Rylands v Fletcher

Rylands v Fletcher is the classic common law instance of liability without proof of negligence for damage caused by the escape from land of a dangerous accumulation. Its practical importance is now modest. Its intellectual importance remains considerable because it forces the law of tort to state how far liability may rest on risk-creation rather than fault, and because it sits uneasily between private nuisance, negligence and statutory environmental regulation.

The modern rule is not that any dangerous thing kept on land makes the occupier an insurer. That is the first error to avoid. The claimant must show an accumulation or bringing onto land for the defendant’s purposes; an escape from land under the defendant’s occupation or control; damage of a relevant kind; a use of land which is extraordinary or unusual, often called non-natural use; and, after Cambridge Water, foreseeability of damage of the relevant type. The House of Lords in Transco confined the rule sharply. Lord Bingham treated it as a sub-species of nuisance, concerned with exceptional hazards to land or interests in land, not as a general doctrine of strict liability for dangerous activities.

For Cambridge Part IA purposes, Rylands is best learned as a coda to Week 11 private nuisance, not as a free-standing alternative to negligence. It shares nuisance’s concern with land, locality, reciprocity and neighbourly relations. Yet it differs from ordinary nuisance because it is triggered by escape and is historically associated with strict liability. The examination problem is usually whether the defendant’s activity crosses the line from ordinary land use into an exceptional risk imposed on neighbours. A water pipe in a block of flats, domestic plumbing, ordinary heating and many industrial activities authorised and regulated by modern society will not normally suffice. Explosives, toxic chemicals, large-scale storage of highly flammable materials, substantial reservoirs and analogous exceptional accumulations are more promising.

The rule’s doctrinal history is also examinable in essay form. In the nineteenth century it appeared as a robust enterprise-liability principle, responding to industrial accidents. In the twentieth century it was restricted by non-natural use, escape, personal injury exclusions and defences. In Cambridge Water and Transco it was rationalised as a narrow land-based tort, requiring foreseeability and confined to exceptional danger. That development has both virtues and costs. It brings the tort into line with modern negligence and nuisance; but it also deprives the common law of a clear response to abnormal risk where proof of fault is difficult.

In supervision essays, the strongest answers do not merely recite the elements. They ask what the rule is for. If it is a principle of corrective justice, it may seem right that a person who introduces an abnormal danger should bear resulting losses. If it is simply nuisance by another name, it should be confined to interferences with land and neighbourly amenity. If it is an embryonic scheme of environmental liability, it has been overtaken by legislation. The modern law largely chooses the second answer.

Rylands v Fletcher arose from the industrial landscape of mid-nineteenth-century England. The defendants constructed a reservoir to supply water to their mill. Unknown to them and to their contractors, disused mine shafts lay beneath the site and connected with the claimant’s working mine. When the reservoir was filled, water escaped through the old shafts and flooded the claimant’s mine. There was no finding that the defendants had personally been negligent. The legal question was therefore whether the common law could impose liability for the consequences of an extraordinary accumulation of water even without proof of want of care.

Blackburn J in the Exchequer Chamber formulated the celebrated rule. A person who, for his own purposes, brings onto land and collects and keeps there anything likely to do mischief if it escapes must keep it in at his peril. That language sounded uncompromising. It was not simply a factual rule about reservoirs. It expressed a principle: where A introduces an artificial danger onto his land, and it escapes onto B’s land, A rather than B should bear the loss. The House of Lords affirmed the result, but Lord Cairns introduced the limiting language of non-natural use. The defendant was liable because he used the land in a special way, not because every accumulation on land created strict liability.

This history matters because the later law is a history of containment. Victorian judges faced mines, canals, reservoirs, gas, railways and factories. They were developing tort law before modern regulation, compulsory insurance, health and safety legislation and public environmental enforcement. Rylands could be understood as a common law answer to industrial risk. The person who gained from a hazardous enterprise should internalise its accidental costs. That idea has continued appeal, especially where the claimant cannot know what precautions were taken on the defendant’s land.

But the common law soon became anxious about the breadth of the rule. Ordinary life involves accumulations capable of causing damage if they escape: water in pipes, electricity in cables, fire in hearths, oil in tanks, domestic waste, tree roots, animals and vehicles. If Rylands applied to all of these, landowners and occupiers would be exposed to very wide liability irrespective of fault. The doctrine was therefore cut down through several devices: the thing must escape; the use must be non-natural; the damage must be of a kind recognised by the tort; some acts of strangers and acts of God provide defences; and the claimant’s consent or common benefit may matter.

The relationship with nuisance also shaped the doctrine. Many Rylands cases could have been pleaded as nuisance if there had been a continuing interference with land. The distinctiveness of Rylands lay in escape and strictness: a single catastrophic escape could suffice. Yet nuisance itself is not purely fault-based. It often imposes liability for unreasonable interference even where care was taken. This made it easier, in the modern cases, to describe Rylands as a branch of nuisance rather than a separate strict-liability tort.

Two late twentieth-century decisions are decisive. In Cambridge Water, chemical spillages at a tannery contaminated groundwater and eventually affected a water company’s borehole. The House of Lords rejected liability because the relevant damage was not reasonably foreseeable when the solvent escaped. This imported into Rylands a remoteness requirement analogous to negligence and nuisance. In Transco, a water pipe serving a block of flats leaked into an embankment and exposed a gas main. The House of Lords held that the ordinary supply of water to domestic premises was not a non-natural use. The speeches stressed that Rylands survives, but only in a restricted form.

