3671 of 1995." "4694 of 1906," "Dist May," "Chinese Writer. Registrar General's Department," "240.00" "55" "Abolition of Ofer." "20th May." "Hand Mistress, Belílios Public School," "2,700,00" "N2" "Age." "ติ" "20th June," "Assistant Warier, Victoria Gnol," "264,00" "48" "III-health."

This is messy. I need to reconstruct rows. Likely each pensioner has a row with multiple columns. The OCR has lost column alignment. I'll need to infer columns from the header.

Let me list the columns as per header:

1. Name of Pensioner
2. Amount of Pension in Sterling, per annum (£ s d?)
3. Amount of Pension in Dollars, per annum ($ c.)
4. Authority under which the Pension was granted (e.g., "2210 of 1901")
5. Date from which the Pension has been paid (e.g., "1901. 6th August")
6. Service for which the Pension was granted (e.g., "Second Shroff, Treasury")
7. Amount of Emolument when last employed in Public Service (maybe in dollars)
8. Present Age of Pensioner
9. Cause of Retirement (e.g., "Ill-health", "Age", "Abolition of Office")

Also there is "Свине of Retirement" maybe "Cause of Retirement".

The "Brought forward" line might be a summary row.

I'll try to parse each pensioner.

First pensioner: "Cheung Wan-tɑai," maybe "Cheung Wan-tai"? Then "Wadowa Singh." Could be two separate pensioners? Or "Cheung Wan-tai, Wadowa Singh"? Actually "Cheung Wan-tɑai," and "Wadowa Singh." appear on separate lines. Might be two names.

Then "J. A. Lowson," then "Fat Ngan," then "Mrs. Jane Ackers (now Mrs, Wahh)," then "John Lee," then "Fonja Singh," then "F. F. Remedioa," then "Fung Fu........" then "Lo Sik-ling," then "†C. Wagner," then "*Sir Wm. Meigh Goodman Kl." then "Janlah Singh." then "Tang Shi-kit," then "Elizabeth Annie Bateman" then "Karwo Dad," then maybe "Chinese Writer. Registrar General's Department," then "Hand Mistress, Belílios Public School," then "Assistant Warier, Victoria Gnol,".

But there are also numbers and dates interspersed.

Better approach: The original document is likely a table with rows for each pensioner. The OCR has read it line by line, mixing columns. I need to reconstruct the table by matching data to columns.

Let me try to group by pensioner.

I see "Cheung Wan-tɑai," then "230.00" "110.83" "J. A. Lowson," "1,634.00" "2210 of 1901," "2501 of 1901." "2062 of 1900," "Fat Ngan," "30,80" "3463 of 1901." "1901. 6th August." "1st September." "26th December." "1902. 1st January." "Second Shroff, Treasury,-" "600.00" "1st Class Assistant Warder, Victoria Gaol," "350,00" "Assistant Surgeon," "5,160,00" "Dispensarymat,..." "132,00" "£ 2 3 4" "80" "Ill-health." "63" "N.J" "Ill-health." "Age." "Mrs. Jane Ackers (now Mrs, Wahh)," "456.16" "John Lee," "1,158,10" "Fonja Singh," "F. F. Remedioa," "Fung Fu........" "Lo Sik-ling," "†C. Wagner," "Sir Wm. Meigh Goodman Kl." "Janlah Singh." "Tang Shi-kit," "Elizabeth Annie Bateman" "Karwo Dad," "87.26" "3203 of 1901," "57 of 1902." "2782 of 1902." "17th February," "Matron, Civil Hospital," "1,190.00" "71" "Ill-health." "1st April." "19th November." "Inspector under the Women & Girls Protection Ord.," "+" "1,878,00" "81" "2nd Class Assistant Warder," "308,00" "59" "Age." "Ill-health" "400.00" "336.40" "318,00" "47 11" "2349 of 1902." "3118 of 1902." "2911 of 1902," "6432 of 1901." "24th December. 1903." "3rd Clerk, Colonial Secretary's Office," "16th February," "Head Master, Sai-Ying-Pan School," "1,592.26" "696.00" "63" "11" "I" "1st July. 1904. 17th May." "Clerk, Education Departineut.." "720,00" "Age." "Ill-health." "1st Clerk, Police Department,." "1,440,00" "76" "Age." ":" "5,400,00" "57.50" "343 of 1905." "849) of 1905." "1905. 2716 April." "Chief Justice, *********" "13,500,00" "81" "17th November, † 2nd Class Assistant Wunder. Vietoria Gaol," "230,00" "H-health." "12.00" "J" "455.00" "1" "66.00" "901 of 190G." "3671 of 1995." "4694 of 1906," "Dist May," "Chinese Writer. Registrar General's Department," "240.00" "55" "Abolition of Ofer." "20th May." "Hand Mistress, Belílios Public School," "2,700,00" "N2" "Age." "ติ" "20th June," "Assistant Warier, Victoria Gnol," "264,00" "48" "III-health."

This is too messy. Perhaps the OCR output is from a multi-page table. The user expects me to output a Markdown table with the data. But I cannot guess the exact rows. However, I must do my best to reconstruct.

Given the instruction: "If data is clearly in table format, reconstruct the table using Markdown table syntax." So I should produce a Markdown table with the columns as per header.

I'll create a table with columns: Name of Pensioner, Pension (Sterling), Pension (Dollars), Authority, Date, Service, Emolument, Age, Cause of Retirement.

I'll try to extract rows from the text. Let's go sequentially.

The text starts with "Brought forward, $ c. 11,477.96 C.S.O. No." That might be a carried forward total.

Then "Cheung Wan-tɑai," maybe first pensioner. Next line "Wadowa Singh." Could be second. But then "230.00 110.83" maybe belong to Cheung Wan-tai? Or to Wadowa Singh? Then "J. A. Lowson, 1,634.00 2210 of 1901, 2501 of 1901. 2062 of 1900," That looks like J. A. Lowson has pension 1,634.00? But 1,634.00 what? Sterling? Dollars? The columns: Amount of Pension in Sterling, per annum; Amount of Pension in Dollars, per annum. So maybe 1,634.00 is in dollars? But there is also "230.00 110.83" two numbers. Could be Sterling and Dollars for Cheung Wan-tai.

Let's assume each pensioner has two amounts: Sterling and Dollars. The header shows "Amount of the Pension in Sterling, per annum." and "Amount of the Pension in Dollars, per annum." So two numeric columns.

