3 $ Mr. de Luc on 
refpect of the thread of deal and quill ; and I have obferved It 
conftantly, in a fmailer proportion,' in all the threads which 
have their Jlationary ftate at the laft period of moijlure , with 
this particular circumftance in all thofe motions backwards and 
forwards, that they are never the fame in two different experi- 
ments. That fymptom already points out a complication of 
caufes ; but we fhall loon fee a more diftindt proof of their 
exigence. 
56. My theory on the march of hygrofcopic threads is 
founded on this general principle, that a retrograde effedt, how 
fmall foever it may be, if it is not produced by a correfpon- 
dent change in the caufe itfelf, is preceded by a Jlationary ftate, 
during which, and the retrogradation, the intenfity of the 
caufe continues to increafe : and this alfo is exemplified in the 
experiment. The Jlationary ftate of the thread of deal begins, 
when its Jhavings have only imbibed from the air a quantity 
of water = 36 ; it is ftill at the fame point when the quantity of 
imbibed water has increafed to 59,9 ; and in the part of its 
retrogradation, which is ftill contained in the regular courfe of 
the experiment, that quantity of imbibed water increafes to 
88,8. The fame, with only fome differences in the degree, is 
feen in the thread of quilh, therefore, as we fee alfo fome, 
though a very fmall, retrogradation in the thread of whalebone , 
as well as in other threads of the fame clafs, we have reafon to 
conclude, that their apparent immobility before they come to 
that point, while th tjlips continue to move, is alfo a Jlationary 
ftate, during which they continue to receive water, by the 
increafe of moijlure in the air. 
57. The experiments I have now analyfed are only one fet 
amongft others which, though made with lefs accuracy, have 
given the fame general refults. Thefe, relating to various 
1 kind 3 
