al Mr. De Luc oft 
creating drynefs. Moreover, they did not follow one another 
in thofe motions contradi&ory to the evident march of moijlure : 
each of them changed direction at different periods, thereby 
often contradicting alfo each other, while the flips conftantly 
agreed together in the direction of their motion, and alfo with 
all the other fymptoms of moijlure. From thefe comparative 
phenomena I firft concluded ; that the motions of the fame 
kind, which I had obferved in the firft- mentioned threads, were 
alfo anomalies , proceeding, only to a fmaller degree, from a caufe 
of the fame nature as that of thefe laft. After which, fimilar 
fvmptoms, which I had formerly obferved in the water thermo- 
meter near the freezing point, made me firft conclude, from a 
general analogy ; that the perceivable modifications of the 
threads, were the compound effeds of two contrary operations 
of moijlure which followed different laws. 
34. Another phaenomenon led me foon after to a more de- 
termined theory in refpect of thofe two oppofite effeds of 
moijlure on threads . I have faid above, that hemp and gut have 
only a very little reirogradation ; their greateft 'difference from 
the flips confifting in their being jlationary, while the flips have 
ftill great motions. But when thefe fame threads are twifled, 
they acquire a very fenfible elongation beyond their point of 
extreme moijlure fucceeded by retrogradation. From feveral 
trials 1 have made in twifting thefe threads more and more, I 
do not confider as impoftible, if fome difficulties, which I 
only could obviate in part, were completely prevented, that 
they might be brought to fuch a ftate, as to have their point of 
extreme drynefs coincide with that of extreme moijlure ; by which 
means, in the progrefsof moijlure from one extreme to the other, 
they would move firft in one direction with decreafing fteps, then 
in the oppofite direction by increafing fteps ; the whole, however, 
5 with 
