In June 1787, two men from Darlington, John Kendrew, a glass-grinder, and Thomas Porthouse, a watchmaker, registered a patent for a new flax-spinning machine. The equipment, was made up of a drawing frame with drums large enough to roll out long flax fibres and a spinning-frame with four spindles. After hearing about this invention, John Marshall, a linen merchant from Leeds, visited Darlington and the men provided him with a demonstration of the machine in operation. Impressed by what he saw, Marshall purchased the right to make his own copies of the flax-spinner.
After obtaining two partners, Samuel Fenton, a Unitarian draper, and Ralph Dearlove, a linen merchant, Marshall leased Scotland Mill near Leeds. Early in 1788, Marshall, Fenton and Company, began spinning flax yarns. At first the machines did not perform well. Breakages frequently occurred and the yarn came out lumpy and hairy.
Although Marshall had little technical experience, he spent the next two years trying to improve its performance. He made little progress until he recruited a talented engineer, Matthew Murray, to help him. Murray, who was later to be involved in the development of the locomotive, took charge, and by June 1790 he had created an efficient machine that produced good quality yarn.
My attention was accidentally turned to spinning of flax by machinery, it being a thing much wished for by the linen manufacturers. The immense profits which had been made by cotton spinning had attracted general attention to mechanical improvements and it might be hoped that flax spinning, if practicable, would be equally advantageous. It would be a new business, where there would be few competitors, and was much wanted for the linen manufacture of this country.
I found the (flax-spinning) machine very liable to be out of order which was in part owing to the badness of the workmanship. In general it only produced about 3 pieces week. I set Matt Murray to work on a new loom.
The heckle seemed to be large enough for the length of fiber but the rollers had not twine enough to pull it out of the teeth of the heckle and it frequently went round with the heckle. Matthew proposed to move the fluted roller in the usual way and cut a part of it away so that the twine from the roving spindle might pass the roller and get hold of the ends of the fibers that were drawing, then the roller would draw them out taking hold of the ends that were twined together, which would answer the same end of twining and drawing at the same time.
Murray placed the hackle roller in a right line between the feeding and drawing rollers. It (the liver) came very well off the heckle roller, and drawing and twisting by hand produced a perfect roving, it therefore appeared certain that the principle of it would answer - and the fault of the drawing roller seemed to be solely its running forward over the twined part and swallowing up the twine.