The historical arc is therefore clear. Rylands began as an ambitious principle of liability for abnormal accumulations. It has ended as a narrow land-based rule for exceptional risks, largely overlapping with nuisance and often displaced in practice by negligence or statute. In Cambridge exams, that arc is often more important than the bare outcome of any single case.

The claimant should structure any Rylands analysis around seven questions.

First, did the defendant bring onto land, or accumulate on land, something likely to do mischief if it escaped? The language of bringing and collecting is not confined to tangible chattels, but the paradigm cases involve physical things: water, chemicals, gas, explosives, fire, sewage, electricity, or large quantities of flammable material. The rule is less apt where the danger consists merely in an activity without a stored or accumulated source of danger. In Stannard the Court of Appeal emphasised that the relevant thing must be the thing that escapes and causes damage. The defendant stored tyres; the fire escaped. Tyres did not escape. That reasoning shows how technical the element can be, and how closely it confines the tort.

Secondly, was there an escape? Escape means escape from land occupied or controlled by the defendant to a place outside that occupation or control. This is a hard boundary. In Read v Lyons, an inspector injured by an explosion inside a munitions factory failed because there had been no escape from the defendant’s premises. The requirement reflects the tort’s kinship with nuisance: it protects neighbouring land interests against externalised hazards, rather than workers or visitors against dangers encountered within the defendant’s premises. If the claimant is injured on the defendant’s land, negligence, occupiers’ liability or employers’ liability will usually be the relevant framework, not Rylands.

Thirdly, was the defendant’s use of land non-natural, extraordinary or unusual? This is the central modern restriction. The phrase non-natural does not mean artificial. Domestic plumbing, ordinary gas supply, electricity, heating and routine commercial uses may be artificial but are not for that reason non-natural. The question is evaluative and contextual. It asks whether the defendant has introduced onto land a special danger, extraordinary in relation to ordinary land use and the circumstances of the locality. Lord Bingham in Transco preferred the language of extraordinary and unusual use because natural and non-natural can mislead. The test is not whether the activity is beneficial or socially desirable; many useful activities are still hazardous. But common benefit and ordinariness in modern life weigh strongly against liability.

The cases form a pattern of expansion followed by restriction.

Rylands itself supplies the starting point: liability for the escape of water from an artificial reservoir into neighbouring mines, despite absence of personal negligence. The Exchequer Chamber’s language is broad, but the House of Lords’ adoption of non-natural use became the principal brake on the doctrine. The case should not be read as imposing liability for every escape of water. It concerned a large artificial reservoir created for industrial purposes over hidden mine workings. Its moral force lay in the defendant’s introduction of an unusual danger for his own enterprise.

Rickards v Lothian is the leading early restriction. Water escaped from a basin after a stranger deliberately blocked it. The Privy Council treated ordinary domestic water supply as natural or ordinary use and stressed that the rule is confined to some special use bringing increased danger to others. This is important for modern cases involving plumbing, tanks and pipes. The mere fact that water can cause serious damage if uncontrolled is insufficient.

Read v Lyons fixes the escape requirement. The claimant, a government inspector, was injured by an explosion within a munitions factory. Even if explosives are plainly dangerous, there was no escape from the defendant’s land. The case also foreshadows the modern reluctance to use Rylands for personal injury. Students should not treat it as a general denial that explosives can found Rylands liability. Rather, it shows that the claimant must be outside the defendant’s sphere of occupation or control and must complain of an escaped danger.

Cambridge Water is the modern remoteness case. Solvent spillages at a tannery migrated through soil and groundwater and contaminated a borehole some distance away. The House of Lords accepted that storage of substantial chemicals might be non-natural, but rejected liability because the relevant type of damage was not foreseeable at the material time. The decision is pivotal because it rejects an absolute version of strict liability. Rylands is strict as to fault once the elements are met; it is not strict as to unforeseeable kinds of harm.

Transco is the leading contemporary restatement. Water leaking from a pipe serving a block of flats damaged an embankment supporting the claimant’s gas main. The House of Lords held that the supply of water to domestic premises was not a non-natural use. The speeches also recast the tort as a sub-species of nuisance, concerned with exceptional danger and damage to land. This decision is indispensable in any essay because it defines the modern, narrow survival of Rylands.

Stannard v Gore shows the Court of Appeal applying that restrictive approach. A tyre-fitting business stored tyres; a fire started and spread to neighbouring premises. The claim failed under Rylands. The thing accumulated was tyres, but the thing that escaped was fire. The storage of tyres in the ordinary course of that business was not, without more, sufficiently exceptional. The decision illustrates that courts will not infer Rylands liability merely from foreseeable combustible risk.

Crown River Cruises v Kimbolton Fireworks is a useful contrast. The conduct of a firework display on a barge caused fire damage to neighbouring property. Fireworks are a classic exceptional danger; their escape and ignition are precisely the mischief that makes them hazardous. Though not a House of Lords decision, it remains a strong problem-question authority where facts involve explosives or pyrotechnics.

Together these cases produce the modern map. Reservoirs, explosives and exceptional chemical accumulations are plausible. Ordinary utilities, domestic water, workplace accidents within the premises, and normal commercial storage are usually not. The controlling question is no longer whether the thing is capable of doing mischief in the abstract. It is whether, viewed in context, the defendant imposed an extraordinary escape-risk on neighbouring land.

See practice questions section below.

See further reading section below.