Then "Authority under which the Pension was granted." e.g., "2210 of 1901". Then "Date from which the Pension has been paid." e.g., "1901. 6th August". Then "Service for which the Pension was granted." e.g., "Second Shroff, Treasury". Then "Amount of Emolument when last employed in Public Service." e.g., "600.00". Then "Present Age of Pensioner." e.g., "80". Then "Cause of Retirement." e.g., "Ill-health".

But the OCR mixes them.

Let's try to parse line by line with possible column breaks.

I'll write a script in my mind to parse. But as a language model, I'll do manually.

First, note that the OCR includes "£ 2 3 4" which might be a Sterling amount: £2 3s 4d? That could be a pension in Sterling.

Also "30,80" maybe 30.80 dollars.

"456.16" maybe dollars.

"1,158,10" maybe 1,158.10 dollars.

"87.26" dollars.

"1,190.00" dollars.

"1,878,00" dollars.

"308,00" dollars.

"400.00" dollars.

"336.40" dollars.

"318,00" dollars.

"1,592.26" dollars.

"696.00" dollars.

"720,00" dollars.

"1,440,00" dollars.

"5,400,00" dollars.

"57.50" dollars.

"13,500,00" dollars.

"230,00" dollars.

"12.00" dollars.

"455.00" dollars.

"66.00" dollars.

"240.00" dollars.

"2,700,00" dollars.

"264,00" dollars.

Many numbers.

Also ages: "80", "63", "71", "81", "59", "47", "63", "69", "76", "81", "55", "48".

Causes: "Ill-health", "Age", "Abolition of Office", "III-health".

Authorities: "2210 of 1901", "2501 of 1901", "2062 of 1900", "3463 of 1901", "3203 of 1901", "57 of 1902", "2782 of 1902", "2349 of 1902", "3118 of 1902", "2911 of 1902", "6432 of 1901", "343 of 1905", "849) of 1905", "901 of 190G", "3671 of 1995", "4694 of 1906".

Dates: "1901. 6th August", "1st September", "26th December", "1902. 1st January", "17th February", "1st April", "19th November", "24th December. 1903", "16th February", "1st July. 1904. 17th May", "1905. 2716 April", "17th November", "Dist May", "20th May", "20th June".

Services: "Second Shroff, Treasury", "1st Class Assistant Warder, Victoria Gaol", "Assistant Surgeon", "Dispensary Matron", "Matron, Civil Hospital", "Inspector under the Women & Girls Protection Ord.", "2nd Class Assistant Warder", "3rd Clerk, Colonial Secretary's Office", "Head Master, Sai-Ying-Pan School", "Clerk, Education Department", "1st Clerk, Police Department", "Chief Justice", "2nd Class Assistant Warder, Victoria Gaol", "Chinese Writer, Registrar General's Department", "Hand Mistress, Belilios Public School", "Assistant Warder, Victoria Gaol".

Now, need to match each pensioner to these attributes.

Let's list pensioners in order of appearance:

1. Cheung Wan-tai (maybe)
2. Wadowa Singh
3. J. A. Lowson
4. Fat Ngan
5. Mrs. Jane Ackers (now Mrs. Wahh)
6. John Lee
7. Fonja Singh
8. F. F. Remedios
9. Fung Fu
10. Lo Sik-ling
11. †C. Wagner
12. *Sir Wm. Meigh Goodman Kt.
13. Janlah Singh
14. Tang Shi-kit
15. Elizabeth Annie Bateman
16. Karwo Dad
17. Chinese Writer (Registrar General's Department) - maybe name not given? Or "Chinese Writer" is the service, name missing.
18. Hand Mistress, Belilios Public School - maybe name missing.
19. Assistant Warder, Victoria Gaol - maybe name missing.

But some of these might be services not names. For example, "Second Shroff, Treasury" is a service, not a name. The name might be "Cheung Wan-tai" for that service.

Let's try to match by looking at the numbers.

The first numbers after "Brought forward" are "230.00 110.83". Then "J. A. Lowson, 1,634.00". So perhaps "Cheung Wan-tai" has pension 230.00 Sterling? 110.83 Dollars? But 230 Sterling is huge. Maybe 230.00 is dollars, 110.83 is sterling? But header says first column Sterling, second Dollars. Could be reversed.

Let's check typical pensions: In 1900s Hong Kong, a pension of £230 per annum is plausible for a senior officer. $110.83 is low. But maybe the dollar column is in Mexican dollars? 1 pound = 4.8 dollars approx. So £230 = $1104. So 110.83 is too low. Maybe the numbers are reversed: 230.00 dollars, 110.83 sterling? 110.83 sterling = $532. Still not matching.

Alternatively, the two numbers might be "Amount of Pension in Sterling" and "Amount of Pension in Dollars" but the OCR has them in wrong order.

Look at "J. A. Lowson, 1,634.00". Only one number. Then "2210 of 1901, 2501 of 1901. 2062 of 1900," three authority numbers. Then "Fat Ngan, 30,80 3463 of 1901." So Fat Ngan has 30.80? Maybe dollars.

Then dates: "1901. 6th August. 1st September. 26th December. 1902. 1st January." These could be dates for multiple pensions.

Then services: "Second Shroff, Treasury,- 600.00 1st Class Assistant Warder, Victoria Gaol, 350,00 Assistant Surgeon, 5,160,00 Dispensarymat,... 132,00 £ 2 3 4 80 Ill-health. 63 N.J Ill-health. Age."

This looks like a list of services with emoluments and ages.

Maybe the table is structured with each row having: Name, Pension (Sterling), Pension (Dollars), Authority, Date, Service, Emolument, Age, Cause.

But the OCR has lost the row structure.

Given the difficulty, perhaps the best is to output the text as a cleaned-up version preserving paragraphs, but the user explicitly wants a Markdown table for tabular data. The instruction: "If data is clearly in table format, reconstruct the table using Markdown table syntax."

The data is clearly a table. I must attempt to reconstruct it.

I'll create a table with the columns I identified. I'll try to assign each pensioner a row based on the order of names and the data that follows.

Let's go through the text sequentially and try to group.

I'll write a pseudo-code:

Initialize list of pensioners.

Current pensioner = null.

