Registration worked in the following way. After a deed had been signed and witnessed, one of the parties to it had a copy made, known as a 'memorial', signed it and had it in turn witnessed by two people, at least one of whom had to have also witnessed the original. The memorial was then sworn before a justice of the peace as a faithful copy of the original and sent to the Registry. Here it was transcribed into a large manuscript volume and indexed. The original memorial was retained and stored, and these are all still preserved in the vaults of the Registry. For research purposes, however, the large manuscript volumes containing the transcripts of the memorials are used. The registration of a deed normally took place fairly soon after its execution, within a month or two in most cases, although delays of up to two years are quite common. If the gap between the execution and the registration of a deed is much longer, this may be significant: it indicates an impending need for one of the parties to the deed, or their heirs, to be able to show legal proof of its execution. The most common reason for such a necessity would have been the death of one of the parties.
The indexing system used by the Registry is complicated and incomplete. There are two sets of indexes: one by grantor's name (i.e. the name of the party disposing of the asset), the other by the name of the townland or street in which the property was situated.
is fully alphabetical and is divided into a number of sets covering different initial letters and periods. Between 1708 and 1833 the Grantors Index records the name of the grantor, the surname of the first grantee, and the volume, page and memorial number. No indication is given of the location of the property concerned-an omission that can make a search for references to a family with a common surname very tedious indeed. After 1833 the index is more comprehensive, listing the county in which the property was situated. In general, the index is remarkably accurate, but there are some mistakes, particularly in the volume and page references. In such instances the memorial number can be used to trace the transcript; several transcribers worked simultaneously on different volumes, and the volume numbers were sometimes transposed. If, for example, Volume 380 is not the correct reference, Volumes 378-82 may contain the transcript. Within each volume the transcripts are numerically consecutive.
is subdivided by county and is roughly alphabetical within each county, with townland names grouped together under their initial letter. This means that a search for deeds relating to, say, Ballyboy, Co. Roscommon, involves a search through all the references to Co. Roscommon townlands that start with the letter 'B'. The information given in the index is brief, recording only the surnames of two of the parties and the volume, page and deed numbers. As with the Grantors Index, the index is divided into a number of sets covering different periods. After 1828 it further subdivides the townlands by barony, making research a good deal more efficient. Alongside the county volumes there are separate indexes for corporation towns and cities. The subdivisions within these are somewhat eccentric, particularly in the case of Dublin, making it necessary to search even more widely than in the rural indexes. It should be pointed out that the Registry does not make it possible for the history of all the transactions in which a property was involved to be traced, because inevitably some of the deeds recording the transactions have not been registered.
In general, of the Registry's two sets of indexes the Grantors Index is the most genealogically useful, because it is strictly alphabetical and lists transactions by person rather than by property. The greatest omission in the Registry is of an index to the grantees: given the distribution of wealth in the country, the social range covered would be enhanced greatly by the production of such an index. Microfilm copies of both the Lands Index and the Grantors Index, amounting to more than four hundred reels, are available at nli, proni and the LDS Family History Library. Volume 1 of Margaret Falley's Irish and Scotch-Irish Ancestral Research, available on open shelves in the nli reading room, gives a complete breakdown of the locations and microfilm numbers of the nli indexes up to 1850 (pp. 71-90). The only online access is by means of the Registry of Deeds Index Project at freepages.genealogy.rootsweb.ancestry.com/~registryofdeeds, a valiant attempt to index all the names in the eighteenth-century memorials using the LDS microfilms.