Bruce J found that Mr Sumpter had abandoned the contract, and said he could obtain money for the value of the materials but nothing for the work.

AL Smith LJ gave the leading judgment: In this case the plaintiff, a builder, entered into a contract to build two houses and stables on the defendant's land for a lump sum.

The Court of Appeal held that there was in that case something from which a new contract might be inferred to pay for the work done by the plaintiff.

The case came before a Divisional Court, consisting of Lord Coleridge CJ and myself, and we said that the decision in Munro v Butt[5] applied, and there being no circumstances to justify an inference of a fresh contract the plaintiff must fail.

That is no evidence from which the inference can be drawn that he entered into a fresh contract to pay for the work done by the plaintiff.

If we held that the plaintiff could recover, we should in my opinion be overruling Cutter v Powell,[6] and a long series of cases in which it has been decided that there must in such a case be some evidence of a new contract to enable the plaintiff to recover on a quantum meruit.

There was nothing new in the decision in Pattinson v Luckley,[7] but Bramwell B. there pointed out with his usual clearness that in the case of a building erected upon land the mere fact that the defendant remains in possession of his land is no evidence upon which an inference of a new contract can be founded.

He says: “In the case of goods sold and delivered, it is easy to shew a contract from the retention of the goods; but that is not so where work is done on real property.” I think the learned judge was quite right in holding that in this case there was no evidence from which a fresh contract to pay for the work done could be inferred.Collins LJ concurred.

I think the case is really concluded by the finding of the learned judge to the effect that the plaintiff had abandoned the contract.

The mere fact that a defendant is in possession of what he cannot help keeping, or even has done work upon it, affords no ground for such an inference.