For each token in text:

• If token looks like a name (capitalized, maybe with title), start new pensioner.
• Then collect numbers, authorities, dates, service, emolument, age, cause.

But doing manually.

Let's split the text into lines as given. The user provided a block of text with line breaks. I'll use those line breaks as they are.

Lines:

1. PENSIONS PAYABLE OUT OF THE REVENUE OF THE COLONY.

2.

1. Amount of the
2. Amount of the
3. Name of Pensioner.
4. Pension
5. in Sterling, per annum.
6. Pension
7. in Dollars, per annum.
8. Anthority under which the Pension
9. Date from which the Pension
10. Servico for which the Pension was granted.
11. was granted.
12. has been paid.

15.

1. Amount of Emolument when last employed in Public Service.
2. Present Age of Pensioner.
3. Свине
4. of Retirement.

20.

1. 320
2. (L4)
3. d.
4. Brought forward,
5. $ c.
6. 11,477.96
7. C.S.O. No.
8. Cheung Wan-tɑai,
9. Wadowa Singh.
10. 230.00
11. 110.83
12. J. A. Lowson, ·
13. 1,634.00
14. 2210 of 1901,
15. 2501 of 1901.
16. 2062 of 1900,
17. Fat Ngan,
18. 30,80
19. 3463 of 1901.
20. 1901. 6th August.
21. 1st September.
22. 26th December.
23. 1902. 1st January.
24. Second Shroff, Treasury,-
25. 600.00
26. 1st Class Assistant Warder, Victoria Gaol,
27. 350,00
28. Assistant Surgeon,
29. 5,160,00
30. Dispensarymat,...
31. 132,00
32. £ 2 3 4
33. 80
34. Ill-health.
35. 63
36. »
37. N.J
38. Ill-health.
39. Age.
40. Mrs. Jane Ackers (now
41. Mrs, Wahh),
42. 456.16
43. John Lee,
44. 1,158,10
45. Fonja Singh,
46. F. F. Remedioa,
47. Fung Fu........
48. Lo Sik-ling,
49. †C. Wagner,
50. *Sir Wm. Meigh Goodman
51. Kl.
52. Janlah Singh.
53. Tang Shi-kit,
54. Elizabeth Annie Bateman
55. Karwo Dad,
56. 87.26
57. 3203 of 1901,
58. 57 of 1902.
59. 2782 of 1902.
60. 17th February,
61. Matron, Civil Hospital,
62. 1,190.00
63. 71
64. Ill-health.
65. 1st April.
66. 19th November.
67. Inspector under the Women & Girls Protection Ord.,
68. +
69. 1,878,00
70. 81
71. 2nd Class Assistant Warder,
72. 308,00
73. 59
74. Age.
75. Ill-health
76. 400.00
77. 336.40
78. 318,00
79. 47 11
80. 2349 of 1902.
81. 3118 of 1902.
82. 2911 of 1902,
83. 6432 of 1901.
84. 24th December. 1903.
85. 3rd Clerk, Colonial Secretary's Office,
86. 16th February,
87. Head Master, Sai-Ying-Pan School,
88. 1,592.26
89. 696.00
90. 63
91. 11
92. I
93. 1st July. 1904. 17th May.
94. Clerk, Education Departineut..
95. 720,00
96. Age.
97. Ill-health.
98. 1st Clerk, Police Department,.
99. 1,440,00
100. 76
101. Age.
102. :
103. 5,400,00
104. 57.50
105. 343 of 1905.
106. 849) of 1905.
107. 1905. 2716 April.
108. Chief Justice, ********
109. 13,500,00
110. 81
111. 17th November, † 2nd Class Assistant Wunder. Vietoria Gaol,
112. 230,00
113. H-health.
114. 12.00
115. J
116. 455.00
117. 1
118. 66.00
119. 901 of 190G.
120. 3671 of 1995.
121. 4694 of 1906,
122. Dist May,
123. Chinese Writer. Registrar General's Department,
124. 240.00
125. 55
126. Abolition of Ofer.
127. 20th May.
128. Hand Mistress, Belílios Public School,
129. 2,700,00
130. N2
131. Age.
132. ти
133. 20th June,
134. Assistant Warier, Victoria Gnol,
135. 264,00
136. 48
137. III-health.

This is still messy. But note that the OCR has split the header across lines 3-19. The actual data starts at line 21.

Line 21: "320" maybe page number? "(L4)" "d." maybe reference.

Line 24: "Brought forward," line 25: "$ c." line 26: "11,477.96" line 27: "C.S.O. No."

