86 TAJINING MATTER, &C. 



... . ' outer extremity of the cock wonld be farther removed 

 of support may from any fixed point in the axis of the pendulum than if 

 be less eflfectii- {he expansion had been uniform throughout, in whieh 

 ^gj.^ " case the screws would have maintained their perpendicu- 



larity. The same reasoning will not apply to the simple 

 irod, the expansion being^ equal in all the concentric 

 strata. 



Mr. BoswelPs errors with regard to the pyrometer 

 have been so ably noticed by you, that it would be ab- 

 surd in me to touch upon them ; but his objections to the 

 gridiron pendulum, on account of the mixed metals of 

 which it is composed^ do not appear tq have been contro* 

 Gridiron pen- verted : Mr. Boswell ought, however, to have been awar^ 



dulums are ad- ^^at an adequate allowance for adjustment is constantly 



justed while j • i 



jictually going, "lade m their construction, to obviate any defects which 



may arise from a faulty estimation of the comparative 



expansions of the different parts of the instrument, as 



described in the account of the excellent seconds pendiv- 



|um of your own construction. 



I am, Sir, 



ypur obedient servant, 



J. G. 



III. 



A Third Series of Experiments on an artificial Suhsianc^ 

 zchich possesses the jmncipal characteristic Properties 

 of Tannin; zcith some Remarks on Coal. By CiiARLts 

 Hatciiett, Esq. F.R.S.* 



(Concluded from page 31.) 



VIT, j 



Experiments, AfTER the tanning substance and the other products 

 &c. on an arti- had been. obtained from the resins, balsams, &c. which 

 kavingthecha- ^^^^ ^^^" mentioned in the beginning of this Paper, the 

 racters of tan- following proportions of coal remained f : 

 ning matter. 



» Philos. Trans. i8c6. For the former Papers sec our Journal. 



f The weight of the coal obtained from each of the abovemen- 

 tioncd substances, was estimated after the complete separation of 

 every other product, and after the moisture had been expelled by ^ 

 ye4 l^at, in close vejsela, 



IGO grains 



