266 
process gives to the mule twist, the peculiar soft property 
that distinguishes it from the throstle or water twist. When 
the Crompton Mule came into general use, several improve- 
ments were made in its construction, hv different eminent 
mechanicians, of whom Mr. John Kennedy and Mr. Peter 
Ewart were the most eminent; but it would form a long list 
to point out the authors of those improvements before any 
notable efforts were made to render it a self-acting machine. 
It was the task of the “hand-spinner” to guide and 
regulate the “putting up” the carriage and winding the 
threads on the spindles, after they had been spun, and to 
perform these operations, the self-acting movements were 
devised, to render the mule independent of his skill. The 
real difficulty to render the mule a self-acting machine was 
chiefly confined to winding the threads on the continually 
changing forms and size of the cops. 
The first notable attempt to construct a self-acting Mule 
was made by Mr. Kiel Snodgrass, about the year 1815, but 
his plan appears not to have attracted much notice. The 
next was the patent obtained by Mr. William Eaton, of Wiln 
Mill (on the Trent in Derbyshire), dated in December 1818. 
Mr. Eaton has laid down with great exactness, the mechanical 
means of giving all the movements required for a self-acting- 
mule, which are separately and minutely described in his 
specification. Mr. Eaton and his sous established a spinning- 
mill, with his patent mules, to the extent of about 3000 
spindles, in Chepstowe-street, Manchester, which were in 
operation in the years 1821 and 1822, where they were open 
to general inspection. 
The next self-acting mule was patented by Mr. Richard 
Roberts, in 1825. 1 1 is winding-on process, however, being 
imperfect, prevented his mule from coming much into use 
until his second patent was brought out, in 1830, which was 
far superior to the former and was more extensively adopted. 
In the meantime, Mr. Maurice de Jongh, of Warrington, 