Then line 28: "Cheung Wan-tɑai," line 29: "Wadowa Singh." line 30: "230.00" line 31: "110.83" line 32: "J. A. Lowson, ·" line 33: "1,634.00" line 34: "2210 of 1901," line 35: "2501 of 1901." line 36: "2062 of 1900," line 37: "Fat Ngan," line 38: "30,80" line 39: "3463 of 1901." line 40: "1901. 6th August." line 41: "1st September." line 42: "26th December." line 43: "1902. 1st January." line 44: "Second Shroff, Treasury,-" line 45: "600.00" line 46: "1st Class Assistant Warder, Victoria Gaol," line 47: "350,00" line 48: "Assistant Surgeon," line 49: "5,160,00" line 50: "Dispensarymat,..." line 51: "132,00" line 52: "£ 2 3 4" line 53: "80" line 54: "Ill-health." line 55: "63" line 56: "»" line 57: "N.J" line 58: "Ill-health." line 59: "Age." line 60: "Mrs. Jane Ackers (now" line 61: "Mrs, Wahh)," line 62: "456.16" line 63: "John Lee," line 64: "1,158,10" line 65: "Fonja Singh," line 66: "F. F. Remedioa," line 67: "Fung Fu........" line 68: "Lo Sik-ling," line 69: "†C. Wagner," line 70: "Sir Wm. Meigh Goodman" line 71: "Kl." line 72: "Janlah Singh." line 73: "Tang Shi-kit," line 74: "Elizabeth Annie Bateman" line 75: "Karwo Dad," line 76: "87.26" line 77: "3203 of 1901," line 78: "57 of 1902." line 79: "2782 of 1902." line 80: "17th February," line 81: "Matron, Civil Hospital," line 82: "1,190.00" line 83: "71" line 84: "Ill-health." line 85: "1st April." line 86: "19th November." line 87: "Inspector under the Women & Girls Protection Ord.," line 88: "+" line 89: "1,878,00" line 90: "81" line 91: "2nd Class Assistant Warder," line 92: "308,00" line 93: "59" line 94: "Age." line 95: "Ill-health" line 96: "400.00" line 97: "336.40" line 98: "318,00" line 99: "47 11" line 100: "2349 of 1902." line 101: "3118 of 1902." line 102: "2911 of 1902," line 103: "6432 of 1901." line 104: "24th December. 1903." line 105: "3rd Clerk, Colonial Secretary's Office," line 106: "16th February," line 107: "Head Master, Sai-Ying-Pan School," line 108: "1,592.26" line 109: "696.00" line 110: "63" line 111: "11" line 112: "I" line 113: "1st July. 1904. 17th May." line 114: "Clerk, Education Departineut.." line 115: "720,00" line 116: "Age." line 117: "Ill-health." line 118: "1st Clerk, Police Department,." line 119: "1,440,00" line 120: "76" line 121: "Age." line 122: ":" line 123: "5,400,00" line 124: "57.50" line 125: "343 of 1905." line 126: "849) of 1905." line 127: "1905. 2716 April." line 128: "Chief Justice, *********" line 129: "13,500,00" line 130: "81" line 131: "17th November, † 2nd Class Assistant Wunder. Vietoria Gaol," line 132: "230,00" line 133: "H-health." line 134: "12.00" line 135: "J" line 136: "455.00" line 137: "1" line 138: "66.00" line 139: "901 of 190G." line 140: "3671 of 1995." line 141: "4694 of 1906," line 142: "Dist May," line 143: "Chinese Writer. Registrar General's Department," line 144: "240.00" line 145: "55" line 146: "Abolition of Ofer." line 147: "20th May." line 148: "Hand Mistress, Belílios Public School," line 149: "2,700,00" line 150: "N2" line 151: "Age." line 152: "ти" line 153: "20th June," line 154: "Assistant Warier, Victoria Gnol," line 155: "264,00" line 156: "48" line 157: "III-health."

Now, I need to map each pensioner to a row. The names appear at lines 28, 29, 32, 37, 60-61, 63, 65, 66, 67, 68, 69, 70-71, 72, 73, 74, 75, and then later maybe the Chinese Writer, Hand Mistress, Assistant Warder are services not names.

But note that some names might be followed by their data. For example, "Cheung Wan-tɑai," then "Wadowa Singh." then numbers 230.00 110.83. Could be that Cheung Wan-tai and Wadowa Singh are two pensioners, but the numbers belong to the first? Or maybe the numbers are for "J. A. Lowson"? But J. A. Lowson appears after the numbers.

Let's look at the pattern: After "C.S.O. No." we have "Cheung Wan-tɑai," "Wadowa Singh." then "230.00" "110.83" then "J. A. Lowson, ·" then "1,634.00". So perhaps the first two names are separate pensioners, but their data is missing? Or the numbers 230.00 and 110.83 belong to Cheung Wan-tai? Then Wadowa Singh has no numbers? That seems unlikely.

Maybe the list is: Name, then Pension in Sterling, then Pension in Dollars, then Authority, then Date, then Service, then Emolument, then Age, then Cause.

But the OCR has interleaved multiple rows.

Another approach: The original table might have multiple columns and the OCR read it column by column? But the text seems to be row-major.

Given the complexity, perhaps the best is to output the text as a cleaned-up version with paragraphs, but the user explicitly wants a Markdown table for tabular data. However, the instruction says: "If data is clearly in table format, reconstruct the table using Markdown table syntax." It is clearly a table. I must attempt.

I'll create a table with the columns and fill in as many rows as I can deduce. For missing data, I'll put "..." or leave blank.

But the user said: "Missing text — if a word or phrase is clearly missing due to OCR damage, insert ... in its place."

So I can put ... for missing cells.

I'll try to reconstruct rows by matching each name with the subsequent data until the next name.

Let's assume each pensioner entry starts with a name. The names I see:

1. Cheung Wan-tai
2. Wadowa Singh
3. J. A. Lowson
4. Fat Ngan
5. Mrs. Jane Ackers (now Mrs. Wahh)
6. John Lee
7. Fonja Singh
8. F. F. Remedios
9. Fung Fu
10. Lo Sik-ling
11. †C. Wagner
12. *Sir Wm. Meigh Goodman Kt.
13. Janlah Singh
14. Tang Shi-kit
15. Elizabeth Annie Bateman
16. Karwo Dad
17. (Chinese Writer) - maybe name not given
18. (Hand Mistress) - maybe name not given
19. (Assistant Warder) - maybe name not given

But some of these might be services. For example, "Second Shroff, Treasury" is a service, not a name. "1st Class Assistant Warder, Victoria Gaol" is a service. "Assistant Surgeon" service. "Dispensary Matron" service. "Matron, Civil Hospital" service. "Inspector under the Women & Girls Protection Ord." service. "2nd Class Assistant Warder" service. "3rd Clerk, Colonial Secretary's Office" service. "Head Master, Sai-Ying-Pan School" service. "Clerk, Education Department" service. "1st Clerk, Police Department" service. "Chief Justice" service. "2nd Class Assistant Warder, Victoria Gaol" service. "Chinese Writer, Registrar General's Department" service. "Hand Mistress, Belilios Public School" service. "Assistant Warder, Victoria Gaol" service.

So the names are likely only the ones that look like personal names. The services are separate column.

Thus, each row should have a Name and a Service.

Now, the data after each name might be: Pension Sterling, Pension Dollars, Authority, Date, Service, Emolument, Age, Cause.

But the OCR has mixed them.

Let's try to parse row by row using the line numbers.

I'll write a table in Markdown with the columns:

I'll fill row by row.

Row 1: Cheung Wan-tai

• Name: Cheung Wan-tai
• Next line: Wadowa Singh (maybe another pensioner). But then numbers 230.00 110.83. Could be for Cheung Wan-tai? Or for Wadowa Singh? Let's see if there is a service for Cheung Wan-tai. Later we see "Second Shroff, Treasury" with emolument 600.00. That might be Cheung Wan-tai's service. But also "1st Class Assistant Warder, Victoria Gaol" with 350.00. That might be Wadowa Singh? Or another.

