Registered and unregistered title

This week concerns the machinery by which land law decides who is bound by proprietary rights. Week 1 identified the estates and interests which English law recognises. Week 2 asks the more forensic question: once an estate or interest exists, how does it fare against a later purchaser, mortgagee, donee, trustee in bankruptcy, or occupier? The answer differs radically according to whether title to the land is registered or unregistered.

The registered-title system is now the ordinary system. Its central statute is the Land Registration Act 2002. It does not abolish equitable interests, beneficial interests, easements, restrictive covenants or leases; rather, it supplies a priority code for deciding when such interests bind later takers. The register is intended to be the principal source of information about the estate. But it is not an exhaustive map of all proprietary burdens. Some interests bind because they are entered on the register, some because the statute treats them as overriding despite non-entry, some because a purchaser is not within the statutory protection, and some because they have been overreached rather than binding the land.

Unregistered title is conceptually older and doctrinally untidier. Legal estates and legal interests generally bind the world. Equitable interests historically depended on the doctrine of notice: a purchaser of a legal estate for value without notice took free; all others were bound. The 1925 legislation partly displaced that doctrine by requiring many equitable interests to be registered as land charges. Failure to register the relevant land charge usually makes the interest void against a qualifying purchaser, even if that purchaser knew of it. Midland Bank Trust Co Ltd v Green is the classic illustration.

The Cambridge difficulty is that students often treat registration as a conveyancing topic rather than as a priority topic. That is a mistake. Registration is the central organising device of modern land law. It determines the practical value of the rights studied in Week 1. An equitable lease, an estate contract, a beneficial interest under a trust, an easement, and a restrictive covenant may all be valid between the original parties; the acute question is whether each survives a disposition of the land.

For Tripos purposes, the topic rewards disciplined structure. Begin every problem by classifying the land: registered or unregistered? Then classify the right: legal estate, legal interest, equitable interest, beneficial interest under a trust, registrable disposition, interest capable of notice, land charge, overriding interest, or mere personal right. Then identify the transaction: sale, mortgage, gift, first registration, registered disposition for valuable consideration, short lease, or transfer of unregistered land. Finally apply the appropriate priority rule. Do not mix the old doctrine of notice into registered land unless the statute leaves room for it. Do not use registered-land language, such as notices and overriding interests, for unregistered land. The two systems overlap in policy, but their operative rules are different.

English conveyancing began from possession, deeds, and private investigation of title. A purchaser of unregistered land had to examine a chain of title, traditionally over a substantial period, to ensure that the vendor could convey what he purported to sell. Legal interests were relatively robust because legal title carried priority. Equitable interests, by contrast, were mediated through conscience. Equity would not permit a purchaser with notice of a prior equitable interest to defeat it; but a bona fide purchaser of a legal estate for value without notice was protected. This doctrine gave equity moral flexibility but imposed heavy informational costs. It also produced awkward distinctions between actual, constructive, and imputed notice.

The Law of Property Act 1925 and associated legislation sought to simplify conveyancing. The 1925 settlement did not produce immediate universal registration of title. Instead it rationalised unregistered conveyancing while encouraging eventual registration. Two devices are especially important. First, overreaching was strengthened: where capital money is paid to at least two trustees or a trust corporation, many beneficial interests under trusts are lifted from the land and transferred to the proceeds of sale. That protects purchasers while preserving the beneficiaries’ economic claims. Secondly, the Land Charges regime required specified equitable interests over unregistered land to be registered centrally against the estate owner’s name. The doctrine of notice was correspondingly displaced for those categories. The reform was powerful but imperfect: it required accurate name-based searches and generated severe consequences for non-registration.

Registration of title has a different ambition. It aims to make title itself a public, state-maintained record. The register is not merely a list of charges; it is intended to state the estate, the proprietor, and protected burdens. The classic policy triad is often described as mirror, curtain, and insurance. The mirror principle says that the register should reflect the state of title. The curtain principle says that some equitable interests, especially under trusts, should be kept off the register and dealt with behind the curtain by overreaching. The insurance principle says that the state should compensate those who suffer loss through errors in the register or certain rectifications. The principles are aspirations rather than absolute rules, but they explain the structure of the modern system.

The Land Registration Act 1925 created the previous modern regime; the Land Registration Act 2002 replaced it. The 2002 Act aimed to move registered conveyancing closer to electronic, disposition-by-registration conveyancing. It tightened the categories of overriding interests, introduced a sharper priority code in sections 28 and 29, and treated registration as constitutive of legal title for many dispositions. The point is not that registration supplies an evidential memorandum of an independently complete transfer. For many registrable dispositions, the disposition does not operate at law until registration. This is a profound doctrinal shift from deed-centred conveyancing to register-centred title.

Unregistered land has therefore become a residual category. It still matters because not all land is registered and because first registration raises transitional problems. It also remains examinable because it reveals why registered land takes its current form. The two regimes embody different answers to the same problem: how should the law allocate the risk that a purchaser does not know about an existing proprietary right? Unregistered land uses a mixture of legal priority, notice, land-charge registration, and overreaching. Registered land uses the register, overriding interests, statutory priority, and indemnity. A Cambridge answer should show the historical movement from private investigation and conscience towards public recording and statutory allocation of risk, while recognising that neither system has achieved perfect transparency.

