The ticket of leave records at the Archives Authority of New South Wales (AANSW) are incomplete.
|DATES||RECORD TYPE||TICKET NUMBERS||DOCUMENT|
|2 July 1810 to 3 October 1814||List or register of tickets||1 to 105/737||COD 18 Reel 601|
|29 July 1824 to 8 March 1827||Register of tickets issued||1227 to 2422||Reel 890|
|16 May 1827 to 1 October 1833||Register of tickets issued||28/77 to 33/887||Reel 890|
|31 March 1827 to 31 Dec 1867||Ticket of leave butts||28/77 to 75/1||Reels 909 to 965 inclusive|
In addition to the above, the convict indents for arrivals after 1800 have three columns for notations of ticket of leave numbers (TL), conditional pardons (CP) and absolute pardons (AP). This information is not noted on the pre-1801 indents.
Extant ticket of leave registers and ticket butts seem very reliable. The register numbers and butt numbers run consecutively. The details called for by the form of the records are generally completed thoroughly. Particularly in the later years, they were obviously used as working documents, with cancellations (e.g. on death, emancipation, renewal or revocation) and alterations (e.g. to district for which the ticket was valid) frequently being noted.
The indents have some deficiencies. Some microfilm copies are of poor quality, and for some pages the copy omits the right hand edge, where the ticket numbers are noted. Some of the original indents, particularly for the later years of transportation, are in very poor condition (see Note 1). There is also a possibility that some entries in the indents have been altered through forgery. However, fresh indents were called for from the UK on a number of occasions to correct the frauds, and errors are probably minimal.
The gaps in the records in Table 11, and duplication in the number series used by Macquarie and Brisbane (see Appendix 10), makes complete confirmation impossible. Nevertheless, where ticket details are available they seem to be fairly reliable. For example, for tickets issued after 1826, when only one numbering system was used, have proven very accurate. A small discovery sample of arrivals in 1801-1802, without ticket details shown on the indents, was compared with muster lists to check reliability of indent data, proving to be at least 95% accurate. The results suggest that the indents are accurate in recording ticket numbers for male convicts, however the results for tickets granted to female convicts in the early years might be understated (see Note 2).
Where ticket numbers given in the indents were in the same number series as the 1810-1814 Ticket of Leave Register but convict names did not match, the Certificate of Freedom Register (see Note 3) for that period was checked. The numbers were not for certificates of freedom.
Tickets of leave details were checked against register and butt records to verify, and to extract issue details. Where register or butt records are not extant, the ticket numbering system (see Appendix 10) was used in conjunction with the musters for 1800-2, 1806, 1811, 1814, 1822 and 1837, and the 1828 census (see Note 8), to determine which ticket series a convict fell into, and calculate the year of issue. In a small proportion of cases (13 out of 274, or 4.7%) there were two or more possible issue dates (see Note 9). These cases were excluded from the sample when calculating the average waiting time for tickets.
Ticket of leave details were extracted from indent sample, and combined with ticket of leave details from the sample of ticket of leave register, and the sample from the ticket of leave butts. Doubtful entries were cross-checked with muster and census records where possible (for example where butts omitted the sentence this could sometimes be picked up from the musters).
The data was then analysed. Results are in Appendix 6.
Note 2 To be statistically reliable for a yes/no response (i.e. accurate/not accurate) a sample of at least 30 is required. Of those sampled, 5% (2 of 38 in sample) showed PRIS TL (i.e. prisoner, ticket of leave) in 1806 Muster. Both convicts were listed as concubines, one with 2 children. As tickets were not always formally granted to female convicts who went off the stores when supported by a man, the errors in the indents are not surprising. The remainder of the sample did not conflict with indent data.
Note 3 AANSW COD 18; Reel 601.
Note 4 Using a computer program which generates random numbers. The numbers were, in order: 10, 30, 79, 9, 83.
Note 5 The series of numbers ran 9, 10, 30, 79, 83, 109, 110, 130, 179, 183, 209, 210, 230 et seq. Five array samples were chosen, rather than simply taking every twentieth entry, because it was not known at the start whether time would permit such a large sample. If time was short, the last random number, viz. 83, could be dropped without jeopardizing the randomness of the remaining four samples. For the importance of randomness to the reliability of the samples see Frederick Mills, Statistical Methods , 3rd ed, Pitman, London, 1955, pp.175 et seq.
Note 6 For convenience, indents for 1796 to 1800 were sampled from the lists contained in Cobley, Sydney Cove , V, 1795-1800.
Note 7 For example, the delay between trial and arrival in NSW dropped to about a year after 1815 (see Appendix 3, Table 340), so I stopped recording trial details (except for sentence) after 1817. I did randomly check the indents and ticket butts to ensure no major increase later.
Note 8 E.g. by comparison of civil status at various times.
Note 9 The muster details did not always enable a definite conclusion.
The next available appendix is Appendix 10: Ticket of Leave Numbering System
© Copyright Sid Hammell 1992: last updated 27 June 2005 Comments to email@example.com