We need to match services to names. The services appear in a list after the dates. The dates: 1901. 6th August, 1st September, 26th December, 1902. 1st January. Four dates. Then four services: Second Shroff, Treasury; 1st Class Assistant Warder, Victoria Gaol; Assistant Surgeon; Dispensary Matron. Then emoluments: 600.00, 350.00, 5,160.00, 132.00. Then "£ 2 3 4" maybe a pension in Sterling for one of them. Then ages: 80, 63. Then causes: Ill-health, Ill-health, Age.

So perhaps the first four pensioners (Cheung Wan-tai, Wadowa Singh, J. A. Lowson, Fat Ngan) correspond to these four services.

Let's test: Cheung Wan-tai -> Second Shroff, Treasury, emolument 600, age 80, cause Ill-health.

Wadowa Singh -> 1st Class Assistant Warder, Victoria Gaol, emolument 350, age 63, cause Ill-health.

J. A. Lowson -> Assistant Surgeon, emolument 5,160, age? maybe 63? But there is "63" after "80". Actually line 55: "63" after "80". Then "N.J" "Ill-health." "Age." So maybe two causes: Ill-health and Age.

Fat Ngan -> Dispensary Matron, emolument 132, age? maybe 63? But there is "63" already used.

But we have four names before the services: Cheung Wan-tai, Wadowa Singh, J. A. Lowson, Fat Ngan. That matches four services.

Now, what about the numbers 230.00 110.83 and 1,634.00 and 30,80? These could be the pension amounts.

For Cheung Wan-tai: Pension Sterling? 230.00? Pension Dollars? 110.83? But 230 Sterling = 1104 dollars, not 110.83. Maybe the columns are reversed: 230.00 dollars, 110.83 sterling? 110.83 sterling = 532 dollars. Not matching.

Maybe the numbers are: 230.00 is the pension in dollars, 110.83 is something else? Or maybe it's 230.00 sterling and 110.83 is the pension in dollars for another? Hmm.

Look at J. A. Lowson: "1,634.00" only one number. Could be pension in dollars. 1,634 dollars = about 340 sterling. That could be.

Fat Ngan: "30,80" maybe 30.80 dollars? That seems low.

But then there is "£ 2 3 4" which is £2 3s 4d = £2.166. That is a small pension. Could be for Dispensary Matron.

And "132,00" is emolument for Dispensary Matron.

Let's check.

Maybe the pension amounts are in the "£ 2 3 4" line for the first four? But there are four pensioners, only one sterling amount.

This is too ambiguous.

Given the time, perhaps the best is to output the text as a cleaned-up version with the table structure approximated, but given the instruction to output only Markdown and not wrap in code fences, I'll produce a Markdown table with the data I can extract, leaving many cells blank or with ... .

But the user might expect a full reconstruction. However, the OCR is too garbled. I'll do my best to create a table with the rows I can identify, using the names and the services that follow.

I'll create a table with 19 rows (for each name). For each, I'll fill the service from the list of services in order. The services list appears in the text in this order:

1. Second Shroff, Treasury
2. 1st Class Assistant Warder, Victoria Gaol
3. Assistant Surgeon
4. Dispensary Matron
5. Matron, Civil Hospital
6. Inspector under the Women & Girls Protection Ord.
7. 2nd Class Assistant Warder
8. 3rd Clerk, Colonial Secretary's Office
9. Head Master, Sai-Ying-Pan School
10. Clerk, Education Department
11. 1st Clerk, Police Department
12. Chief Justice
13. 2nd Class Assistant Warder, Victoria Gaol (again)
14. Chinese Writer, Registrar General's Department
15. Hand Mistress, Belilios Public School
16. Assistant Warder, Victoria Gaol

That's 16 services. But we have 19 names. Some names might not have services? Or some services correspond to multiple names? Actually, the list of names includes some that might be the same person? For example, "†C. Wagner" and "*Sir Wm. Meigh Goodman Kt." are likely pensioners with high ranks. "Chief Justice" is a service for Sir Wm. Meigh Goodman.

Let's match known historical figures: Sir William Meigh Goodman was Chief Justice of Hong Kong. So he should have service "Chief Justice". That appears later.

"†C. Wagner" might be a police officer? "2nd Class Assistant Warder, Victoria Gaol" appears twice.

"Mrs. Jane Ackers (now Mrs. Wahh)" might be "Matron, Civil Hospital".

"Elizabeth Annie Bateman" might be "Inspector under the Women & Girls Protection Ord.".

"Hand Mistress, Belilios Public School" might be a woman.

"Chinese Writer, Registrar General's Department" might be "Tang Shi-kit" or "Lo Sik-ling".

"Assistant Warder, Victoria Gaol" might be "Karwo Dad" or "Janlah Singh".

But this is speculation.

Given the difficulty, I think the appropriate response is to output the text as a cleaned-up version preserving the paragraph structure, but the user explicitly said to reconstruct tables. However, the instruction also says: "Your most important task is to fix the text flow while preserving the original paragraph structure." But this is a table, not paragraphs. The primary task is to reconstruct sentences and paragraphs, but for tables, reconstruct the table.

I will output a Markdown table with the columns and as many rows as I can, using the data in order. I'll include the "Brought forward" as a row maybe.

Let's create a table with the following columns: Name, Pension Sterling, Pension Dollars, Authority, Date, Service, Emolument, Age, Cause.

I'll parse the text sequentially and assign each piece to the next empty cell in the current row, but that's complex.

Given the constraints, I'll produce a table that captures the visible data in a structured way, even if not perfectly aligned. I'll use the header row as given.

I'll write:

But this is largely invented.

The user might not want me to invent. But the instruction says: "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." So I should not invent rows. I should only use the text as is.

Perhaps the best is to output the text as a series of paragraphs, but the table is the main content. The user said: "Return the proofread text in standard Markdown: Use #, ##, ### for headers found in the original. Use bold for labels, titles, and emphasized text. Use Markdown table syntax to reconstruct tabular data."

So I must use Markdown table syntax. I'll create a table with the header row as in the original, and then for the data, I'll try to keep the original order of cells as they appear in the OCR, but in a table format. However, the OCR doesn't have clear cell delimiters.

Given the difficulty, I'll output the text as a Markdown table with a single column "Raw Text" and put the whole text there? That would not be reconstructing the table.

Alternatively, I can output the text as a cleaned-up version with line breaks, but not a table. But the instruction says to reconstruct tabular data.

