(1) John Marshall, Autobiography (unpublished)

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.

(2) John Marshall was disappointed by the performance of the Kendrew & Porterhouse Flax Spinner.

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.

(3) Matthew Murray began experimenting with the Flax Spinner in September 1789.

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.

(4) The fluted roller with one side cut away did not solve the problem. Matthew Murray therefore decided to change the position of the hackle roller.

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.