1. Priority is distinct from validity. A proprietary right may be valid between the parties who created it but fail against a later purchaser. Conversely, an entry on the register may protect priority without proving the substantive validity of the interest. This distinction is essential. A notice on the register protects the priority of an interest; it does not guarantee that the interest exists. Equally, an unregistered estate contract over unregistered land may be perfectly enforceable against the contracting vendor yet void against a qualifying purchaser if it should have been registered as a land charge.

1. Always classify the title. Registered and unregistered land have different priority codes. In unregistered land, legal estates and most legal interests bind the world. Equitable interests are dealt with either by the doctrine of notice or by the land charges system. In registered land, the key questions are whether the disposition is registrable, whether it has been completed by registration, whether the competing interest is protected by notice or restriction, whether it overrides under Schedule 3, and whether overreaching has occurred.

1. In unregistered land, legal rights are generally binding without registration. If A grants B a legal lease for a term not exceeding the statutory maximum capable of existing at law, B’s legal lease will bind later purchasers. Legal easements similarly bind if properly created. The purchaser’s ignorance is usually irrelevant. This reflects the older idea that legal proprietary rights have inherent priority. There are qualifications, but the broad proposition remains central.

1. Equitable rights over unregistered land require separate treatment. Some are registrable as land charges under the Land Charges Act 1972. Estate contracts, restrictive covenants, equitable easements, puisne mortgages, and certain matrimonial home rights are examples of interests which may require land-charge registration. If the relevant interest is registrable but unregistered, the statutory consequence can be draconian: the interest is void against the specified class of purchaser. It is not merely postponed. The identity of the qualifying purchaser depends on the class of land charge. The traditional doctrine of notice is excluded where the land charges code applies.

1. Midland Bank Trust Co Ltd v Green is the warning case. A father granted his son an option to purchase the farm. The option was an estate contract and should have been registered as a land charge. It was not. The father then sold to his wife for a small sum, with knowledge of the son’s option. The House of Lords held that the unregistered option was void against the purchaser. Actual knowledge did not rescue the son. The case demonstrates that land-charge registration is a priority rule, not a notice rule in disguise.

The cases should be learned as illustrations of statutory priority, not as free-standing moral anecdotes. Midland Bank Trust Co Ltd v Green belongs to unregistered land. It demonstrates that the land charges system deliberately sacrifices equitable notice-based reasoning to certainty of search. The son’s option was vulnerable because it was not registered. The purchaser’s knowledge was irrelevant once the statutory conditions were met. A Cambridge answer should use the case to show the displacement of notice, not merely the harshness of the result.

Williams & Glyn’s Bank Ltd v Boland is the starting point for actual occupation. A wife’s beneficial interest under a trust, coupled with her occupation of the matrimonial home, bound the mortgagee as an overriding interest under the previous registered-land legislation. The case established that domestic occupation could have proprietary consequences against a lender. It also explains why mortgagees adjusted practice by requiring occupiers to consent, postpone, or obtain independent advice.

Abbey National Building Society v Cann qualifies Boland through timing and indivisibility of acquisition. The mother’s alleged interest could not override the mortgage because the acquisition and mortgage formed one indivisible transaction, and actual occupation had to exist at the relevant time. The case is central to purchase-money mortgages. It prevents an occupier from interposing an interest in the scintilla temporis between acquisition and the lender’s charge where the loan funds the purchase.

City of London Building Society v Flegg is the principal authority on overreaching in registered conveyancing. The beneficiaries were in occupation and had beneficial interests. Nonetheless, because the mortgage money was paid to two trustees, their interests were overreached and attached to the proceeds. The mortgagee therefore took free of those beneficial interests. The case gives practical content to the curtain principle. Beneficiaries behind a trust are protected by the trust machinery, not necessarily by binding the land.

Strand Securities Ltd v Caswell is a caution about actual occupation. The right-holder’s occupation must be occupation by the person claiming the interest, or through an appropriate agent or representative, not merely occupation by someone whose presence is unrelated to the claimant’s right. The case is useful because students often equate physical presence on the land with occupation by every person who has a proprietary claim.

Malory Enterprises Ltd v Cheshire Homes and Swift 1st Ltd v Chief Land Registrar expose the difficult border between registration’s conclusiveness and the correction of mistakes or fraud. Malory suggested that registration following a forged transfer did not necessarily destroy the former owner’s beneficial claim. Swift 1st, under the 2002 Act, reflects a stronger conception of title by registration, though still within a statutory scheme of alteration and indemnity. The broader lesson is that registration does not simply reproduce pre-registration conveyancing concepts. It creates a statutory estate, but the Act also contains mechanisms for correcting the register and allocating loss.

Thompson v Foy illustrates the limits of actual occupation protection where the occupier’s conduct and the timing of the relevant disposition are problematic. It is a useful modern example because it prevents over-generalisation from Boland. Actual occupation is intensely factual, but the right claimed must be proprietary, must exist at the relevant time, and must fall within the statutory terms.

See practice questions section below.

See further reading section below.