I think I need to make a reasonable attempt. I'll create a table with the columns as per the header, and then for each line in the OCR that seems to be a row, I'll create a row. But the OCR lines are not rows.

Maybe the original scan had a table with vertical lines, and the OCR read it as text with spaces. The numbers align in columns? In the provided text, there are no clear columns.

Given the time, I'll produce a Markdown table that includes all the text in a single column, but that's not right.

Wait, the user might be testing my ability to handle messy OCR. I should do my best to reconstruct the table by logical grouping.

I'll write a script-like approach in my reasoning, but final output must be Markdown.

Let me try to parse the text into a list of tokens and then assign to columns based on patterns.

But as an AI, I can simulate that.

I'll assume the table has 9 columns. The OCR text is a linear stream of cells. The header row gives the order. The data rows follow. The "Brought forward" might be a row with only some columns.

I'll read the text after the header as a sequence of cells. The cells are separated by line breaks in the OCR, but actually each cell might be on a new line. In the provided text, each line seems to be a cell. For example, line 28 "Cheung Wan-tɑai," is a cell (Name). Line 29 "Wadowa Singh." is another Name? Or maybe it's the next cell for the same row? But the header has only one Name column. So each row has one Name. So line 28 and 29 are two separate rows? But then line 30 "230.00" would be Pension Sterling for first row? But then line 31 "110.83" would be Pension Dollars for first row? Then line 32 "J. A. Lowson, ·" would be Name for second row? But then we have only two numbers for first row, but we need more columns.

If each line is a cell, then the table would have many rows with only one cell each. That's not right.

Perhaps the OCR output has each cell on a new line, but the table has multiple columns, so the cells are ordered row by row. For example, row1: cell1, cell2, cell3,... row2: cell1, cell2,...

If that's the case, we need to know how many columns. The header has 9 columns (Name, Pension Sterling, Pension Dollars, Authority, Date, Service, Emolument, Age, Cause). But the header lines are split: "Amount of the" "Amount of the" "Name of Pensioner." "Pension" "in Sterling, per annum." "Pension" "in Dollars, per annum." "Anthority under which the Pension" "Date from which the Pension" "Servico for which the Pension was granted." "was granted." "has been paid." "Amount of Emolument when last employed in Public Service." "Present Age of Pensioner." "Свине" "of Retirement." That's 15 lines for header. But some are continuations. Actually, the header likely spans two rows: first row: "Amount of the Pension in Sterling, per annum." "Amount of the Pension in Dollars, per annum." "Name of Pensioner." "Authority under which the Pension was granted." "Date from which the Pension has been paid." "Service for which the Pension was granted." "Amount of Emolument when last employed in Public Service." "Present Age of Pensioner." "Cause of Retirement." That's 9 columns.

The OCR split each header cell into multiple lines. So the header cells are not each line.

Thus, the data lines are also not each a cell.

Given the impossibility, I'll output the text as a cleaned-up version with the table represented as a Markdown table with the header and then a single row containing the entire data as text? No.

I think the best is to output the text as a series of paragraphs, but the user explicitly said to use Markdown table syntax for tabular data. I'll create a table with two columns: "Field" and "Value" and put each line as a row? That would be a key-value table.

But the original is a multi-column table.

Given the instructions, I'll produce a Markdown table with the 9 columns and fill in the data as I can, using ... for missing.

I'll base it on the order of names and the subsequent data blocks.

Let's list the names in order of appearance with the text that follows until the next name.

1. Cheung Wan-tai

Followed by: Wadowa Singh. 230.00 110.83 J. A. Lowson, · 1,634.00 2210 of 1901, 2501 of 1901. 2062 of 1900, Fat Ngan, 30,80 3463 of 1901. 1901. 6th August. 1st September. 26th December. 1902. 1st January. Second Shroff, Treasury,- 600.00 1st Class Assistant Warder, Victoria Gaol, 350,00 Assistant Surgeon, 5,160,00 Dispensarymat,... 132,00 £ 2 3 4 80 Ill-health. 63 » N.J Ill-health. Age. Mrs. Jane Ackers (now Mrs, Wahh), 456.16 John Lee, 1,158,10 Fonja Singh, F. F. Remedioa, Fung Fu........ Lo Sik-ling, †C. Wagner, Sir Wm. Meigh Goodman Kl. Janlah Singh. Tang Shi-kit, Elizabeth Annie Bateman Karwo Dad, 87.26 3203 of 1901, 57 of 1902. 2782 of 1902. 17th February, Matron, Civil Hospital, 1,190.00 71 Ill-health. 1st April. 19th November. Inspector under the Women & Girls Protection Ord., + 1,878,00 81 2nd Class Assistant Warder, 308,00 59 Age. Ill-health 400.00 336.40 318,00 47 11 2349 of 1902. 3118 of 1902. 2911 of 1902, 6432 of 1901. 24th December. 1903. 3rd Clerk, Colonial Secretary's Office, 16th February, Head Master, Sai-Ying-Pan School, 1,592.26 696.00 63 11 I 1st July. 1904. 17th May. Clerk, Education Departineut.. 720,00 Age. Ill-health. 1st Clerk, Police Department,. 1,440,00 76 Age. : 5,400,00 57.50 343 of 1905. 849) of 1905. 1905. 2716 April. Chief Justice, ********* 13,500,00 81 17th November, † 2nd Class Assistant Wunder. Vietoria Gaol, 230,00 H-health. 12.00 J 455.00 1 66.00 901 of 190G. 3671 of 1995. 4694 of 1906, Dist May, Chinese Writer. Registrar General's Department, 240.00 55 Abolition of Ofer. 20th May. Hand Mistress, Belílios Public School, 2,700,00 N2 Age. ти 20th June, Assistant Warier, Victoria Gnol, 264,00 48 III-health.

This is one big block. The names are embedded. So the OCR has not separated rows.

Given this, I think the only feasible output is to present the text as a cleaned-up paragraph, but the user said to use Markdown table for tabular data. However, the text is not clearly tabular in the OCR output; it's a jumble. The instruction says: "If data is clearly in table format, reconstruct the table using Markdown table syntax." The data is clearly from a table, but the OCR output is not clearly in table format. So maybe I should not force a table, but instead output the text as a cleaned-up version with the header and then the data in a list.

But the user might expect a table. I'll compromise: I'll output the header as a Markdown table header, and then put the entire remaining text in a single cell? That seems wrong.

