290 Joseph E. Harris
(1964) who averaged reports for ports and noted a likely high count of 1 million
for the nineteenth century. For the southern (Swahili) coast, nineteenth-
century data came from British consular reports in Zanzibar and the navy
patrols. Most estimates have been based on those reports, which are criticized
as reflecting abolitionist proclivity for high figures to justify greater vigilance
for suppression. Some observers have argued further that abolitionists also
did not take into account the demand for slaves for plantation labour in East
Africa and thus assumed that all slaves were for export when in fact some were
kept in Zanzibar and Pemba. A distinction should therefire be made between
local and export trade for Zanzibar which seems hard to establish at this point,
especially since the Zanzibar archives have been closed to researchers since
1964.
The problem of quantifying becomes greater at the Asian points of entry.
In the Persian Gulf for example, one of the most active ports, Sur, had no
customs office, and Muscat's records are scattered and difficult to obtain. In
addition, quantification is hampered by the inability to distinguish between
sales and resales. Slaves were frequently moved from Mecca, Jidda, Sur,
Muscat and other points to Basra, Bushire, and on to India which itself received
some direct shipments.
Indian shipments came through several small areas in the north-west,
Gujarat in particular, but estimates are virtually non-existent. For at least two
depots, Surat and Bombay, there are scattered documents in the East India
Company records which are available in India; police records are also useful
since the Commissioner of Police was charged with the responsibility of
monitoring the import of slaves and later with the distribution of freed slaves
among families in India or placing them in mission stations or on government
farms.
The argument here is for the organization of a concerted approach to
these problems in a number of key locations known to have been important
slave depots, markets, etc. Critical aspects of this would be the identification
of all available official and unofficial records in East Africa and across the
length of Asia. This obviously would require the involvement of researchers in
the affected areas; and the search should not be limited to coastal regions.
Indeed, my research in India, Hyderabad in particular, revealed the importance
of the slave trade and documentary evidence in an interior centre. Repositories
in Hyderabad contain numerous uncounted stacks of uncatalogued documents
on political, economic, and military matters written in Arabic and Urdu. A
cursory examination of a sample of those documents by an informant confirmed
their relevance to the slave trade and the African presence in the area. A some-
what similar situation exists in Iran, Kerman province and the area around
Bandar Abbas in particular, where evidence in Arabic and Farsi needs to be
identified and made available to interested researchers.^3