Given the time, I'll output a Markdown table with the columns and then a few rows that I can confidently extract, and for the rest, put a row with "..." indicating missing.

But the instruction: "Do not add or remove substantive words." So I shouldn't add rows that aren't there.

I think the best is to output the text as a cleaned-up version preserving the original line breaks, but with corrected words, and not as a table. However, the user explicitly said: "Use Markdown table syntax to reconstruct tabular data." So I must attempt a table.

I'll create a table with the header row and then one row per line of the OCR after the header, but that would be many rows with one column each. Not good.

Maybe the OCR output is already in a table format if we consider the original had columns. The user might have provided the OCR text as a block, but the original image is a table. My job is to reconstruct that table.

I'll assume the table has 9 columns. I'll write a program in my mind to parse the text into 9 columns by using the fact that the header mentions the columns. But without the original alignment, it's guesswork.

Given the difficulty, I'll output a Markdown table with the header and then a single row containing the entire data as a string, but that's not reconstruction of text. That would be a 1x9 table? No.

I'll do this: Provide the header row, and then for each pensioner name I can identify, create a row with the name and the service that seems associated, and leave other columns blank.

But that adds substantive words (the service assignments).

Given the constraints, I'll output the text as a cleaned-up version with the table represented in Markdown as best as I can, using the pipe syntax, but with the data in the same order as the OCR, grouped by the header.

I'll write:

But this is largely fabricated.

Given the instruction not to add substantive words, I should not fabricate the service assignments. However, the services are in the text, so I'm not fabricating, just associating.

But the association is guesswork.

Perhaps the correct approach is to output the text as a series of paragraphs, each representing a row, with the fields separated by commas. But the user said Markdown table.

I'll check the instruction again: "Return the proofread text in standard Markdown: - Use #, ##, ### for headers found in the original. - Use bold for labels, titles, and emphasized text (e.g., RESTRICTED, CONFIDENTIAL, MEMORANDUM). - Use Markdown table syntax (| col | col |) to reconstruct tabular data. - Wrap file references and codes in backticks. - Do not wrap your response in code fences. - Do not output HTML tags. - Do not add any commentary, notes, or explanations."

So I must use Markdown table syntax for tabular data. The data is tabular. I must reconstruct it.

Given the OCR is from a known historical document, maybe I can recall the exact table? But I don't have external knowledge.

I'll do my best to reconstruct the table by using the OCR text as a sequence of cells in row-major order. Assume the table has 9 columns. The header row is 9 cells. Then the data rows follow. The OCR text after the header is a list of cells. But the OCR has line breaks that may correspond to cell boundaries. In the provided text, each line might be a cell. Let's count lines after the header.

Header ends at line 19. Then line 20 empty. Line 21 "320" line 22 "(L4)" line 23 "d." line 24 "Brought forward," line 25 "$ c." line 26 "11,477.96" line 27 "C.S.O. No." line 28 "Cheung Wan-tɑai," line 29 "Wadowa Singh." line 30 "230.00" line 31 "110.83" line 32 "J. A. Lowson, ·" line 33 "1,634.00" line 34 "2210 of 1901," line 35 "2501 of 1901." line 36 "2062 of 1900," line 37 "Fat Ngan," line 38 "30,80" line 39 "3463 of 1901." line 40 "1901. 6th August." line 41 "1st September." line 42 "26th December." line 43 "1902. 1st January." line 44 "Second Shroff, Treasury,-" line 45 "600.00" line 46 "1st Class Assistant Warder, Victoria Gaol," line 47 "350,00" line 48 "Assistant Surgeon," line 49 "5,160,00" line 50 "Dispensarymat,..." line 51 "132,00" line 52 "£ 2 3 4" line 53 "80" line 54 "Ill-health." line 55 "63" line 56 "»" line 57 "N.J" line 58 "Ill-health." line 59 "Age." line 60 "Mrs. Jane Ackers (now" line 61 "Mrs, Wahh)," line 62 "456.16" line 63 "John Lee," line 64 "1,158,10" line 65 "Fonja Singh," line 66 "F. F. Remedioa," line 67 "Fung Fu........" line 68 "Lo Sik-ling," line 69 "†C. Wagner," line 70 "Sir Wm. Meigh Goodman" line 71 "Kl." line 72 "Janlah Singh." line 73 "Tang Shi-kit," line 74 "Elizabeth Annie Bateman" line 75 "Karwo Dad," line 76 "87.26" line 77 "3203 of 1901," line 78 "57 of 1902." line 79 "2782 of 1902." line 80 "17th February," line 81 "Matron, Civil Hospital," line 82 "1,190.00" line 83 "71" line 84 "Ill-health." line 85 "1st April." line 86 "19th November." line 87 "Inspector under the Women & Girls Protection Ord.," line 88 "+" line 89 "1,878,00" line 90 "81" line 91 "2nd Class Assistant Warder," line 92 "308,00" line 93 "59" line 94 "Age." line 95 "Ill-health" line 96 "400.00" line 97 "336.40" line 98 "318,00" line 99 "47 11" line 100 "2349 of 1902." line 101 "3118 of 1902." line 102 "2911 of 1902," line 103 "6432 of 1901." line 104 "24th December. 1903." line 105 "3rd Clerk, Colonial Secretary's Office," line 106 "16th February," line 107 "Head Master, Sai-Ying-Pan School," line 108 "1,592.26" line 109 "696.00" line 110 "63" line 111 "11" line 112 "I" line 113 "1st July. 1904. 17th May." line 114 "Clerk, Education Departineut.." line 115 "720,00" line 116 "Age." line 117 "Ill-health." line 118 "1st Clerk, Police Department,." line 119 "1,440,00" line 120 "76" line 121 "Age." line 122 ":" line 123 "5,400,00" line 124 "57.50" line 125 "343 of 1905." line 126 "849) of 1905." line 127 "1905. 2716 April." line 128 "Chief Justice, *********" line 129 "13,500,00" line 130 "81" line 131 "17th November, † 2nd Class Assistant Wunder. Vietoria Gaol," line 132 "230,00" line 133 "H-health." line 134 "12.00" line 135 "J" line 136 "455.00" line 137 "1" line 138 "66.00" line 139 "901 of 190G." line 140 "3671 of 1995." line 141 "4694 of 1906," line 142 "Dist May," line 143 "Chinese Writer. Registrar General's Department," line 144 "240.00" line 145 "55" line 146 "Abolition of Ofer." line 147 "20th May." line 148 "Hand Mistress, Belílios Public School," line 149 "2,700,00" line 150 "N2" line 151 "Age." line 152 "ти" line 153 "20th June," line 154 "Assistant Warier, Victoria Gnol," line 155 "264,00" line 156 "48" line 157 "III-health."

If each line is a cell, then we have 137 cells after header. 137 / 9 = 15.2 rows. So maybe 15 rows of 9 columns, with some extra.

But the header might have more than 9 columns? Let's count header cells from the lines: The header lines 3-19 are 17 lines. But some are continuations. For example, "Amount of the" and "Pension" and "in Sterling, per annum." might be one cell. So the header cells are not per line.

Given the difficulty, I'll output the text as a cleaned-up version with the table represented as a Markdown table with the header and then the data rows as they appear in the OCR, but I'll group every 9 lines as a row? That would be arbitrary.

I think the only way to satisfy the user is to output a Markdown table that contains the entire OCR text in a single column, but that's not reconstructing.

Given the time, I'll produce a Markdown table with the header row and then a single row with the entire remaining text in the first cell, and other cells empty. But that's silly.

I'll instead output the text as a cleaned-up version with paragraphs, and for the table, I'll create a table with the header and then note that the data is too garbled to reconstruct accurately, but the instruction says not to add commentary.

I'm stuck.

Perhaps the user expects me to use the OCR text to create a table by aligning the columns based on the header. Since I cannot do that perfectly, I'll do a reasonable approximation.

I'll write a Markdown table with the 9 columns and fill in the data by reading the text sequentially and assigning to columns based on the type of data (name, number, date, service, etc.). I'll do this manually for the first few rows and then put ... for the rest.

But the instruction: "Preserve original content — do not add or remove substantive words." So I must include all the words from the OCR. If I put them in a table, I'm not removing, just arranging.

I'll create a table with two columns: "Original Line" and "Content" and put each line as a row. That would preserve all content and be a table. But that's not reconstructing the original table.

The user said: "reconstruct tabular data". The original data is a table with multiple columns. Reconstructing it means producing a similar table.

Given the difficulty, I'll assume the table has the columns as in the header. I'll parse the text into those columns by using the fact that the OCR might have read the table row by row, but the line breaks in the OCR correspond to row breaks? Look at the text: after "C.S.O. No." we have "Cheung Wan-tɑai," then "Wadowa Singh." then numbers. That could be two rows? But then "J. A. Lowson," appears later.

Maybe the table is arranged with two columns of names? No.

I'll look at the original document: "PENSIONS PAYABLE OUT OF THE REVENUE OF THE COLONY." This is likely from the Hong Kong Blue Book. The table lists pensioners. Typically, such tables have columns: Name, Pension (£), Pension ($), Authority, Date, Service, Emolument, Age, Cause.

The OCR has captured the data but scrambled.

Given the constraints, I'll output a Markdown table with the header and then for each name I can identify, a row with the name and the subsequent data until the next name, but that would be messy.

I'll do this: I'll output the text as a cleaned-up version with the table represented in Markdown as a table with the header and then a single row per pensioner, using the data I can extract. I'll include all the text in the table cells, but split by the columns as best as I can.

I'll write a Python script in my mind to parse, but I'll do it manually for the first few.

Let's try to split the text into rows by detecting names. The names are: Cheung Wan-tai, Wadowa Singh, J. A. Lowson, Fat Ngan, Mrs. Jane Ackers (now Mrs. Wahh), John Lee, Fonja Singh, F. F. Remedios, Fung Fu, Lo Sik-ling, †C. Wagner, *Sir Wm. Meigh Goodman Kt., Janlah Singh, Tang Shi-kit, Elizabeth Annie Bateman, Karwo Dad. That's 16 names.

Then there are services: Second Shroff, Treasury; 1st Class Assistant Warder, Victoria Gaol; Assistant Surgeon; Dispensary Matron; Matron, Civil Hospital; Inspector under the Women & Girls Protection Ord.; 2nd Class Assistant Warder; 3rd Clerk, Colonial Secretary's Office; Head Master, Sai-Ying-Pan School; Clerk, Education Department; 1st Clerk, Police Department; Chief Justice; 2nd Class Assistant Warder, Victoria Gaol; Chinese Writer, Registrar General's Department; Hand Mistress, Belilios Public School; Assistant Warder, Victoria Gaol. That's 16 services. Good! So each name corresponds to a service in order.

Thus, the table has 16 rows. The columns: Name, Pension Sterling, Pension Dollars, Authority, Date, Service, Emolument, Age, Cause.

Now we need to assign the other data to these rows.

The data between the names and services includes numbers, authorities, dates, emoluments, ages, causes.

Let's list the data in order as they appear in the text after the names.

The text after "C.S.O. No.":

Cheung Wan-tɑai,

Wadowa Singh.

230.00

110.83

J. A. Lowson, ·

1,634.00

2210 of 1901,

2501 of 1901.

2062 of 1900,

Fat Ngan,

30,80

3463 of 1901.

1. 6th August.

1st September.

26th December.

1. 1st January.

Second Shroff, Treasury,-

600.00

1st Class Assistant Warder, Victoria Gaol,

350,00

Assistant Surgeon,

5,160,00

Dispensarymat,...

132,00

£ 2 3 4

80

Ill-health.

63

»

N.J

Ill-health.

Age.

Mrs. Jane Ackers (now

Mrs, Wahh),

456.16

John Lee,

1,158,10

Fonja Singh,

F. F. Remedioa,

Fung Fu........

Lo Sik-ling,

†C. Wagner,

*Sir Wm. Meigh Goodman

Kl.

Janlah Singh.

Tang Shi-kit,

Elizabeth Annie Bateman

Karwo Dad,

87.26

3203 of 1901,

57 of 1902.

2782 of 1902.

17th February,

Matron, Civil Hospital,

1,190.00

71

Ill-health.

1st April.

19th November.

Inspector under the Women & Girls Protection Ord.,

+

1,878,00

81

2nd Class Assistant Warder,

308,00

59

Age.

Ill-health

400.00

336.40

318,00

47 11

2349 of 1902.

3118 of 1902.

2911 of 1902,

6432